Lecture notes

in

Epidemiology and research methodology

Prepared, edited and compiled by:

Dr JZJ Killewo
Dr SH Kapiga
Dr I Semali
Dr G Kwesigabo

Muhimbili University College of Health Sciences

Institute of Public Health
Department of Epidemiology and Biostatistics
Dar es salaam, 1995
 2

Contents -----------------------------------------------------------------------Page

Table of contents --------------------------------------------------------------------------- 2

Foreword ------------------------------------------------------------------------------------ 3

I. PRINCIPLES OF EPIDEMIOLOGY

1. Introduction to epidemiology -------------------------------------------------------- 5

2. Ecology of disease and models of disease causation ----------------------------- 8
3. Natural history of disease and levels of prevention------------------------------16
4. Principles of disease transmission--------------------------------------------------19
5. Theories and strategies of disease Control ----------------------------------------23
6. Measurement of health, morbidity and mortality --------------------------------25
7. Sources and uses of morbidity and mortality data -------------------------------33
8. Rates and standardization of rates--------------------------------------------------38
9. Testing and screening for disease --------------------------------------------------42
10. Epidemiologic approach to causation----------------------------------------------49

II. METHODS OF EPIDEMIOLOGY

1. Introduction to epidemiological methods and research methodology ---------59

2. Descriptive studies (Ecological, Cross-sectional etc.)---------------------------67
3. Case-control studies ------------------------------------------------------------------71
4. Cohort studies -------------------------------------------------------------------------76
5. Experimental studies -----------------------------------------------------------------80

III. RESEARCH METHODOLOGY

1. Investigation and control of epidemics --------------------------------------------91

2. The research proposal and the research process----------------------------------97
3. Writing of scientific papers for publication------------------------------------- 103
4. Evaluation and critique of a scientific articles---------------------------------- 109
5. Sources of error in epidemiologic studies and their findings ----------------- 113

IV. EPIDEMIOLOGY OF SPECIFIC DISEASES

1. Epidemiology of major diseases in Tanzania ------------------------------------ 118

 3

Foreword

At last we have come up with a manual of lecture notes in Epidemiology and Research
Methodology after many attempts since 1985. The motivation to do this work was a result of a
chronic shortage of books and even when these became available they were very few compared to
the number of students who needed to use them. There was therefore, a scramble for books with
irresistible urge and temptation by many students to ‘beg, borrow or steal’ the few books in order
not to return them. These ‘lost’ books became a hot cake among the students and were being passed
from one batch of students to another until a fresh supply was ‘begged borrowed or stolen’ again.

As years passed by, the situation became worse, and at times we were not even able to provide
handouts for every lecture given due to the meager resources provided to the department in terms of
stationery and photocopying services. With the advent of cost sharing in which students were
required to contribute towards acquisition of lecture handouts the department provided one copy of
notes for further reproduction by the students themselves. This has been satisfactory to the
department but perhaps not equally so for the students since the notes have been loose and scattered
here and there making revision for examinations difficult.

This set of lecture notes is the answer to some of the problems mentioned above. However, it must
be properly utilized. It is not to be taken as a replacement for books and lectures. Relevant books
and other literature must be consulted as indicated at the end of most notes. It is recommended that
the notes are read before a scheduled lecture is delivered so that students can participate fully during
the lecture by asking questions about concepts which might not have been clear to them during the
time they first read the notes.

The department will endeavour to up-date the notes from time to time to keep students abreast with
new developments in the field of epidemiology and research methodology. We sincerely hope that
our students will enjoy the subject and that this set of notes will become a useful guide towards their
goal.

Dr. J. Z. J. Killewo
HEAD OF DEPARTMENT
SEPTEMBER 1995
 4

1
PRINCIPLES OF EPIDEMIOLOGY
 5

Introduction to epidemiology
Epidemiology may be defined as the study of the distribution and determinants of disease
frequency. These three closely interrelated components - distribution, determinants and frequency -
encompass all epidemiologic principles and methods.

Epidemiology is a multidisciplinary subject which has borrowed from demography, statistics,

sociology, and other sciences to become a distinct discipline with its own philosophy, concepts and
methods during the nineteenth century. You will have the opportunity to learn more about the
concepts and methods of epidemiology in the course of your programme to study epidemiology. In
order for you to appreciate the concepts and methods in this module we shall learn about the
history, the scope and uses of epidemiology. The module will conclude by a brief statement on the
analogy of the epidemiologic method to the scientific method on which the whole philosophy of
science is based.

History of Epidemiology

The concepts of epidemiology have their origin in the works of Hippocrates "On Airs, waters and
places" which stress the importance of considering the variety of environmental influences on
diseases in humans, e.g. the hot, the cold, the winds, the water quality, and whether the inhabitants
are fond of drinking and eating to excess and given to indolence or are fond of exercise and labour
and not given to excess in eating and drinking. This concept provides for appropriate definition,
description and analysis of a variety of situations in efforts to investigate causal relationships in
disease occurrence with the ultimate aim of preventing disease.

In an attempt to investigate and control epidemic diseases in historic times two theories evolved in
the course of time: The contagion vivum and the miasma theories.

The Contagion vivum Theory.

According to this theory a living contagion was thought to be involved. This theory necessarily
depended on two other concepts:
(i) The specificity of both disease and its cause
(ii) The existence of microscopic organisms

Although this theory is relatively modern (mid 1800's, and about 25 years before the time of Louis
Pasteur) some practices commonly associated with it are quite old. For example, the custom of
isolating people with contagious diseases is recorded in the bible and was continued by the Roman
Catholic Church during the Middle Ages. Girolamo Fracastoro (1478-1553) was the first to state
explicitly that specific diseases directly result from specific contagions, but he did not have any
notion of micro-organisms. One of his contemporaries, Gilamo Cardano, stated in 1557 that the
seeds of disease were minute animals capable of reproducing their kind. You might describe this
stage of development as the beginning of the science of microbiology.

The Miasma Theory.

This theory, dating back to the early 1700's, offered an alternative explanation for the origin of
epidemics. The idea was based on the notion that when the air was of a bad odour or quality the
persons breathing that air would become ill. For example, "malaria" means "bad air". This theory
led to the total abandonment of the contagion vivum theory, but after the 1720 Marseilles plague
epidemic which could only be scientifically explained by the contagion vivum theory and the
discovery of the fungal origin of muscardine, a silkworm disease, by Agostino Bassi in 1816 the
contagion vivum doctrine was adopted by many European scientists. and the miasma theory
subsequently abandoned. Indeed, the contagion vivum theory was the scientific basis for the
 6

initiation of the filtration of part of the London water supply in 1829 to remove possible disease
contagions.
In essence, the current era of epidemiology began in the mid 1800's when the contagion vivum
theory and the evolving epidemiologic study merged.

The essence of the epidemiologic study is the comparison of groups of people with regard to a
characteristic of interest in order to determine relationships in disease occurrence. The earliest
account of such a comparison is found in the Old Testament in the first chapter of the Book of
Daniel. Today an epidemiologist reading this account will have much to say regarding the
composition, selection and allocation of the comparison groups into the different food allocation
categories (in modern epidemiology - the treatment groups). Bias resulting from the way the
outcome of treatment was assessed in this biblical account forms yet another subject of
epidemiology we shall study. In epidemiology, comparisons of groups are so common that one can
not avoid the use of the tool even in their simplest situations. For this reason investigators should
be aware of the misleading use of numerators in making epidemiological comparisons.

John Graunt, inspired by Bacons writings collected the bills of mortality which had been initiated in
1603 by the parish clerks in London. After organising the published bills he derived from them
inferences about mortality and fertility in the human population. He noted the following inferences:

1. excess of male births

2. high infant mortality
3. seasonal variation in mortality

From the collected data Graunt constructed the first known life table, summarising the mortality
experience in terms of the number, percent or probability of living or dying over a lifetime.

James Lind used the idea of comparative groups in studies to elucidate the aetiology and treatment
of scurvy. From the results of his study, Lind inferred that citrus fruits cured the scurvy and that this
would also provide a means of prevention.

William Farr was the first compiler of Abstracts of the Registrar General's Office and organised the
first modern vital statistics system.

John Snow conducted a series of classical studies on cholera (1848 - 1854) and inferred the
existence of a "cholera poison" transmitted by polluted water. This led to a legislation mandating
that all of the water companies in London filter their water by 1857. However, It was not until 1883
that Robert Koch identified the cholera vibrio.

Sir Benjamin Ward Richardson, a pioneer of the sanitary movement, was among the first to develop
a disease reporting system and to propose the teaching of preventive medicine in medical schools in
the 1860's.

Goldberger demonstrated in 1914 by comparative experimentation that pellagra was not an

infectious disease - the then prevailing concept - but rather a result of a dietary deficiency of a
vitamin - later determined to be nicotinic acid (vit. B-complex)

Much of our epidemiological knowledge on measles is credited to P.L. Panum who studied the
measles epidemic on the Faroe Islands in 1846.

Sir A. Bradford Hill pioneered the application of epidemiologic methods to non-infectious diseases.
 7

The scope and uses of epidemiology

A. Description of the distribution of diseases in a population

B. Identification of etiological factors in pathogenesis
C. Provide and analyse information for the planning, implementation and evaluation of health
services
D. Study the natural history of disease

The epidemiological and scientific method

Stages in the scientific method

A. Identification of the problem, its importance and feasibility

B. Description of the problem
C. Formulation of a hypothesis
D. Testing of the hypothesis
E. Adaptation of the theory
F. Identification of new problems

Steps in the epidemiological approach to studying a problem of disease aetiology :

A. Initial observation in laboratory or clinical findings

B. Definition of the disease or process by;

- pathology
- common clinical characteristics
- specific etiological agent

C. Descriptive epidemiology : studying the distribution of disease in relation to time, place, and
person

D. Analytical epidemiology : identifying associations of disease with possible etiological factors

E. Further refinement and testing of hypotheses regarding aetiology

F. Experimental epidemiology : evaluate the effectiveness of modifying etiological factors in

order to control or prevent the disease in a population

References for further reading;

1. Epidemiology - An Introductory Text. Mausner JS and Kramer S. WB Saunders

Company, London 1985 pg 26-44.
2. Foundations of Epidemiology Lilienfeld and Lilienfeld Oxford University Press
3. Epidemiology: Principles and Methods MacMahon, B. and Pugh, T.F. Little, Brown Co.,
Boston
4. Basic Epidemiology R. Beaglehole, R. Bonita and T. Kjellstrom, World Health
Organization, Geneva, 1993.
5. Modern Epidemiology. Rothman KJ 1986. Little Brown and Company. Boston/Toronto
 8

Ecology of disease and models of disease causation

Introduction:

The epidemiologic patterns of infectious diseases depend upon factors that influence the
probability of contact between an infectious agent and a susceptible person, known as the
host. This probability of contact is determined by an intricate interplay between factors related
to the triad of agent, host and environment. It is important to approach disease causation in this
manner as it gives us an insight necessary for developing rational preventive measures.
Examples of factors and associated diseases are listed below:

I. Factors related to the Agent

A. Nutritive elements
Factor Disease
Excess or deficiency of cholesterol atherosclerosis
calories Kwashiorkor
protein Marasmus
iodine goitre
iron, folic acid anaemia

B. Chemical Agents
Presence of poisons or allergens
pesticides poisoning
drugs, alcohol intoxication/dependency
allergens eczema
pollen hay-fever

C. Physical Agents
energy, speed accidents
solar radiation sun burns
radioactivity neoplasms
D. Infectious Agents

Bacteria e.g. Mycobact.tuberc. tuberculosis

Vibrio cholerae cholera
Viruses e.g. measles virus measles
polio virus poliomyelitis
smallpox virus smallpox
Rickettsia e.g. Rickettsia prowazeki typhus
Rickettsia conorii tick bite fever
Protozoa e.g. Plasmodium malariae malaria
Entamoeba histolytica amoebiasis
Trypanosoma trypanosomiasis
Helminths e.g. Schistosoma haematobium urinary schistosomiasis
Ascaris lumbricoides ascariasis
Fungi e.g. Trychophyton spp. tinea corporis
 9

II Factors related to the host

A. Age paralytic polio, the ratio of paralytic cases to infections

increases with age. 1:1000 among young children and 1:75
among adults.
B. Sex Males usually have higher poliomyelitis attack rate than
females
C. Genetic Persons with sickle cell trait are associated with a decreased
risk of malaria due to Plasmodium malaria. Blood group A
have increased risk of gastric cancer while group O have
increased risk for duodenal ulcers.
D. Ethnicity certain ethnic groups have increased risk for keloid and
Gastro-intestinal tract cancers
E. Physiology pregnancy-candidiasis, puberty-goitre, stress, nutritional state ,
fatigue
F. Immunology hyper-sensitivity, allergy
- active prior infection, immunisation
- passive maternal antibodies,
gamma globulin
G. Existing pathology pre-existing disease may initiate another by interfering with
immunity, e.g. malaria and herpes simplex or other intercurrent
disease
H. Behaviour personal hygiene, religion, customs, habits, utilisation of
health resources and related diseases

III Factors related to the Environment

A. Physical climate, geology, radiation, heat, light, air pollution associated

with chronic respiratory disease
B. Biological
- human population density,
- flora food source, influence on disease agents/vectors
- fauna influences presence of hosts/vectors and agents

C. Socio-economic
- occupation occupational hazards
- urbanisation crowding, stress
- development education, poverty, availability of health services
- disruption civil and other wars - Rwanda 1994/5, natural
disasters- earthquake in Kobe Japan, 1995.

Agent, host, environment and their interaction

If the characteristics of the agent, host and environment are studied in isolation, we will never
reach a full understanding of why diseases are distributed the way they are. The interaction
between agent, host and environment is the object of the study of ecology. The study of this
interaction will give clues to answers of questions like such as:

How and why do new pathogens and/or new diseases emerge ?

Why does the spectrum of disease change over time ?
Is it possible to control and finally eradicate infectious diseases ?
 10

Definition and basic concepts of ecology

Ecology is the study of the relationship of organisms to each other as well as to all
other aspects of the environment

Climax state: the final constant condition of flora and fauna. The equilibrium or balance
reached varies, depending on climate and soil. The climax state is not stable,
but a dynamic condition.

The main factor responsible for precluding stability is human activity. Human activity can
change the balance much more rapidly, profoundly and unpredictably than changes in weather,
climate or natural events. Examples are: irrigation projects, behavioural changes (e.g. sexual
behaviour), wars etc.

Food chain: from the smallest plants or animals up to large predators, plants and animals serve
as food for other species higher up in the food-chain. Predators (lion, cheetah and hyena) and
birds of prey (hawks, eagles) are at the end of the food chain. They can be exposed to
dangerously high levels of toxins and chemicals, originating from the environment, and
accumulated through the food-chain.

Habitat: This is the typical environment in which a certain species usually lives. The habitat
of different species is very much interdependent. Some species require very strict conditions,
and are restricted to limited geographic areas. Humans have adapted to an unusually wide
habitat, partially as a result of technology. The habitat of domestic animals naturally coincides
much with the human habitat. Psittacosis and toxoplasmosis are typical zoonoses originating
from pets.

Population density: fluctuations in animal and plant populations affect other species higher in
the food-chain. This may become relevant to human diseases, especially zoonoses (plague,
rabies). The density of human populations is an important variable too, as it affects food
availability, exploration of new areas (possible habitats for vectors: onchocerciasis,
trypanosomiasis, jungle yellow fever), increased numbers of domestic animals, in turn leading
to overgrazing and erosion/desertification. High human population densities (urbanization)
together with human patterns of socialisation provide good opportunities to diseases of close
contact (droplet transmission, sexual transmission).

It has been found, that the probability to contract certain diseases (e.g. measles) does not
depend on the proportion of susceptibles in a community, but rather on the absolute number of
susceptibles. In the USA measles transmission at present is confined to larger cities (over 0.5
million inhabitants.), where adequate numbers of susceptibles are available.

Socialisation: refers to the ways in which humans socialise or come in contact with each other.
Within each community/culture a great number of patterns exist: through work, school,
religion, recreation etc.

Adaptation: A heritable component of the phenotype which confers an advantage in survival and
reproductive success. In another perspective, it is the process by which organisms adapt to
environmental conditions. Two organisms co-existing in a host-parasite relationship, usually show the
phenomenon of adaptation: the parasite becomes less virulent, and the host develops resistance. Due to
selection pressure, more sensitive hosts will be weeded out, or relatively resistant mutants will develop.
This is in the interest of the parasite as well : the worst that might happen to the parasite, is extinction of
its obligate host. If man is only an occasional host, there is no pressure towards a milder course of the
disease in question, as long as an abundant reservoir of infection remains available elsewhere in nature
(e.g. rabies).
 11

Development of relative resistance has been described for measles on Fiji after 1875, when
20% of the population died, and this has been postulated as an explanation for the decreasing
mortality from tuberculosis in Europe early this century. The emergence of cholera in 1870 as a
hitherto unknown disease, has been explained by the incorporation of plasmids (small
corpuscles of genetic material outside the nucleus) coding for an enterotoxin, originating from
E. Coli, by innocent vibrios in the estuaries of Bengal. Also Vibrio cholerae seems to adapt to
less virulent strains.

Endemic syphilis, previously highly endemic in many arid regions like Botswana and Saudi-
Arabia, has been disappearing. As the treponemes are very fragile, and require very close
contact between infected and uninfected to get transmitted, increasing personal hygiene and
wearing of clothing might have forced treponemes in a specific 'niche' (special limited habitat
of a particular organism): its only refuge was intimate sexual contact.

Poliomyelitis has showed a remarkable change in its spectrum: only during this century it has
changed towards much feared epidemics of infant paralysis. Presumably due to improved
hygiene and social-economic conditions and opportunities to spread in early infancy (when
polio only causes mild diarrhoea, rarely followed by paralysis). From an endemic disease of
early childhood, it became epidemic amongst school-age children, with frequent sequelae of
paralysis.

Herd-immunity: an inverse relationship exists whereby, as individual immunity decreases

there is an increase in the probability that an individual will develop a particular disease when
exposed to an infectious agent. High herd immunity indicates the decreased probability of a
group or community to develop an epidemic upon introduction of an infectious agent, although
there may be a certain number of persons who are individually susceptible to the disease agent.
This decreased probability (resistance) is a product of the number susceptible and the
probability that those who are susceptible will come into contact with an infected person.

At which percentage of immunes herd immunity occurs depends on the disease concerned, and
on socialisation patterns. Measles transmission still occurs in places like Tanzania and other
tropical developing countries. In these places the measles virus may be present in the
community, The number of cases depends upon the number of susceptible persons (children)
exposed to the virus. If there is a cyclic pattern with peaks occurring at intervals of every 2-3
year, this suggests that every 2-3 years there is a sufficient number of new susceptibles, hence
low herd immunity.

Childhood diseases: diseases do occur at that particular age when they are highly infectious.
Such diseases usually provide lifelong immunity and socialisation patterns provide ample
opportunity for transmission at a certain age (e.g. school entrance). As they typically occur in
epidemics, which cease when the number of susceptibles is almost exhausted. They are
anthroponoses and require a large human population, subdivided in fairly large communities. A
disease like measles cannot persist on smaller islands; it only (re-)occurs when imported.
 12

Epidemiological Models of Disease Causation

The occurrence of a disease in a person or community can seldom be attributed to a single

cause or factor. It is often due to a number of factors whose interplay is responsible for the
disease process. Disease causative factors (aetiological factors) may be divided into two types:

1. Essential or necessary factors

These are referred to as factors that must be present (a sine qua non) for the disease to occur.
These are the Agents of disease such as bacteria and viruses in infectious diseases, and fire,
nutrition, radiation or various poisons in non-infectious diseases. Many, but not all of the
known agents of disease are located in the biological and physical environment. A possible
agent in the social environment is maternal deprivation. Many studies have demonstrated that
the quality of parental care in the early years of life is intimately related to normal physical,
emotional, and mental development. From an ecological point of view, an agent is considered
to be a necessary but not sufficient cause of disease because suitable conditions of the host
and environment must be present for the disease to develop.

2. Contributory factors:

These are host and environmental factors which are associated with increased likelihood of
disease occurrence. e.g.: Host factors: immunity, sex, age. Environmental factors: social
(poverty, ignorance), physical (rainfall, temperature) or biological (presence of vectors, animal
reservoir). The interplay of these factors - also called ecology - can be presented in the form of
a model.

(Ecology: study of the inter-relationship between living things (agent and host) and the
environment. Human ecology : the study of human groups as influenced by environmental
factors often including social and behavioural factors). (Model: simplified framework which
can be used to represent more complex systems)

The Triangular Model

In this model, also called the epidemiological triad, three main components are important in the
chain of disease transmission.

These are: AGENT, HOST and ENVIRONMENT

AGENT

ENVIRONMENT HOST

Fig 1 The Epidemiologic Triangle

 13

Under stable ecological conditions, the epidemiological triad is in a balanced state, and the
disease can be said to be absent or endemic. If this balance is disturbed and becomes
unfavourable to the agent, the incidence of the disease will decrease. If the situation remains
unfavourable the disease may become sporadic or even disappear. If, on the other hand, the
balance alters to favour the agent, then an epidemic may occur. Although a lot may be known
about agents of disease, the role of host and environmental factors is incompletely understood.
For example, we do not know why only some of the people exposed to large doses of X-rays
develop leukaemia or why not all heavy smokers do develop lung cancer.

Some diseases such as schizophrenia, coronary heart disease, rheumatoid arthritis and essential
hypertension have not been linked to any specific causative agent. In this respect, new models
have been developed which de- emphasise the role of the agent and rather stress the multiplicity
of interactions between host and environment. These model are:
1. the wheel model
2. the web of causation model

The wheel model

In this model, the wheel consists of the hub, (the host) which has genetic make-up as its core.
Surrounding the host is the environment, schematically divided into the three sectors viz:-biological,
social and physical. The relative sizes of the different components of the wheel depend upon the specific
disease problem under consideration. For hereditary diseases the genetic core would be relatively large.
For a condition like measles, the genetic core would be of lesser importance; the state of immunity of the
host and the biological sector of the environment would contribute more heavily.

PHYSICAL BIOLOGICAL
ENVIRONMENT ENVIRONMENT
- temperature - reservoir
- rainfall HOST - vectors
- altitude - heredity - microbes
- radiation - sex (agents)
- chemicals - age

SOCIAL ENVIRONMENT
- socio-economic conditions
- occupation
- education
- culture
- customs
- beliefs

Fig 2 The Wheel model of host-environment interactions

 14

If we consider road traffic accidents as one of the diseases, there are a number of factors which
influence the occurrence of an accident :

1. Host factors:
- age : young drivers are more prone to accidents than older ones.
- driving experience: less experience, more accidents
- drinking habits : a driver who is under the influence of alcohol is more prone to causing
accidents than the one who is not.

2. Social and environmental factors:

- national policies on licensing: only the fully qualified should be granted licenses.
- policies on speed limits and zebra crossing
- availability of programs for drivers education.
- sign posts to indicate danger

3. Physical environmental factors:

- mechanical or electric conditions of the vehicle
- highway designs - tarmac, rough or dusty
- weather - e.g. rain, fog or snow

4. Biological environment:
- animals on the road may cause accidents
- pets and insects (tsetse or bees) in the vehicle may cause accidents

Intervention for control: all of the above factors should be intervened in order to reduce the
incidence of accidents.

In coronary heart disease or essential hypertension, the most important role is played by host
and social environmental factors.

1. Host factors:
- hereditary: probable genetic disposition
- age: usually adults are more affected
- behaviour: consumption of large quantities of cholesterol containing foods.

2. Social environmental factors

- high socio-economic status
- availability and cheapness of cholesterol containing foods
- occupation - sedentary work

3. Biological environmental factors

- presence of animals containing large amounts of cholesterol.

4. Physical environmental factors

- climatic factors

The Web of Causation Model

In this model many factors are associated with the occurrence of disease. Effects do not depend
on a single isolated cause but rather develop as a result of chains of causation in which each
link is a result of a complex genealogy of antecedents. The large number of antecedents creates
a condition that may approximately be conceptualized as a web.
Examples:
 15

1. Web of causation of cardiovascular diseases, smoking, hormones, stress, diet (fat, total
calories, salt), physical activity, heredity, hyperlipidemia, obesity interact to bring about
atherosclerosis, hypertension and coagulation / clot lysis which in turn lead to coronary heart
disease, cerebral vascular disease and hypertensive disease.

2. Treatment of syphilis and the development of jaundice (serum hepatitis): before the advent of
antibiotics, the treatment of syphilis was by intravenous injection of arsenical compounds.
Many of the syphilitic patients who were so treated later on developed jaundice. In 1967
viral antigens (necessary factor) became associated with hepatitis for the first time.

References:

1. Mausner JS and Kramer S Epidemiology - An Introductory Text WB Saunders

Company, London 1985 pg 26-44.

2. Roht LH, Selwyn BJ, Holguin AH and Christensen BL Principles of Epidemiology A

self teaching guide Academic Press Inc (A subsidiary of Harcourt Brace Jovanovich
Publishers) London, New York 1982 pg 26-44

3. Natural History of Infectious Diseases by Sir MacFarlane Burnett

4. Last JM A dictionary of Epidemiology Second Edition Oxford University Press 1988

5 Krieger N. Epidemiology and the web of causation: Has anyone seen the spider ? Soc
Sci Med 39 (7) 1994 pg 887-903.
 16

Natural history of disease and levels of prevention

By "natural history" of a disease we refer to the course of a disease over time, unaffected by
treatment. The following are the stages in the natural history of diseases:-

1. Pre-pathogenesis:
• In this stage, disease has not developed but the groundwork has been laid by the
presence of factors that favour its occurrence.
• The stage of pre-pathogenesis involves the period of interaction of agent, host and
environment. Ultimately this interplay may create a disease stimulus, which in turn
might produce disease in man.
• Agent refers to that aetiological factor, which is necessary to bring about a particular
disease, disability or pathological state in a susceptible host.
• Host refers to the organism capable of being infected by the agent.
• Factors whose presence is associated with an increased probability that disease will
develop later are called risk factors.
• Risk factors may be immutable (e.g. age, sex, race) or susceptible to change (e.g.
smoking habit, alcoholism, high cholesterol levels in serum etc.).

2. Pathogenesis:

• The period of pathogenesis is the natural course of the disease in man from the first
interaction with the disease provoking stimulus, until equilibrium is reached, or recovery,
death, defect or disability ensues.
• The following phases can be distinguished:-
(a) Incubation period: this is the period from the moment of infection until the
disease becomes manifest (i.e. discernible early lesions appear).
(b) Pre-patent period: period before any lesions are discernible by any available
laboratory means.
(c) Induction period: the period of time from causal action until disease initiation
(d) Latent period: the time interval between disease occurrence and detection. The
latent period can be reduced by improved methods of disease detection.
• The stage of pathogenesis can result in three distinct outcomes namely; recovery, death, or
disability.

(3) Post-Pathogenesis:
This is the period of deformity or disability, resulting from the pathogenesis of disease.
The deformity or disability may be temporary or permanent.
 17

Reasons for studying the natural history of diseases:

1. Identification of preventive measures.

Understanding the natural history of a disease, enables an orderly listing of factors which
predispose to and/or cause a given disease, and of factors which may lead to improvement
or deterioration in the disease process. This information is essential to prevent a disease
from occurring or, once it has occurred, to prevent the disease process from causing
deterioration, defect, disability or death. These factors are called risk factors.

2. Prediction of prognosis.
It enables us to predict the prognosis of a given patient, i.e. to estimate the probability that a
patient with a given disease will die, recover, or suffer various complications. We can use
this knowledge to develop a prognostic model.

3. Evaluation of preventive measures.

The knowledge of the natural history of disease allows us to evaluate various methods of
treating the disease. For example, if we know that, untreated, 50% of patients with a certain
disease will die, whereas if a certain treatment is given, only 10% of them will die, then we
have a good basis for deciding to use the latter treatment.

Levels of prevention:

In a narrow sense, prevention simply means inhibiting the development of a disease before it
occurs. However, in current usage, the term has been extended to include measures that interrupt or
slow the progression of the disease. For this reasons several levels of prevention are said to exist.

1. Primary prevention

Refers to prevention of healthy people from becoming sick. It includes:-

• Immunization
• Health education; e.g. wearing shoes to prevent hookworm; safe disposal of human
excreta; personal hygiene.
• Promotion of nutrition e.g. adequate intake of calories, proteins, minerals and vitamins.
• Vector control
• Provision of adequate housing

2. Secondary prevention

Refers to detection of people who already have a given disease at the earliest possible stage,
in order to stop the disease from developing further (i.e. early detection and prompt
treatment). examples:-
• Improved medical care e.g. new therapies, specialized care facilities, improved access
to care.
• Screening for early diagnosis and specific treatment. Screening can be carried out at
two stages:-

(a) Detection of subclinical disease using laboratory tests. e.g. screening of syphilis by
VDRL tests or screening of anaemia by Hb estimation.
(b) Detection of early clinical disease, using clinical examinations with laboratory tests
if necessary. e.g. screening for breast cancer by palpation and/or mammography or
screening on tuberculosis by chest X-ray
 18

3. Tertiary Prevention

Prevention of advanced disability and death in a patient who has already developed the
disease, and cannot be cured. It includes rehabilitation.

EXAMPLE: NATURAL HISTORY AND LEVELS OF PREVENTION OF

POLIOMYELITIES:

STAGE OF DISEASE PREVENTIVE METHOD

1. Pre-pathogenesis Primary prevention

Risk factors:
-No immunity -Immunization with polio vaccine
-Overcrowding -Improved housing
-Poor hygiene -Health education
2. Pathogenesis Secondary prevention
-Early symptoms (e.g. fever, headache, vomiting) -Health education during epidemics to recognize
early symptoms
-Late symptoms (e.g. paralysis) -Treatment of paralysed patients to prevent
permanent paralysis.
3. Post-pathogenesis Tertiary prevention
-Permanent paralysis -Physiotherapy to prevent contractures
-Vocational training
-Provision of walking aids

REFERENCES:

1. Mausner JS, and Krammer S. Epidemiology - An introductory text. W.B. Saunders

Company 1985.
 19

Principals of disease transmission

Introduction:
Transmission of diseases commonly refers to communicable or infectious diseases. Transmission of
diseases in man requires the following components:
1. an agent, capable of infecting man
2. a source: an infected host or reservoir of infection
3. a portal of exit from the source
4. a suitable means of transmission
5. a portal of entry into a new host
6. a susceptible host

A detailed description of each of the above components is presented in the following sections:

1. Agent
The agent refers to that aetiological factor, which must be present for the disease, disability or
pathological state to occur in a susceptible host. An agent may be defined as the presence, absence,
excess or deficiency of a certain factor. (Refer to the section on ecology of disease and models of
disease causation)

Characteristics of infectious agents

Resistance : this refers to the ability of the agent to survive under adverse environmental conditions. Some
agents are remarkably resistant, such as Mycobacterium tuberculosis, Clostridium tetani etc. Others are
extremely fragile, such as gonococci and influenza viruses. The term resistance is also used with a different
meaning as a characteristic of the host.

Infectivity : this is the capacity of a micro-organism to enter a susceptible host. It can be expressed as the
proportion of the susceptible population, which is infected by a particular organism. Experimentally it can be
thought of as the minimum number of particles or agents necessary to cause infection in 50% of a group of
hosts of the same species. The so called Infective dose to 50% of the hosts (ID50). Susceptibility of an
individual can be measured by serological means (absence of protective antibodies). For certain diseases
susceptibility may be assumed, if a person has no previous history of overt disease.

Pathogenicity : this refers to the capacity of a micro-organism to cause overt disease in the infected host. It
can be expressed as the proportion of the infected population (to be established by serological or other
laboratory techniques), which develops overt disease.

Antigenicity : refers to the ability of the agent to induce antibody production in the host. These antibodies
may not necessarily be protective.

Toxigenicity : refers to the capacity of the agent to produce a toxin or poison that can cause pathogenic
effects on the host. The pathogenic effect of agents in diseases such as botulism and shellfish poisoning
depends on the toxin produced by the micro- organism rather than on the direct effect of the micro-organism
itself.

Virulence : refers to the severity of the disease. It is the degree of pathogenicity of an infectious agent.
One measure of virulence is the Case Fatality Rate (CFR) which can be expressed as follows:
 20

The number of persons dying of a disease

during a stated period of time
CFR = ------------------------------------------------------------------ X 100
Total number of persons with the disease
during the same period

2. Source: An infected host or reservoir of infection

The infectiousness of a host differs from disease to disease, but is usually consistent for any given
disease. Many diseases are transmissible even during periods when the disease is not (yet) clinically
manifest. For instance mumps during the incubation period, and gonorrhoea in asymptomatic cases.
Some diseases can be transmitted years after recovery, as in a chronic carrier state: e.g. typhoid and
hepatitis B, or by reactivation of a latent infection e.g. herpes.

Reservoir (of infectious agents) refers to any human beings, animals, arthropods, plants, soil or
inanimate matter or a combination of these in which an infectious agent normally lives and
multiplies, and on which it primarily depends for survival and reproduction in such a manner that it
can be transmitted to a susceptible host. A reservoir of infection may also be termed as the natural
habitat of the infectious agent.

It is important to make a distinction between those diseases in which humans are the only reservoirs
- anthroponoses (e.g. smallpox, measles, cholera etc.) - and those diseases which involve other
animal reservoirs - zoonoses (e.g. plague, rabies).

3. Portal of exit in the human host

Portals of exit in man include the respiratory passages, the alimentary canal, the openings in the
genito-urinary system, and skin lesions. Additional portals of exit may be made available through
insect bites, the drawing of blood, surgical procedures, accidents and other injuries.

For the chain of transmission to be continued, the portal of exit must be appropriate to the
particular agent. To produce infection, the agent must exit from the source in sufficient quantity to
survive the environment and the defences surrounding the portals of entry in the new host.

4. Suitable means of transmission

Transmission of infectious agents is any mechanism by which a susceptible host is exposed to an

infectious agent. It may be either direct or indirect.

A. Direct transmission

Direct transmission means that the agent is transmitted directly from the infected host (man or
animal) to the new host. e.g. influenza, gonorrhoea etc.. Direct transmission may be horizontal
(between people of the same generation) or vertical (between mother and child).

Examples of horizontal transmission of diseases

include:
- droplet infection
- faeco-oral route
- genital
- direct skin contact
Examples of vertical transmission of diseases
include:
- transplacental
- genital tract

B. Indirect transmission
 21

Indirect transmission requires a vehicle or vector to carry the agent of disease from one host to
another.

Vehicles may be substances such as water, air, food, blood used in transfusion, or formites
(inanimate objects used by an infectious host, such as clothing, handkerchiefs, doorknobs etc.).
Multiplication may or may not take place in or on the vehicle (e.g. in influenza, hepatitis and
streptococcal disease).

Vectors may be either mechanical or biological, but are always living. Mechanical vectors are such
animals as insects, which can carry agents from place to place on their feet, proboscis or other body
parts. (e.g. flies in Shigellosis). Biological vectors must have growth or multiplication of organisms
occurring within the vector (e.g. fleas in plague). Some of these vectors may be necessary to the
life cycle and therefore to the survival of agents of certain diseases (e.g. malaria).

1. Vehicle mediated transmission 2 Vector mediated transmission

- soil, air, water, food, formites - mechanical vector
- biological vector

C. Complex cycles
- helminths : ovum > miracidium > cercaria

5. Portal of entry in the human host.

A portal of entry must be available in the new host. This portal of entry must be appropriate to the
particular agent. Most natural portals of entry in the new host provide barriers to infectious agents,
such as the epithelia in the respiratory passages, acid gastric secretion, etc. The intact skin also
provides an excellent defence against entry of agents. However, as in the case of the portals of exit,
insect bites, other skin lesions, blood transfusions, surgical procedures and accidents may result in
additional portals of entry for disease agents.

Summary of portals of exit / entry in the human host :

Portal of entry Portal of exit

respiratory tract measles, flue, tuberculosis measles,, flue, tuberculosis, leprosy
gastrointestinal tract enteric fevers enteric fevers, hookworm
urinary schistosomiasis
genital STD STD
cutaneous impetigo diphtheria hookworm ? leprosy
schistosome
placental syphilis
conjunctival trachoma trachoma
umbilical tetanus
haematogenic hepatitis-B congenital syphilis hepatitis-B

6. Host

Host refers to the organism capable of being infected by the agent. One can distinguish between
different types of hosts : Obligate host: if man is necessary for the life cycle or for the continued
existence of an agent, he constitutes the obligate host for this agent. Man is the obligate host for
such agents as Salmonella typhi, measles and influenza virus. Incidental (occasional or accidental)
hosts are not usually involved in the natural cycle of transmission.
 22

Characteristics of the host:

Infected: an infected person harbours an infectious agent and has either manifest disease or
inapparent infection. He may or may not be infectious.

Infectious: is a person, from whom the infectious agent can be naturally acquired. A person or his
articles or clothing may also be merely contaminated with an infectious agent, without being
infected.

Immune: is a person who possesses specific protective antibodies or cellular immunity as a result
of previous infection or immunisation. Alternatively, these may result from such previous
experiences as to respond adequately with production of antibodies sufficient to prevent clinical
illness following exposure to the specific infectious agent. Immunity is relative: an ordinarily
effective protection may be overwhelmed by an excessive dose of the agent. It may also be
impaired by immuno-suppressive drug therapy or concurrent disease such as AIDS.

Inherent resistance: is the ability to resist disease, independent of antibodies or specifically

developed tissue response. It commonly resides in anatomical or physiological characteristics of the
host, and may be genetic or acquired, permanent or temporary.

Susceptible: a person is considered susceptible if he does not possess sufficient resistance (inherent
and/or acquired) against a particular pathogenic agent to prevent contracting a disease when
exposed to the agent. This definition includes the fact that the person or animal must belong to a
species that is biologically capable of being an efficient host to the agent in question. Susceptibility
of a host may be modified by characteristics, such as age, sex, race, genetic make-up, physiological
state, habits and customs, pathological state and his previous experience with the agent (immunity).

References:

1. Mausner JS and Kramer S Epidemiology - An Introductory Text WB Saunders Company,

London 1985

2. Roht LH, Selwyn BJ, Holguin AH and Christensen BL Principles of Epidemiology - A

self teaching guide Academic Press Inc (A subsidiary of Harcourt Brace Jovanovich
Publishers) London, New York 1982

3. Last JM A dictionary of Epidemiology Second Edition Oxford University Press 1988

 23

Theory and strategies of disease control

Disease Control Theory

As infectious diseases are due to interaction between agent, host and environment, methods of
control should attempt to modify these factors whenever possible. For this reason, a detailed
understanding of a disease and its interactions is required, in order to develop an effective approach
to its control.

In general, the following methods for prevention and control of infectious diseases are available:

A. Increase host resistance

1. Active immunisation by use of vaccines and toxoids.
2. Passive immunisation by use of immuno globulins or vaccination of the mother
3. A specific: proper nutrition, exercise

B. Interrupt chain of transmission

1. Rapid case detection and specific chemotherapy to limit infectivity
2. Isolation of infectious cases
3. Quarantine of contacts
4. Chemoprophylaxis before or after exposure
5. Environmental control of air, dust and dirt
6. Asepsis in handling patients and their excretions
7. Control of biological vectors
8. General sanitation: food, water, sewage
9. Personal measures to avoid exposure/limit spread

- personal hygiene
- proper food-handling procedures
- protective clothing/repellents for insect bites
- avoid suspect water, foods, animals and insects

C. Inactivate infectious agent

1. Physical methods
- heat: cooking, pasteurisation & sterilisation
- cold: store food at low temperatures
- radiation: ultraviolet light, gamma radiation

2. Chemical methods
- Chlorinate water supplies and sewage effluents
- disinfect (potentially) infectious material

Objectives of Immunisation

1. Produce a good humoral, cellular and local immune response, similar to natural infection
2. Produce protection against clinical disease and re-infection
3. Give protection over several years, preferably a lifetime.
4. Result in minimal immediate side reactions or mild disease and with no delayed effects such as
late reactivation, CNS involvement or cancer
5. Can be administered simply in a form acceptable to the public
6. Cost and benefits of administration should clearly outweigh the cost and risk of natural disease
 24

*Strategies of disease control

Control can be defined as the reduction of the occurrence of a certain disease, as far as this is
feasible with present means

Case - finding
For certain diseases, case-finding and adequate treatment is an important means of source reduction
(e.g. tuberculosis). Case-finding may be active or passive. In passive case-finding, only cases
reporting spontaneously to be facility are treated. In active case-finding, activities are undertaken to
screen a population for early signs of a certain disease.

Surveillance
The term surveillance. may be defined as a system of close observation of all aspects of the
occurrence and distribution of a given disease, through the systematic collection, tabulation,
analysis and dissemination of all relevant data pertaining to that disease. Although the methodology
of surveillance is basically descriptive, surveillance must be dynamic, current and purposeful. It is
fundamental tall to prompted effective control and prevention of disease.

Elimination
Elimination may be defined as the purposeful reduction in prevalence of a communicable disease, to
the point of continued absence of transmission, from a certain geographical area. For instance,
malaria has been eliminated from Europe. As elimination is confined to a limited area, continuous
vigilance against re-introduction of the disease remains important.

Eradication
Eradication means the complete and universal removal of ac certain disease from the world. So far
only eradication of smallpox has been successful. Smallpox virus presently only persists in a few
laboratories around the world. The advantage of eradication over control is the cost of a time-
limited capital investment as against the ever-recurring cost of control measures.

To be eradicated, a disease should fulfil the following conditions:

1. Availability of a simple, inexpensive told of proven efficacy in breaking transmission.

2. Epidemiological features, facilitating effective case detection and surveillance in the later stages
of the campaign. This includes: few in apparent or atypical cases.
3. Disease of recognised socio-economic importance, national or international.
4. A specific reason for eradication (rather than control)
5. Adequacy of financial, administrative and operational in resources.
6. Absence of unfavourable socio-economical conditions

The process of eradication cab be divided in distinct phases:

1. Attack phase: aimed at the elimination of sources of infection and prevention of possibly
infectious relapses.
2. Consolidation phase: aimed at the absence of infectious cases for a specified period of time.
The end point of this phase is epidemiological eradication.
3. Maintenance phase: full eradication will be achieved when all old infections have naturally died
out. This is particularly important for diseases where the agent tends to persist (tuberculosis,
syphilis, leprosy).
The phases have to be worked out according to the specific requirements of a given disease. It is
crucial that methods, operational standards and end-points are clearly defined.

References and further reading

1. Foundations of Epidemiology Lilienfeld and Lilienfeld

Oxford University Press
2. Epidemiology: Principles and Methods MacMahon, B. and Pugh, T.F.
Little, Brown Co., Boston
 25

Measurements of health, morbidity and mortality

Epidemiological inquires are based on the ability to quantify the occurrence of disease and a clear
definition of what is a case. Measurements of disease are important in hypothesis formulation and
permits fair comparison of frequencies of disease occurrence between two populations.

The commonest used measurements of health are:-

1. Numbers
Number of new cases of active Tuberculosis in some regions of Tanzania 1965

Table I
Region:
Dodoma 900
Coastal 850
Singida 700
Tabora 870
Mwanza 450
Other region 8,280
TOTAL 11,750

The limitation of using numbers is, that they do not allow any comparison, as the populations they
refer to are unknown. The numbers in table I above (fictitious data) give the false impression that
Dodoma is the most affected. Nevertheless numbers are useful for national allocation of resources
for instance to determine the amount of tuberculostatics required per region.

2. Ratios:
A ratio is the result of the division of two numbers; it is the basic measure of the relative magnitude
of two numbers. For example

numbers of males
Sex ratio = -----------------------
number of females

Ratio is a fraction whose denominator is independent of the numerator. Two distinct types of ratios
are commonly used; one has dimension while the other is dimensionless. The number of hospital
beds per 1000 population is a ratio possessing dimension; which male to female ratio, is
dimensionless. Still birth per 1000 live births is another example of a ratio with a dimension.

3. Proportion
A proportion is the result of the division of a number by the population it refers to. Unlike a ration
it is a fraction whose numerator is included in the denominator. It is a dimensionless quantity whose
value must range from 0 to 1. A proportion can be expressed as a fraction
( 0.5) or as a percentage (50%).
.
For instance: the number of new TB cases in Tabora Region is 870, and if the population of Tabora
is 870,000 people, the proportion of new TB cases in 1965 is, 870 / 870,000 = 00010 or 0.010% or
0.1 per 100,00. This allows better comparison between the regions regarding the magnitude of the
TB problem in 1965:
 26

1
By using this kind of measurement one can know which of the regions had the biggest problem of
new TB cases when the regional population is taken into account. To illustrate this point table II
below (tuberculosis data) shows that despite differing number of new cases, the magnitude of the
problem of new cases of TB was the same in the regions in 1965.

Table II

Number of new cases of active tuberculosis. Tanzania 1985

Number Proportion
Population of cases per 1000

Dodoma Region 90,000 900 10

Coastal Region 85,000 850 10
Singida Region 70,000 870 10
Tabora Region 87,000 870 10
Mwanza 450,000 450 10
Other 8,280,000 8,280 10
Total 12,050,000 12,050 10

4. Frequency distribution and proportional distribution.

Classification of population into categories such as region, age or sex, yields a tabulation of these
new TB cases according to these characteristics in the total population, which is a frequency
distribution. For example the column of number of cases in the above table is a frequency
distribution of new TB cases according to regions. However the proportions making up the total
population constitute a proportional distribution. The following table shows the proportional
distribution of new TB cases in Tanzania in 1965 according to regions.

Table III
Region New Case Proportional distribution
%

Dodoma 900 7.5

Singida 700 5.8

Tabora 870 7.2
Mwanza 450 8.9
All others 8,250 68.7
TOTAL 12,050 1000

5. Proportional mortality ratio (PMR)

number of deaths in a specified group *
PMR = -------------------------------------------
total number of deaths in the population
in a given period of time

PMR shows the relative importance of a disease or factor in causing mortality e.g. Deaths
due to AIDS expressed as a ration of total deaths.

Deaths due to AIDS

PMR = --------------------------------------------------------------------------------------
Total deaths from all causes in a given area or a health facility in a year

1
*the specified group may be categories of disease, age or sex group etc..
 27

Advantages of the PMR:

- simplicity of calculation
- no need to know the population at risk
- data frequently available for crude (inter)national comparison.

Disadvantages of the PMR:

- ignores population structure and population at risk
- difficult to interpret: Variation may be due to changes in numerator or denominator

6. Index
An index is a percentage of a standard value. In the following table IV the IMR index is calculated
using the 1970 IMR as a reference is observed to decrease with decreasing time which clearly
indicate improvement in the health of the children (infants)

Table IV

Year IMR Index

1970 150.0 100%(IMR1970/IMR1970 x 100)
1975 135.0 90% (IM1975/IMR1970 x 100)
1985 120.0 80% (IMR1980/IMR1970 x 100)

In this case we set the 1970 IMR at 100%

2
The term index is also used for more complex indices, e.g. Quatelet Index = W/H (Weight divided
by Height squared) which is a measure of obesity.

7. Rates
Is basically a proportion that includes specification of time period or a ratio in which there is a
distinct relationship between the numerator and the denominator with time being and intrinsic part
of the denominator. Rate refers to the instantaneous change (or potential for change) in one
quantity per unit change in another quantity, in a period of time. A rate indicates the frequency of
events (like births, deaths, illness), occurring in a population per unit of time.

The scale of this measure (percent, per 1000 or per 100,000) should always be stated, as well as the
unit of time (per year, per month). For example a rate can generally be expressed as follows.

Number of events during a specified period of time x 1000

Rate = -------------------------------------------------------------------------------
Total Population at risk for this event at the beginning of the period.

The denominator must be clearly defined and the numerator completely counted.
2
Examples of rates are: Crude Birth Rate, Crude Death Rate and Incidence Rate

Number of deaths in a defined population per year

Crude Death Rate (CDR) = -------------------------------------------------------------
Mid-year population

Many measurements are erroneously called rates, while they are actually proportions; e.g.: Case
Fatality Rate, Prevalence Rate, maternal mortality rate, etc. In many such rates the denominator is
not exactly the population at risk e.g.: Infant mortality rate, where the denominator is composed of

2
Note: Prevalence rate when unspecified will always refer to number of cases existing in a given population
at a given point in time, divided by the total population.
 28

live births during the year, which is strictly not the population at risk of death under one year during
the same year.

Measurements of morbidity are:

1. Prevalence rate
Prevalence rate describes the proportion of a given population that has a given disease or a
characteristic at a point in time. Since in most cases prevalence refers to state of affairs at a given
point in time, the term "Prevalence" rate is always used to denote point prevalence. For example
the prevalence of leprosy on 1st January 1989 may be expressed as follows:

Number of leprosy cases as on 1st Jan. 1989

_____________________________________________ x 100
The total population on that date under consideration

To avoid dealing with fractions, prevalence rate is often expressed as per 100, per 1000, per 10,000,
or per 100,000 depending on the frequency of the disease. Occasionally reference is made to a term
called period prevalence. This refers to the proportion of individual who ever suffered from the
given condition (old cases + new cases) during a period of time.

Point prevalence rate is used to measure the magnitude of a disease in a community. The following
general formula is used.
Number of people in a group with a certain characteristic
Prevalence = --------------------------------------------------------------------------x 100
Total number of persons in the group

So a prevalence is a proportion. When we say "In 1985 the prevalence of malnutrition in under-5's
was 12%, We means that 12% of all children under the age of 5 years were malnourished in 1985.
Strictly speaking, it is incorrect to speak of 'prevalence rate'.

2. Incidence rate

The incidence rate indicates the number of new events in a certain population at risk of such events
during a specified period of time.

Basic incidence measures: (risk and rate)

Incidence refers to the propensity of a given population for producing new cases of a certain
disease. i.e. number of new cases of disease occurring amongst a susceptible population at risk over
a given period of time. Typically, incidence measures are estimated from cohort studies. For most
chronic diseases, particularly incurable conditions, we are concerned with the first occurrence of
disease for each subject at risk. When studying acute diseases, however, especially in the absence of
acquired immunity, we might be interested in multiple occurrences of diseases within individuals.
Such occurrences are called episodes and must be taken into account when studying disease
incidence and its control.

There are two distinct measures of disease incidence RISK and RATE which must be differentiated.
Risk (Cumulative incidence -CI) is defined as the probability of a disease-free individual
developing a given disease over a specific period of time, conditional to that the individuals’ are not
dying of any other disease during that time. As a conditional probability risk varies from 0 to 1, and
is dimensionless. For example a 5 year risk of developing coronary heart disease (CHD) for an
individual aged 55 could be obtained as:
• Population of healthy individual aged 55 years at risk at the beginning of the observation period
is 15,000. This population is followed up regularly for a period of five years. During the five
years all new cases of CHD were identified and counted, which were 500.
• Therefore the 5 year risk of CHD = 500/15000=0.033 or 3.3 percent.
 29

A related frequency measure is the risk odds which is the conditional probability of developing the
disease divided by the probability of not developing the disease during a given period of time.
From the above example odds ratio = 0.033 = 0.034
(1-0.033)

Calculation of incidence density(ID) or rate:

The incidence rate (is also called incidence density -ID) is the instantaneous potential for change in
disease status per unit time relative to the size of the disease free population.
In many studies the study members will have different times of observation. This is because time to
study entry and censoring are different for each participant. The denominator in incidence density
accounts for the different times of observation for each individual.

In figure 1. the first subject contributed 2.5 person years of observation prior to the onset of illness.
Similarly the contribution made by each of the subject can be calculated and added to form the total
person time of observation as the denominator. The incidence density can therefore be calculated
thus:

ID = I/PT
Where I = the number of new cases
PT = the total person time of observation
N
PT = Σ d(ti)
1:1

Where d is the individual observation period.

3
In our example of 12 subjects the ID is equal to 5/36 = 138.5 per 1000 person years. If the disease
does not confer lasting immunity the estimation of PT need to be made taking that aspect into
consideration.

Calculation of risk (cumulative incidence - CI)

In the case of a fixed cohort with hardly any withdrawals, the calculation of risk will be as follows:

R = I/No
Where I = the number of new cases
No = the number of disease free individuals at risk at the beginning of the study.
Numbers of persons in a group developing a disease during a given time period
Incidence Rate = ----------------------------------------------------------------------------------- x1000
Total number at risk in the group at the beginning or midterm population

3
: Midterm or mid year population is used when calculating a rate for large population such as a city or
a country.
 30

When determining incidence consider :

(a) The method of classifying individuals as having disease or not must be clearly stated i.e.
whether it was by detailed examination, screening, or examination of routine sources of health
statistics, etc.
(b) One should also be able to determine date of onset thus identifying incidence cases from
prevalent.
(c) Denominator must be clearly defined and enumerated.
(d) Period of observation must be long enough for the numerator to stabilize.
(e) The denominator or the mid year population which should be enumerated during the period of
the study.

Relation between incidence (CI)and prevalence:

The amount of disease in a population at a given moment in time, depends on:
1. new cases of disease entering the population (Incidence).
2. Old cases leaving the population by:
- recovery
- death
- emigration of cases and susceptibles
3. Improved diagnostic facilities
4. Improved reporting
Under stable conditions prevalence rate depends on the incidence rate and the duration of the disease which is
from disease onset to termination. The incidence rate and the disease duration should be expressed in the
same units. When these factors are constant the prevalence equals the incidence (CI) times the average
duration of the disease i.e. P = ID. Diseases with equal prevalence (as found during a survey) can differ
considerably in incidence, as is shown by the following example.

Disease Prevalence Incidence per year Duration

Common cold 4.0/1,000 200/1,000 1 week
Tuberculosis 4.0/1,000 2/1,000 2 years

As the incidence is expressed in years, the average duration of a common cold should be expressed in years
as well: 0.02/ year i.e. cumulative incidence and average duration of disease should be expressed in the same
time unity.

Prevalence provides information for:

(1) Establishing the importance of the disease as public health problem.
(2) Planning for
(a) Drug requirement
(b) Personnel needed
(c) Equipment needed i.e. beds, laboratory facilities etc.
(3) Monitoring of disease control programmes.
(4) Tracking changes in disease pattern

Incidence is used to
(1) study disease aetiology
(2) study efficacy of an intervention
(3) when disease is of short duration due to quick recovery or death incidence is a better measure
than prevalence.

Other measures of health:

1. Secondary attack rate:

Number of exposed people developing the disease of interest within the range of the incubation period
= -------------------------------------------------------------------------------------------------------------- x 1000
Total number of susceptibles exposed to the primary case.

Gives information on infectivity of the agent, efficacy of a prophylaxis measure, or whether a disease of
unknown aetiology is communicable.
 31

Measurement of mortality
Infant mortality rate: number of live born infants under the age of 1 year that died during a year/number of
live births dying during a year, are actually born in the preceding year, the denominator is not exactly the
population at risk, except when the number of births remains constant.

The following measures of mortality of the fetus or infant are in use:

Measure time interval/enumerator denominator

infant mortality rate birth - 1 yr. live births
perinatal mortality rate postpartum 28 wks gest. -7 dys live births + stillbirths
neonatal mortality rate birth - 28 days live births
post neonatal mortality rate 28 days -1yr live births
childhood mortality rate 1 yr -5yr live births
stillbirth rate . 28 wks gest-birth live births + stillbirths
early neonatal mortal, rate birth -7 days live births
late neonatal mortal, rate 7 dys 28 days live birth
fetal death rate conception birth live births +stillbirth

Crude death rate: total deaths in defined population in a given time period divided by the total population.
Uniform application of the definition of birth and liverbirth are essential for obtaining reliable and
comparable data.

Maternal mortality rate:. All maternal deaths occurring during pregnancy or within 42 days often
termination of the pregnancy in a year divided by the total number of live births in that year (per 100,000.

Other measurements of health

1. The average number of years an individual of a given age is expected to live if current mortality
rates continue.

2. Median survival time

Refers to the time at which fifty percent of individuals with a certain diagnosis or having had a
certain intervention would have died or would have died if there was no intervention.

3. Person years of life lost (PYLL).

PYLL is the number of years an individual would have lived should this individual not have died at
this age. It is a measure of premature mortality. Areas experiencing high (PYLL) are less healthy
than those with low PYLL. The upper limit could be life expectancy at that age in that area or in
another area i.e. a developed country e.g. USA as a standard for developing countries. If an
individual dies at the age of 30 years, the person years of life lost will be the difference between the
age at death and the life expectancy at that age in that area. The life expectancy in Tanzania is
estimated at 58 years therefore PYLL = 58-30 = 28 for this individual dying at 30 years of age.

4. Potential productive years lost

You might also hear of potential productive years lost(PPYL), this refers to the number of years an
individual would have remained productive if the individual didn't die at this age i.e. an individual
dying at the age of 34, the number of potential productive years of life lost would be 21 if we take 55
years as the upper limit of productivity( the age of retirement), similarly it would be 26 years if we
say that an individual retires at 60 years of age.

Disability:
It is a measure of the presence of consequences of disease i.e. impairment, disability, and handicap. Several
levels are known and defined as follows.
• Impairment; any loss or abnormality of psychological, physiological, or anatomical structure or function.
• Disability; any restriction or lack of ability to perform in the manner or within the range considered
normal for a human being.
• Handicap; a disadvantage of a given individual resulting from an impairment or a disability, that limits or
prevents the fulfillment of a role that is normal for that individual.

5. Risk
 32

Risk difference(attributable risk): Relative risk shows the likelihood of disease amongst exposed compared
to those not exposed. Attributable risk (AR) or a risk difference is a measure of association that provide
information about the absolute effect of exposure on disease occurrence amongst those exposed compared to
those not exposed.

Attributable fraction (AR %): Attributable risk fraction is the fraction or proportion of disease that could be
prevented by eliminating the exposure. It is calculated as:

Incidence rate of disease in exposed - incidence rate of disease in unexposed

AR % = ___________________________________________________
Incidence rate of disease in exposed -

Population attributable fraction (PAR)

This is the excess rate of disease in the population due to the exposure. PAR is equal to Disease incidence rate
in population minus disease incidence rate in the unexposed group. OR multiplying the attributable risk by
the proportion of individuals exposed in the population. Population attributable risk fraction (PAR %) PAR %
is obtained by dividing the population attributable risk in by the rate of the disease in the population.

FIGURE 1
____┌────────────────────────────────────────┐
____│_________________X______________________│
1___│_____________________________________ __│2.5
2___│_______________________X________________│3.5
3___│________________________________________│1.5
4___│____________________________X___________│2.5
5___│________________________________________│4.5
6___│________________________________________│.5
7___│________________________________________│.5
8___│________________________________________│2.5
9___│__________________0_____________________│2.5
10__│__________________X_____________________│2.5
11__│______________0_________________________│1.5
12__│______________X_________________________│1.5
____│________________________________ _______│
____└────────────────────────────────────────┘
0 1 2 3 4 5
Time (years)
Key
X first occurrence of disease X
O Death
- Censored
 33

Sources and uses of morbidity and mortality statistics:

A. Sources of routine morbidity statistics
Morbidity statistics are available at all levels of health care. These are
available at various documentations usually kept at the health delivery centres. The following
records are usually kept at hospitals.
• Admission registers, records
• clinical records
• discharge summaries
• investigation laboratory records
• investigation request forms
• Burial permits and other mortuary records.

Registers :
Information can also be available from registers when available, the registers can be institutional or
community based. Community based registers include those for surveillance of diseases, industrial
safety etc.
Institutional based registers include cancer registers (in Tanzania it is currently being maintained at
Muhimbili, department of pathology). The tuberculosis and leprosy registers are maintained at each
district. Registers for birth malformations, accidents, toxicity, and other infectious diseases can be
established

Records of special health programs

1. MCH
Maternal and child health services are available in all health care delivering units. Special
cards are available for the mother and the child, the state of health, type of treatment or
procedure for each mother and child is recorded regularly.

2. Essential Drug Program (EDP)

The essential health programme maintains records of drugs supplied, how long they lasted, how
many patients were treated and their respective health problems. This is recorded at all dispensaries
and health centres. This gives a good indicator of morbidity pattern at the community level.

3. Expanded Program on Immunization (EPI)

Generate and maintain information on all vaccinations done, type of vaccination and their batch
numbers. Can provide information on vaccination coverage.

4. Mental health
Maintains records of people who are mentally ill

5. Oral & dental health

Maintains records of people found with or developing oral or dental health problems.

6. Eyes
Eye disease control programmes keeps records of people having eye problems. This is currently
being done in areas with high incidence of infectious diseases of the eye i.e. Onchocerciasis in
Morogoro, Mtwara, and Iringa regions, and trachoma in Dodoma
 34

7. Nutrition
The nutrition programmes maintains records of individuals having/developing malnutrition, this
includes special nutritional deficiencies like iodine.
8. Others sources
(a) Health requirements
Recruitment to new employment require someone to undergo medical examination
prior to being recruited. The same is needed for school entrance, insurance, military
conscription etc.

(b) Disease screening programmes

Programmes/projects may require screening for disease in communities as a basis of
evaluation and disease control programme.

(c) Census
Census may collect information on people who are sick, disabled or quality of
living.

(d) Drug surveillance programmes:

These programmes routinely collect information on drug supply, drug reactions,
type of reaction, outcome of the reaction and drug interaction. An example is
TADATIS situated at Muhimbili. These are also important sources of morbidity
statistics.

B. Routine sources of mortality statistics

Death certificates
Death certificate is a legal document completed by the doctor certifying the death of a
person. Ideally every individual dying a death certificate has to be completed and further
transmitted to the Registrar of Births and deaths. From the Registrar, death certificates are
available from various places in the country, a copy remains with the completing authority.

Death certificate contain the following information about the deceased:(see death certificate
specimen below)

It has two parts (Part I and II) It has information on:

• Name of the deceased
• Age of the deceased
• Place of death
• Date of death
• Sex
• Address
• Name of Certifying doctor

Part I of the death certificate collects information on:

• Immediate cause of death
• Intermediate cause of death and

Part II collects information on:

• Underlying cause of death and
• Other significant conditions

Other sources of mortality statistics are:

Disease control programs
Control programmes for diseases which always end fatally will provide information on people who
died from the disease i.e. plaque, and rabies etc.
 35

Census
Census may aim at collecting information on mortality for the entire population or part of the
population.

One of the uses of routine mortality statistics will be to compare mortality between different areas.
Such comparison is only possible if the criterion for acertaining cause of death is uniform, such
uniformity is achieved by the use of the International Classification of Diseases and Injuries (ICD).
This is internationally used and it is revised every 10 years to accommodate developments in
disease diagnosis and occurrence of new diseases.

Identification of cause of death from death certificates

From the death certificates causes of death are identified and are selected for tabulation. When only
one cause of death is entered it is selected for tabulation of mortality statistics. When more than one
causes are recorded the following rules should guide the selection of cause of death.

General:
Select the condition entered alone in the lowest used line of part I.

Rule I
If there is a reported sequence terminating in the condition first entered on the certificate select the
underlying cause of this sequence. If more than one such sequence select for the first.

i.e. Coronary embolism, asteriosclerotic heart disease and Influenza.

In this case select: arteriosclerotic heart disease

Rule II
If there is no reported sequence terminating in the condition first entered on the certificate, select
the first e.g.
• Pernicious anaemia and gangrene of foot
• Arteriosclerosis
• Select: pernicious anaemia
• Rule III

If the condition selected by the general rule or rule 1 or rule II can be considered a direct sequel or
another reported condition whether in part I or part II select this primary condition. If there are two
or more of these, select the first mentioned of these. e.g. cerebral hemorrhage, hypertension, chronic
pyelonephritis and prostatic obstruction.

e.g. IVa) Pulmonary oedema

Secondary anaemia and chronic lymphatic leukemia
Select chronic lymphatic leukemia. Surgical complications occurring in 4< wks they are
considered as direct sequel to an operation.

Uses of routine morbidity and mortality statistics:

• Monitoring of public health
Establish disease/mortality trends by
- person
- time
- place
 36

Increase in mortality with age could be due to:

1. Accumulation exposure of an individual during life time to environmental insults
2. Decrease with age in immunological defense
3. Increase in time with mutations or chromosomal abnormalities with age due
cumulative effects of environmental exposure, decreased efficiency of such biological
mechanism as mitosis
4. Hormonal changes throughout life time.
5. Exposure to an agent early in life may impair immunological status of the aging person.
6. Non specific genetically determined weaning out

Variation by Place
Where are the rates of disease highest or lowest? Characteristics of place could explain variation of
disease occurrence or mortality.

Variation by time
• When does the disease occur commonly or rarely?
• Is the disease frequency at present different from the corresponding frequency in the past?.
On answering such questions will lead to development of hypothesis. Rarely one will be able to
prove a hypothesis using routine data.

Possible reasons for changes of morbidity/mortality trends

A. Artifactual
1. Errors in the numerator due to
(a) changes in the recognition of disease.
(b) changes in the rules and procedures for classification of causes of death
(c) changes in the classification code of causes of disease.
(d) changes in the accuracy of reporting age at death.

2. Errors in the denominator and or the numerator, under reporting of cases or

inaccuracy in counting the population at risk.

B. Real
1. Changes in age distribution in the population
2. Changes in survivorship due to treatment.
3. Changes in incidence due to
(a) Genetic factors
(b) Environmental factors
Other uses of routine morbidity/mortality statistics
• Administration & planning
• Disease monitoring
• Epidemiological
• Generation of hunches

Disadvantages of routine morbidity/Mortality data:

Completeness
Some information is not filled, or is wrongly filled for reasons that the information was not
collected, or the informant could not provide it.

Delay
Sometimes there is delay in transmitting such information to centres, hence being out of date.
 37

Accuracy
Information though filled may be inaccurate, i.e. recording of sex as male instead of female,
recording of age as adult instead of actual age.

Coding
In the process of coding different coding systems may be used or not coded.

Retrievability
It is always not easy to retrieve all routine data due to missing data for various reasons. One would
desire to get information more than what is available on the form, this is not possible.

Publication
Publishing information from routine data is not easily acceptable by some readers.

Confidentiality
Routine data contains information on individuals, retrieval and publication might be divulging
information of the diseased , you might require ethical legal clearance to overcome that. Errors
could occur at any of these stages, diagnosis, recording of particulars, certification, coding, data
processing, interpretation and reporting.

Uses of Morbidity/Mortality statistics:

Uses of mortality/morbidity statistics will be by analysis of how it varies with time, person and
place. Variation by person: Who is getting the disease? age is one of the factors mostly considered
in this case.

References:

1. Michael, Alderson: 1983 An introduction to epidemiology. Macmillan press London.

2. WHO International classification of diseases 1975 revisions Vol. 1 1977.
 38

Rates and standardization or rates (adjustment of rates)

Crude rates:
Crude rate, is a summary measure calculated by dividing the total number of cases of the outcome
in the population by the total number of individuals in that population in a specified time period.

An example is crude death rate (CDR) which is expressed as below:

Total events (deaths) in given period

CDR = ───────────────────────────────────── x a factor (e.g. 1000)
Total population giving rise to the events (deaths)
in the given period usually a calendar year.

The crude death rate in the table below is 14.2 per 1000 population, which the age-specific rate
range from 10 to 20 per 1000 population.

Table 1: Age specific death rate of population X

Age group Population Number of Rate/1000
deaths
<20 10,000 100 10
21-29 8,000 120 15
30-39 5,000 75 15
40-49 3,000 51 17
50+ 4,000 80 20
Total 30,000 426 14.2

Crude rate is weighted average of the individual category (specific) rates, weight being the
proportion of the population in each category.

CDR =Σ (Category specific rates X proportion of population in the category)

From the table above
CDR= 10,000 x 10 +8,000 x 15 +4,000 + 3000 x 75 + 3000 x 17 + 4000 = 14.2
30,000 3000 30,000 30,000 30,000 30,000

Crude rates assume that all individuals in the population giving rise to the event of interest are at
equal risk of the event.
i.e. CDR - all individual are at equal risk of death.
CBR - all individuals are at equal risk of giving birth.

Two factors contribute to crude death rate:

(1) Age distribution of the population

(2) Probability of dying for each individual other than due to age.

One way to evaluate crude rates of two populations or more is to work out the population
proportion distribution and strata specific rate and find out how they differ between the two
populations. Another way is to compare crude death rates.

However if you are comparing crude death rates, differences in age composition virtually always
occur, thus confounding the difference observed between two populations being compared. If a
 39

frequency distribution is given, comparison can be made easier by calculating a proportional

(percentage) distribution as follows;
Proportional (percentage) distribution of population in Districts A and B by age groups
Age group District A. District B.
populat. % dist populat. % distr..

0- 1 5,674 1.5 18,819 2.0

1- 4 22,167 5.9 74,554 8.0
5 – 14 51,932 13.9 162,633 17.4
15 – 24 32,565 8.7 108,310 11.4
25 – 34 33,877 9.0 126,938 13.3
35 – 44 41,633 11.1 140,768 15.0
45 – 54 41,670 11.1 118,013 12.6
55 – 64 51,985 13.9 93,058 10.0
65 – 74 65,785 17.6 67,994 7.3
75 + 27,379 7.3 25,960 2.8
all ages 374,665 100.0 935,047 100.00

From the table above 50.1 percent of the population in district A are 44 years or younger while in
district B 52.1 % are 34 years or younger. Hence the population in district A is more aged than that
of B. This difference in the age distribution between the two districts could account for an
observed difference in rates between the two. Two ways to account for a confounding effect of a
characteristic/variable between two populations being compared are:

1. stratification : specific rates, which involves comparing many rates

2. standardization (also called adjustment), this one involves comparing summary rates
adjusted for the confounders.

When comparing crude rates of two populations difference observed could be due to (other than true
difference) differences in the population age structure, and sex composition etc. or sampling error of
the study population. The best ways to account for differences in population characteristics are one,
comparing category specific rates which is more accurate than crude rates, however this requires
comparing many numbers and at times can be very cumbersome. This then calls for a summary rate
which will take into consideration the differences in the characteristics of populations being
compared. This is achieved by standardizing (adjusting) for those characteristics hence
standardized (adjusted) rate. When comparing rates adjusted for a age, any remaining observed
difference between the groups cannot be attributed to the confounding effect of age.

Standardized(adjusted) rates:
A standardized (adjusted) rate is a statistically constructed summary rate which accounts for the
difference between populations being compared with respect to other variables.
Standardization is either by the direct or indirect, results in one summary measure, which is adjusted
for differences in population composition i.e. sex, age etc. and thus removes the confounding effect
of that variable adjusted for in the observed difference between the two populations being
compared.

Direct standardization
Direct standardization by age is used when:
- age specific rates in the index population are known
- the index populations and number of deaths are large enough to calculate stable
age specific rates.
 40

Procedure for direct standardization:

• Define standard population (bigger of the two, both combined, national, or a international,)
• Apply the age (or other category) specific rates of the index population to the numbers in each
age group of the standard population to obtain the number of cases or deaths we would expect in
the standard population if the index rates applied to the standard population:
• Add the expected cases or deaths over the age groups to obtain the total number of expected
cases or deaths:
• Divide the total expected cases or deaths by the total standard population to get a standardised
rate. In studies of morbidity standardized rate my be prevalence incidence rate, while in studies
of mortality it will be standardized mortality rate.

Comparative mortality index (CMI) Or

Comparative incidence index (CII) = observed rate x100
Standardised rate
example:
Age group standard age-specific death rate expected deaths
population of index populations*
Distr. A Distr.B Distr.A Distr.B
0- 1 23,000 28.2 28.2 649 60
1- 4 90,000 1.4 1.2 126 10
5 - 14 199,000 0.6 0.4 119 8
15 - 24 134,000 0.8 1.2 107 10
25 - 34 127,000 0.4 1.7 179 20
35 - 44 134,000 3.0 3.3 402 40
45 - 54 114,000 7.7 8.2 878 90
55 - 64 87,000 15.9 16.7 1,383 1,45
65 - 74 61,000 28.9 31.1 1,763 1,89
75 + 31,000 82.5 84.2 2,558 2,61
all ages 1,000,000 15.3 8.9 8,164 8.56

* per 1,000 population

The directly standardized death rate=-- expected deaths= 8,164 = 8.16 per
standard population 1,000,000 1000 population

For district B= 8545 = 8.5 per 1000 population

1,000,000
Indirect standardization
Indirect standardization is used when:
-age specific death rates for the index population are unknown
-the index populations or number of deaths are too small for calculating stable age specific
rate

Procedure of indirect standardization:

Indirect method:
• Define the set of standard age specific rates
• Apply the standard age specific rates to the population in the corresponding age group of the index
population to get the expected number of cases or deaths in each group
• Add the expected cases or deaths over the age groups
• Divide the total observed cases or deaths by the total expected to get (SMR)
• The crude rate in the standard population multiplied by the SMR gives the standardized rate in the index
population.
 41

Example:
Age group standard index - populations expect Nr. of deaths
rates per Distr.A Distr.B Distr.A Distr.B
1,000

0- 1 27.0 5,674 18,819 153 508

1- 4 1.1 22,167 74,554 24 82
5 – 14 0.5 51,932 162,633 24 81
15 – 24 1.1 32,565 108,310 36 119
25 – 34 1.5 33,877 126,938 51 187
35 – 44 3.0 41,633 140,768 125 422
45 – 54 7.6 41,670 118,013 317 897
55 – 64 17.4 51,985 93,058 905 1,619
65 – 74 38.2 65,783 67,994 2,513 2,597
75 + 106.0 27,379 25,960 2,902 2,752
all ages 9.5 374,665 935,047 7,052 9,264

Advantages of the direct method of standardization

1. The method is simple to apply
2. You do not need to know the strata specific rate of the standard population.
3. To compare two index population simply calculate the ratio of the two CMIs or CIIS

Advantages of the indirect method of standardization

1. Because of precision it is preferred when dealing with small numbers.
2. You do not need to know the strata a specific rates of the index population.

Reading References

1. Hennekens, C. H. and Buring J. E.

Epidemiology in Medicine Eds Mayrent, S.L. Little Brown Co. Ltd 1987.

2. Hill, A. B (Sir)
Short textbook of medical statistics 11st ddn Holder and Stoughton London in association
with The Lancet 1984.
 42

Testing and screening for disease

Screening:

Screening is defined as 'the presumptive identification of unrecognized disease or defect by the

application of tests, examinations, or other procedures which can be applied rapidly to sort out those
who probably have a disease from those who probably do not. The screening procedure itself does
not diagnose illness.

The basic purpose of screening for disease is to separate from a large group of apparently well
persons those who have a high probability of having the disease under study, so that they can have a
diagnostic work up and if diseased, brought to treatment. Screening is carried out with the
assumption that early detection of disease before development of symptoms will lead to a more
favourable prognosis because treatment began before the disease becomes clinically manifest will
be more effective than later treatment.

Types of screening and their specific aims:

Mass screening: This involves screening of a whole population

Multiple or multiphasic screening: Involves the use of a variety of tests on the same occasion for the
same condition
Targeted screening: Involves screening of groups with specific exposures
Case-finding or opportunistic screening: Screening of patients visiting a health care delivery point
for some other purpose.

What should be the measurement properties of a screening test?

We would ideally want this test to identify correctly all those with the disease under investigation
and to exclude this disease amongst all non-diseased.

The test should give:

• true measurement of the attribute under investigation (accuracy). Accuracy may be defined as a
measurement process yielding values that are equal on average to the true underlying value for
the diagnostic variable being measured)
• consistent results in repeated trials (precision sometimes called reproducibility) Precision may
be defined as the degree to which a series of measurements fluctuates around a central
measurement. The central value may or may not be the true value of the variable. Precision is
independent of accuracy. Measurements may vary from inaccurate and imprecise to inaccurate
but precise, to accurate but imprecise to the ideal measurements which should be accurate and
precise.
• In addition a screening test should have high sensitivity and specificity for the disease in
question (also called validity properties of a screening test).

Describing the performance of a new diagnostic test:

Often times one is faced with the task of evaluating the merit of a diagnostic test. Assessing a new
diagnostic test begins with the identification of a group of patients known to have the disorder of
interest, using an accepted reference test known as the gold standard. However the gold standard
has the following limitations:
(a) The gold standard is often the most risky, technically difficult, expensive or impractical of
the available diagnostic options. (e.g. post-mortem brain biopsy, the gold standard for the
diagnosis of Alzheimers disease).
 43

(b) For some conditions (e.g. angina pectoris) no gold standard is available
(c) Comparisons with an imperfect gold standard may lead to the erroneous conclusion that the
new test is worse when in fact it is better. The following are examples:
(i) If the new test detects diseased individuals more accurately than the gold
standard, these patients will be mistakenly labelled false positives.
(ii) If the new test is negative in more disease free individuals these patients will be
mistakenly labelled false negatives

The results of a screening test and disease status, can be examined conveniently by use of the four
fold contingency table.

DISEASE STATUS
SCREENING TEST Positive Negative

True False Total Test

Test positive Positives (TP) positive (FP) positives (T+)

Test negative False True Total Test

Negatives (FN) Negatives (TN) Negatives (T-)

Total Disease Total Disease Total

Positive (D+) Negative (D-) population (N)

From a population of N people, D+ have the disease, while D- do not have the disease.

D+
---- Denotes the prevalence of disease in this population.
N

From this population, T+ persons are positive on the screening test, while T- persons are negative
on the screening test.

Diseased individuals with a positive screening test are the True Positives TP
Healthy individuals with a positive screening test are the False Positives FP
Diseased individuals with a negative screening test are the False Negatives FN
Healthy individuals with a negative screening test are the True Negatives TN

An ideal screening test has as few false positives and false negatives as possible. The capacity of a
screening test to identify correctly the diseased against the non-diseased is expressed by its
SENSITIVITY and SPECIFICITY which are a measure of the test's validity.

Sensitivity is the ability of a test to give a positive finding when the tested person truly has the
disease under the study. In other words it is the probability that the test will be reactive in a diseased
individual. A test with a high sensitivity will detect a high percentage of diseased individuals.
TP
Sensitivity = ---------X 100 %
D+

Specificity is the ability of a test to give a negative finding when the tested person is truly free of
the disease under study. In other words it is the probability that a test result will be non-reactive in
an individual who is not diseased.
 44

TN
Specificity = ---------X 100 %
D-
In practice, sensitivity of a test is usually determined on a group of proven cases of a given disease,
while its specificity is determined on a group of healthy people, or people with diseases other than
the one under investigation.

Sensitivity and specificity are test characteristics and are not very much influenced by properties of
the population such as age, sex, etc. However, the pattern and prevalence of other diseases will
influence specificity of the test. The mix of subclinical and clinical cases in a population might
affect the sensitivity of a test.

Predictive Value

When interpreting a screening test result, or a sign or symptom, under practical circumstances, it is
important to measure its diagnostic accuracy in the population in which it will be used. This is the
predictive value which is used to measure the accuracy of a test in classifying individuals as
diseased or not diseased when the test is positive or negative respectively.

The predictive value of a positive test (PVP), denotes the proportion of diseased amongst all
positive tests. In other words, it signifies the probability that a given patient with a positive test,
indeed does have the disease. PVP is also known as the posterior probability or post-test probability
of disease.

The predictive value of a negative test (PVN), denotes the proportion of healthy individuals
amongst all negative tests. In other words, it signifies the probability that a given individual with a
negative test, indeed does not have the disease.

The predictive value is influenced by the sensitivity and specificity of the test as well as the
prevalence of the disease in the population. The influence of prevalence on predictive value, and the
independence of sensitivity and specificity on prevalence is illustrated by the following table:

LOW PREVALENCE (10%) HIGH PREVALENCE (50%)

DISEASE DISEASE

Pos Neg Total Pos Neg Total

Test +ve 9 9 18 Test +ve 45 5 50

Test -ve 1 81 82 Test -ve 5 45 50

10 90 100 50 50 100

In both cases sensitivity and specificity are 90% but the predictive value of a positive test is
different and depends on the prevalence of the disease in the study sample.

LOW HIGH
PREVALENCE (10%) PREVALENCE (50%)
9 45
-- x 100 = 50 % -- x 100 = 90 %
18 50

The concepts of qualitative and quantitative screening tests.

In a qualitative test, the test result is categorical, i.e. it can only assume certain values, e.g. normal
or abnormal, positive or negative, present or absent, yes or no. For example, an X-ray might
confirm or disapprove the existence of a fracture.
 45

In a quantitative test, however, the test result is a continuous variable, which can assume all values
within a certain range, with a typical distribution (normal, log-normal, Poisson, bimodal etc.).
Classifying individuals as diseased or disease free is on the basis of whether they fall above or
below a preselected cut-off value known as the positivity criterion. This cut-off value is also
referred to as the critical value or referent value. The same applies to semi-quantitative tests like
serological tests (e.g. VDRL) which use titre values.

Selecting a positivity criterion:

A positivity criterion for a particular diagnostic test may be selected in one of the six ways:
1. Gaussian distribution method.
2. Percentile method
3. Culturally desirable method
4. Therapeutic method
5. Risk factor method
6. Diagnostic or predictive value method
(For details of these methods reference should be made to ‘Describing the performance of a
diagnostic Test’ chapter in Clinical epidemiology and biostatistics by Knapp RG and Miller III MC
pg 31-33)

Depending on the chosen cut-off point, the sensitivity and specificity of a test will vary as
illustrated in the following table :

DIABETICS (N=70) NON-DIABETICS (N=510)

-------------------------------------------------------------
No Test No Test Sensitivity No Test No Test Sensitivity Blood Positive
positive negative in % positive negative in % glucose in predictive
gm % value in %
70 0 100 504 6 1 80 12
69 1 99 473 37 7 90 13
68 2 97 381 129 25 100 15
65 5 93 263 247 48 110 20
62 8 89 162 348 68 120 28
60 10 86 90 420 82 130 40
52 18 74 45 465 91 140 54
45 25 64 20 490 96 150 69
39 31 56 7 503 99 160 85
37 33 53 2 508 100 170 95
35 35 50 1 509 100 180 97
31 39 44 1 509 100 190 97
26 44 37 0 510 100 200 100

Which cut-off point one wishes to choose, depends on several considerations:

1. Is it harmful or serious to miss a case ? If this is so then choose a value for the positivity
criterion that minimizes the false negatives and hence sensitivity should be high, usually at
the expense of specificity. (e.g. neonatal phenylketonuria screening).

2. Is treatment risky? (e.g. side-effects, or operative mortality), then the number of false
positives should be low, and specificity high usually at the expense of sensitivity. This also
applies when a false positive diagnosis may have deleterious effects on a patient's lifestyle,
self-image, or financial situation (e.g. AIDS, mental illness or learning disorders).
 46

3. Is a highly specific and sensitive confirmatory test available? Then one would go for a very
high sensitivity in the preliminary (screening) test.

Sensitivity however is always inversely proportional to specificity: by increasing sensitivity one has
to accept a loss in specificity and vice-versa.

Serial (consecutive) vs parallel testing

In serial testing we first apply one test to a certain population and all people identified as having a
positive test are then submitted to a second test (so called re-testing of reactive individuals with
another test). As an example we will take the VDRL test as a screening test for syphilis, and the
TPHA (Treponema Pallidum Haemoagglutination Assay) as a confirmatory test. We will assume a
population with a prevalence of syphilis of 20 %.

SYPHILIS
VDRL Positive Negative Total

Positive 180 80 260 P.P.V. =69.2%

Negative 20 720 740 Sensitivity =90 %
Specificity =90 %
Total 200 800 1000

In the serial testing the 260 positive on the VDRL (above table) will then be subjected to a
second series of testing with TPHA which is known to have the following test characteristics:
sensitivity 95 % and specificity 99 %. The result will be as shown in the table below.

SYPHILIS
TPHA Positive Negative Total

Positive 171 0.8 171.8 P.P.V. =99.53%

Negative 9 79.2 88.2 Sensitivity =95 %
Specificity =99 %
Total 180 80.0 260

The overall sensitivity = 171 / 200 = 85.5 %

The overall specificity = (720 + 79.2) / 800 = 99.9 %
Overall predictive value = 171 / 171.8 = 99.53 %

The overall sensitivity does decrease, while overall specificity increases when using tests in series.
Most importantly is that the predictive value always improves because the prevalence is increased
for the second test to 69.2% (180/260x100).
In general one can use the following general formulae to calculate overall sensitivity and
specificity:

Overall Sensitivity =[Sens A x Sens B]

Overall Specificity =[1 - (1-Spec A) x (1 - Spec B)]

As far as sensitivity and specificity are concerned, it does not matter in which order the tests are
carried out, but efficiency and total cost may differ considerably. In general, one benefits most from
serial testing if the most sensitive test is used for screening, and the most specific test as
confirmatory.
 47

Summary:
1. Parallel tests result in:
a. Greater sensitivity
b. Increased PVN (decreased False Negative Rate)
c. Decreased specificity (increased False positive Rate)

2. Serial tests result in:

a. Lower sensitivity (higher False Negative Rate)
b. Increased specificity
c. Increased PVP (lower False Positive Rate)

Screening Criteria for initiating a programme for screening

Disease:
• Should be serious and of public health importance
• The natural history should be well understood
• There should be a long period between first sign and overt disease
• High prevalence of pre-clinical stage
Diagnostic Test:
• Sensitive and specific:
• Simple and cheap
• Safe and acceptable to both the public and the professionals
• Reliable

Diagnosis and Treatment:

• Facilities should be available and adequate
• Treatment should be effective, safe, acceptable and available
• There should an agreed policy on whom to treat as patients including
management of borderline cases.

Costs of a screening programme must be balanced against the number of cases detected and the
consequences of not screening, the cost too, should be economically balanced in relation to the total
expenditure on medical care.

In a screening programme differentiate between lead time bias and length bias:

Lead time is defined as the interval between the diagnosis of a disease at screening and when it
would have been detected due to development of symptoms. In other words it represents the time by
which the diagnosis has been advanced as a result of screening. It is important to take into account
lead time when evaluating a screening programme since survival may erroneously appear to be
increased among screen detected cases simply because the diagnosis was made earlier in the course
of the disease. See Fig 1

Age (years)
35 40 41 43 46
_______|_____________|_________________|_______________|_______________|______

Biological onset Disease detectable Woman A Symptom develop Women A and B

of disease by screening diagnosed at woman B both die from
screening diagnosed breast cancer

Fig 1: Apparent increase in the duration of survival due to lead time in hypothetical screened and
symptom-diagnosed cases of breast cancer.
 48

Length bias is the overrepresentation among screen detected cases of those with a long preclinical
phase and hence favourable prognosis. This occurs because diseases with long preclinical phase are
readily detected than rapidly progressing cases with a short preclinical phase. A programme may
seem successful when in fact observed differences in mortality were a result merely of the detection
of less rapidly fatal cases through screening while those more rapidly fatal are diagnosed after
development of symptoms.

Examples of generally accepted screening methods, diseased and target populations:

Test Disease Target group

weight malnutrition under-fives
HB malnutrition, malaria, hookworm,
sickle cell disease
VDRL syphilis during pregnancy pregnant
mothers
Widal typhoid fever food handlers

References:

1. Griner, PF., Mayeuski, RJ., Mushlin, AI., Greenland, P. Selection and interpretation of
diagnostic tests and procedures Ann. Intern. Med. 1981: 94 557-93.
2. Wilson JMG and Juguer F. Principles and practice of screening for disease (Public Health
papers No 34) WHO, Geneva, 1968.
3. Galen and Gambino Beyond Normality. The predictive value and Efficiency of medical
Diagnoses. John Wiley and Sons (1975).
4. Gurick DS. Efficacy of screening of cervical cancer; a review Am. J. Public Health 68: 125,
1978.
5. Elveback LR. How high is high? A proposed alternative to the normal range. Mayo Clinic
Proceedings 47: 93. 1972.
6. Charles H Hennekens and Julie E Buring
Epidemiology in Medicine
Little Brown and Company
Boston/Toronto 1987 pg 327-347
7. Lewis H Roht, Beatrice J Selwyn, Alfonso H Holguin JR
Bobbe L Christensen
Principles of Epidemiology: A self teaching Guide
Academic press Inc, Harcourt Brace Jovanovich, Publishers
San Diego, New York, London 1982 pg 217-230.
8. Rebecca G. Knapp, M Clinton Miller III
Clinical Epidemiology and Biostatistics
NMS National Medical Series
Harwal Publishing Company, Malvern, Pennsylvania
Baltimore, Hong Kong, London Sydney 1992. pg 31-50
9. Sackett, DL. Holland, WW. Contraversy in the detection of disease. The Lancet August 23
1975 pg 357-359
10. Loefler IJP. Screening for cancer. Is it useful ? East Afr Med J 71 (5) May 1994 pg 323-7
 49

Epidemiological approach to causation

Introduction

When disease occurs in a population there is always an attempt for investigators to determine the
cause for such occurrence. One of the first steps towards finding the cause is to collect information
that will enable the investigator to associate suspected or potential causal factors with the disease. In
epidemiology a variety of methods are available to determine associations between suspected causal
factors and disease. However, such associations even though they may be statistically significant do
not by necessity mean that they are causal. A causal association is defined as an association
between a factor and a disease in which an alteration of the factor is followed by a change in the
incidence of the diseases.

The Concept of Cause

We shall illustrate the concept of cause by using the electric circuit with which most of you are
familiar. We all know that the flick of a light switch will make the lights go on, but the unseen
causes that also operate to produce the effect are usually unappreciated. The need for a good light
bulb in the socket, appropriate wiring from the switch to the bulb, and adequate voltage to produce
a current when the circuit is closed. To achieve the effect of turning on the light each of these is
equally as important as moving the switch because absence of any of these components of the
causal constellation will prevent the effect from occurring.

In the health field, it is important to understand the causes of disease in order to look for rational
means of preventing disease. Prevention here is taken to be broad and includes primary, secondary
and tertiary prevention. In secondary prevention for example, correct treatment can only be given
when the cause is known from early diagnosis.

In clinical sciences a cause can be defined as an event, condition, characteristic or their combination
which plays an important role in producing the disease in an individual. Logically therefore, a
cause must precede a disease in time.

A cause is said to be sufficient when it inevitably produces or initiates a disease in an individual. A

cause is said to be necessary if a disease cannot develop in its absence.

Risk factor versus cause

The term "risk factors" is commonly used to describe factors that are positively associated with the
risk of developing a disease but that are not sufficient to cause the disease. The concept has been
found useful in a number of practical prevention programmes. Some risk factors (e.g. tobacco
smoking) are associated with several diseases, while some diseases (e.g. coronary heart disease) are
associated with several risk factors. Epidemiological studies can measure the relative contribution
of each factor to disease occurrence, and the corresponding potential reduction in disease from the
elimination of each risk factor. The concept of attributable fraction (aetiological fraction) with
which you should be familiar is key to this contribution.

The concept of models of causation

In epidemiology, two models of disease causation have been described. The single factor model
and the multifactorial model.

Single factor model:

The disease has either a single necessary and sufficient cause, or there is a single necessary cause,
but this will not be sufficient to cause disease unless the individuals resistance is inadequate. This
model is often applicable to communicable diseases.
 50

Multifactorial model:
The disease may have neither a single sufficient cause nor a single necessary one. The sufficient
cause is not usually a single factor but often comprises several components. Each sufficient cause
has a necessary cause as a component. For example, there are different components in the causation
of tuberculosis but the tubercle bacillus is a necessary cause. The disease therefore, is a result of a
complex interaction of factors, some tending to promote the disease and others tending to oppose it.
This model is consistent with current views on chronic disease aetiology.

In general, it is not necessary to identify all the components of a sufficient cause before effective
prevention can take place since the removal of one component may interfere with the action of
others and thus prevent the disease. For example cigarette smoking is one component of the
sufficient cause of lung cancer. Smoking is not sufficient in itself to produce the disease; other
factors, mostly unknown, are required. However, the cessation of smoking reduces the number of
lung cancer cases in a population even if the other component causes are not altered.

The following rules were formulated by Henle and Koch in an attempt to determine whether a
specific living organism causes a particular disease. Anthrax was the first disease demonstrated to
meet these rules. Henle and Koch formulated these rules after they had been spurred by Pasteur's
work on micro-organisms.

Henle and Koch’s postulates on causation:

•The organism must be present in every case of the disease.
•The organism must be able to be isolated and grown in pure culture.
•The organism must, when inoculated into a susceptible animal, cause the specific disease.
•The organism must then be recovered from the infected animal and identified.

Even though the above rules may be applicable to some diseases, they are inadequate for
determining causation for most diseases, both infectious and non-infectious. Many causes are
usually operating in the causation of a disease, while a single factor such as cigarette smoking may
be a cause of many diseases. In addition, the causative organism may disappear when a disease has
developed making it impossible to demonstrate the organism in the sick person. Rules such as
Koch's are of value only when the specific cause is an overpowering infectious agent, a situation
which is uncommon. Susceptibility due to other factors, and a sufficient dose of the agent are
usually required before clinical disease ensues.

Types of associations

In order to determine whether an association is causal or not, it is necessary to have a thorough

knowledge of the types of association which have been described and to rule out those which are
not causal. In epidemiology, associations are of two main types:

1. Non-causal association which may occur in three ways:

(a) Artefactual or spurious association may arise because of bias in the study. Examples of
artefactual association are abundant. In determining the association between use of oral
contraceptives and development of cardiovascular disease, cases should not be recruited
from a cardiac clinic. Under such a situation, any association found will most likely be
due to selection bias.
(b) When a factor A is causally associated with both category B and C, an association may
then be found between B and C, but this would be non-causal since a change in B
would not produce a change in C.
(c) When a disease causes exposure (rather than the exposure causing disease) e.g. iron
deficiency anaemia and geophagia (craving for soil). In this situation, disease precedes
the cause.
 51

2. Causal association which is said to be present between factor A and disease B, when a change
in A produces a change in B.
For example, in situation (a) above a positive statistical association has been demonstrated between
CHD mortality rates and coffee drinking habits. (See Table 1)

Table 1: CHD mortality rates for males aged 55-64 years by coffee drinking habits.

Coffee consumption CHD mortality

(cups/day) (Per 1,000 / year
None 6
1-5 8
6 or more 12

From the above table it can be seen that as coffee consumption increases the CHD mortality also
increases. However, it has been shown that people who drink coffee also tend to be cigarette
smokers, while cigarette smoking is known to be strongly associated with CHD mortality. (See
Table 2).

Table 2: CHD mortality rates for males aged 55-64 years by cigarette consumption.

Cigarette consumption (packs CHD mortality

/day per 1,000/ year
None 4
1-2 10
3 or more 15

Thus, to isolate the effect of coffee drinking, from that of cigarette smoking we cross-tabulate CHD
mortality according to both variables as shown in Table 3. For this process, additional analysis of
the raw data would be required.

Table 3. CHD mortality rates (per 1,000 per year) for males aged 55-64 years by coffee drinking
habits and cigarette consumption.

Coffee consumption Cigarette consumption

cups per day (packs/day)
0 1 - 2 3+ Total
None 4 9 15 6
1-5 6 10 13 8
6 or more 5 9 16 12
Total 4 10 15

Table 3. shows that when cigarette smoking is held constant the effect of coffee drinking on
mortality disappears. Thus, the association between coffee drinking and CHD mortality is non-
causal (Figure 1). In general when an outcome is affected by multiple variables, in order to
examine the influence of a single one it is necessary to adjust for the effects of the others
 52

Cigarette
smoking
Causal Non-causal

CHD coffee
mortality Non-causal association drinking

Figure 1: Causal and non-causal association.

Causation or merely association?

Causal inference is the term used for the process of determining whether observed epidemiological
associations are likely to be causal; the use of guidelines and the making of judgements are
involved. Before an association is assessed for the possibility that it is causal, other explanations,
such as chance, bias and confounding, have to be excluded. The steps in assessing the nature of the
relationship between a possible cause and an outcome are shown in Fig. 2.

OBSERVED ASSOCIATION

Could it be due to selection

or measurement bias?

NO

Could it be due to
confounding?

NO

Could it be a result of
chance?

PROBABLY NOT

Could it be causal?

APPLY GUIDELINES
AND MAKE JUDGEMENT

Figure 2. Assessing the relationship between a possible cause and an outcome

 53

A systematic approach to determining the nature of an association was used by the United States
Surgeon General to establish that cigarette smoking caused lung cancer. This approach was further
elaborated by Hill (1965) and on the basis of these concepts, a set of "guidelines for causation" has
been prepared. In Table 4 the concepts are listed in the sequence of testing that the epidemiologist
should follow to reach a conclusion about a cause of disease.

Table 4: Guidelines for causation

Temporal relationship
Does the cause precede the effect? (essential)
Plausibility Is the association consistent with other knowledge? (mechanism of
action; evidence from experimental animals)
Consistency Have similar results been shown in other studies?
Strength What is the strength of the association between the cause and the
effect? (relative risk)
Dose-response Is increased exposure to the possible cause associated with
relationship increased effect?
Reversibility Does the removal of a possible cause lead to reduction of disease
risk?
Study design Is the evidence based on a strong study design?
Judging the evidence How many lines of evidence lead to the conclusion?

To establish whether or not an association is causal may not be easy. If feasible or ethical this could
be investigated by an experiment, i.e. a randomized controlled trial. However, this may not be
feasible in many situations, and therefore an indirect approach is used by asking the following
questions:

1. What is the strength of the association?

A strong association between possible cause and effect, as measured by the size of the risk ratio
(relative risk), is more likely to be causal than a weak association, which could easily be influenced
by confounding or bias. The stronger the association (high relative risk) the more readily we can
accept direct causation as the likely explanation of the observed association.

Relative risks greater than 2 can be considered strong. For example, cigarette smokers have an
approximately two-fold increase in the risk of acute myocardial infarction compared with non-
smokers. The risk of lung cancer in smokers, compared with non-smokers, has been shown in
various studies to be increased between four-fold and twenty-fold. However, such very strong
associations are rare in epidemiology.

The fact that an association is weak does not preclude it from being causal; the strength of an
association depends on the relative prevalence of other possible causes. For example, weak
associations have been found between diet and risk of coronary heart disease in observational
studies; and although experimental studies on selected populations have been conducted, no fully
satisfactory trials have been completed. Despite this, diet is generally thought to be a major
causative factor in areas with high rates of coronary heart disease as in many industrialized
countries.
 54

2. Is the association consistent?

Consistency is demonstrated by several studies giving the same result. If the same association has
been repeatedly observed by different researchers in different places, under different circumstances
and at different times, it is very likely that the association is causal.

For example, the association between smoking and lung cancer was found in 29 case control studies
and 7 cohort studies in the USA. Broadly, the same answer has been reached in quite a wide variety
of situations and techniques. In this case it is very likely that the association between cigarette
smoking and lung cancer is causal. Consistency is particularly important when a variety of designs
are used in different settings, since the likelihood that all studies are making the same mistake is
thereby minimized. However, a lack of consistency does not exclude a causal association, because
different exposure levels and other conditions may reduce the impact of the causal factor in certain
studies. Furthermore, when the results of several studies are being interpreted the best-designed
ones should be given the greatest weight.

Techniques are available for pooling the results of a number of studies that have examined the same
issue, particularly randomized controlled trials. This technique is called meta-analysis, and it
combines the results of a number of well-designed trials, each of which may deal with a relatively
small sample, in order to obtain a better overall estimate of effect (Sacks et al., 1987).

Fig. 3 illustrates the results of 11 trials on the use of β-blockers for the prevention of death after
myocardial infarction. One important reason for the apparent inconsistency of the results is that
several of the early studies were on small samples. The estimated relative risk in each study is
marked by a cross; the horizontal lines indicate the 95% confidence intervals. For the aggregated
data from all the trials, covering a very large number of events, the 95% confidence interval is very
narrow. Overall, treatment with β-blockers after myocardial infarction is seen to reduce the death
rate on average by about 35%; the 95% confidence interval shows that the reduction in death rate is
at least 20% and could be as much as 50%.

x
x
x
Individual x
x
trials x
x
x
x
x
x

All trials x

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Figure 3: Meta-analysis of selected randomized trials of beta-blockers in the prevention of deaths

following a myocardial infarction.
 55

3. Is there a temporal relationship?

"Which is the cart and which is the horse?" Exposure to the factor must necessarily precede
development of the disease in order to consider a causal association. Temporal relationship is
therefore crucial in studies of cause and effect. This is usually self-evident, although difficulties
may arise in case-control and cross-sectional studies when measurements of the possible cause and
effect are made at the same time and the effect may in fact alter the exposure. In cases where the
cause is an exposure that can be at different levels, it is essential that a high enough level be reached
before the disease occurs for the correct temporal relationship to exist. Repeated measurement of
the exposure at more than one point in time and in different locations may strengthen the evidence.

4. Is there a biological gradient?

If a dose-response relationship can be demonstrated, then the likelihood that the exposure is causal
increases. A dose-response relationship occurs when changes in the level of a possible cause are
associated with changes in the prevalence or incidence of the effect. Table 5 illustrates the dose-
response relationship between noise and hearing loss in which the prevalence of hearing loss
increases with noise level and exposure time. The demonstration of a clear dose-response
relationship in unbiased studies provides strong evidence for a causal relationship between exposure
or dose and disease.

Table 5: Percentage of people with hearing loss.

Noise level per

working day (decibels) Exposure time in years
5 10 40

<80 0 0 0
85 1 3 10
90 4 10 21
95 7 17 29
100 12 29 41
105 18 42 54
110 26 55 62
115 36 71 64

5. Is there biological plausibility?

If an association is biologically plausible, i.e. consistent with other knowledge, it is more likely to
be causal. For instance, laboratory experiments may have shown how exposure to the particular
factor could lead to changes associated with the effect measured. However, biological plausibility
is a relative concept, and seemingly implausible associations may eventually be shown to be causal.
For example, the predominant view on the cause of cholera in the 1830s involved "miasma" rather
than contagion. Contagion was not supported by evidence until Snow's work was published; much
later, Pasteur and his colleagues identified the causative agent. Lack of plausibility may simply
reflect lack of medical knowledge. The scepticism that still exists about the therapeutic effects of
acupuncture and homeopathy may be at least partly attributable to the absence of information about
a plausible biological mechanism.

The study of the health consequences of low-level lead exposure is an example of the opposite
situation. Animal experiments indicate an effect of lead on the central nervous system. Similar
effects in an epidemiological study of children are therefore plausible but, because of potential
confounding factors and measurement difficulties, epidemiological studies have shown conflicting
results. However, assessment of all the available epidemiological data leads to the conclusion that
effects do occur in children at a low level of exposure to lead.
 56

6. Reversibility
When the removal of a possible cause results in a reduced disease risk, the likelihood of the
association being causal is strengthened. For example, the cessation of cigarette smoking is
associated with a reduction in the risk of lung cancer relative to that in people who continue to
smoke. This finding strengthens the likelihood that cigarette smoking causes lung cancer. If the
cause leads to rapid irreversible changes that subsequently produce disease whether or not there is
continued exposure (as with HIV infection), then reversibility cannot be a condition for causality.

7. Study design
The ability of a study design to prove causation is the most important consideration (Table 6). The
best evidence comes from well-designed, competently conducted randomized controlled trials.
However, evidence is rarely available from this type of study, and usually only relates to the effects
of treatment and prevention campaigns. Other experimental studies, such as field and community
trials, are seldom used to study causation. Evidence comes most often from observational studies,
almost all the evidence on the health consequences of smoking comes from observational studies.

Table 6 Relative ability of different study designs to ‘prove’ causation

Type of study Ability to ‘prove’

causation

Randomized controlled trials Strong

Cohort studies Moderately strong
Case-control studies Moderately strong
Cross-sectional studies Weak
Ecological studies Weak

Cohort studies are the next best design because, when well conducted bias is minimised. Again,
they are not always available. Although case-control studies are subject to several forms of bias,
the results from large well-designed investigations of this kind provide good evidence for the causal
nature of an association; judgements often have to be made in the absence of data from other types
of study. Cross-sectional studies are less able to prove causation as they provide no direct evidence
on the time sequence of events.

Ecological studies provide the least satisfactory type of evidence on causality because of the danger
of incorrect extrapolation to individuals from data on regions or countries. However, for certain
exposures that cannot normally be measured individually (such as air pollution, pesticide residues in
food, fluoride in drinking water), evidence from ecological studies is very important. Yet only very
rarely has it been considered adequate for establishing causation. In 1968 the sale of
bronchodilators without prescription in England and Wales was stopped because the increase in
asthma deaths in the period 1959-66 had been shown to coincide with a rise in bronchodilator sales.
Despite the fact that only very limited evidence was available linking the use of bronchodilators
with death in asthmatics, the ecological evidence was deemed sufficient; two decades later this
relationship continues to be debated and has relevance to a recent increase in asthma deaths among
young people in New Zealand Crane et al., 1989).

8. Is the association specific?

A particular exposure should produce one specific disease otherwise there is a weak argument in
favour of causation. If the association is limited to specific workers and to particular sites and type
of disease, while there is no association between the type of occupation and other diseases, then
clearly, that is a strong argument in favour of causation. However, causes of a given effect cannot
be expected to be without other effects on any logical grounds. Single events may therefore have
many effects.
 57

9. Is there coherence?
Coherence implies that a cause and effect interpretation of an association does not conflict with
what is known of the natural history and biology of the disease. Any cause and effect interpretation
of, for example , cigarette smoking and lung cancer should be coherent with the histopathologic
effects of smoking on the bronchial epithelium. In this regard, the difference between coherence
and consistency is unnecessarily fine.

10. Analogy
The premise that, if one drug can cause birth defects, perhaps another one can also, could
conceivably enhance the credibility that an association is causal.

Judging the evidence

Regrettably, there are no completely reliable criteria for determining whether an association is
causal or not. Causal inference is usually tentative and judgements must be made on the basis of the
available evidence: uncertainty always remains. Evidence is often conflicting and due weight must
be given to the different types when decisions are being made. In judging the different aspects of
causation referred to above, the correct temporal relationship is essential; once that has been
established, the greatest weight may be given to plausibility, consistency and the dose-response
relationship. The likelihood of a causal association is heightened when many different types of
evidence lead to the same conclusion. Evidence from well-designed studies is particularly
important, especially if they are conducted in a variety of locations.

References and further reading.

1. Foundations of Epidemiology Lilienfeld and Lilienfeld Oxford University Press
2. Epidemiology: Principles and Methods MacMahon, B. and Pugh, T.F. Little, Brown Co.,
Boston
3. A study guide to epidemiology and Biostatistics Morton, R.F. and Hebel J.R. University Park
Press Baltimore 1979.
4. A short textbook of medical statistics, Hill, A.B. Hodder and Stoughton, London, Sydney,
Toronto, 1977.
5. Basic Epidemiology, R. Beaglehole, R. Bonita and T. Kjellstrom, World Health Organization,
Geneva, 1993.
6. Crane J. et al 1989, Prescribed fenoteral and death from asthma in New Zealand 1981-1983: a
case-control study. Lancet 1:917-922
7. Sacks HS et al 1987, Meta-analysis of randomised controlled trials. New England journal of
medicine 316: 450-455
8. Rothman KJ 1986, Modern Epidemiology, Little Brown and Company, Boston/Toronto
 58

2
METHODS OF EPIDEMIOLOGY
 59

Introduction to epidemiological studies and research

methodology
Introduction to research methodology

Research:
Activity aimed at the advancement of knowledge (scientific or non-scientific)

Methodology:
The body of methods in a discipline

Purpose of Research

Collection of information that will contribute to the solution of a problem and therefore, provide a
basis for action whether immediately or in the long run. Research begins when an investigator
perceives a problem which in his/her view requires a solution. He/she then decides that a particular
study will contribute to this end, and embarks upon it through a research process..

Study- a project designed to yield evidence for the advancement of knowledge (in science).
Once the problem has been identified, it should be described clearly and appropriate hypotheses
formulated and methods for testing them designed. This is important particularly if funds or
facilities are applied for from funding agencies, in order to justify expenditure of public funds. In
order to do this well the following must be done:
• Review of previous work on the subject
• Describe the present state of knowledge
• Explain the significance/rationale of the proposed investigation

In identifying the problem preconceived ideas must be avoided at any cost lest they introduce bias
in the findings. If you think a service is bad you should not collect data that will help you to
condemn it. You should be honest and ensure objectivity in the collection and interpretation of
information

Formulating the topic for research

With a clear purpose in mind (why you want to conduct the study) you can now formulate the topic
of study in general terms.

If the reason for the investigation is that Infant Mortality is unduly high in a given population and
there is insufficient information on its causes for the planning of an action program the topic can
broadly be stated as:

“The causes of I.M. in a defined area in a given time period”.

At this stage the topic may be regarded as provisional and may frequently change as more
information becomes available. Study objectives are then listed and research methods identified
and described.
Quantitative methods in research are the back bone of epidemiology. They generate data that are to
be analysed quantitatively. They are also known as epidemiological methods because they are
designed to determine the extent and distribution of diseases and their determinants (causes) in
human populations with the aim of identifying effective management and preventive measures.
 60

Due to the quantitative nature of epidemiologic studies, statistical techniques for measuring these
‘quantities’ have been developed and extensively used. Examples of measurements commonly used
in epidemiology are prevalence and incidence. These are used to measure the magnitude and
‘speed’ of spread of the disease respectively.

Prevalence represents the proportion of the population that is affected by the disease under study at
a given point in time, while incidence represents the rate of new cases of disease occurring in a
defined population within a specified period of time.

In quantitative research, measures of comparison are used to compare the risks of disease in
different population groups. These are either absolute or relative measurements. Of particular
interest to disease aetiology are the relative measurements which are used to determine whether or
not suspected causes are related to the disease (or health problem) being investigated. These
measurements are the relative risks (RR) or odds ratios (OR) which are used to measure the degree
of association such that the further away the calculated quantity is from unity (RR=1) the stronger is
the evidence for causation.

Types of Research Studies/Methods/Investigations

Depending on the type of health problem to be investigated a variety of study designs have been
developed and grouped into two broad categories. These are observational and experimental studies.

Observational studies involve only observational techniques and no manipulations like those
employed in experiments are used to study the population. Figure 1 depicts a common classification
of the methods.

Research methods discussed in this module are those which fall within the realm of epidemiology.
Other research methods e.g. socio-anthropological techniques will be discussed in the relevant
disciplines. Epidemiological studies are designed to determine the extent and distribution of
diseases and their determinants (causes) in human populations with the aim of identifying effective
management and preventive measures.

Cross sectional
Descriptive
Longitudinal

(1) Surveys Cross sectional

Analytical
Longitudinal

Surveys are investigations in which information is systematically collected but in which the
experimental method is not used. These are also called observational studies.

A) Descriptive studies

Descriptive surveys set out to describe disease situation in a particular area. e.g. the
distribution of a disease in a population in relation to age, sex, time, place and other
characteristics. Studies of this kind are not necessarily time limited. Registries of diseases in
a particular area are examples of descriptive studies that are not time-limited. They record
various characteristics of the affected individuals including age, sex, occupation, duration
of symptoms etc. Cross-sectional studies are examples of time-limited descriptive studies
but can sometimes be used as analytical studies. Descriptive studies are useful in studying
the natural history of diseases.
 61

B) Analytical studies

Analytical surveys (or explanatory surveys) try to explain the situation. These are studies designed
specifically to explain the determinative processes of diseases. They answer the questions; why
does the disease occur in the persons having it and not in the persons not having it? Why do certain
persons fail to make use health services? Can the decreased incidence of the disease be attributed to
the introduction of preventive measures? To answer these questions hypotheses are formulated and
tested that may help to explain the situation. Examples of analytical studies are ecological, cross-
sectional, cohort and case-control. These are briefly described below:

Ecological studies

These studies are often referred to as correlational studies. They frequently initiate the
epidemiological process and allow for more detailed analysis of observed correlations. The units of
analysis are groups of people rather than individuals. For example, in one country, a relationship
was demonstrated between average sales of an anti-asthma drug and the occurrence of an unusually
high number of asthma deaths. Such relationships may be studies by comparing populations in
different countries at the same time or the same population in one country at different times. The
latter approach may avoid some of the socio-economic confounding.

Although simple to conduct and thus attractive, ecological studies are often difficult to interpret
since they usually rely on data collected for other purposes and hence essential exposure data may
not be available. In addition, since the unit of analysis is a population or group, the individual link
between exposure and effect cannot be made. One attraction of ecological studies is that data can be
used from populations with widely differing characteristics e.g. correlation between oesophageal
cancer rates in communities with different patterns of salt consumption.

Ecological fallacy or bias may result from ecological studies if one erroneously infers that
relationships established between two or more variables, measured at an aggregate level, will also
hold at the individual level.

Cross-sectional studies

Cross-sectional studies or prevalence studies are carried out at a certain point in time and in a given
population or geographical area. They depend on a single examination of a cross-section of the
population in which sick and healthy, or exposed and unexposed are not distinguished until results
are examined.

Information is collected through a survey by means of questionnaires, and/or laboratory or physical

examination of individual members of the study population.

The prevalence of a risk factor or a disease is expressed as the proportion of the affected individuals
in the study population in a given geographical and area at a given point in time. An important
limitation of cross-sectional studies is their inability to sort out cause and effect relationships since
both are found in the study population at the same time.

Cohort studies

Cohort studies are synonymous with prospective studies, longitudinal studies, follow-up studies or
incidence studies. They can be carried out prospectively or historically ("retrospectively").

These studies are carried out on a sample of the population to determine the rate at which groups of
the population develop disease or die from it when differentially exposed. This is one way of testing
a hypothesis in disease causation.
Basic data from cohort studies are represented in Table 1
 62

Table 1
Disease status
diseased healthy Total incidence
Exposure status
exposed a b a+b a/a+b)
not exposed c d c+d c/c+d)

When the incidence of disease among the exposed (a/a+b) is divided by the incidence of disease
among the unexposed (c/c+d), a measure of effect of exposure (RR) is obtained and an association
is said to exist between exposure and development of disease if the measure is significantly
different from unity. Standard statistical techniques are available to test this difference. Although
cohort studies may take long to accomplish, they have the advantage that they are more reliable in
providing evidence for causation than other analytical studies.

Case-control studies

Synonyms are case-referent studies. These studies involve the comparison of cases of the disease
under study with comparable controls for levels of exposure. The effect of exposure in such studies
is measured by the odds ratio (OR = ad/bc) as shown in Table 2 which is an approximation to the
relative risk (RR).

Case-control studies are by far the simplest in determining cause and effect relationships. They take
a short time to complete and have the advantage that they use cases of the disease under study and
are especially useful for studying rare diseases. Important disadvantages of case-control studies are
that they are more susceptible to selection bias and that information on exposure is less accurately
ascertained than in cohort studies.

Table 2
Disease status
Cases Controls Total
(diseased) (healthy)
Exposure status
exposed a b a+b
not exposed c d c+d

Odds ratio = ad/bc

Surveys can be cross sectional or longitudinal

Cross sectional surveys provide information concerning the situation that exists at a given time. e.g.
weights of children at a given time.

Longitudinal surveys provide data concerning events during a period of time. e.g. Weights of
children as they grow older, each child being weighed more than once during the study period. In
determining the natural history of a disease e.g. tuberculosis or AIDS individuals are followed up
through acquisition of the disease and appearance of signs and symptoms to recovery chronicity,
development of various complications or death.
 63

Types of Analytical Surveys

Group based:
In these surveys groups of populations are compared e.g. groups of countries. Group Information is
obtained from individual countries or from the same country at different time periods.
Example: National Heart disease rates and consumption of animal fat in the country.
AIDS case rates and importation of used clothes.

Inferences from such surveys are usually regarded as hints because they often
suffer from Ecological fallacy: e.g. Proportions affiliated to certain religions
and suicide rates in a country. Proportion of the population in mining
occupations and lung cancer rates in the country..

Individual based:
These are also surveys of groups but utilize information from individuals. Such surveys are
performed to test hypotheses that a specific factor is related to a specific disease.

e.g. cross sectional - unselected population (prevalence study)

Case - control or cohort - These studies require data referring to more than one point in time and
hence are longitudinal or time span studies.

1. Experimental studies
An experimental study is an investigation in which the researcher, wishing to study the
effects of exposure to or deprivation of a defined factor, decides which subjects (persons,
animals, communities etc.) will be exposed to, or deprived of, the factor. If the investigator
compares subjects exposed to the factor with subjects not exposed to it, the study is a
controlled experiment. Intervention studies and clinical trials are examples of experimental
studies.

2. Other studies
A Evaluative Studies:-These are carried out to appraise the value of health care. They set out
to measure how good care is.

(i) Programme review:-

• It evaluates the care given to specific patients community or population.
• Programme - any enterprise organised to eliminate or reduce one or more problems.
• Programme review may evaluate a particular programme that operates in a defined
setting with well defined aims and goals.
• Immunization coverage in an area, iodination of salt for goitre prevention and
fluoridation of water supplies for the prevention of dental caries.

(ii) Inbuilt evaluation - is that evaluation which is planned in advance and the required
information is collected in a systematic way as an integral part of the provision of
service.
(iii) Medical Audit: another type of evaluation in which the quality of a service is
evaluated by appraising the quality of care given to individuals. If a medical audit
discovers that some useful procedure is not being done in the course of patient care,
then it recommends that it should be done. e.g. testing of blood for Hepatitis B
virus before transfusion.

(B] Surveillance - Systematic collection analysis, and use of information for the control of a
specific disease. Generally it is the maintenance of an ongoing watch over the health status
of a community.
 64

(C] Pilot Study: - A dress rehearsal of an investigation performed in order to identify defects in
the study design.

(D] K.A.P Studies:- Studies of knowledge, attitudes and practices towards health care. These
are important in health education methods

(E] Operational Research - Research concerned with organizational problems.

i.e. How best can service be provided given all the possible constraints.

Example: We know vaccination against measles is effective and could eradicate measles, but
there are many constraints.
Cold chain maintenance problems
Faulty immunization techniques
Lack of proper supervision
Inadequate facilities and resources
Lack of staff motivation
Lack of community motivation

Operational research determines which one of these constrains is most important in the control of
measles

Types of epidemiological studies. (see figure 1).

(a) Observational studies
• descriptive studies (descriptive in terms of person, place and time)
• analytical studies. (Ecological studies, cross-sectional studies, case-control studies and
cohort studies).

(b) Experimental studies

• classical experimental design
• Natural experiments
• randomized controlled trials e.g. clinical trials, prophylactic trials, intervention studies
 65

EPIDEMIOLOGICAL
STUDIES

OBSERVATIONAL EXPERIMENTAL
STUDIES STUDIES

Classical design
Natural Exper'mts
Randomized
controlled trials

DESCRIPTIVE ANALYTICAL
Clinical
STUDIES STUDIES
trials
Ecological
Person (Who) Prophylactic
Cross-sectional trials
Place (Where)
Case-control
Time (When) Interventions
Cohort

Figure 1: Classification Of Epidemiological Studies

NB: The above figure is a very simplified one and is meant to assist the reader in understanding the
possible relationships in epidemiological methods. In practice there will be overlaps between certain
methods.
 66

Determine
Cross-sectional study
prevalence

Determine
Cases
Past Case Control study
Controls
Exposure

Exposed

Retrospective cohort study Prospective cohort study

Unexposed

Determine Incidence

PAST NOW FUTURE

Figure 2 The time relationship of analytical studies:

Reference and further reading:

1. Foundations of Epidemiology Lilienfeld and Lilienfeld

Oxford University Press
2. Epidemiology: Principles and Methods MacMahon, B. and Pugh, T.F.
Little, Brown Co., Boston
3. A study guide to epidemiology and Biostatistics Morton, R.F. and Hebel J.R.
University Park Press Baltimore 1979.
4. A short textbook of medical statistics
Hill, A.B. Hodder and Stoughton, London, Sydney, Toronto, 1977.
 67

Descriptive studies
Descriptive studies describe pattern of disease occurrence in relation to, place, time and person.
There are three types of descriptive studies, namely ecological (correlational) and cross-sectional
studies (which can be considered partially analytical), case reports, case series, registries, etc.

Correlational studies (ecological studies):

Measures that represent characteristic of the entire population are used to describe disease in
relation to some factor of interest i.e. per capita food consumption, per capita cigarette consumption,
infant mortality, mean annual rainfall etc. The two variables are correlated and the measure of
correlation is correlation coefficient ® . Correlating per capita cigarette smoking and the
occurrence of lung cancer.

Correlation coefficient(r)
Quantifies the extent to which there is linear relationship between exposure and disease (it ranges
from -1 to +1)

Advantages of correlational studies:

They are quick to be done,(most often you use routine data already available)
They are cheap to conduct.

Limitations:
• Unable to link exposure with disease in particular individuals, i.e. increasing pap smear rate with
decreasing mortality from cancer of cervix. In this case it is not easy to prove that those who
underwent pap smear are the same who benefited.
• Lack of ability to control for the effects of potential confounding factors.

It has been shown that there is a very strong correlation between per capita number of colour
television sets and coronary heart disease in various countries. Owning a colour television is an
indicator of economic well being and related to other factors which are more likely to increase
the risk of CHD than ordering a colour television

• Lack of correlation does not mean lack of statistical association, some relationships can only be
detected by analytical studies.

• Correlational data represent average exposure levels rather than actual individual values.
e.g. CHD showed a strong negative correlation with alcohol consumption, countries with high
alcohol consumption showed a low risk of coronary heart disease, from analytical studies it
was shown that the relationship was J-shaped. Thus consuming moderate amounts of alcohol
was beneficial.

Case reports:
Describes the experience of a single patient or a group of patients with similar diagnosis.
Case reports: document unusual medical occurrences and can represent the first clue in the
identifications of new disease or adverse effects of exposure: e.g. drugs

Case series:
Collection of individual case reports which may occur within a fairly short period of time.
e.g. Pneumocystis carinii pneumonia was initially described in young homosexuals who were
initially healthy, otherwise it was previously described amongst elderly patients suffering from
debilitating disease like cancer. This then led to the idea that this was a probably new disease, and
so was Kaposi sarcoma occurring amongst young people.
Limitation:
. Cannot test for the presence of valid statistical association.
 68

. Based on the experience of one individual.

Uses:
Useful for hypothesis formulation.

Cross-sectional studies (prevalence).

Cross-sectional studies or prevalence studies are carried out at a certain point in time and in a given
geographical area. They depend on a single examination of a cross-section of the population in
which sick and healthy are not distinguished until results are examined.
From a well defined population, disease status and exposure are assessed simultaneously.
The point in time could be:

1. Calendar year,( mid year, mid month,) or

2. fixed point in the course of events which varies in real time from person to person, e.g.
menarche, adolescence, military recruitment, school age etc.

Information is collected through a survey by means of questionnaires, and/or laboratory or physical

examination of individual members of the study population. The prevalence of a factor or a disease
is expressed as the proportion affected in a given geographical or area at a given point in time.

Uses of Cross-sectional studies:

Cross-sectional studies can be used for the following purposes:

1. To determine the magnitude of disease or disease determinants in a community in terms of

their prevalence.
2. To study preliminary associations between disease and possible aetiological factors by
comparing the characteristics of the sick with those of the healthy.
3. To screen for undiagnosed disease in a community.

Limitations of Cross-sectional studies:

Not possible to determine whether the exposure preceded or resulted from the disease.
Reflects determinants of survival as well as aetiology.

Stages in the design and conduct of cross-sectional studies

1. Specification of the aim of study. Aim should be relevant to disease control or management.
2. Definition of the study population including sample size determination. Sample size should be
adequate for valid estimates.
3. Sampling, recruitment and management of sample in order to achieve a high response rate
which is important for valid estimates.
4. Examination, interviews and record retrieval methods.
5. Data handling and management.
6. Analysis, interpretation of the results and report writing.

Hypothesis formulation starts with descriptive characteristics of a disease

- person
- place
- time

Descriptive studies tell us about the distribution of disease and disease determinants in human
populations. In descriptive studies we commonly describe disease distribution by:

1. Person (who ?): The kind of persons afflicted by the disease; what are their characteristics?
e.g. age, sex, race, socio-economic status, genetic constitution, immunological status etc.
These are then related to those of the population to which such persons belong, i.e. census
data to determine if there is any predisposition to the disease.
 69

2. Place (where ?): The kind of places where diseased individuals are found. Important
characteristics of place are: geographical placement, altitude, latitude, climate, vegetation
etc. A classical example of the use of place in descriptive studies is found in the studies of
Burkitt lymphoma in Africa. The influence of place in determining disease occurrence can
also be studied by using migrant populations. Such studies are called migrant studies. e.g.
Occurrence of cancer by site for Japanese of different migratory status. Relative risks of
mortality from cancer of the stomach, liver and colon among Japanese men in Japan,
Japanese immigrants to California and sons of Japanese immigrants compared to while men
in California (aged 45-664)

Ca Japanese Japanese in America Sons of Japanese in

Japan America
Stomach 8.4 3.8 2.8
Liver 4.1 2.7 2.2
Colon 0.2 0.4 0.9

3. Time (when ?) : Does disease have any time trend? Many infectious disease occur during
certain periods of the year (seasonal distribution). At some instances there are clear-cut
explanations for the seasonality, especially when this can be related to the development of
breeding sites for disease vectors, and hence increased disease transmission. Other diseases
have a secular trend, i.e. changes in disease incidence occurring over periods of many years.
If disease trend can be predicted it may be possible to institute preventive measures. Other
diseases shows a cyclic trend, this is change in disease occurrence over a short period of
time usually two to three years. This is a characteristic of infectious diseases which confers
long lasting immunity, when it occurs all people who are susceptible develops the disease if
they get infected. The disease will die out after infecting all those who are susceptible and
will recur again when they are sufficient number of susceptibles. It is then possible to
predict disease outbreak, and also evaluate preventive measure like vaccination

4. Space-time clustering: disease incidence can very simultaneously with both place and
time. If disease can be described by both place and time, it will be possible to determine the
dynamics of spread in a population. Many infectious diseases occurring in one geographical
area tend to occur at around the same time. This phenomenon is called space-time
clustering. Such a phenomenon can occur by chance in non-infectious diseases. Cluster
analysis techniques are used to determine whether the degree of space-time clustering
observed for any disease is pertinent to infectious disease aetiology. Time-space clusters are
therefore a useful due to the aetiology of infectious diseases.

Interpretation of changes or differences in disease incidence.

Geographical as well as temporal changes or differences in disease incidence should be interpreted
with caution because such changes, apart from being the result of influences by aetiological factors,
may be due to the following influences:

- changes or differences in the level of arsertainment of cases (are all cases found ?)
- changes or differences in diagnostic criteria between two points in time or place.
- changes or differences in population structure between two points in time or place.
Secular trends can be due to one or more of the following:
- Changes in diagnostic technique.
- Changes in the accuracy of enumerating the population at risk of developing the disease.
- Changes in the age distribution of the population between two time periods.
- Changes in survival
- Changes in actual incidence due to
 70

(a) changes in environmental exposure

(b) changes in individual characteristics i.e. lifestyle,
(c) changes in individual susceptibility.

Hypothesis formulation

There are three methods of hypothesis formulation:

(a) Method of difference; if the disease frequency is significantly different between two sets of
circumstances the disease might a causal association with a particular factor that differs
between the two

(b) Method of agreement; if a single factor is common in a number of circumstances in which the
disease occur causal association can be suspected.

(c) Method of concomitant variation; if a factor varies in proportion to the frequency of disease
causal association can also be suspected.

References and further reading:

1. Foundations of Epidemiology Lilienfeld and Lilienfeld

Oxford University Press
2. Epidemiology: Principles and Methods MacMahon, B. and Pugh, T.F.
Little, Brown Co., Boston
3. A study guide to epidemiology and Biostatistics Morton, R.F. and Hebel J.R.
University Park Press Baltimore 1979 Person
4. Epidemiology in Medicine. Hennekens, CH. Buring, JE. eds. Mayrent, SL. 1987
5. Epidemiology in Medical Practice Baker, D.J.P. and Rose, G. Churchill
Livingistone,1979.
6. Smoking, Air pollution and bronchitis in Britain.
Lambert, P.M. and Reid, D.D.
The Lancet, 1,7652 : 853.
 71

Case-control studies
(Also called case-referent, case-comparison or case history studies)
A case control-study is an epidemiologic investigation which involves comparing the characteristics
of diseased persons (the cases) with those of non-diseased persons (the controls). The purpose of
this comparison is to identify factors which occur more (or less) frequently in the cases as compared
to the controls and hence provide clues regarding the role of such factors in elevating (or reducing)
the risk of the disease under investigation.

Advantages
- Efficient in time and cost.
- Efficient for the study of rare diseases.
- Efficient for the study of chronic diseases.
- Tends to require a smaller sample size as opposed to other designs.
- Allows exploration of a large number of exposures for the disease under investigation.
- May be completed more rapidly than alternative designs.
- Ethical problems are minimal.
- Subjects need not be volunteers.
- Attrition problems are minimal.
- Used to test the hypothesis that 'disease in the cases is or is not associated with exposure to
some factor' hence can provide suggestive evidence of a causal relationship that warrants
public health intervention to reduce exposure to the risk factor

Disadvantages:
- Both the exposure and disease have already occurred at the time the participants are
recruited into the study, hence the design is susceptible to bias arising from differential
selection of cases or controls into the study on the basis of their exposure status as well as
from differential reporting or recording of exposure information among study groups based
on their disease status
- Inefficient in ascertainment of rare exposures
- Rates can not be computed directly unless a modified design is adopted
- Temporal relationships usually difficult to ascertain

Retrospective and prospective case control designs:

Retrospective design:
All the cases have been diagnosed by the time the investigator initiates the study (prevalent cases
are used) Disadvantage: (1)Sampling bias (systematic differences between the study population
and the target population that pose a threat to external validity) Why ? Because patients who die or
recover rapidly, as well as those with mild symptoms whose illness goes undetected are not
represented in the study sample. As a result, only survivors are studied. These survivors are unlikely
to be representative of the total patient population. (2) Temporal relationship is difficult to establish.

Prospective design
The study is began and all new cases diagnosed within a specified period of time are included into
the study. This design tries to avoid studying survivors and hence addresses etiological factors as
close as possible to the commencement of the disease process.
 72

Design and conduct of case control studies:

CASES:
Cases should represent as far as possible a homogeneous disease entity as e.g. cervical cancer or
cancer of the body of the uterus and not uterine cancer (which includes both types), to ensure that
the cases selected represent a homogeneous disease entity :Establish a strict diagnostic criteria for
the disease. Depending on the certainty of the diagnosis and the amount of information available it
is often useful to perform the analysis and present the results separately for cases classified as
definite, probable or possible.

Selection of cases: Should be selected from a well defined population called: “source population”.
The following are possible sources of cases:

1. Hospital or health care facility:

Commonly involves identifying people with disease who have been treated at a particular health
facility during a specified period of time. Although such cases may be identified easily, the
underlying source population may not be well defined.

2. General Population:
All diseased individuals or a random sample from a defined population . Avoids bias from selection
factors that led the affected individual to utilise a certain health care facility.

CONTROLS:
Controls are necessary in order to allow the evaluation of whether the frequency of an exposure or a
specified characteristic observed in the case group is different from that which would have been
expected based on the experience of a series of comparable individuals who do not have the disease.
To select controls consider:

The source and characteristics of the cases.

Practical and economic factors.
The controls should be selected to represent the population of individuals who would have been
identified and included as cases had they also developed the disease.

Sources of controls:
1. Hospital/Health care facility
These may be patients admitted at the same hospital as the cases for conditions other than the
disease being studied

Advantages
- easily identified
- readily available and minimizes assembling costs
- minimizes potential for recall bias in both cases and controls
- provides for similar hospital selection factors for cases and controls (if it is not a referral
hospital for that particular disease-patients may be drawn from a wider area in comparison
to the controls)
- likely to co-operate than healthy individuals hence minimizing non-response
Disadvantages
They are ill by definition and differ from the healthy in a number of ways that may be
associated with the exposure. Thus the experience of these patients may not accurately
represent the exposure distribution in the population from which the cases are derived.
 73

2. General population

This is the source to be used when the cases have been selected to represent affected individuals in a
defined general population, and also when hospital based controls are not scientifically desirable or
feasible

Recruitment may be done by

- random sampling methods using the available sampling frame (administrative).
- selection of individuals from population registers, voting lists,
- census records, salary lists, etc.

Though controls from the general population may represent the non-ill individuals in the
community they are not short of problems
- usually costly and time consuming
- sampling frame may not be available
- difficult to get hold of busy people with a lot of scheduled activities
- may not recall exposures at the same level as the cases
- individuals who have not experienced an adverse health effect, are less motivated to
participate, hence non-response may be higher than in hospital based controls

3. Special groups:

These include friends, relatives, spouses and neighbours of the cases.

Advantages:
They are healthy like other members of the population.
Likely to be co-operative due to the interest they have in the health of the case.
May offer some degree of control of confounders such as ethnic background, socio-economic status,
environmental factors and the like.

Disadvantages:
Due to the closeness of the controls to the cases, distribution of the factor under study, may be
similar in the cases as in the controls, and hence an underestimate of the true effect of the exposure
may occur

How many controls per case:

One case to one control according to logistics and time. As the number of controls increases, the
power increases too. Usually one stops at a case:control ratio of 1:4 as there is little gain in
statistical power with each additional control beyond this point.

Ascertainment of exposure status

After the cases and controls have been defined in terms of characteristics and sources, the
information on exposure status must be obtained. This may be done by: interview, mailing
questionnaire, from surrogates i.e. spouses or mothers of children and from records. Procedures
must be similar for cases and controls. Interview settings should be the same, interviewer bias
should be minimized where possible -the interviewers should not know which is the case and which
is the control. They should not know the hypothesis under investigation.
 74

Measure of effect from a case control study

Results from a case control study can only provide an estimate of the relative risk, and this can be
done under the following assumptions;
- The controls are representative of the general population with respect to the frequency of
exposure.
- The assembled cases are representatives of all the cases of the disease
- Frequency of the disease in the population is small.
Under these assumptions the odds ratio can be used to estimate the relative risk and can be
calculated as the ratio of odds of exposure among the cases to that among the controls (for
unmatched designs).

Case Control
Exposed a b
Not exposed c d

Odds ratio=
a/c ad
----- = ------
b/d bc

In interpreting results from case control studies and indeed from any other epidemiological study,
the following should always be considered:
* Bias (minimized by adopting adequate design and implementation)
* Chance (evaluated by statistical methods)
* Confounding (may be minimized by adopting adequate design and adjusting for it in the
analysis).

When the above are taken care of, the ratio may be interpreted as follows: An odds ratio of 1
signifies lack association between exposure and disease. Odds ratios different from 1 indicate the
possibility of an association between exposure and disease. If the odds ratio is greater than 1
exposure to the factor will lead to an increased risk of disease and if the odds ratio is less than 1 it
shows a protective effect of the exposure under investigation.

Matching in case control studies

Matching is done in case control studies to control for confounding.
Variables known to be confounders are distributed in an identical manner in both the study groups,
however if matching is done in the design of the study, then matched analysis should (may) be
carried out during the analysis phase of the study.

Basic data from 1:1 matching may be represented in the following way:

CONTROLS
Exposed Not exposed
CASES
Exposed a b
Not exposed c d

b and c are discordant pairs while a and d are concordant pairs

 75

Odds ratio estimate = number of pairs with unexposed control and exposed case
number of pairs with exposed control and unexposed case
b
Odds ratio= ------
c
i.e. only the discordant pairs contribute to the association between factor and disease.
The interpretation of the odds ratio is the same as already mentioned earlier.

References
1. Lewis H Roht, Beatrice J Selwyn, Alfonso H Holguin JR
Bobbe L Christensen
Principles of Epidemiology: A self teaching Guide
Academic press Inc, Harcourt Brace Jovanovich, Publishers
San Diego, New York, London 1982 pg 217-230.
2. Charles H Hennekens and Julie E Buring
Epidemiology in Medicine
Little Brown and Company
Boston/Toronto 1987 pg 327-347
3. Rebecca G. Knapp, M Clinton Miller III
Clinical Epidemiology and Biostatistics
NMS National Medical Series
Harwal Publishing Company, Malvern, Pennsylvania
Baltimore, Hong Kong, London Sydney 1992. pg 31-50
 76

Cohort studies
Below is a list of hypothetical studies which illustrate the main characteristics of cohort studies:-

1. Alcohol and coronary heart diseases (CHD):

=> Identify study population and then interview to obtain information about alcohol taking. Classify
people into alcohol drinkers and non-alcohol drinkers. Follow them up for a period of time (say 5
years) and identify those who will develop CHD among both groups and compare the incidence
rates.

2. High risk behaviour and HIV infection:

=> Identify high-risk behaviour you might be interested to investigate (e.g. other STDs, multiple sex
partners, etc.). After identifying the study population, you classify people into either "high" risk
group or "low" risk group. Then follow them up for a specified time period. At the end of follow-
up, identify those who will have become HIV infected in the two groups and compare the
incidence rates.

3. Oral contraceptives (OC) and obesity:

=> To investigate the association between use of OC and obesity, investigators selected users of OC
and comparable non-users of OC and invited them to participate in the study. Consenting subjects
were weighed at enrollment and later yearly for 3 years. At the end of the study the proportion of
subjects who were detected to be obese in the two groups was compared.

- From these studies we can identify three key characteristics of cohort studies:-
• Start with "healthy" subjects (i.e. before they develop outcome of interest) who are free of the
disease of interest.
• Classify them on the basis of exposure (or risk factor).
• Follow subjects and assess the occurrence of outcome of interest among exposure groups
• Compare incidence of outcome in both groups

The term "cohort" was originally a Roman military term: a cohort was one tenth of a legion. The word
"cohort" has come to be used in epidemiology to designate individuals sharing similar experiences
which we observe in order to learn about the occurrence of disease. These studies involve assembling
healthy cohorts which are followed forward in time for development of disease. A cohort is the group
of persons who share a common experience within a defined time period. For example, a birth cohort
consists of all persons born within a given period of time. A marriage cohort would consist of all
persons married within a certain period of time.

Group or groups of individuals are defined on the basis of presence or absence of exposure to a
suspected risk factor for a disease. At the time exposure status is defined, all potential subjects must be
free from the disease under investigation, and eligible participants are then followed over a period of
time to assess the occurrence of that outcome.

Issues for consideration in assembling the cohort:

The group of individuals selected to form a cohort can come from a variety of sources. The choice of a
particular group will depend on the following scientific and feasibility considerations:
a) Frequency of the exposures under study.
- For relatively common exposures, (e.g. cigarette smoking or coffee drinking) a sufficiently large
number of exposed individuals could probably be identified from a number of possible populations.
 77

- For rare exposures, however, such as those related to particular occupation in specific locations, it is
more efficient to choose a group specifically because they have undergone some unusual exposure or
experience, the effects of which are to be evaluated. This allows to obtain sufficient numbers of
exposed group within the shortest possible time.

- The key issue is to have variability of exposure with sufficient numbers in various exposure groups.

b) Accessibility of cohort members for measurements and follow-up because of the need to obtain
complete and accurate exposure and follow-up information on all study participants. Most cohort
studies are therefore conducted in populations where loss to follow-up is likely to be minimal.
Incomplete follow-up affects validity of the study.

c) The cohort selected should assure sufficient number of outcomes.

d) An attempt should be made to make the results of the study generalizable. This requires assembling
of the cohort which is representative of the population where the study is being conducted.

Selection of comparison group:

Once the source of exposed subjects has been determined, the next consideration is the selection of an
appropriate comparison group of non-exposed individuals. The major principle underlying this decision
is that the groups being compared should be as similar as possible with respect to all other factors that
may be related to the disease except the exposure under investigation. This means we expect to observe
the same disease rates in the two exposure groups if there is no real association between exposure and
disease.

When several risk factors (or exposures) are being considered simultaneously, the non-exposed group
should be defined as those with none of the risk factors (or exposures) under evaluation. In many cohort
studies, it may be useful to have multiple comparison groups, especially when no single group appear
sufficiently similar to those who are exposed to provide assurance about the validity of the comparison.

Example: In evaluating potential adverse health outcomes associated with the use of oral
contraceptives (OC), choice of an appropriate comparison group is of critical importance.
Some studies have used for comparison a group of women not using OC, while others have
selected a group using some form of contraception other than OC. Women not using
contraception may differ substantially from users of any type of contraception in terms of
ability or desire to become pregnant or the nature of their sexual practices. On the other hand,
women using various forms of contraceptives are likely to differ from OC users with respect to
religion, socioeconomic status, and other life style factors. Thus, it may be that no one
comparison group is clearly superior, and information on the role of OC and disease can best be
elucidated by comparing the results from various comparison groups.

NOTE:
Consistent results from multiple comparison groups reinforces the believe that exposure under
observation is related to outcome or disease under investigation.
 78

Selection of subjects:

- Options:-
a) Random samples of populations
- not useful when exposure is rare
b) Special exposure groups (e.g. atomic bomb survivors)
- groups with high exposure prevalence
- useful for occupational or environmental exposures
c) Within special information groups (e.g. doctors, students)
- groups with high anticipated quality of information

Data analysis:

Basic analysis of data from a cohort study involves the calculation of rates of the incidence of a
specified outcome among the cohorts under investigation. The basic layout of data at the end of the
study is in the form of a 2x2 table as shown below:-

Diseased Not diseased TOTAL

Exposed a b a + b = Ne
Unexposed c d c + d = No
TOTAL a+c b+d a+b+c+d

- Cumulative Incidence of disease among exposed (Ie) = a/(a+b) = a/Ne

- Cumulative Incidence of disease among unexposed = Io = c/(c+d) = c/No

Measure of effect:

- Effect of exposure is measured by the ratio Ie/Io, known as Relative Risk (RR).

Relative risk is the measure of strength of the association between exposure and disease. It measures
the risk of a disease in exposed group when compared to the unexposed group. A relative risk of 1
shows no increased risk in the exposed group (i.e. no association between exposure and disease); while
a RR greater than 1 indicates increased risk among exposed group. A RR of less than 1 indicates a
decreased risk in the exposed group.

NOTE:
There is another measure of association which is commonly used. It is called Risk difference (RD) or
absolute difference which is the difference of the incidence rate in the exposed and that in the unexposed
group.

Suppose;
Ie = Incidence of disease in exposed
Io = Incidence of disease in unexposed

RD = Ie - Io

Risk difference estimates the amount of the disease which is attributable to a certain exposure
(sometimes called "attributable risk"). Therefore, it specifies the amount of disease which can be
reduced from the study population if certain exposure is removed.
 79

Advantages of cohort studies:

1. Can elucidate temporal relationship between exposure and disease. This is possible as the cohort is
classified in relation to exposure to the factor before the disease develops. Development of a disease
after exposure to a certain factor may indicate causal relationship.
2. Permits direct measurements of exposure specific incidence of the disease. Therefore, the absolute
difference in disease incidence rates between groups (attributable risk) and also the true relative risk
can be measured.
3. Allows for evaluation of multiple outcomes of the same exposure. For example, although
prospective studies of smokers and nonsmokers were originally designed to detect association of
smoking with lung cancer, they also showed that smoking is associated with the development of a
host of additional ailments such as emphysema, coronary heart disease, peptic ulcer, cancer of
esophagus, and cancer of urinary bladder.
4. Particularly useful when exposure is rare.
5. Less prone to selection bias for prospective cohort studies.

LIMITATIONS OF COHORT STUDIES:

1. Expensive and time consuming and therefore takes a long time before completion. This is true for
most prospective cohort studies.
2. Liable for attrition or loss to follow-up among subjects in the study. This can affect validity of the
results obtained.
3. Inefficient for rare diseases.
4. Not suitable for diseases with long latency period as these require long period of follow-up.

References:

1. Beaglehole R, Bonita R, and Kjellstrom. Basic Epidemiology. WHO, 1993.

2. Greenberg RS. Medical Epidemiology. Appleton & Lange 1993.
3. Mausner JS, and Krammer S. Epidemiology - An introductory text. W.B. Saunders Company
1984.
 80

Experimental studies:
Introduction:

An experiment is a deliberate application or withholding of a supposed cause, and observation for

subsequent appearances or lack of appearance of an effect. It is the final full proof of a cause
relationship.

How do experimental studies relate to observational studies:

Initially descriptive studies offer a tentative explanation of relationship between a suspected cause
and an effect. This relationship is further analysed in cheap and rapid analytical studies i.e. case
control studies. Based on the strength of the relationship established, decision is made to embark on
more analytical studies such as cohort studies. Should the relationship still be strong and convincing
decision is made to embark on an experiment. An experiment on human beings should only be
based on removal or addition of a factor which will eventually improve health of an individual or a
community. In the course of an experiment an investigator should not cause suffering to any one in
order to prove cause and effect relationship.

Types of experiments

Types of experimental studies:

- Classical experiments
- Natural experiments
- randomized control studies

Classical experimental design.

The experiment is usually carried out on healthy laboratory animals and not on humans because of
ethical issues. These studies are typically those involving laboratory animals to test drug toxicity
etc.

Natural experiments:
These are observations made by comparing two similar populations in which one of them has been
exposed to some disease factor as a result of some natural disaster such as famine, earthquakes and
wars. The circumstances of exposure and observation for effects have been likened to experiments.
Differences in physical environment i.e. geochemical characteristics leading to low or high levels of
iodine and development of goitre, thus providing a proof that low levels of iodine or no iodine will
lead to development of goitre. This condition will be rectified by improving the supply of iodine
thus proving the cause and effect relationship. NOTE: It is unethical to deprive people of iodine to
establish the relationship between iodine and goitre, hence nature alone has been able to provide the
clue.

Randomized control studies:

These are studies to determine the efficacy of a drug, vaccine (or any other methods of treatment or
prevention) in the treatment or prevention of a disease. These are the commonest forms of
experiments.
4
Randomized control study is a standard against which all other designs can be compared.
Randomized control studies are comparative with an intervention group and a control group. In
randomized control studies there are known or unknown factors related to disease progress or
4
Randomization is different from sampling, where the process of sampling ensures that each sampling unit
has equal chances of being sampled
 81

outcome. As such subjects should be assigned to intervention or control group by a way which
mostly control for those factors known or unknown so as to produce comparable groups other than
for the intervention. The comparability between the two groups is achieved by
RANDOMIZATION. This procedure ensures that all individual have equal probability to be
assigned to either control or intervention group.

Randomization process:
Randomization produces two or more study groups comparable with respect to known and unknown
risk factors. It also removes investigators bias in allocating of subjects to treatment groups, and also
guarantees that statistical tests will have valid significance levels.

Randomization methods:

Simple Randomization.
The simplest method is to toss unbiased coin each time a subject is eligible to be randomized. You should
decide before hand which of the sides i.e. head or tail will represent which trial group, and hence whichever
side appears the subject will be assigned the respective treatment group. This procedure is not feasible for
large studies where you will need to randomize a large number of subjects, in such cases the use of random
number tables is more convenient. When using random number tables you should use even and odd numbers
to decide which subjects will be assigned to which treatment group. However you should decide before hand
whether even or odd numbers should represent which group. When you are starting the randomization
process, record the page of the random number tables, the column, and the starting number, for validity of the
assignment and objectivity. The advantage of this simple method is that it is easy to carry out. The
disadvantage is that there could be a substantial imbalance between the two groups, the solution to such
problem is the use of Blocked Randomization.

Blocked randomization:
Blocked randomization (permuted block randomization) ensures that at no time during randomization there
will be imbalance, and at certain points in time the numbers in each group are equal, if they are unequal the
difference does not exceed the block length. When assigning subjects on equal probability to two treatment
groups, given blocks of even sizes (4, 6 or 8 etc.), one half will be assigned to control and the second half to
the other. The order in which the interventions are assigned in each block are randomized. One of the
disadvantages of blocked randomization is that if staff happen to know the blocking factor (block length) they
can predict the next subject and thus bias the allocation, one way to get around that, is to vary the blocking
factor(block length) randomly.

Stratified randomization:
The major reason for doing randomization is to achieve comparability between groups as much as
possible. However at base line there may be factors which are crucial in determining the prognosis
(outcome) of the disease process. In such situation stratifying the subjects according to those
prognostic factors is done, thus subjects with certain prognostic factors are pulled together and
randomized. This ensures that there is a balance of prognostic factors between the intervention
groups. These prognostic factors need to be measured during the baseline.

Advantages of randomization
-Removes potential bias in allocation of subjects
-Produces comparable groups
-Guarantees validity of statistical tests.
 82

Randomized control studies:

Type of randomized control studies:

• Therapeutic trials: These are trials done to determine the efficacy of a therapeutic agent, or a
procedure. There are four phases of therapeutic trials which are:

Phase I
This phase involves clinical trials which are usually performed on human volunteers. The objective
of such trials is to determine how much of the drug can be given without causing serious side
effects. Secondly it is done to understand drug metabolism, and availability. Thirdly it is to
determine appropriate dose schedules. The sample size is usually small involving 20-30
individuals.

Phase II
It requires close monitoring of each patient. The objective is to determine the effectiveness and
safety of a drug. A second objective is to select drugs with genuine effect from those without
beneficial effect or are seriously toxic. The sample size ranges from 100 to 200 patients on each
drug.

Phase III
After establishing that a drug is safe and effective it is then compared with a standard treatment for
the same condition. This is usually a large trial involving a substantial number of patients. It is a
very vigorous and extensive type of clinical investigation of a new drug. Many times when people
take of clinical trials they are referring to trials in this phase..

Phase IV
These are also referred to as postmarketing surveillance. This phase is conducted after the trials in
the previous phases having established the suitability of the drug for marketing. They are usually
large studies and of long term on morbidity and mortality. The major objective is monitoring for
adverse effects of the drug. It is also an exercise to promote the drug by introducing it to a large
number of clinicians.

• Intervention trials: Intervention is done before a disease develops in those with disease risk
factors to reduce the risk for disease.
• Prophylactic trials: Determine the efficacy of prophylactic agents such as vaccines to prevent
occurrence of disease.

Comparability in assessment
One of the requirements of experimental studies is that all groups must be assessed comparatively.
Such requirement can be achieved at two levels:

A. At the planning stage either by;

1. Relying on objective criteria e.g. length of survival, laboratory findings etc. or
2. In case of subjective assessment
(a) It should be done by someone who doesn't know to which intervention group the subjects belong.
(b) If the assessor is the patient he/she must not know to which group he/she belongs (pain intensity
etc.)
(c) If the assessor is a clinician he/she must not know which treatment has been given to whom.

B. When assessing disease progress

Study groups must not be assessed differently, one should take the following into consideration;

(a) Changes in behaviour, diet. auxiliary treatment etc.

(b) Clinician or other involved in the care may affect similar factors.
 83

(c) Autosuggestion, beliefs that one is receiving an effective treatment can influence the
outcome. Placebo can also demonstrate beneficial effect and is referred to as placebo effect
(d) Clinicians effect by way of heterosuggestion

To avoid all this and ensure comparability in assessment and disease progress employ Blinding.
5
Blinding
In any clinical trial, bias is one of the major concerns. Bias may be defined as a systematic error or
difference between the true value and that actually observed due to all causes other than sampling
variability. Its causes can be due to conscious factors, subconscious factors or both.

Bias can occur at a number of places in a clinical trial from the initial design, through data analysis
and interpretation. A general solution to the problem of bias is blinding. Blinding is to keep the
subject, observer, or data analyst or all unaware of the assigned intervention.

There are four types of blinding :

1. No blinding(unblinded) :
The investigator and the subject know to which intervention group one belongs. Some studies can
only be done that way, i.e. most surgical procedures, and changes in life styles.

Advantages
-It is simpler
-Easier to design
-Easier to carry out

Disadvantages:
There is much risk that allocation of subjects to control or intervention group or assessment will be
biased, secondly subjects may drop out if they know that they are receiving
placebo.

2. Single blinding:
Only the investigator knows to which group the subject is assigned thus the patient is blinded. This
is very common in surgical procedures i.e. chest skin incision as placebo versus a therapeutic
surgical procedure like coronary vascular surgery, or comparing two therapeutic procedures i.e.
4
Bilroth I anastomosis as compared to Bilroth II anastomosis etc. In such situation the surgeon
knows to which surgical procedure a subject has been assigned, but the subject is not aware of the
intervention.

Advantages
-simpler than double blind
-easier to design and carryout
-knowledge of the intervention may help investigator while caring for patients
-enables participation since some investigators are reluctant to participate in studies which they
don't know the study group assignment.

Disadvantage:
-It increases the likelihood that the investigator will be biased in assigning subjects to intervention
group, as well as when doing the assessment, and analysing the results.
-concomitant and compensatory therapies, could be imbalanced between the two groups

5
Bilroth I II are both surgical treatments for peptic ulcer
 84

3. Double blind
Neither the subject nor the investigator responsible for follow up knows to which group the subject
is assigned.
Assigning subjects to intervention and control group is done by a person other than the investigator.

Advantages are;
The risk of bias is much reduced.
The effect of preconceived ideas by the investigator is much reduced, so is the subject.

Disadvantages;
- Mainly ethical, a placebo- control study is justified if there is no standard treatment superior to
placebo, and that subjects should be informed that a placebo is in use and what their probability of
being on placebo is.
-some investigators are not willing to participate in trials which they are blinded.

4 Triple blind:
In a triple blind study the committee monitoring response variables is not told which group is
allocated to intervention or control. It gives the advantage that the committee will evaluate the
response variable more objectively. On the event that the monitoring committee has an ethical
responsibility to monitor subjects safety, such design may be unfavourable. Many will be unlikely
to participate if they don't know to which intervention each study group is assigned, since many will
like to intervene in case the subject is not benefiting or is getting worse.

-it is rare that one accepts beneficial outcome unless statistical significance has been reached, but
rarely an investigator will continue a study in order to achieve a significant difference in an adverse
direction.

Therapeutic study design:

1. Historical control design:

There is no random allocation of patients to treatment and control groups. Results are compared to
those of previous series of comparable subjects.

Strengths
• No subject is denied the new treatment
• same subjects are on new treatment, as a result size of study is cut by half.
• Historical data easily obtainable
• Rapid & inexpensive
• feasible when large benefits are expected i.e. for a disease which always ends fatally and you
expect survival.

Disadvantages:
• Many historical studies claim benefit but? consider such benefit could be attributed to:
(a) Changes in factors affecting survivor
(b) Changes in patient management
(c) Excluding seriously ill patients
(d) Migration of patients.
(e) Limitation of data collected historically.
Could explain the observed benefit.

2. Non randomized, concurrent control studies:

Controls are subjects treated with a different treatment at approximately the same time. i.e.
comparing outcome of patients treated at two institutions.
3. Parallel group design:
 85

The comparison groups run parallel and members of each group are given treatment at
the same time under similar conditions.

The subjects are initially recruited and then randomly allocated to control or intervention group. For both
groups initial evaluation or measurement is done. Intervention group then receives the intervention, while the
control group receives a placebo or another treatment. The strength of this design is that it ensures validity.

4. Sequential design:
Patients enter the study one pair after the other. Treatment A and B are allocated to either individual
of the pair and results of treatment are analysed as, and when they become available. When a pair
gives similar treatment outcome the pair is ignored in the analysis. Not suitable for chronic illness.

5. Cross-over design:
This design is started as in the parallel group design. Treatment is exchanged between the groups in the
course of study. Suitable for chronic diseases and is unsuitable if there is carryover effect.. There should also
be no order effect. When there is no carryover effect it has the advantage of requiring small sample size. If
you are not certainly sure that there is no carryover effect avoid the use of cross-over design.

6. Factorial design:
Several trials are combined into one, in order to study the effect of several treatments for the same
disease. Assuming that there is no interaction between the drugs tested. It has the advantage that
two or more trials can be combined in one with a small increase in sample size.

Treatment A Control C1
Treatment B a b
Control 2 c d

Those in cell (a) are receiving both treatments (A and B), cell (b) are receiving treatment B and
control, cell (c) are receiving treatment A and a control, while those in (d) are receiving control
only.

7. Withdrawal design;
A subject doing well on an intervention the treatment is withheld or the dosage reduced to find the
outcome. This is done to asses the response to discontinuation or dose reduction. The benefit will be
to evaluate the duration of an effect already known to be effective. This is only feasible for chronic
illness, and investigators should seriously consider ethics before such a study is done.

8. Hybrid design.
Instead of randomizing equally the number of subjects to be assigned to the control and intervention groups,
fewer are assigned to placebo or the alternative treatment, while more subjects are randomized to the new
intervention. This becomes feasible only when information on controls is fairly recent. It should be done by
the same individual or clinic using the same entry criteria and evaluation factors.

Advantage
- It requires smaller sample size
Disadvantage
Bias from the use of historical controls

Analysis of data from therapeutic trials:

- Percentage success or failure (placebo Vs treatment)
- X2 test
- Survival analysis methods
 86

Intervention Trials:

Design of intervention trials.

Such trials are useful in confirming associations found to exist between a factor and a disease in
analytical studies. The investigator intervenes before disease has developed in individuals with
disease risk factors.

Individuals with disease risk factors are randomized into two groups and whilst an attempt is made
to remove or reduce the risk factor in one of the groups, the other group is not intervened.
Observations are then made on both groups, to determine the effect of such intervention. Such trials
are limited to circumstances where ethical issues are unlikely to arise. e.g. To confirm the causal
association between smoking and lung cancer, one could select a group of smokers and randomize
them in two groups. The first group continues to smoke and an attempt is made to discourage the
second from smoking. If lung cancer mortality in the second group is lower than in the first, the
causal implication is obvious.

Prophylactic Trials

Design of prophylactic trials.

In prophylactic trials, there are four types of subject allocation, depending on the requirements of
the particular trial.

1. Individual allocation: Individuals are randomly allocated to the vaccination and control groups.
2. Group allocation (Community trials): communities such as villages, towns, schools or
factories are allocated wholly either to the vaccination or to the control group. Group
allocation is the method of choice when testing some live vaccines, in which persons in the
control group might inadvertently become vaccinated by contact with members of the
vaccinated group, if individual allocation was used.
3. No allocation: This method is used when there are ethical objections to leaving certain
individuals or groups of individuals unvaccinated. In such a situation, before - and-after
vaccination comparisons are made.
4. Matched allocation: Careful matching of the comparison individuals is done for all possible
factors except for the vaccination status. This method is commonly used in laboratory animal
experiments.

Assessment of effectiveness (protection)

Two methods are available.

1. Use of disease incidence
2. Use of immunological tests.

Disease incidence: by comparing the incidence of disease among the vaccinated with that among the
unvaccinated:

(C - V)
% Protection = ------- * 100
C

C = Incidence of disease among the control group

V = incidence of disease among the vaccinated group
Immunological tests:
• Measurement of antibody titres, e.g. measles, polio
• Measurement of antitoxin, e.g. tetanus, diphtheria
• Local reactions to vaccination, e.g. smallpox
• Skin tests, e.g. tuberculosis, diphtheria.
 87

In immunological tests cut-off points are established and individuals are categorized as immune or
non-immune to the disease. It should be understood that these cut off points are often arbitrary and
may not be synonymous to protection against the disease. Protection is therefore determined by
comparing the proportion non-immune to the disease among the vaccinated with that among the
unvaccinated.

Important Issues To Be Considered While Planning An Intervention Study:

A. Which is the primary question and secondary questions if any

• Each clinical trial must have primary question clearly
• selected
• defined and
• stated in advance

Primary question, is the one the investigator is most interested to answer.

It is also upon which sample size is based.
Secondary questions- should be limited in number and related to the primary question.

B. Define the response variables:

• response variables are outcomes measured during trial
• single response variable is chosen to answer the primary question
• State the question and response variable in advance
• i.e. In defined subjects is drug A at daily dose x more efficacious in reducing z over a period of
time than B at daily dose of y.
• Primary response variable must be capable of being assessed in all subjects.
• Occurrence of primary response variable is the end
• Should be capable of unbiased assessment.
• Should be able to be ascertained as completely as possible.
• unless there is combination primary response variable in which the subject remains at risk of an
additional event the occurrence of the primary response is the end.

C. Sample size
• Before calculation of sample size one should initially identify the primary response variable
which will be used to judge effectiveness. Primary response variable are of two types
a)qualitative (dichotomous) response variables and b) quantitative (continuous) response
variable

Method for qualitative outcome:

An outcome is said to be qualitative when one can classify the outcome into success or failure e.g.
cured and those not cured, death in a year versus those who survived. In this situation you have
been able to stratify the population into those who have experienced success and those
experiencing/experienced failure. When we are comparing two intervention populations or more we
will consider the proportion of success for each intervention group. If an intervention is more
effective than a placebo or a standard intervention it will be judged by the size of the difference
between the proportion who are successful amongst the intervention and control group is, as well as
what is the probability that such a difference between the two proportions could be observed by
chance alone.
 88

The formula is; n = [P1 x (100 - P1) + P2 x (100 - P2)] x f(aB)/(P2 - P2) 2

P1 = % of success expected on one treatment

P2 = % success on the other treatment (different from P1)
a = Type I error (risk of false positive result).
B = Type II error (risk of a false negative result)
n = sample size

Method for quantitative outcome.

An outcome is said to be quantitative if it is measurable or else it is provided by variables which
generate measurement data, e.g. serum cholesterol, weight etc., interventions in this case will
compare difference between two or more means.

N= 2σ2 x ƒ(&, B)
(M2-M1)2

D. Before instituting a clinical trial think about;

1. Is the trial feasible?
2. Do you have the necessary tools and knowledge
3. What is the knowledge about the safety of the intervention
4. What are the outcomes to be assessed
5. Do you have the technique to do so
6. Trials are costly and should be done when there is enough evidence of success
7. Timing: Are other research findings likely to make the intended intervention outdated in a
short time.

Planning a clinical trial:

• Rationale and background of the study
• Specific objectives of the study
• Concise statement of the study design including:-

a) Sample size
b) Blinding procedure
c) Randomization procedure
d) Types, dose, and duration of treatment.

• Criteria for excluding or including a patient (case definition)

• Outline of intervention procedure
• Definition of all clinical and laboratory methods to be used to asses progress and outcome.
• Define method of ensuring integrity of the data
• Define and choose major and minor end points
• Provision for observing and recording side effects
• Procedure for handling problem cases
• Procedure for obtaining informed consent of subjects
• Procedure for analysing the results.

Ethical considerations.
Every proposed trial must be exhaustively weighed in the ethical balance - each according to its
own circumstances and its own problems. On every occasion certain general questions must be
answered within the specific and particular circumstances of one trial he wishes to undertake.
 89

Such questions are:

1. Is the proposed treatment safe and therefore unlikely to bring harm to the patient?
2. For the sake of a controlled trials can a treatment ethically be withheld from any patient in the
doctors care?
3. What patients may be brought into a controlled trial and allocated randomly to different
treatments.
4. Is it necessary to obtain the patients consent to a controlled trial?
5. Is it ethical to use a placebo or dummy treatment?
6. Is it proper for a trial to be "double blind"?

References and further reading.

1. Foundations of Epidemiology Lilienfeld, A.M. and Lilienfeld, D.E.
2. A short textbook of medical statistic Sir Bradford Hill.
3. Fundamentals of clinical trials, second ed.
Lawrence M. Friedman, Curtm L.DeMets. 1985.
4. Clinical Trials, A practical approach. Stuart J. Pocock
 90

3
RESEARCH METHODOLOGY
 91

Investigation and control of epidemics

Introduction

The magnitude of a particular disease present in a specific population may remain stable for long
periods of time or it may alternatively rise and fall due to fluctuations in the number of susceptible
individuals and the nature and extent of their exposure to disease agents. Diseases
continually/habitually present in populations are considered to be endemic to that population.
Endemicity denotes the habitual presence of a disease in a community. An epidemic is defined as
the occurrence of a disease in a community clearly in excess of the expected incidence
characteristically revealing itself in a relatively short period of time. In order to say that there is an
epidemic, it is necessary to know the level of endemicity of the disease. Does a single case
constitute an epidemic ? The number of cases that constitute an epidemic will vary with the type of
disease, the population, and its previous experience with that disease. An epidemic cannot be
determined strictly in absolute terms. Rather its a relative concept. In the USA a disease such as
cholera is not normally present in the population and many persons would be susceptible if exposed
to the cholera organism. If the disease reappeared, the source of the outbreak would be investigated
in order to prevent rapid and serious spread of the disease. Therefore, even one case of cholera
would constitute an epidemic in the USA. But in a country like India where cholera is always
present in some population subgroups, a few hundred cases a year may be the usual or expected
incidence. For cholera to be considered an epidemic in India several hundred or thousands of cases
(i.e. cases above the endemic level or frequency) would have to occur. Diseases like Ebola (Ref:
Zaire May 1995) do not occur habitually in human populations: a single case will also constitute an
epidemic in any part of the world.

Modern epidemiology arose out of the study of the so-called classical epidemics, such as plague,
smallpox, cholera, typhus, typhoid fever and dietary deficiencies. Some of these epidemics remain
an important threat to many tropical countries. Other frequently encountered epidemics are:

Poliomyelitis Streptococcal infections

Measles Meningococcal meningitis
Mumps Hepatitis A
Rubella Food poisoning etc.
Hepatitis A

Although infectious diseases are the most important cause of epidemics in Tanzania and other
tropical countries, road accidents, drug addiction, poisoning, suicide, coronary heart disease and
lung cancer are important examples of epidemics elsewhere.

Disease outbreaks such as those of organic mercury poisoning with resultant deaths and
neurological disability have been reported from Iraq, Pakistan and Guatemala as a result of
ingestion of wheat products treated with methyl- and ethyl-mercuric compounds. The wheat was
intended only for use as seed and was so treated to prevent fungus growth.

Other disease outbreaks involving the nervous system (Konzo) have been reported from
Mozambique and Tanzania and were later found to be associated with the consumption of certain
types of cassava with a high content of cyanide.
 92

Epidemics of unrecognized diseases.

(New) diseases such as Lassa fever, Legionnaire's or Veteran's disease continue to pose problems
from time to time in certain parts of the world, and now AIDS is a recognized world-wide problem.

Lassa fever: a viral disease transmitted from rodents and was first recognized in 1969. Three nurses
contracted it in Nigeria and two of them died.

Ebola, which is a viral disease was first recognized in Southern Sudan and Zaire in 1976, and
subsequently in Southern Sudan again in 1979. An Ebola virus Haemorrhagic Fever outbreak has
also been reported in Zaire in May 1995

Legionnaires disease: outbreak of pneumonia at a convention of the American Legionnaires in

Philadelphia in 1976. There were 29 deaths. A gram-negative bacillus was identified as the
causative agent (Legionella pneumophila).

AIDS: An immune deficiency disorder (Acquired Immuno deficiency Syndrome) brought about by
infection with the Human Immunodeficiency Virus (HIV). It was described for the first time in
1981 among homosexual men and intravenous drug abusers in USA. Later the epidemic was found
to spread among heterosexual populations in Africa and elsewhere. The disease has a long
incubation period (5-10 years) and is transmitted through sexual contact, blood transfusion and
unsterile skin piercing instruments including injections. Vertical transmission is also an important
route.

Principles of investigation:
In a clinical case, investigation and treatment must go side by side for the successful management of
the patient. Likewise an epidemic is always an emergency where action to counteract it must begin
even before complete investigation. For example, in 1854 Dr. John Snow showed that the outbreak
of cholera around the Broad Street in London resulted from contamination of drinking water with
excrement from cholera sufferers. The epidemic was well managed and put under control 30 years
before the identification and description of the cholera vibrio, the causative organism.

Types of epidemics.
1. Common Source. Occurs when a group of people are exposed to the same causative agent. If the
period of exposure to the agent is brief and essentially simultaneous for all persons contracting the
disease, the epidemic is called a point source epidemic. All persons are affected by the same source
and person -to-person transmission does not occur. Common source epidemics are not necessarily
caused by infectious agents, they may also result from common exposure to noxious agents in the
environment. The Bopal disaster involving gaseous explosions, the Chenobyl outbreak involving
radiation leakage, children swimming in a chemically polluted river or factory workers exposed to
extreme heat or volatile chemicals are some of the examples which do not involve infectious agents.

2. Propagated (progressive) epidemic. Propagated epidemics result from transmission of an

infectious agent from an infected host to a susceptible one. The transmission can either be direct
(e.g. infectious hepatitis or measles) or indirect through a vector, as in malaria and yellow fever.
Transmission of the infecting organism continues until the number of susceptibles is depleted or
susceptible individuals are no longer exposed to infected persons or intermediary vectors. There are
three important aspects of person-to-person transmission of disease, and they include: generation
time, herd immunity and secondary attack rate.

Generation time: Time interval between receipt of infection and maximal infectivity for both
clinical and subclinical infection.

Herd Immunity: The decreased probability of a group of people or community to develop an

epidemic upon introduction of an infectious agent, the decreased probability is due to the
 93

presence of a high proportion of immunes although there may be a certain number of persons
who are individually susceptible to the agent.

Secondary attack rate: Proportion of contacts who get a communicable disease as a consequence
of contact with the index case within the accepted incubation period.

Number of new cases in a group minus the initial (index) cases

Secondary attack rate: ----------------------------------------------
Number of susceptible persons in the group minus initial cases

In general the two types of epidemics (common source vs propagated) can be differentiated by
plotting the distribution of cases by time of onset of symptoms. This is the epidemic curve - the rate
at which new infections occur.

Common Source Epidemics contrasted from Propagated Epidemics

The curve of onsets for a common source epidemic shows a rapid rise and fall within one incubation
period, whereas new cases in a propagated epidemic continue to develop beyond one incubation
period. If you look at the epidemic curve you can see that those affected have different times of
onset of symptoms. This is because of individual differences in the level of immunity and the taking
of different doses of the infective agent.

In the curve of a propagated epidemic, usually a gradual rise to a peak may be observed, followed
by a gradual fall in the number of new cases. This is because, as the number of cases increases, the
number of susceptibles falls below a critical level so that the number of new cases begins to fall.
The shape of the epidemic curve in this type of epidemic reflects several factors including the
population size and composition, the proportion of susceptibles in the population, the number of
cases at the start of the epidemic, the contact rate between the infected persons and the susceptible
individuals, the infectivity or pathogenicity of the disease agent and the incubation period of the
disease.

Sometimes it may be difficult to identify the nature of an epidemic from the shape of the epidemic
curve alone. The typical common source epidemic curves may be affected by the continued
development of cases through persistent contamination of the source, or exposure occurring
repeatedly or by a long and variable incubation period. The shape of the curve may also vary
depending on the size of the population exposed, the type of source distribution and extent of use or
the extent of contact with the susceptible population. The typical shape of a point source epidemic
may be modified by presence of more than one disease agent, each with a different incubation
period, or if secondary cases (person to person transmission) follows exposure to the original point
source.

Conversely, a propagated epidemic can create a rapidly rising and rapidly falling epidemic curve
similar to that of a common source epidemic. This is especially so when the disease has a short
incubation period and is highly infectious. (e.g. cholera).
 94

Outline of the steps taken during an investigation:

A successful investigation of an epidemic requires painstaking accumulation of information in the

field, and careful analysis of the data as well as the ability to make relevant observations. There is
no "rule of thumb" as to the order in which the steps should fall. Every epidemic will be assessed
according to its particular circumstances. The following are crucial steps in the investigation of an
epidemic.

A. Verification of the diagnosis:

Investigation of an epidemic starts with a study of the affected as well as the unaffected persons. For
this reason it is necessary to establish criteria for labelling persons as “cases” or “non-cases”.

Next clinical and laboratory studies should be done to confirm the diagnosis. Always consider
whether the initial reports are correct, otherwise you may start a large scale investigation of a "false
epidemic". You should always consider new, enthusiastic or inexperienced staff and the enthusiastic
use of new equipment. Also consider the possibility of using faulty reagents in the laboratory which
can produce false positive results.

B. Establishment of the existence of an epidemic:

Existence of an epidemic is established by comparing the current disease incidence with past levels
of the disease. If this is clearly in excess of the expected incidence, then there is an epidemic.
Sometimes it may be difficult to determine whether an epidemic exists due to the following reasons:
(a) Recent and marked fluctuations in the number of cases and populations although the
incidence rate may remain constant
(b) Gross exaggeration in the number of cases due to misdiagnosis, duplicate reporting by
hospitals, health centres, etc.
(c) Normal seasonal variation in the occurrence of a disease may give an impression of an
epidemic when few or no cases occurred until recently. Comparison of incidence rates
during the same season in previous years will reveal the expected frequency of disease and
clarify whether the observed frequency is unusual.
(d) Completeness of recording
(e) Manufactured epidemic - special surveys and efforts to find cases and innaparent infections

C. Description of the epidemic:

The description should answer the questions:

- who is affected? (in terms of age, sex, occupation etc.)

- where does that individual live? (in terms of geographic location)
- when was that individual affected? (time of onset of symptoms)

The cases are then plotted by time of onset of symptoms to determine the epidemic curve; and by
location to determine their geographical distribution (spot map). One may also draw/plot occurrence
of cases in relation to presumed antecedent exposure, e.g. place where lunch was taken, or where
they went for recreation etc.
 95

D. Formulation and testing of hypothesis:

(i) From the description of the epidemic; i.e. the epidemic curves, spot maps and the personal
characteristics, identify the type of epidemic - common source against propagated
epidemics.

(ii) Consider possible source or sources from which disease may have been contracted, e.g.
contaminated food or water supplies, breeding sites of insect vectors, etc.

(iii) Design an epidemiological study to test the hypothesis that the source is/or is not a
particular factor. Usually a case control study or a retrospective cohort study design is used
for this type of investigation depending on the nature of the epidemic.

(a) Use a case-control study design when only some of the population members of the
affected community can be identified and interviewed. e.g. large communities such as
villages, districts etc.

(b) Use a retrospective cohort study design when all the population members of the
affected community can be identified and interviewed, e.g. small and closed
communities such as schools, hostels, army barracks etc.

(iv) Compare cases with controls with regard to exposure to the postulated source by using a
case-control study design, or compare exposed populations with non-exposed populations
with regard to attack rates by using a retrospective cohort study design.

(v) Carry out statistical tests to determine the most probable source (Odds ratio or relative risk
and their 95% confidence intervals)

(vi) When appropriate confirm the epidemiologic findings by laboratory tests: i.e. on samples of
blood, faeces etc. or samples of suspect food, water etc.

E. Further investigations and analysis:

Up until this point in time the epidemic has been studied passively by the use of reports and
incoming information. For proper control of the epidemic the next step should include active
investigation of the cases in more detail and of the remainder of the population.

(i) Case finding - search for unreported cases

(ii) Detection of subclinical cases.
(iii) Detection of carriers or reservoirs of infection, if any

These three groups can be important sources of continued spread of the disease agent to new
susceptible hosts. Further investigation may utilize the following methods, note that the methods are
not exhaustive:

- stool culture and examination

- blood smears for parasites
- serum for antibody titres
- food and water for toxic substances
- surveillance for infection in wild and domestic animals
 96

F. Control of the Epidemic

This is done right from the time the existence of the epidemic is established.
- The principles of control are:
- controlling the source of the agent
- interrupting the methods of transmission
- enhancing the defence mechanisms of the host

G. Report writing for the investigation

- scientific report for publication if possible

- preliminary report for the ministry of health for action
- simplified report for community health education

References and further reading:

1. Mausner and Bahn; Epidemiology, An Introductory Text Mausner and Kramer

WB Saunders Company, Philadelphia, London etc. 1985
2. Benenson A.S. Ed Control of communicable Diseases in man. 11th Edition. Amer. Public
Health Assn. New York 1970.
3. Killewo J.Z.J., Do Amsi D, and Mhalu FS. Investigation of a cholera epidemic in Butiama
village in Tanzania. Journal of Diarrhoeal Disease Research, 1989; (1&2): 13-17.
4. Practical Epidemiology Barker G P Churchill Livingstone London ELBS Edition
5. Epidemiology and the community control of disease in warm climate countries. Edited by
Derek Robinson Second Edition Churchill Livingstone, Edinburgh, London, Melbourne
New York 1985.
6. Lewis H Roht, Beatrice J Selwyn, Alfonso H, Holguin JR, Bobbe L Christensen Principles
of Epidemiology: A self teaching Guide Academic press Inc, Harcourt Brace Jovanovich,
Publishers San Diego, New York, London 1982 pg 217-230.
 97

The research proposal and the research process

Introduction:

There are as many research proposal formats as there are donors who support research activities,
and the format and information required usually varies considerably. Many of the larger donors
insist on strict adherence to given formats, which must be filled in. Some donors distribute
guidelines which the applicant follows in preparing the proposal. Yet some donors, especially the
smaller ones offer little or no guidelines.

Despite the considerable variation in formats, there are however, key sections which are
encountered in most research proposal formats. Even where there are no strict formats or
guidelines, it is advantageous to be familiar with the key elements of research proposal
development, so that essential information is not excluded and, the proposal is presented in a logical
sequence and is comprehensible to the reviewers who may often represent the views of the donor.

The Table below (from WHO/EDUC/87.187) presents modules leading to a comprehensive project
proposal. Where a donor agency requires less detail it usually suffices to only include the problem
statement, goals and objectives, methods or strategy and the resources required.

Modules For Project Proposal Writing

1. Summary What is the proposal all about?

2. Title What is the topic
3. Problem Statement Where are we now in terms of Background, Needs Analysis & Literature
what is known? search
4. Goals & Objectives Where do we want to go? Goals & Objectives
5. Methodology Which route shall we take? Components ,assumptions, methods outputs,
feasibility
6. Implementation Plan How shall we Travel there? Resource inputs workplan organization
7. Monitoring And Evaluation How shall we know when we Monitoring & evaluation
arrive?
8. Budget and justification How much will it cost? Capital cost, recurrent cost
9. References What is the literature source? From what publications?
10. Profile Of The Applicant And who are we? Organization, record of achievements
11. Appendices What additional information Questionnaires, permits, ethical clearance,
should be appended? CV’s etc.

*Adopted from WHO/EDUC/87.187

THE FORMAT

1. Summary

The summary of a research proposal, which in most formats appears at the very beginning, is
perhaps the most critical section of any proposal. It must be short (often a page or less), precise, yet
interesting and must state why the research project should be carried out and not what will be done.
The summary aims at convincing the donor agency that the proposal is relevant to his concerns and
yet meets the concerns, needs, and interest of the investigator, the institution and even the country.
Although short, the summary must also clearly state the problem, the objectives, strategy or
methods, and the total budget.

The summary among other things, guides executives receiving the proposals whether it is relevant to their
funding priorities, where to refer it to for detailed evaluation, etc. A busy executive may reject a proposal
 98

merely on the basis of the summary. Later at committee level, some members may only read the summary.
Considerable effort must therefore be invested in preparing the summary, which although it usually comes
first, it must be written last.

2. Title: What Is The Topic?

The proposal must have an interesting but relevant heading. It must be clear and short and must
refer to the key variables of the proposal.

3. Problem Statement: Where Are We Now?

This section covers the general background as well as an analysis of the present situation which
may for example need to be changed. It would allow the investigator to see the importance of the
problem, its magnitude, and relevance, to describe it systematically, and to discuss why it should be
researched. A brief description of earlier work on the problem should be included; so a literature
search is essential. This section too must be concise as it describes what is known so far.
Subsequent sections must be traceable to this section.

4. Goals And Objectives: Where Do We Want To Go?

There is sometimes confusion of terms used in defining goals, objectives, purpose, aims, etc.

One author says, "Goals are broad statements of what is ultimately to be accomplished. Objectives
are more specific aims which the project is to achieve" another author says, "Research objectives
are the goals to be achieved by a research project. The general objective or purpose of the research
identifies in general terms what is to be accomplished by the research project and why". "The
specific research objectives identify, in greater detail, the specific aims of the research project, often
breaking down what is to be accomplished into smaller clear, logical and measurable components."
Therefore, research objectives are stated as questions, or positive statements, or hypotheses

5. Methodology: Which Route Shall We Take?

Methodology is the project strategy and refers to the design to be followed in order to achieve the
set objectives and eventually the overall goal of the project. A strategy may be broken down into its
components, each of which uses certain methods. In laboratory studies there are such aspects as
experimental design, statistical design and analysis, laboratory methods, resources to be used
including personnel required ( and whether they require retraining or not).

In clinical studies attention is paid to detailed descriptions of study protocols, criteria for selection
of study subjects, description of investigations to be made, patient care, patient follow up and a
description of the clinical facilities available. Clinical trials are even more demanding in that the
protocols are much more detailed - they include for example a description of the type of protocol
(e.g. controlled, single, double or even triple - blind, etc.), the number of treatment groups, the
proposed trial agent or drug, the criteria for selection, admission procedures, criteria for
discontinuation, follow up, etc.

Field Studies naturally must describe the study area (including sites, geography, climate, seasons,
etc.) and the human population, its activities and socio-cultural background. In such studies, one
must also describe the epidemiological method to be used (descriptive, analytical or experimental),
relevant epidemiological information available, and whether control measures (of the diseases to be
studied) have been undertaken and their impact if any.
 99

Whatever the type of study, the following areas must be well documented in the methodology
section of the research proposal..
• Study design
• Study population
• Sample size
• Research instruments
• Definition of terms (i.e. what is a case and what is a control?)
• Data collection and management
• Ethical issues
• Perceived constraints, biases and limitations in achieving the set objectives

6. Implementation Plan: How Shall We Travel There?

This section describes what is actually going to be done. Here are specified all the required inputs,
e.g. manpower, facilities, equipment, supplies and operating costs. Usually all required inputs are
specified, then after pointing out what is already available, you specify the inputs being requested
from the donor.

Then a workplan which outlines the activities, targets and schedules is presented. Workplans,
especially in large projects, may be subdivided into their component activities, and a timetable or
schedule is drawn showing when each activity will be carried out. Gantt charts are often used to
show when particular activities are to be done. You may for example show the periods for: pre-
testing of questionnaires, carrying out pilot studies, data collection, data analysis, etc. Remember to
consider allocating time for staff recruitment, staff training, acquisition of supplies and equipment.

Make sure you also include a description of the project organization (e.g. project organization
chart), explaining lines of linkage to other units or external institutions. Bio-data or CV of principal
investigators and their co-investigators is essential to increase the donors’ confidence in the strength
of the team.

7. Monitoring And Evaluation: How Shall We Know When We Arrive?

A well conceived research project must include an on-going process for gathering information
concerning its implementation and evaluation. Monitoring for example, keeps track of resources,
the quality and quantity of the activities undertaken, so as to continuously assess the status of the
project, identify problems and take remedial measures as is necessary, and to provide a feed back to
all those concerned.

Monitoring is undertaken by all concerned, especially the principal investigator, the members of the
research team, representatives of collaborating units or institutions, and a project officer from the
donor agency. Monitoring is often based on reports, field visits and regular meetings of the project
team. Workplans are invaluable in project monitoring and should always be prepared and utilized.

Evaluation on the other hand involves a systematic collection and analysis of data to assess whether
the objectives (i.e. goals) of the research project have been met. Data gathered in monitoring may
be used in final evaluation. Evaluation often focuses on the methodology used (was it appropriate,
valid, reliable), the design, data analysis, implementation, dissemination of research results, use of
findings (by government, communities, health workers, research team itself, etc.) financing (was the
investment worth it?), and follow up (has it led to further research activities, implementation etc.).
Evaluations are often scheduled in the middle and towards the end of the project period, and may
involve donor representatives.
 100

8. Budget and justification: How Much Will It Cost?

In this section the anticipated costs of planned inputs and activities are systematically enumerated.
All budget items in the proposal must be justified relating them directly to the activities to be
carried out. Unjustified budget items tend to be rejected by funding agents. Justification may
include statements to support the need for each budget item to accomplish the project activities. For
example, if you will need a four-wheel drive vehicle, state the reasons. First consider if a two-wheel
drive vehicle or a motorcycle will suffice before you go for costly means of transport. Could hiring
a car be a cheaper alternative? If you feel a car is necessary, who will drive it and why may it be
necessary to employ a driver?

Factors determining the type of budget include:

• donor agency (or your institution's) requirements;
• the type of budget items needed for the particular project;
• workplace activities (see section 5 above) - make sure each request is directly related to the
workplan

Some donors stipulate listing of other contributions, especially those by your government or
institute to ensure commitment. Examples here will include personnel already employed in
government service who may not need salaries to do the study. Other recipient government
contributions may include office space and utility bills such as electricity, water and other well
established services e.g. accounts and general administration.

Common entries in budgets may for example include:

(i). Personnel (salaries - not very popular with most donors/governments)
-professionals (researchers)
-technicians (laboratory, data analysis, etc.
-interviewers
-secretarial and administrative costs
-drivers
-others

(ii). Consultants
-fees and honoraria may be required for consultants.

(iii). Allowances
-Subsistence allowances, hardships, housing, mileage, etc. Will be required by project staff.

(iv). Training for project personnel

Training specific to project is usually necessary- workshops for field workers or research
assistants. Institutional research capacity strengthening involving technical and degree
programmes (masters, PhD. Etc.) must be included in large projects. Lack of such
forethoughts tend to frustrate project personnel.

(v). Facilities (purchase or rental)

• for office work
• for laboratory work
• for training
• for lodging of researchers, interviewers, etc.
• costs for renovation or construction where allowed.
 101

(vi). Transportation
• for consultants
• for researchers and their assistants, etc.
• for project managers and administrators
• for participants to seminars on the research project
• fuel, maintenance and repair of project vehicles.

(vii). Supplies and Equipment

• laboratory equipment
• computers, calculators and the relevant software
• typewriters, photocopiers and other office equipment
• supplies - e.g. drugs, chemicals and reagents
• audio-visual equipment
• survey forms
• stationery

(viii). Dissemination of research results

• cost of publication of reports
• postage and other forms of communication e.g. faxes
• seminars (facilities, transportation, per diems, consultant fees, preparation and
production of documents, secretarial assistance, rental of audio-visual equipment,
nutritional breaks, receptions etc.)
• newsletters or circulars
• mass media
• posters

(ix). Institutional Overheads

This must be included in any large project to cover the institution's expenses in administering
the grant. Depending on what the donor can offer, this may range between 5% and 20%.
However, some donors do not recommend this item if they are not asked.

(x). Contingency funds

If allowed by the donor agency this item could for example include funds for:

• unexpected expenses and events

• treatment for study subjects (service component)
• emergency treatment of project staff

Depending on the type of project and donor, 5-10% contingency budget item is reasonable. Given
the ever escalating costs of living, funds for any recurring budget item should be increased every
year to cater for inflation.

Most donor agencies insist on dollar budgets, yet others want local currency budgets. Some require
two columns one for local and the other for foreign currency. It is therefore essential to consult the
particular guidelines of your target donor agency.

Where more than one donor is requested to fund a proposal, it is advisable to present a
comprehensive budget showing the amount you are requesting from each donor as well as that to be
contributed by your institution (and government), even if the latter are in kind. A budget Check List
from WHO/EDUC/87.187 is attached for further guidance.
 102

9. References
A reference list must be prepared and included in the proposal. All references cited in the text must
be listed here in one of the acceptable formats. Alternatively and in the early stages of a proposal a
reference list must be kept for future use during report writing or manuscript preparation.

10. Profile of the applicants

This will include a description of the applicants as well as the institutions to which they belong.
This may be presented in the form of appendices as follows:
-Bio-data or CVs of investigators
-Institutional profiles to describe its capability
-Sample data sheet (questionnaires)
-Informed consent form
-Ethical clearance
-etc.

BUDGET CHECK LIST (WHO/EDUC/87.187)

Capital Expenditure
1. Construction Office buildings; staff housing; training schools;
health units; other construction. (Few donors support
this budget item)
2. Commodities Equipment; furniture; vehicles etc.

Recurrent Expenditure
3. Personnel Technical staff; support staff; consultants; casual
labourers
4. Training Fellowships; study travel; workshops; refresher
courses; correspondence courses; extension courses;
other courses.
5. Travel costs Train/bus/air fares; per diem, etc. for personnel.
6. Supplies Medical supplies; training supplies; vehicles; office
supplies; printing and photocopying
7. Maintenance Equipment; furniture; buildings
8. Vehicles running /maintenance Operation and maintenance of vehicles
9. Other costs/ institutional overhead Utilities (electricity, fuel, water, rent);
communications (telephone, telex, fax, postage);
licences and permits
10. Contingency Usually 5% of the budget
11. Inflation
12. Budget justification
 103

Writing a scientific paper for publication

Introduction

Some "scientists" prepare articles for publication just because they feel it is time their name
appeared in print, or because there is a forthcoming staff review, or they think they have hit a new
finding and can therefore, submit a preliminary report. Whatever the circumstances, scientific
papers should describe significant findings, that extend scientific knowledge or advance application
of existing tools or known principles. Remember "you are what you write" - many who will read
your paper will judge you so, and you will usually never have the chance to correct your paper.

Planning the Article

It is advisable to first write down your conclusions clearly, precisely and economically, and relate
them to the statement of the problem or the hypothesis that was being advanced at the time you
were developing the research proposal. Measure these conclusions against existing reports, and
discuss them with experienced scientists for advise before deciding whether your findings are worth
publishing. Remember papers likely to be read are those that are useful to readers because they
answer questions for them. An effective paper deals with an important question clearly answered
with an adequate evidence for the answer. Sometimes it may be important to apply the ‘So what
test’ How important will the message be ? What if the paper’s message is correct, ‘So what’. The
‘so what’ look at the message of the paper, is a powerful, but crude sounding measure of its
importance. What effect will the message have ? Will it change concepts or practices ? If so, How ?
If you are satisfied with the answers to these questions then proceed.

Only submit for publication previously unpublished work (except in a review article) nor work
being considered for publication elsewhere. But originality is not invalidated by previous
publication of a preliminary communication, although early release of results (e.g. to the press, in
conference proceedings etc.) may lead to their rejection in scientific journals by some editors.

At this early stage, consider who will read the paper, and this will determine the "target journal" to
which you submit the manuscript. The author should draw up a list of journals that seem to be
suitable choices, the list should be short enough to allow for considering each journal carefully.
Apply these questions to each journal.
• Is the topic of my proposed paper within its scope ?
• Is the topic represented in the journal frequently or only rarely ?
• Would it offer the best match of audience with that topic ?
• What formats does it accept ?
• Does it publish an information for authors page or issue a similar sheet or booklet that may give
better answers to these questions ?

On narrowing down the possible journal after applying the above questions, read their purpose and
scope and/or their instructions to authors. Your final choice may also be influenced by such
considerations as: the scientific quality of their previous publications, the quality of their print, the
time taken for articles to appear in print, their reach (e.g. number published per issue), appearance in
major abstracting journals, etc. On choosing the target journal, carefully study its most recent
"Instructions to Authors" so that you can start preparing the article according to their style.
 104

Preparing the article

You should decide on authorship (unless you worked alone!) at a very early stage. Include as
authors all those who contributed substantially to the theoretical or practical aspects, the following
four criteria may be considered when justifying authorship:

1. An author should have generated at least a part of the intellectual content of the paper: initially
conceived of the study it reports, if it is a research paper or case report; or developed the plan for
the paper, if it is a review or an editorial.
2. An author should have collected reported data (including clinical observations) and interpreted
them for the papers message.
3. An author should have taken part in writing the paper or revising its intellectual content.
4. An author should be able to defend publicly in the scientific community that intellectual content
of the paper for which he or she can take responsibility.

Despite meeting any of the above criteria, never include anybody before obtaining his/her approval.
Arrangement of authors' names usually reflects the amount of contribution to the work reported,
starting with the researcher who contributed most. If several researchers made equal contributions,
arrange their names alphabetically. If you work in a team, and publish several papers together, take
turns at being first author.

Besides style, many journals have other requirements, especially relating to ethics where humans
and other animals are investigated. You should be familiar with the policy of your "target journal",
or write and ask the editor. Now compose a working title and draft abstract (i.e. a short version of
the article). The working title must cover only one topic. In the working abstract express concisely
the research hypothesis, the approach adopted, the findings and the conclusions reached.

Next determine the organization of the paper - again refer to the "Instructions to Authors". Most
biomedical articles run through four sub-headings viz.: Introduction, Materials and Methods,
Results and Discussions. Other journals, especially in a descriptive field may run; "Introduction,
Population, Materials and Methods, Data Analysis, Results, Discussion, and Conclusion. A paper
describing a new method may run: Introduction, Description of Procedures, Test of New Method,
and Discussion. These will form the first-order headings of your article.

Write each first-order heading at the top of a sheet of paper, and on each write all the relevant
findings or thoughts as they occur to you, referring to your laboratory, field, or other notes and
records. Then add the tables, illustrations, and all the reprints (or relevant bibliographical cards)
necessary for the article. Then take the tables and illustrations, and improve their design so that each
item conveys a clear message. Pay close attention to the title, legend, and design; make sure they
provide maximum information in minimum space. The title of the table should reflect the contents
of the table and should guide the reader on how to interpret them. Model the tables in accordance
with the style of its "target journal".

Next number the tables in the order they are expected to appear. Then on a separate sheet list in
order, the tentative titles of the tables. Within each table, arrange the columns so as to allow readers
to easily understand the contents of the table.

Remember, in tables; do not give numbers implying greater accuracy than the methods used; give
numbers to the nearest significant figure; do not give percentages to more than one decimal place;
eliminate surplus zeroes by choosing appropriate factors in the column heading; and avoid dashes in
columns (make a notation, e.g. *, then explain in footnote). Make sure that tables are not crowded,
and can preferably fit across a column or at most across one journal page only. Try to omit certain
information from an over-large table, or even split such a table; many journals reject large tables.
 105

Do not include columns of data that can be calculated easily from data given in other columns. For
example, never include percent positive and percent negative in the same table.

For illustrations most journals prefer line drawings (diagrams, histograms, graphs) to photographs.
Illustrations must be functional, they should not repeat materials presented in tables or text. Since
illustrations are expensive, only submit them if essential. Many journals will accept colour
illustrations only at the author's expense. In making reference to tables and figures always use upper
case as in, “see Table 3 or see Figure 1; not see table 3 or see figure 1”.

In case material is borrowed from other people’s work, the author must request and get written
permission for its reproduction, from the copyright holder. Even when material is quoted from
unpublished work, permission to reproduce or quote it must be obtained.

Writing the First Draft

At this moment you already have drafted a working title and abstract, tables and illustrations and the
outlines. The notes collected earlier are then arranged to form topic and sentence outlines. Quickly
write the first draft following the already prepared outlines. At this stage use the name-and-date
system for citing references, and do not worry much about the grammar, but try to use the
appropriate person, tense and choice of verbs.

Introduction
The introduction must be brief - two or three paragraphs - indicating the aim and scope of the paper.
The introduction must, therefore, open with a succinct statement of the subject, orient the report in
its field, and should explain the purpose. The introduction thus serves as a link between the past
and the present. Generally all aspects of the literature cannot be covered for a relatively brief
research paper. Therefore the review of past research is done with a bias towards only those aspects
of the problem which are of direct relevance to your report. A good introduction starts out by
making a few general statements about the field of research, leading down logically to a narrow and
specific statement of the aim or hypothesis. The last paragraph of the introduction should state the
precise aims or the problem being investigated or the hypothesis being tested.

Methods
Details of how research was carried out include description of the study population, its size and
selection. Description of research instruments such as questionnaires, surgical and laboratory
equipment is essential in this section. Adequate description of research procedures must be given
here to provide the reader with enough information to allow replication (of course, replications may
not be feasible for a unique event such as a case study of a specific individual). Do not describe
familiar procedures, use references instead. Follow a logical order in describing the methods -
many authors follow a chronological order . For chemicals and drugs use systematic chemical
names, or international non-proprietary names. Where laboratory animals are used, give the genus,
species, race, strain, breed or line name. Ethical considerations must be discussed in this section.
Unusual equipment must be described.

Results
The results section must be comprehensive on its own, and results must be presented in a logical
order. This section presents the findings of the investigation and draws attention to points of
interest. Raw data and statistical calculations are not presented in this section. Rather, we use the
principles of descriptive and inferential statistics to present the summerized and analysed data: i.e.
graphs, tables and outcomes of statistical tests and also provide their interpretations.

Discussion
In the discussion section the validity of the present work is assessed, these are comments on the
significance of the results, which are also compared to previous work. In writing a good discussion
section consider the following questions,
 106

"Did your research find what you expected?" "How do the present results relate to those of
previous research?". If unexpected results were obtained, possible reasons for the outcome should
be discussed such as faulty designs and inappropriate selection of controls in previous studies.

In the conclusion, usually the last paragraph of the Discussion section, you summarize your main
findings and make recommendations on action to be taken as a result of your findings. Here you
also point out new lines of study and suggestions concerning further research.

A summary is included where the "target journal" wants it. The summary then states an outline of
the methods, the main findings and conclusions, referring briefly to new hypotheses and future
work - it may (unlike the abstract) include references to figures and tables. Some journals request
for a non-English summary.

Remember to acknowledge any substantial help received e.g. from a donor agency, from individuals
who gave money, materials, specimens, referred patients, assayed materials, advice on the execution
of the study and for typing the manuscript.

References and appendices

The references and any appendices must be checked for completeness. It is expected that all the
literature cited in your paper is listed in the reference section. This enables your reader to evaluate
your literature sources. You should refer to appropriate style manuals for information on how
references should be listed. Basically, you must give sufficient information for an interested reader
to be able to identify and retrieve your sources. You must read and keep copies of the contents of
all your references in case some authors should react to your article.

Revising your paper

First examine the draft to make sure that paragraphs and sentences are in the right order, that all
essential points are included, superfluous ones removed, and that arguments run logically. Check
the paragraphs - they should not be too long (usually not more than 125 words) nor too short to be
adequately descriptive. In the results, make sure that no sentences begin with a number; e.g. 25% of
the population....! And that all numbers below 10 are written in words; e.g. four, six, nine etc.
Instead of 4, 6, 9 etc. Check the tables and illustrations. Some may need modifications, some may
be superfluous, some may be combined etc. Now have the tables and illustrations made in their
final form. The paper size used for illustrations must not exceed A4 paper (210 x 297 mm). Follow
the instructions given by the "target journal" and the "Uniform Requirements for Manuscripts
submitted to Biomedical Journals".

Check the legends against the illustrations and make sure they agree. Now re-read (completely) all
the papers cited to make sure you do not misquote or misrepresent the work cited. Next check for
consistency. Make sure you adhere to the style of the "target journal". Try to reduce the number of
abbreviations and footnotes.

Now examine the first draft and from it check the style. Make sure you use concise language; be
precise, clear and avoid very lengthy construction. Also avoid verbosity and pomposity; you should
write (and speak) to express, not to impress. To avoid misuse of English words, use a thesaurus at
this stage. After reviewing and editing the first draft (for style and structure) prepare the second
draft by adding more materials to it and editing it. Have the paper reviewed by critical but
sympathetic colleagues and co-authors.

Refining your paper

Now go over carefully the comments from the co-authors and colleagues, and make changes where
you feel they are needed. Then polish the article, paying very particular attention to the title and
abstract, and prepare the final reference list. The title must be concise, accurate and informative.
 107

Some journals specify the permitted or preferred title length. Aim at keeping the title below 15-20
words or 120 characters (including spaces). It is advisable to omit such phrases as "Observations
on.....", "A study of ....", "An approach to ...", and the like.

After the title, supply a short title called the "running head" which usually consists of 45 to 60
characters. The final abstract must summarize the contents and conclusions of the paper, pointing
out new information and indicating the relevance of the work. The abstract should be completely
self-explanatory and intelligible in itself. Do not write footnotes nor quote references in the abstract
itself. This abstract, which is on a separate page, will be submitted by the editor to an abstracting
journal. Next select "index terms", i.e. key works or phrases suitable for use in the journal's subject
index and for information retrieval systems.

The next stage is to assemble the completed manuscript. The title page bears the new title and the
by-line (i.e. names of the authors and the addresses where the work was done). The abstract follows
on a separate page; depending on the target journal's instructions, index terms may be on the title
page or on the page with the abstract.

Now assemble the text, adhering to the target journal's instructions. In the text use symbols for
units of measure whenever a number precedes them, but spell out the name of a unit if the number is
spelt out, for example at the beginning of a sentence. Alternatively you may recast the sentence so
that the number falls in the middle of the sentence. Never begin a new sentence with a number.

In case there is an appendix, it should go either after the text and before the acknowledgements and
references, or after the list of references. Place the acknowledgements on a separate page at the end
of the paper, before the references. References should conform to the guidelines of the target
journal. Assemble the tables and illustrations (and legends) making sure each is lightly marked in
pencil with the name of the first author. This will prevent mixing of illustrations of one paper with
those of another during publication. Make sure that references in the text and the list of references
follow the instructions in the target journal, and that they are correct.

Then have the whole paper typed or edited by a competent typist. In the advent of computers or
word processing equipment this is only a simple editing task. Aim at creating a good first
impression (although ultimately the article will be judged by its contents). The article must be neat,
clean and accurately typed, on good quality paper. Where journals are photographically reproduced
directly from the author's typescript, perfect typing is absolutely essential. Only use white A4 size
(210x297mm) paper. The original copy must be on strong good quality paper to withstand frequent
handling. Do not get disheartened when you find that you are typing and re-typing your article
many times. This is the only way to obtain a final and refined manuscript. When the typescript is
ready, read it word for word, and make necessary corrections on all the copies.

Regulations regarding permission to publish differ widely. Some academic institutions, lack such
regulations. But other organisations, require their staff to obtain clearance before the article is
submitted to the editor. It is probably more convenient to submit the manuscript for clearance
before the final typescript is made. Under all circumstances it is wise to obtain approval from a
responsible official in order to safeguard the interest of the staff member, and his/her institution;
this is more so with technical data, and where sensitive issues are the subject of the report.

Submitting the Manuscript for Publication

Now the final typescript is ready for submission for publication in the target journal. Write a short
covering letter to the editor, following the instructions given in the "Uniform Requirements for
Manuscripts Submitted to Biomedical Journals" and the instructions to authors from a recent issue
of the target journal.

Responding to the Editor

 108

On receiving your manuscript, the editor will submit it to referees, who will review it and return it
to the editor with their comments. The editor will then compile the comments, and mail them to the
author. Your article may on very rare circumstances be accepted without revision, it may be
rejected outright ! In case the editor rejects the article, study his comments carefully, and consider
submitting it to another journal; contesting the editor's decision is often fruitless.

Where the editor requests for revision, consider the suggestions carefully, and modify the article if
you agree with him/her. If you do not agree, write a polite letter explaining your position.

Proofs
After you have corrected your manuscript the editor will then send it to the printer, through a
publisher. The printer will usually prepare galley proofs, send them to the editor who will in turn
send them to you (the author) with appropriate printer's and editor's correction marks. Read the
proof carefully, word for word against the typescript, to detect errors, omissions or other mistakes.
Mark all corrections and instructions in the margins of the proof, with corresponding marks in the
text. You must, therefore, be familiar with the basic symbols used in proof corrections. Make sure
the proofs are returned to the editor immediately.

References
1. Council of Biological Editors (1972) CBE Style Manual 3rd Edition, American Institute of
Biological Sciences, Washington, D.C.
2. International Committee of Medical Journal Editors (1986) Uniform Requirements for
Manuscripts submitted to Biomedical Journals. Annals Internal Medicine 108 (2): 258-65.)
or BMJ 296: 401-405
3. O'Connor, M. and Woodford, F. P. (1975) Writing Scientific papers in English: An HLSE-
Ciba Foundation Guide for Authors Associate Scientific Publishers, P. O. Box 211,
Amsterdam.
4. Introduction to research in the Health Sciences, Stephen Polgar and Shane A. Thomas
Churchill Livingstone, Melbourne 1992.
5. Huth E,J., How to write and publish papers in the Medical Sciences, Second edition,
Williams and Wilkins, London 1990.
 109

Evaluation and critique of a scientific article

Introduction

It is imperative that in order to evaluate and critique a scientific article (or a research proposal) you
must know how a scientific article is written. We have already had a session on how to write a
paper for publication. Knowledge of how to write a scientific paper for publication is therefore a
pre-requisite for evaluating critically such communication. This section of the lecturer notes
therefore assumes that you have this knowledge. Before proceeding with this set of notes you are
therefore advised to recapitulate on the steps in writing a paper for publication.

This section deals mainly with scientific articles which are a result of original research work.
General knowledge in this area will also enable you to evaluate research reports, case reports,
scientific reviews and editorials including letters to editors of scientific journals.

We may scan journals as we do newspapers to keep in touch with what is going on in our fields.
But in reading some papers closely we are looking for more. i.e. answers to questions; solutions to
problems. For example, a clinician may wish to know the best antibiotic for treatment of a
pneumonia. A nursing instructor may wish to know the best way to teach the surveillance of
obstetrical labour. The microbiologist may wish to know what characteristics of Legionella
pneumophila justify creating a new genus and species in bacterial taxonomy. The physiologist may
wish to know how substance X mediates renal vasoconstriction. The clinical pharmacist wants to
know the pharmacokinetic variables by which the dose of drug Y can be adjusted when its clearance
by the kidney is reduced by half.

Scientific readers are looking for more than just answers to questions and solutions to problems. If
they only want to get the answers and the solutions, they need not read entire papers because the
answers and solutions are usually stated clearly in a summary at the head of the paper, or at the end
of the paper’s Introduction or Conclusion section. What we need as readers is to be convinced that
the message of the paper - its answer to our question, its solution for our problem- is correct, is
valid, is reliable. We are not going to adopt the message into our way of thinking and working
unless we are persuaded that we should. Scientific papers are not just baskets carrying unconnected
fats, like the telephone directory; they are instruments of persuasion. Scientific papers argue you
into believing what they conclude; they are built on the principles of critical arguments. Rather,
“argument” is defined as “a coherent series of reasons, statements, or facts intended to support or
establish a pint of view.” And “critical” means, of course, assessing evidence carefully for its
validity, what evidence to accept and what evidence to reject.

Consider a typical paper reporting clinical investigation of antibiotic treatment for the infectious
disease Q. The investigators know that antibiotic A is often effective but that it has a high rate of
adverse effects. A new antibiotic, antibiotic B, comes along and gives in vitro evidence of
bactericidal effect against the causative organism. What is the question to be answered? “Is
antibiotic B more effective that antibiotic A in the treatment of disease Q?”
The investigators draw up a protocol for a study properly designed to answer the question. The
study gets under way. Antibiotics A and B appear for a while to have the same efficacy, but then
antibiotic B begins to look better. The investigators carefully monitor results of treatment with an
appropriate statistical procedure so that the comparative trial can be stopped when a statistically
significant answer is reached. Finally, antibiotic B is found to be better, a conclusion reached with
statistically supported evidence; the trial is brought to an end. Then the investigators write and get
published a paper reporting the trial. The paper makes clear what was studied (what question was to
be answered), the findings (evidence), their validity (statistical statements, assessment of possible
alternative conclusions), and the answer, “Antibiotic B is better.” The reader who wants of know
 110

the answer and to know whether the answer is reliable gets in the paper all that he or she needs to
know. The reader is persuaded; the reader has been argued into accepting the answer. This is what
we call critical reading.

In its simplest form the basic structure of a scientific paper is summarised by the acronym
IMRAD, which stands for

Introduction (What question was asked?)

Methods (How was it studied?)
Results (What was found?)
And
Discussion (What do the findings mean?)

However, when evaluating a scientific article the following aspects should be critically examined.

1. Title
2. Abstract
3. Introduction
(a) Review of literature
(b) Objectives of the study
4. Methodology
5. Results and analysis
6. Discussion
7. Conclusion(s)
8. References

The process of evaluating a scientific article must naturally follow the order of the sections above.
Below is a set of questions that keep reminding you what you should do to evaluate properly the
article.

1. Title
The title should have the following characteristics. If it lacks any of them you should be in
position to suggest a good title for the article after reading through the entire paper when you
know what it is all about.

Is the title of reasonable length?

Does it have unnecessary leads?
Does it indicate concisely and clearly the following:-
-key variables included in the study
-the proper relationship (if any) between the variables.
-Is the target or reference population stated.
2. Abstract
Does the abstract contain the essence of the whole paper?
Does it say: Why the study was done, What was done, What was found
What was concluded. Is it clear and concise with unnecessary detail?

3. Introduction
(a) Review of literature
Has the author outlined the background of the problem?
Has he reviewed prior knowledge on the problem?
Has he identified properly the relevant variables selected and not selected in the
preceding studies?
Has he outlined the significance or public health importance of the problem and
therefore the purpose of this particular study?
 111

(b) Objectives
Broad objectives
Specific objectives
-free of value statements not amenable to testing
-meet the purpose of the study
-clear and unambiguous
-stated in measurable terms

4. Methodology
Study design
Type of study design; prevalence survey, case-control study controlled trial case reports etc.
Is study design appropriate for the study objectives?

Study population
is the study group defined?
-what criteria were used for selection?
-was any kind of control or comparison group necessary
-Does this control group adequately control for variables which are not being studied?
-was any form of sampling used?
-type and sample size:- how representative is it of the target population?
-any factors which might have biased the selection of study and/or control group?

Methods
Type of measurements: what measurements were used? were they appropriate?
-notified cases
-attitudes
-laboratory tests
-examination

instruments used to collected the data are a variety of specific equipment, registers,
observation records and questionnaires

Type of errors to be expected in the measurements and how they were dealt with:
-observer variation
-subject variation
-instrument variation

Steps taken to reduce these errors or to quantify them?

Were methods used appropriate to the objectives of the study?

5. Results and analysis

Are the results clearly set out?
Tables and graphs:- understandable without reference to the text: do they add up properly?
Discrepancies between results in text/tables and graphs?
Losses to follow-up or non-response
-If so, might these have biased the results in any way?
Are the data given adequate in terms of number and quality to draw conclusions?
Statistical tests: are they appropriate is it clear how they were done?
If they were not done, should they have been done?
Any other analysis that should have been done?

6. Discussion
Has the author assessed properly how the results have demonstrated the evidence (or lack of
it) for a new scientific contribution. Does author show value of this contribution in disease
 112

control programmes? What are the recommendations if any? Do they really come from this
piece of research work? And if so what kind of studies and what possible methods?

Conclusions
Are the conclusions justified from the data and evidence given in view of any problem you
have encountered with methods, results or analysis? Can you draw any alternative
conclusions from the same data? To what extent can these conclusions be generalized?

Have we really gained any important new knowledge or insights from this study, and are they
of immediate practical value, or are more studies required?

References
Are references properly cited? Are they relevant? Are they up to date? Are they retrievable?
i.e. are they complete and properly written according to requirements for manuscripts
submitted to biomedical journals?

Further reading

1. Huth E. J. How to write and publish papers in the Medical Sciences. ISI Press
Philadelphia 1982.

2. Hall G.M ed. How to write a paper BMJ Publishing group London 1994
 113

Sources of error in epidemiologic studies

By this stage you have been exposed to the general principles of epidemiologic research. As you might
have noted, epidemiologic research is conducted in human populations and that unlike basic research
(conducted in the laboratories), many factors may influence the findings obtained from such studies.
Because of this, it is important to remember that while study results might reflect the true effect of an
exposure on a certain disease, it is also possible that the findings may have alternative explanations.

There are three alternative explanations which we need to consider before we conclude that there is a
real association between exposure and a certain disease. These includes:-

a.) Chance - observed association may be due to chance, or the "luck of the draw".

b.) Study results may be explained by some form of bias or systematic errors due to
problems related to the way subjects were selected into the study or the way information was
obtained.

c.) Findings of the study may be due to other factors associated with both exposure and
outcome of interest or confounding.

It is therefore important to consider all these alternative explanations before we can conclude that there
is a real association between exposure and outcome of interest. Most errors in epidemiological studies
results from the failure to consider these alternative explanations during the period of study design and
data analyses.

Before focusing in detail on various sources of error, we must broadly distinguish between the terms
"validity" and "precision", especially since different approaches to the assessment of each are
advocated. Validity is the extent to which the study measures what it is intended to measure. Precision is
the reproducibility of a study result, that is, the degree of resemblance among study results, were the
study to be repeated under similar circumstances. In general, these terms are concerned with two
different sources of inaccuracy that can occur when estimating an effect: systematic error (a validity
problem) and random error (a precision problem).

TYPES OF ERRORS:

I: Random Errors:

Source of random errors:

Refers to part of our experience that we cannot predict (chance variations). In epidemiologic studies,
random errors has many components, but a major contributor is the process of selecting the specific
study subjects. This process is usually referred to as sampling; the attendant random error is known as
sampling error. Case-control studies literally involve a sampling process, but all epidemiologic
studies, including follow-up studies, which do not involve actual sampling of subjects, are said to have
sampling errors.

Strategies to reduce random errors:

Precision in epidemiologic measurements corresponds to the reduction of random errors. Precision can
be improved in two ways:-
 114

a) The primary means is to increase the size of the study population (increase sample size).
b) Precision can also be improved by modifying the study design to increase the efficiency with which
information is obtained from a given number of study subjects. Study efficiency can be assessed as:-

i) the amount of information obtained per number of study subjects, OR

ii) amount of information obtained about the effect in relation to the cost of obtaining it
regardless of the number of subjects in the study.

II: Systematic Errors (Bias):

Bias may be defined as a systematic error in an epidemiologic study that results in an incorrect estimate
of the association between exposure and risk of disease.

Various types of biases affect validity of the study findings. There are a lot of possible biases that can
distort the estimation of an epidemiologic measure. The distinction among these biases is occasionally
difficult to make, but three general types can be identified:-

a) Selection bias
b) Information (observation) bias
c) Confounding

These categories are not always clearly demarcated. A useful practical distinction between confounding
and other biases is to consider a bias as confounding if it can be controlled during data analysis stage.

SELECTION BIAS:

Can occur wherever the procedures used to select individual subjects for inclusion into the study on the
basis of either exposure (cohort) or disease (case-control) status leads to an effect estimate among
subjects included in the study being different from the estimate obtainable from the entire population
theoretically targeted for the study.

Selection bias can result from a number of circumstances related to the way in which individuals are
ascertained and selected for the study. Some of the major forms of selection bias include:-

a) Self-selection bias:

- Self-referral of subjects is ordinarily considered a threat to validity, since the reasons for
self-referral may be associated with the outcome under study.
- e.g. Centers for Disease Control (CDC) investigated subsequent leukemia incidence among
troops working in the atomic plants. 76% of the troops identified as members of that cohort
were traced; out of which 18% of the subjects contacted the investigators on their own initiative
in response to publicity about the investigation. In this study, there were 4 leukaemia cases
among the 18% of subjects who referred themselves and 4 among the 82% of subjects traced by
the investigators, indicating that self- referral was indeed biased towards leukaemia cases.

b) Differential surveillance, diagnosis, or referral:

Selection bias can occur when there is differential surveillance, diagnosis or referral of
individuals into the study. e.g. When the first hospital-based case-control studies of the
relationship between oral contraceptives (OC) use and thromboembolism were reported, there
was some concern that since physicians were already aware of a possible relationship of these
agents with thromboembolism;
i) OC users were more likely to be hospitalized than non-users for evaluation of this condition.
ii) Among cases, those with history of OC use were more likely to be subjected for intensive
investigation to rule out this condition than those without such a history (i.e. detection bias).
 115

Hence increased frequency of current OC use among women hospitalized for thromboembolism might
actually be due, at least in part, to the fact that hospitalization and the determination of the diagnosis
were both influenced by a history of OC use.

Similarly, criticism concerning the increased risk of uterine cancer associated with use of exogenous
oestrogens observed in early case-control studies of this hypothesis was based on the argument that
women who used oestrogens experienced uterine bleeding and were, therefore, more likely to seek
medical attention and have a diagnostic evaluation than women not on oestrogens.

c) Loss to follow-up or non-response:

- Loss to follow-up or non-response may be associated with exposure (case-control studies) or

outcome of interest (cohort studies).

d) Selective survival:

- This is a common type of selection bias in case-control and cross-sectional studies.

- Concerns bias that may result from the use of prevalence data to draw conclusions about incidence measures.
- Determinants of survival may be associated with outcome of interest.

Strategies to avoid/reduce selection bias

1. Choice of appropriate comparison groups:

- Comparison group needs to be similar to the exposed (cohort studies) or diseased (case-
control) group with respect to the important predictors of outcome of interest.

a) In follow-up study, the comparison group serves to provide an estimate of what the disease
incidence would have been in the exposed group had the exposure been absent or without effect.
- e.g. assessment of cardiovascular mortality among factory workers.
? Compare mortality of workers with that in the general population?
? Compare mortality of sub-groups of workers based on level of activity?

b) In case-control studies, control group provides an estimate of the exposure distribution in the
source population from which the cases originate.
• Since completely identical groups will never be found, the process of determining which
factors are relevant in choosing a comparison group is a critical feature of epidemiologic
study design.
• Use of two or more comparison groups drawn in different ways. If the effects do not differ,
it means there is no bias in selection. Consistent results from several comparison groups
reinforce belief about the results.

2. Ensuring complete follow-up:

- Selection of a study group which is likely to have minimal loss to follow-up (cohort studies)
or high response rate (case-control studies).

3. Diagnostic surveillance:
- Uniform procedures in ascertainment of exposures and diseases.

4. Selection of subjects independent of exposure (case-control) status.

Information (observation) bias:

Refers to a distortion in the estimate of effect due to measurement error or misclassification of subjects
on one or more variables. Major sources of information bias include:-
 116

- Invalid measurements
- Incorrect diagnostic criteria
- Omissions, imprecisions, or other inadequacies in previously recorded data.
- Unequal diagnostic surveillance among exposure groups (cohort studies). Note that in case-control
studies, selective surveillance can lead to selection bias rather than information bias.

NOTE:

RECALL BIAS is the commonest type of information bias in case-control studies. It occurs if there is
differential recall of exposure in the cases and controls. e.g. in a study to assess exposure to certain
drugs during antenatal period, mothers of malformed infants recall exposures more thoroughly than
mothers of healthy infants.

Recall bias is a possibility in any case-control study that uses an anamnestic response since the cases and
controls by definition are people who differ with respect to their disease experience, and this difference
may affect recall.

Strategies to reduce/avoid information (observation) bias:

1. Precise measurements
- well defined methods for taking measurements.
- Intensive training of data collection personnel

2. Diagnostic tests
- Uniformity in criteria used in diagnosis
- Methods with high sensitivity and/or specificity required.

3. Reduce Inter-and Intra-observer variations

- Use same observers for the entire study
- Common methodology used/common training
- Random quality checks

4. Use of appropriate tools for data collection (research instruments which provides information
required).

5. Recall bias can be reduced by:

- Choose controls who, although they lack the disease under study, nevertheless have had an
equal recall stimulus.
- Blinding the interviewers and study subjects about the objectives of the study.

Confounding

Confounding refers to mixing of the effect of the exposure under study (first factor) on the disease
(second factor) with that of a third factor. The third factor must be:-

a) associated with the exposure in the source population.

b) independent determinant of the outcome (or a risk factor for the disease under investigation
independent of exposure of interest).

Example: In a study it was found that subjects with a high consumption of alcohol (first factor) had a
higher risk of developing lung cancer (second factor) than those who did not drink alcohol. The reason
for this association was that subjects with a high consumption of alcohol were often heavy smokers.
Confounding from smoking explained the observed association between alcohol consumption and lung
cancer.
 117

References:

1. Hennekens CH and Buring JE. Epidemiology in Medicine. Little, Brown and Company, 1987.

2. Beaglehole R, Bonita R, and Kjellstrom T. Basic Epidemiology. WHO, Geneva 1993.

3. Abramson JH. Survey Methods in Community Medicine (4th Edition). Churchill Livingstone, 1990.
 118

Epidemiology and control of major diseases in Tanzania

Introduction

1. You have now completed the part of the curriculum on principles and methods of
Epidemiology. You are now ready to learn how to apply this knowledge and skill to
describe and investigate major public health problems in Tanzania. In order to do this the
class will be divided into the usual number of discussion groups so that you can gain some
practical skills on the subject. This exercise will begin during the third term of the
academic year and will stop a few weeks before the final university examination.

2. Each discussion group will be assigned or will select a specific disease for class
presentation. Members of each group must consultant the relevant instructor for references
and relevant literature and seek advice on how to prepare and give the presentation on the
selected disease. The value of this exercise is to expose you to the kind of experience
which is hard to lean from a library. It is expected to provide you with the skills of
preparing scientific materials for public presentation. Talking with confidence in a
scientific forum is an experience you can only get through this kind of exercise.

3. Students will be required to use all available knowledge on the disease and discuss its
epidemiology using the format below. Students in a group are advised to assign each other
particular sections of the presentation and consult relevant specialists and institutions on the
subject. The library is expected to be particularly useful for up todate knowledge on the
epidemiology of diseases. Copying text books alone without relating the materials to your
presentation will not be accepted. The Epidemiology unit and the planning section of the
Ministry of Health headquarters as well as disease specific national control programmes eg.
TB & Leprosy, AIDS etc. are also good sources of information on health problems in
Tanzania. You should also consult these to improve your presentations. In this case, letters
of introduction from the department may be useful. It must be stressed that getting
information from busy government officials is often difficult and so you have to be patient
and show that you are really interested in the information you are looking for. Remember
that it is easier to obtain information when you know exactly what you are looking for than
when you only have a vague idea of your problem.

4. The presentations must be brief and to the point. Students will be encouraged to use
overhead projections during presentations. Each group will be allocated a maximum of 10
overhead sheets and water soluble overhead felt pens from the department all of which must
be returned at the end so that they can be re-used by other groups. The group leader should
organize the presentations so that they are interesting and not boring. In order that each
group gets a sufficient audience, attendance for these sessions will be mandatory and those
who will be absent shall be penalized. The current penalty is 5 marks each time you don’t
appear except when it is your turn to make a presentation.

Presentation format

A. Description and discussion of the Natural History of the Disease(s)

Describe the various stages of the disease(s) you have selected.
Classify the epidemiological factors into, host and environmental factors and in all
their sub-classes (e.g biological, physical and social factors) and show the disease
distribution relative to each of the main factors (e.g. show and discuss the age
 119

distribution of measles or TB in Tanzania show the differences in age and sex

distribution etc). Support your account with quotations from recent research data,
demonstrate with tables and graphs on board or screen. Transparencies for
overhead presentation may be available on request.

B. Levels of Prevention

Classify the preventive measures into primary, secondary and tertiary levels of
prevention, showing the effect each measure will have on the disease(s) natural
distribution (natural history). This, like in “A” above, must be supported by the
most current research findings properly quoted and where possible by
demonstrations with tables and graphs.

C. Position of Control in Tanzania:

Indicate the country’s policy towards this disease and discuss it relative to the
previous discussion on the disease’s epidemiology: What is the current stage (s) of
control in Tanzania? What control options are in operation in Tanzania?

And what is the basis for the options taken? If control is indicated or exists, what
would be the ideal control strategy in the preparatory, attack, consolidation and
surveillance phases for this particular disease considering the natural history of the
disease and available control measures. In this section participant’s comments,
speculations and suggestions are appropriate in addition to basic factual
information. Participants may differ in many of these aspects, though differences
ought to be governed also by scientific evidence.

D. Case Presentation and critique of a given research paper on the disease:

Together with the presentation on the epidemiology and control of major diseases in
Tanzania, students will be required to critique a published paper reporting on research
results relating to epidemiology and control of diseases. To achieve this, one of the
participants in a group should look for a relevant published paper on the subject and
critique it. The participant should present a summary of the paper and then critique it as
follows:

(i) Was the research problem clearly defined?

(ii) Was the problem relevant for investigation at the time of the study?
(iii) Has the title been properly stated?.
(iv) Are the objectives properly stated?
(v) Were the methods proper for the objectives?
(vi) Were results according to objectives?
(vii) Any contribution to knowledge?
(viii) Is it pointed out, what further hypothesis(ses) have been or can be formulated or
what programmes can be initiated or discontinued, as a result of this study?
If not what would have been pointed out.
 120

List of major diseases in tanzania

The following is the preliminary list of major diseases assigned to groups. More
diseases and other relevant problems may be added whenever it becomes necessary.

(i) Diarrhoea Diseases - group A

(ii) Tuberculosis - group B
(iii) Leprosy - group C
(iv) Malaria - group D
(v) Acute Respiratory Infections - group E
(vi) Schistosomiasis - group F
(vii) AIDS - group G
(viii) Trypanosomiasis - group H
(ix) Filariasis - group I
(x) Dental caries - group J

At the end of the sessions all presented materials should be handed in for assessment and grading.
Groups must ensure that all members participate and any materials handed in must contain a list of
all participating members. A hard working group will be awarded high marks and a lazy group will
get low marks. These marks will contribute to the continuous assessment grade. Remember that
the work must be neat and tidy. Typing of materials will be encouraged but not essential.