Case control design

Editor's Notes

• #12: Limitations Exposure cannot be linked with disease as whole population is represented. Presence of a correlation does not necessarily mean a statistical association. Correlation data represent average exposure levels rather than actual individual levels
• #13: analyze data collected on a group of subjects at one time rather than over a period of time. Cross-sectional studies are designed to determine “What is happening?” right now. Subjects are selected and information is obtained in a short period of time Because they focus on a point in time, they are sometimes also called prevalence studies. Surveys and polls are generally cross-sectional studies, although surveys can be part of a cohort or case–control study. Cross-sectional studies may be designed to address research questions raised by a case–series, or they may be done without a previous descriptive study. Subjects are selected and information is obtained in a short period of time Because they focus on a point in time, they are sometimes also called prevalence studies.
• #18: In early 1940s, Alton Oschner, a surgeon in New Orleans, observed that virtually all of the lung Ca patients which he has operated were smokers. Although this relationship is established now but at that time it was controversial! He hupothesised that smoking is linked with lung cancer- was this valid? Again in 1940s, Sir Norman Gregg, an Australian Ophthalmologist, observed that a number of infants and young children presenting with unusual form of cataract have been in utero at the time of rubella outbreak!!! – he hypothesized that there is an association (that time it was not known that virus is teratogenic!!!!)
• #19: In November 1989 CDC Atlanta and local health departments published the first description of EMS – characterized by incapacitating myalgias, elevated eosinophil counts, and in some patients arthralgias, hair loss, skin thickening and interstitial lung disease. L-tryptophan is an essential amino acid available without prescription EMS occur predominantly in women and is relatively rare. Nation wide surveillance identified 1500 cases, including 40 fatalities. Nearly all cases occurred btw mid 1988- to end of 1989. case control studies established strong association of ingesting L-tryptopan and EMS. After recall of L-tryptophan from market reported cases fell to near zero. Case control studies showed that nearly everyone with EMS and less than 50% without it consumed L-tryptopan produced by one particular manufacturer. Risk of getting EMS after consuming L-tryptopan of this manufacturer was 20-40 times more than the risk among those who took L-tryptopan of other manufacturers. There were many contaminants but exact one couldn’t be found bcz of ethical issues- no animal testing! Astute (judicious) observations and quick public health response led to timely recall of L-tryptopan and prevented larger outbreak.
• #21: Case–control studies begin with the absence or presence of an outcome and then look backward in time to try to detect possible causes or risk factors (Exposure) that may have been suggested in a case–series report. The cases in case–control studies are individuals selected on the basis of some disease or outcome; the controls are individuals without the disease or outcome. The history or previous events of both cases and controls are analyzed in an attempt to identify a characteristic or risk factor present in the cases' histories but not in the controls' histories. Figure illustrates that subjects in the study are chosen at the onset of the study after they are known to be either cases with the disease or outcome or controls without the disease or outcome. The histories of cases and controls are examined over a previous period to detect the presence (shaded areas) or absence (unshaded areas) of predisposing characteristics or risk factors, or, if the disease is infectious, whether the subject has been exposed to the presumed infectious agent. In case–control designs, the nature of the inquiry is backward in time, as indicated by the arrows pointing backward to illustrate the backward, or retrospective, nature of the research process. We can characterize case–control studies as studies that ask “What happened?” In fact, they are sometimes called retrospective studies because of the direction of inquiry. Case–control studies are longitudinal as well, because the inquiry covers a period of time. Investigators sometimes use matching to associate controls with cases on characteristics such as age and sex. If an investigator feels that such characteristics are so important that an imbalance between the two groups of patients would affect any conclusions, he or she should employ matching. This process ensures that both groups will be similar with respect to important characteristics that may otherwise cloud or confound the conclusions. Deciding whether a published study is a case–control study or a case–series report is not always easy. Confusion arises because both types of studies are generally conceived and written after the fact rather than having been planned. The easiest way to differentiate between them is to ask whether the author's purpose was to describe a phenomenon or to attempt to explain it by evaluating previous events. If the purpose is simple description, chances are the study is a case–series report.
• #22: Population based case control studies select newly diagnosed cases and controls from a well defined population Hospital base case control studies usually use prevalent rather than incident cases and take controls from conveniently available patients of other diseases, in the same institution. Ex.- case control study of Reye’s Syndrome (a condition characterized by encephalopathy associated with fatty degeneration of liver, occurs almost exclusively in children and typically follows a viral illness). Cases were selected from children admitted with Reye’s Syndrome in any hospital of the city and controls from same institution from children admitted with some other illness preferably viral, to find out any association btw any exposure (here – aspirin intake) and Reye’s syn. 26 out of 27 cases and only 6 out of 22 control took aspirin. In hospital based study patients are more willing to participate. Factors like SES are balanced as both cases and control are from same hospital BUT… distortion arise as source population is not defined moreover controls are patients suffering from other diseases which may confound the results (if exposure history of controls differ from unaffected people in the population). To rule out this select controls from various diagnostic groups and early acute conditions so that earlier exposures are not affected by dis condition. Also donot select pts with multiple conditions and controls with condition related to the disease of interest
• #25: Case definition – e.g. lung cancer confirmed by biopsy • Prevalent vs. incident cases Prevalent: • No need for waiting • RFs may be more related to survival than incidence. If many people die soon after diagnosis, may over-represent long term survivors Incident: • Recruit new cases at time of disease occurrence • Better for making inference about association between risk factor and developing the disease Controls should have same levels of exposure as the unaffected person in the source population. Should be comparable to cases. • Should have the potential to become cases (must be susceptible to the disease of interest) • Possible control sources: population, neighborhood, friend, hospital knowing! - Publicity of concerning the suspected association, Low participation – if exposure histories are differrent in participants and non participants, error in selecting controls- e.g. pts of insomnia (more chance of taking L-tryptophan) In 1929 Raymond Pearl, professor of biostatistics at John Hopkins conducted a study to test if TB is protective for Cancer – from 7500 consecutive autopsies he identified 816 cancer cases and selected 816 other autopsies randomly and determined proportion of TB in cases and controls – 6.6% vs 16.3% (since it was less in cases he concluded that TB had antagonistic or protective effect). Was it justified? – only if proportion of TB in non-cancer patients is same as it was in controls. Which was not the case!!! actually at those times TB was quite prevalent and he selected controls from other patients other than cancer admitted and died in the hospital during that time. And many died due to TB Carlsons and Bell repeated Pearl’s study – compared pts who died of cancer to pts who died of heart disease. They found no difference!!!
• #26: Controls were cases of disease other than cancer pancreas admitted by same gastroenterologist - usually of esophagitis and peptic ulcer (they are prone to consume less coffee)
• #27: 15 to 20 yr!! May be too broad to loose relevance to any comparision
• #28: To remove recall bias other methods should be employed to verify – e.g. see drug packages in case of L-tryptophan use (manufacture could also be verified) Medical record, occupational records, etc Biomarkers though objective but are costly, invasive and not always available
• #32: 118-110= 8, z = 8/ 4= 2.0 Area under normal curve at z score 2.0 = at z table at x axis 2.0 and on y axis 0.0 = 0.9772 if we multiply it with 100 it is 97.72% i.e. only 2.3% (100-97.72 = 2.3) chance is that this person having BP od 118 mm of Hg is not from the same population having average BP as 110 mm of Hg (normotensive)
• #41: Now, turning to the case-control study design, cases arise from a clearly defined source population and the investigator then chooses controls from this same population. Thus to conduct a case control study - cases i.e. diseased persons- incident cases only are taken. Both cases and controls are selected without regard to exposure so that proportion of cases in the study that are exposed should, on average, equal the proportion of cases in the full cohort.