This table describes the key characteristics of research, service evaluation, audit and health surveillance projects in order

to assist with deciding how a project should be managed.

Each of these project types has their own separate governance requirements which you will need to arrange before starting the work. A programme of work may involve more than
one project type, but each individual project within the programme should sit clearly under one column. If you find that your planned project spans more than one column, it is
likely that its scope and purpose is not defined clearly enough. Consider making revisions to the project design to ensure that it clearly sets out what you want to achieve, and the
methodology you will use.

SERVICE EVALUATION / IMPROVEMENT CLINICAL/ NON-FINANCIAL

RESEARCH* HEALTH SURVEILLANCE
/ DEVELOPMENT AUDIT
A key feature of research is that it is intentionally
planned and designed using documented
methodology which will allow results to be
extrapolated or applied from the study sample to Designed and conducted to assess
a larger population. This extrapolation / priorities, evaluate interventions, and
Designed and conducted to produce
application is what the terms ‘generalisable’ and Designed and conducted solely to define or judge detect and manage threats to health and
information to inform delivery of
PURPOSE ‘transferable’ refer to. In the case of quantitative current care or service, or to deliver and measure
best care.
adverse health status (including incidents,
research, statistical methods are used to achieve improvements in quality of the current service. risk factors, hazards, outbreaks and
results that are ‘generalisable’ from a sample to epidemics, may also address health
the sampled population. In the case of qualitative inequalities).
research, the context and findings are described
and defined so that the conclusions can be
applied or transferred to other settings.
Service evaluation is designed to answer the Designed to answer the questions: “Is
question: “What standard does this service Designed to answer the question: there a need to start, continue or stop
Aims to generate a new hypothesis or test a
achieve?” This is normally addressed by asking “Does this service reach a defined public health interventions”, or
QUESTION/ hypothesis. The approach to this may be
those in receipt of the service. Service predetermined, recognised or pre- “Is there need for further investigations”,
HYPOTHESIS quantitative, qualitative or both.
development or improvement seeks to find out established standard?” or “What is the cause of this outbreak
what improvement can be achieved within that (often of a disease) or incident and how
service only. do we manage it?”
Measures current service without reference to a
Has clearly defined aims and objectives. The Measures against historical (or
standard.
project seeks to answer a specific research geographical) comparators and/or
(In the case of service improvement / Measures against a standard.
AIM question or questions.
development the current service may be
defined levels (triggers) for action.
Systematic, quantitative or qualitative
compared to the previous service).
methods may be used.
Service evaluation involves an intervention or Involves an intervention or service Intervention (if relevant) in use only. Any
service already in use only. Service improvement already in use only. The choice of choice of intervention, treatment, care or
May involve evaluating or comparing or development involves a new intervention or treatment, care or services is that of services is based on best public health
interventions, particularly new ones. Not all service, or one that is new to that context. The the care professional and evidence or professional consensus, but
INTERVENTIONS research involves interventions. choice of treatment, care or services is that of the patient/service user according to may also be used to assess the need for
care professional and patient/service user guidance, professional standards an intervention when none is being taken
according to guidance, professional standards and/or patient/service user currently.
and/or patient/ service user preference. preference.
Published October 2022 © Health Research Authority 2022. Copyright and other intellectual property rights in this material belong to the HRA and all rights are reserved. The HRA authorises UK healthcare organisations to
reproduce this material for educational and non-commercial use.
 May involve analysis of existing routine
data supplied under licence, agreement
Usually involves collecting data that are or administration of interview or
Usually involves analysis of existing
additional to those for routine care or service Usually involves analysis of existing data but may questionnaire to those in the population
data but may include administration
(but not always). May involve comparing data on also include administration of interview(s) or of interest. This includes collection of
DATA treatments, samples or investigations additional questionnaire(s).
of simple interviews or
data on hazards, exposures and other
questionnaires.
to routine care. May involve data collected from data to enable interpretation of issues
interviews, focus groups and/or observation. relevant to the population rather than the
individual. May also require evidence
review.
Quantitative research study design may involve No allocation to intervention: the care Not applicable. Collects data on issue of
No allocation to intervention: the
allocating patients/service users/healthy professional and patient/service user have chosen concern in situ. May involve allocation to
PARTICIPANT care professional and patient/service
volunteers to an intervention. intervention independently of the service control group to assess risk and identify
ALLOCATION Purely qualitative research does not usually evaluation / improvement / development.
user have chosen intervention
source of incident, but no allocation to
before audit.
involve allocating participants to an intervention. intervention.
May involve randomisation for
May involve randomisation for sampling, but not May involve randomisation for sampling,
RANDOMISATION May involve randomisation. for treatment/ care/ intervention.
sampling, but not for treatment/
but not for treatment/ care/ intervention.
care/ intervention.
Normally requires NHS REC review but not
NHS REC review
always. Refer to [Link] Does not require REC review. Does not require REC review. Does not require REC review.
required? [Link]/ethics/ for more information.

Published October 2022 © Health Research Authority 2022. Copyright and other intellectual property rights in this material belong to the HRA and all rights are reserved. The HRA authorises UK healthcare organisations to
reproduce this material for educational and non-commercial use.
*The UK Policy Framework for Health and Social Care Research defines research as:

“3.1 For the purpose of this policy framework, research is defined as the attempt to derive generalisable or transferable1 new2 knowledge to answer or refine relevant
questions with scientifically sound methods3. This excludes audits of practice and service evaluations. It includes activities that are carried out in preparation for or as a
consequence of the interventional part4 of the research, such as screening potential participants for eligibility, obtaining participants’ consent and publishing results. It also
includes non-interventional health and social care research (i.e. projects that do not involve any change in standard treatment, care or other services), projects that aim to
generate hypotheses, methodological research and descriptive research. Projects whose primary purpose is educational to the researcher, either in obtaining an educational
qualification or in otherwise acquiring research skills, but which also fall into the definition of research, are in scope of this policy framework. Activities that are not research
according to this definition should not be presented as research and need not be conducted or managed in accordance with this framework. A decision tool that provides a
definitive answer about whether a project counts as research under this policy framework is available at [Link]/research.

1NB This definition involves an attempt at generalisability or transferability, i.e. the project deliberately uses methods intended to achieve
quantitative or qualitative findings that can be applied to settings or contexts other than those in which they were tested. The actual
generalisability or transferability of some research findings may only become apparent once the project has been completed.
2 Including new knowledge about existing treatments or care.
3 Projects that are not designed well enough to meet this definition are not exempt from this policy framework – see paragraph 9.10.a.
4 This means the part of the research where a change in treatment, care or other services is made for the purpose of the research. It does not
refer to other methodological ‘interventions’, e.g. issuing a postal survey.

Published October 2022 © Health Research Authority 2022. Copyright and other intellectual property rights in this material belong to the HRA and all rights are reserved. The HRA authorises UK
healthcare organisations to reproduce this material for educational and non-commercial use.