Best practice in health outcomes research

Nicola Bonner clarifies the guidelines, and which are relevant, to ensure practitioners maintain high and consistent standards. Published on Pharmaphorum 26th June 2015

Qualitative and quantitative non-interventional health outcomes research studies are on the increase, but there are misconceptions about where this relatively new field sits in relation to other types of pharmaceutical research and hence inconsistencies in standards.

Outcomes research encompasses a broad range of studies, including those involving the development and implementation of patient-reported outcomes (PRO) and clinical outcomes assessments. The approaches taken, and the level of rigour, generally sit somewhere between market research and interventional clinical trials.

Work that involves patients and/or their data and which will ultimately support regulatory submissions or publications should be thorough and follow industry standards. However, while outcomes research typically involves patients, their caregivers or healthcare professionals, there is no medical intervention and so the same requirements regarding safety concerns do not apply.

This seeming ‘no-man’s land’ between clinical trials and market research often leads to questions regarding the requirements for outcomes research according to the principles of good clinical practice (GCP). Confusion can result in extensive discussions between clinical and operational teams in pharma companies and researchers at public or healthcare institutions over what standards should be followed. The outcome is rarely the same and it is clear that in this case one size does not fit all.

So how do we achieve consensus on what the requirements should be for GCP and outcomes research?

The ICH guideline, published in 1996, define GCP as an ‘international ethical and scientific quality standard for designing, conducting, reporting and recording trials that involve the participation of human subjects’.

Interestingly, this definition simply refers to trials involving human subjects. The confusion arises with the many references to ‘clinical trials’ within the guidelines. So what does this mean for outcomes research? Should a large, non-interventional, real-world study involving observational data be held to all of these standards? How about qualitative patient research conducted in smaller samples of patients (e.g. n=20), such as open-ended qualitative interviews?

The ICH guideline presents 13 principles of GCP, so let’s consider how these relate to outcomes research.

1. Ethics

When it comes to ethics, outcomes research definitely lies closer to the clinical trial end of the spectrum. To ensure that study participants are not put at unnecessary risk, ethical approval should be sought. However, in some countries, such as the UK, this does create delays and challenges due to the fact that this observational research is typically reviewed in the same way as clinical research.

Of note is that outcomes studies often deviate from clinical trials in their use of Principal Investigators who are usually experienced outcomes researchers with Master’s degrees or PhDs, rather than physicians.

2. Risk/benefit

For any study involving human participants, consideration of risk vs benefit is a universal factor for clinical, market and outcomes research. In outcomes research the risk is primarily in terms of asking potentially sensitive or personal questions and the distress than can result, rather than risk of serious adverse events.

3. Rights and safety of trial subjects

Again, this universal consideration is pertinent in outcomes research but the safety risks are generally of a much lower order of magnitude, given that there is no interventional treatment. The additional element of safety of researchers should be considered in outcomes research; something not usually associated with clinical trials.

4. Available investigational product information

A clear example of where GCP doesn’t directly translate to outcomes research.

5. Trials should be scientifically sound and described in a clear protocol

To support the diligence of outcomes research a clear scientific rationale should be presented in a study protocol. This is a clear differentiator from more informal market research, where detailed protocols are not typical. However, protocols for outcomes research are generally shorter than those required for clinical studies. Sponsors can sometimes require a qualitative research study protocol to be drafted using a clinical trial protocol template, within which many of the sections are not applicable. It is here that some differences in methodology should be recognised and procedures adapted accordingly.

6. Trials should be conducted according to a protocol that has received ethical approval

As noted above, ethics approval is highly relevant for outcomes research. However, in some countries and/or institutions a ‘waiver’ can be applied to research with no medical intervention or a less rigorous, ‘proportionate’ review is required.

7. Medical care given to, or decisions made on behalf of, subjects should be the responsibility of a qualified physician

Outcomes research does not typically involve any form of medical intervention or change to a patient’s care. Therefore, no medical care or decisions should be provided by the study staff, or taken within the remit of the study.

8. Individuals involved in conducting the trial should be qualified to perform the task

An entirely relevant guideline: all individuals working on outcomes research studies should have appropriate qualification and training. However, this training is more likely to be in the form of experience of running outcomes research studies or having the relevant educational qualifications, rather than medical qualifications.

9. Freely-given, informed consent should be given by every participant

As a cornerstone in any research study, it is critical that informed consent is obtained for all participants; the need for this is as important as in clinical research.

10. ‘Trial’ information should be recorded, handled, and stored in a way that allows accurate reporting, interpretation and verification

In accordance with data protection rules, all data should be carefully managed and reported. However, the requirements are not quite to the strict level of clinical trials.

11. The confidentiality of subject records should be protected

All data should be handled according to local data protection laws. However, in comparison to market research, which is usually conducted on a ‘double-blind’ basis, in the interests of transparency outcomes research subjects may know the name of the sponsor commissioning the research without jeopardising the validity of the research.

12. Investigational products should be used according to good manufacturing practice

Another element of GCP not directly relevant to outcomes research.

13. Systems with procedures that assume the quality of every aspect of the trial should be implemented

Quality is a key element of outcomes research and an important part of upholding the standard of the field.


This whistle-stop tour of GCP hopefully highlights the fact that, while most principles apply to outcome research, the extent and manner in which they should be considered must be tailored to the study.

So where next? Highlighting these points shows there is a need for further discussion. Arguably there would be value in creating some guidelines for the application of GCP principles in outcomes research studies, to ensure a consistent approach. Currently, while outcomes research is generally conducted to high ethical and clinical standards, lack of clarity regarding the degree to which processes must be implemented can lead to inefficiencies and delays.

Related categories

Thought Leadership