FDA guidance on patient reported outcomes

The value gained from understanding health outcomes from the patient’s perspective has been acknowledged increasingly in recent years.1 2 National surveys of patient experience are now a feature of NHS regulation,3 and the drugs industry now recognises the use of such outcomes in labelling (efficacy) claims to promote market access.4 5 Release of the long awaited definitive guidance from the US Food and Drug Administration (FDA) on the use of patient reported outcome (PRO) measures to support labelling claims6 marks a coming of age for PRO measurement.

After publication of the draft guidance in 2006,5 there was much speculation about the meaning and implications of the FDA’s stance on the development and use of PRO measures. The draft offered guiding principles to those conducting and supporting industry sponsored clinical trials, and it was largely welcomed as encouraging the adoption of scientific standards. However, several aspects of the draft guidance were seen as contestable (for example, the use of ”minimally important differences”) or lacking in detail (for example, recommendations for use of ”conceptual frameworks” and “endpoint models”). Consequently, the draft guidance generated more questions than answers about the nature of PRO research and the level of evidence needed to demonstrate good scientific practice. Indeed, detailed responses were submitted by more than 50 concerned stakeholders (academics, health professionals, policy makers, professional organisations, and industry representatives7).

In its revised 2009 guidance, the FDA has clarified firstly that PRO data can be used to measure the risk of treatment (for example, side effects, inconvenience) as well as benefit (risk was not previously mentioned); secondly, that the development of PRO measures is an iterative process with no single correct approach; and thirdly (by conspicuous omission) that certain details included previously lacked an evidence base (such as the use of minimally important differences to help interpret changes in PRO scores).

In the detail of the final guidance, several key themes emerge.6 Firstly, emphasis is placed on defining the role that a PRO end point is intended to play in the clinical trial, so that chosen outcome measures are matched appropriately to specific treatment objectives. Secondly, the fundamental role of a PRO measure’s content validity (the extent to which an instrument measures the concept of interest) is expanded and clarified.6 The FDA acknowledges that different approaches may be needed to ensure that existing, modified, or newly created PRO measures are appropriate to support product labelling claims. The guidance also states that sponsors need to supply evidence that a PRO instrument adequately measures what is claimed by documenting its development history. This would detail item (question) design and modification, measurement properties (including reliability, validity, and responsiveness), how well patients understand questionnaire items and response (answer) options, and the appropriateness of the PRO measure to the patient group. In addition, sponsors need to specify the method of collecting the data (for example, pencil and paper v electronic) and other details such as the recall period (the timeline over which the respondent is asked to reflect).

Finally, the FDA recommends submission of a PRO evidence dossier (a file including full details of the rationale for the use of the PRO measure, its development history, and validation). Previously, recommendations about the content of a PRO evidence dossier were vague but are now detailed fully in the appendix, offering clear and practical guidance about the evidence to be submitted in support of a PRO based labelling claim.

So what are we to make of the FDA guidance? Some have warned about the dangers of PRO measures being used to adjust hospital reimbursements,8 and commercial forces driving the inclusion of PRO measures in drug labelling. However, any initiative that puts patients at the centre of treatment decisions is welcome providing these measures are selected and interpreted appropriately and related precisely to a priori hypotheses regarding treatment outcome. Moreover, the FDA guidance will end the unscientific practice of including any vaguely relevant PRO instrument in a clinical trial at the eleventh hour. Instead, a clear strategy for the inclusion of PROs in clinical trial programmes will be required, just as for other clinical end points. If drug companies believe there are valid PRO claims to be made for their compounds, they will need to gather evidence to support the inclusion of PROs as early as phase I-II trials and give careful consideration to the selection and development of appropriate PRO measures.

The number of successful PRO based product labelling claims fell in the years after the release of the draft guidance (compared with the preceding five years),9 which may indicate increased challenges faced by drug companies wanting to show treatment benefits from the patient perspective. Furthermore, successful claims were largely symptom based, possibly highlighting the difficulty (or perceived difficulty) of securing a claim based on more contentious (though, arguably, more patient centred) concepts, such as wellbeing and treatment satisfaction. Careful recrafting means that the 2009 guidance now offers greater clarity and flexibility regarding what constitutes acceptable supporting evidence. In so doing, the FDA has simultaneously raised the stakes and improved the odds of drug companies securing a greater number and variety of successful PRO based labelling claims.

The FDA’s guidance explicitly states that it does not cover the use of PRO measures beyond evaluation of medical product claims, but the basic principles are undoubtedly relevant for a range of stakeholders: know what you want to measure, why you want to measure it, and be sure that the PRO instrument you choose measures what is intended. Most importantly the guidance promotes good scientific research practices that should be adopted by any researcher or clinician, thus ensuring that outcomes reported from the patient’s perspective are robust and meaningful.