The FDA's draft guidance, "Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims,"
marks a formal positioning of the agency's stance regarding patient reported outcomes (PROs) in clinical trials. In publishing
its guidance, the FDA has implicitly acknowledged the role of PROs and does so explicitly with statements such as: "Patients
provide a unique perspective on treatment effectiveness," "Some treatment effects are known only to the patient," "Self-completed
questionnaires...are often preferable to the clinician administered interview and rating." In addition, much like the EMEA,1 the FDA has acknowledged the importance and advantages of the collection of electronic patient reported outcomes (ePRO).
The guidance is an important document in that it provides a single point of reference for the agency's current thinking on
the topic. As such, it should be essential reading for anyone engaged in the development or implementation of PROs. Nevertheless,
a poll of delegates participating in a Web seminar on the topic in late October 2006 revealed that only 19.4% of participants
had read it. Although the results were conditioned by the mix of delegates, they do suggest that it would be valuable to begin
by putting the document into context.
In this article, we will explore the role of electronic modalities in collecting PROs and consider the implications for validation
when changing from paper-based to electronic administration.
Context The draft was published together with an invitation for public comment and it is anticipated that the guidance will be finalized
sometime in 2007. An initial deadline of April 2006 was set for public comment, but the FDA has indicated that it still welcomes
observations. Further, if there is a desire to use PROs while the guidance is still in draft form, the FDA indicates it is
willing to discuss PRO use and promises a timely and constructive review, within the context of the planned study protocols
and intended label claim(s).
The FDA is keen to note that the draft and final versions are intended as guidance only. It recognizes that alternative methods
and opinions do exist, and these will be acceptable when presented with suitable and appropriate analysis and evidence. The
agency also indicates that it does not plan to approve individual PRO scales as such, certainly not in isolation of the intended
conceptual framework (line 190), protocol, and indication.
All methods of capturing PROs are referenced within the guidance: paper-based or electronic site-based questionnaires; home-based
assessments; standard instruments; and assessments administered by interview (line 322). In doing so, the guidance goes further
than any previous document from the FDA in recognizing the role of interactive voice response (IVR) and other electronic modalities.
The document does reference what Sponsors should do and should avoid in relation to ePRO:
- Control of source documents: Exclusive control by the Sponsor is expressly discouraged.
- Database backup: To reduce the risk of data loss or data corruption.
- Investigator accountability: The investigator remains responsible for the subject's data. PRO data should not be modified
without the investigator's involvement.
- Access to unblinded data: Investigators must have access to patient care and safety data and be able to monitor the activity/timeliness
of patient reporting. Precautions must be taken to prevent unblinding, such as through access to patient reported side effects
data, during the study.
- Inability of FDA inspector to inspect data: Systems should be fully 21 CFR Part 11 compliant, with a clear audit trail.
Role of electronic modalities
Clearly, FDA is not saying explicitly that paper is not an acceptable method for collecting PROs, but it has indicated several
advantages associated with ePRO. Indeed, electronic modalities such as IVR address two major concerns previously expressed
by regulatory agencies regarding PRO data:
Conflicting/missing/ambiguous data. The limitations of paper-based diaries have been well documented and can result in erroneous data collection, invention of
data, and low (true) compliance levels.2,3
