When Clinical Endpoints Aren’t Enough
For many indications there are clear objective clinical endpoints, such as changes in antibody levels, tumour size and chemistry or haematology lab parameters, which can be utilized to measure the efficacy of a new treatment. However, for a number of indications the endpoints of primary interest are concerned with the individual’s impression of how the treatment is affecting them, whether that be in evaluating symptoms such as pain or itching, or evaluating how the treatment impacts a patient’s day-to-day life. These are all typically measured via Patient Reported Outcomes (PROs). Careful design and analysis are critical for these types of endpoints, often benefiting from specialist statistical expertise early in development.
PROs take a wide variety of forms from the SF-36 used to measure quality of life, and the C-SSRS used to ascertain the presence and severity of suicidal ideation, to the single question Patient Global Impression of Severity (PGI-S) or Change (PGI-C). While there are a lot of questionnaires available, often it is difficult to identify one that accurately or completely reflects the evaluation of the disease or symptom of interest and the decision could be made to develop a new questionnaire to better capture this.
Regulatory Expectations for PROs
The FDA guidelines on the use of PROs to support claims for a medical product state that ‘A PRO instrument, like physician-based instruments, should be shown to measure the concept it is intended to measure, and the FDA will review the evidence that a particular PRO instrument measures the concept claimed.’
As part of meeting these requirements, any new instrument used in the evaluation of products should be formally validated and shown to provide a reliable and consistent measure of the concept of interest.
Building Validated PROs with Patients as Priority
It is recommended that PRO instrument development and validation involve the patients, through interviews and focus groups, to help inform the language of the questions and the ease of use of the PRO instrument. Statistics can also be useful in the development of these instruments and can be used to help define the optimum recall period, identify redundant questions, or classify domains within the instrument.
The Role of Statistics
The next stages of PRO development rely heavily on statistics to examine the reliability and validity of the PRO instrument, and evaluate its ability to detect change. If these elements cannot be demonstrated satisfactorily, the PRO instrument will not be accepted as an endpoint.
Reliability and Consistency
It is important to evaluate the stability of responses over time when no changes are expected, to ensure that the instrument provides a reliable result. Similarly, if the instrument consists of multiple questions it is important that the responses within a questionnaire are consistent when multiple questions are intended to measure the same concept.
Validity
Additional considerations are whether the responses to the new instrument are consistently related to other instruments used in the field of interest and whether the relationship is as expected. This could be looking at the overall instrument, or domains within the instrument. Examining responses in known groups (which can be defined clinically or via a different validated PRO) can be used to look at whether the response differs between the groups as expected.
Change
Assessment of the ability to detect change focuses on evaluating the ability of the instrument to measure any degree of change when a known change in the concept of interest has occurred. The minimally important clinical difference (MCID) can be identified using anchor-based methods and instruments for which an MCID has already been established.
How We Can Help
If you are considering developing a new PRO instrument or are already using one which has not been validated, we can help. We have experience with the various statistical aspects of PRO validation, in both the design of studies (observational and interventional) and the methods that can be used to demonstrate the validity of the PRO instrument and ensure that it will be acceptable for use as a key endpoint in your studies.