Fundamental Outcome Measurement: Selecting Patient Reported Outcome Instruments and Interpreting the Data they Produce.

Over the past 40 years literally thousands of generic and disease specific patient reported outcome (PRO) instruments have been developed. While most were developed for a specific study and were never used again, there is still the question of how manufacturers and others should select a PRO instrument for a study. These studies may be clinical pivotal trials or observational tracking studies to support therapy response. Formulary committees also need to be able to interpret PRO data to make decisions about whether to accept claims for therapy response. It is possible to argue that the many different approaches to outcome measurement have resulted from the lack of agreed methodologies. However, a more likely explanation is that the authors have failed to apply the axioms of fundamental measurement when creating their measures. The result is a plethora of ordinal PRO instruments that inform little about the impact of interventions. Clinical trials rarely report PRO data. Where they do, analyses are generally restricted to (for example) changes in the experimental group's scores. Comparisons between the treatment and placebo groups or between active groups are infrequently reported, most likely due to the failure of the instrument to show differences or changes in outcome. This is unfortunate as it means no assessment is made of the value that patients gain from the intervention. This commentary is intended to make researchers and formulary committees aware of the issues that need to be addressed when selecting PRO instruments for a study or evaluating publications and claims for therapy response. The latter is crucial as reported data influence the selection of medicines and healthcare products. In the latter case a particular concern is with PRO claims embedded in simulation models.