Cognitive Testing for the Pain Quality Assessment Scale (PQAS)

In recent years, a growing number of investigators have noticed a curious finding when summarizing the research literature on pain treatment: Most effective pain treatments show very similar effects on global pain intensity, despite vast differences in patient samples, and in presumed mechanisms of action (e.g., Collins et al., 2000; Gutierrez-Alvarez, 2007; Wiffen et al., 2005). As the list of available treatments for various pain conditions grows, and to the extent that only global measures of pain are used to assess outcomes, it is becoming increasingly difficult for any one treatment to stand out from the others; to understand when it might be chosen over other treatments for the management of any one patient's pain problem.

One way to better understand how treatments differ is to determine the effects of those treatments on pain quality. That is, to determine if treatment A is more effective for "aching" and "deep" pain than treatment B, which might be more effective for "electrical" and "surface" pain than treatment A. By systematically measuring the effects of pain treatments on different pain qualities, it becomes possible to begin to distinguish the effects of different treatments from one another.

To do so requires investigators to include measures of pain quality as secondary outcome measures in clinical trials. The NPS and PQAS (which includes the 10 NPS items) are increasingly used in clinical trials to detect the effects of pain treatments on pain qualities. Moreover, a growing body of research supports the validity of these measures for identifying the specific pain qualities impacted by different pain treatments. However, it is not entirely clear that either measure could be used for making labeling claims. The most recent draft of the FDA's guidance for industry for patient-reported outcomes specifies a number of criteria that measures must meet in order to be able to use them for making labeling claims. Although the NPS and PQAS meet many of those requirements, they do not meet three critical ones. First, the guidelines specify that a measures' items need to be generated with patient involvement. The NPS and PQAS items were generated from (1) the clinical experience of the measures' authors and (2) reviews of the literature concerning the pain qualities most often identified by patients with various chronic pain conditions. However, these items have not yet been directly checked using patients input (to clarify that the items reflect the most important and most common pain quality domains). Second, the guidelines specify that patients should be interviewed to help determine the readability and understanding of the items. These interviews then should be analyzed, and actions taken to delete or modify items in accordance with those interviews. Finally, the FDA recommends that the instrument development process include "... the generation of a user manual that specifies how to incorporate the measure into a clinical trial in a way that minimizes administrator burden, patient burden, missing data, and poor data quality." To date, no manual has been written for the NPS or PQAS.

To address these concerns, the current proposal seeks to address the limitations of the PQAS (and because the PQAS incorporates the NPS items, this would also address the limitation of the NPS) by performing cognitive testing of the PQAS instructions and items in two samples of patients with chronic pain, modify the PQAS/NPS as needed, and write a manual for the PQAS/NPS. The procedures would allow for critical testing and improvement of the PQAS and NPS, making these measures even more useful for understanding the impact of different pain treatments. In this way, these measures could be even more useful than they already are for identifying the unique advantages of new, and already developed, pain treatments.