W-9: Exploring Accuracy of Abdominal Pain Reporting with and Without Specific Instruction
Poster Presenter
Alyssa Peechatka
Clinical Science Advisor
ERT United States
Objectives
Regulatory authorities such as the FDA and EMA are increasingly seeking data that are provided in the form of reliable patient reported outcomes (PRO), however results obtained using PRO instruments can vary according to instructions given to patients.
Method
905 participants completed an online survey in which they were asked to rate abdominal pain based on a presented scenario. Demographic data were also collected.
Results
Participants were presented with the following scenario: “You are participating in a clinical trial related to stomach conditions and are asked to rate the level of your average abdominal pain over the past 24 hours, on a scale from 0-10 (where 0=no pain and 10=worst pain you can imagine). During the past day, your abdominal pain changed. In the morning you experienced no pain, but by the evening (when you complete your report) your abdominal pain had increased to an 8.” 37% (343 participants) chose the correct answer representing average pain (4) and 63% chose an incorrect answer. Specifically, 48% chose the worst pain experienced/pain experienced at the time of report (8), 3% chose the minimum pain expressed in the scenario (0), and the remaining 12% selected another numerical answer between 1-10.
A small amount of educational information was then presented: “When reporting your average abdominal pain, do not report the worst pain you experienced that day or your pain severity at the time you complete the report. Rather, report the average level of pain you experienced over the course of the entire day; the average is about midway between the highest and lowest level of pain you experienced that day.”
The same question was asked again and 303 participants changed their responses from an incorrect response to the correct response (4), increasing the percentage of participants who chose the correct answer to 67%. Thirty-six participants changed their responses from correct to incorrect. An exact McNemar’s test demonstrates that this represents a significant increase in correct responses, following the provision of additional information (p< 0.0001). Only 10% of participants continued to select the worst pain experienced/pain experienced at the time of report (8).
Conclusion
PROs continue to be the industry standard for collecting endpoint data, however the reliability of the data is dependent on the instructions provided. This can be particularly problematic when capturing ratings that might vary across a 24-hour recall period, including abdominal pain. These data indicate that, without specific instruction on how to rate average abdominal pain, 63% of participants in this study would have provided inaccurate data. Given that providing even brief instruction significantly increased the accuracy of responding, these data show that it is not inherently obvious to patients how to rate “average pain.” However, this problem can be remedied in the form of interactive training videos on the trial device, where context can be provided to questions at any time. It is theorized that with further instruction and training, patients would better understand the PRO task asked of them.
