Patient-Reported Outcomes: A New Era in Research and Clinical Care (April 12, 2018)
Earlier this month, patient-reported outcomes (PROMs), a topic which has seen a timely resurgence, was crowned this year’s NephMadness champion. Given that PROMs relate to my own research interests, I was excited to learn from current researchers and thought leaders in the field, fill in my own knowledge gaps, and refine my future areas of focus.
Dr. Michelle Richardson opened the session by defining PROMs as “any report of a patient’s health status that comes directly from the patient, without interpretation of the patient’s response by a clinician.” She highlighted the numerous benefits of incorporating PROMs into our clinical care and research, such as enhanced physician-patient communication, improved shared-decision making, nuanced predictions of disease trajectory, and even increased physician satisfaction in work.
Dr. Richardson provided a helpful overview of familiar and less familiar PROMs used by the nephrology community, including the strengths and limitations of each measure. While the 36-item Short Form Survey (SF-36) has comparative data in other chronic illnesses such as diabetes and lung disease, it is static. The Patient-Reported Outcomes Measurement Information System (PROMIS) measures from the National Institutes of Health have yet to be extensively used in kidney disease and have an increased respondent burden.
However, these measures are gaining attention due to fewer floor and ceiling effects and their availability as computer adaptive testing, meaning that subsequent questions are given to the patient based on his or her prior responses. Dr. Richardson also highlighted the utility of solitary questions, such as the “time to recovery” question, which asks patients the length of time it takes them to recover from a full dialysis session. Responses to this question have been shown to be stable over time and correlate well with other measures such as fatigue.
The next talk, led by Dr. Matthew Rivara, focused on psychometrics, or the methods used in the development and validation of PROMs. I was struck by the time-intensive nature and iterative process required to create a new PROM, ranging from creating a conceptual framework based on literature review, defining the target population, using qualitative analyses and focus groups for concept elicitation, generating question items, using parallel language development for diverse populations, using cognitive debriefing techniques to administer the questions, and field testing. It seems that the level of rigor needed to develop a PROM worthy of being a clinical trial endpoint often takes years.
Dr. Rivara also reviewed ways that the accuracy and precision of PROMs are tested. He defined validity as the extent to which an instrument measures what it intends to measure, and further subdivided validity into content validity and construct validity. Content validity, or the extent to which a survey item represents a broad variety of concepts (weakness, fatigue, lack of energy) requires an initial, extensive, qualitative phase. Construct validity, or the extent to which scores are indicative of the theoretical construct on which the instrument is based, requires that measurements be taken over time and that these measurements correlate well with existing, related PROMs. Dr. Rivara defined reliability as the ability of a measure to deliver the same result in a similar set of circumstances over time. This requires internal consistency, or an agreement between questions that make up a multi-item scale, as well as test retest reliability, or agreement when the same test is administered more than once over time.
The final portion of the session was led by Dr. Jennifer Flythe, who emphasized the need to move beyond ‘quality of life’ as the main PROM used in trial endpoints in kidney disease. She highlighted upcoming initiatives from the Centers for Medicare and Medicaid Services Technical Expert Panel to develop PROMs that focus on patient life goals and patient safety, as well as the Meaningful Measures Initiative, which aims to identify the highest patient-centered priorities for quality measurement and improvement.
The session closed with an open forum for discussion with Derek Forfang (@ForfangDerek), a kidney patient and passionate advocate for patient-centered kidney disease care. This was the most illuminating part of the session for me, as Mr. Forfang revealed that though he had scored poorly on his own Kidney Disease Quality of Life (KDQOL) assessment, he personally felt that he had an excellent quality of life. He expressed interest in including questions related to travel, social relationships, and dialysis-related infections as part of a new quality of life measure, and he stressed that it was not the length of the instrument but rather its degree of meaningfulness and alignment with personal goals that determined whether patients would be willing to participate in its use.
Despite the “promise” (no pun intended) of PROMs, significant challenges remain. We need to be cognizant of floor and ceiling effects, the burden of measure length and time, recall bias, time-sensitive changes in measurement, cultural nuances in our patient population, the need for more disease-specific measures, and most importantly, the degree of meaningfulness of each measure to the daily lives of our patients. We do have our work cut out to rigorously develop and test new and improved PROMs, but given their clear, wide-ranging implications in clinical care, research, and policy, they should continue to be utilized in clinical trials and made a part of patient-centered kidney care.