Study: Patient-Reported Outcomes Require Planning, Patience

New Journal of Clinical and Translational Science publication outlines implementation strategies

When implementing Patient-Reported Outcome (PRO) measures in the clinical setting, a set of planning tools holds the potential to maximize success, according to a new study led by researchers at Northwestern University, the University of Chicago, the University of Florida, and the University of Illinois at Chicago.

The findings, published in the Journal of Clinical and Translational Science, highlight the lessons learned from a collaborative team of 20 technology experts and physician-scientists.

“PRO measures are so important because they provide uninterpreted information directly from the patient, allowing providers to track patient symptoms and function over time and elevating the voice of patients in their own care,” says Therese Nelson, AM, LSW, director of sociotechnical innovation at the Northwestern University Clinical and Translational Sciences (NUCATS) Institute and lead author of the paper. “What we were able to show is that highly structured planning tools informed by sociotechnical conceptual models are an effective strategy for enabling clinics to successfully create a PRO implementation plan.”

The researchers identified a number of gaps in available PRO implementation strategies and developed three standardized planning tools. They then applied those tools systematically at each of four clinics to create pilot implementation plans. The finished collection of documents includes approximately 90 questions to ask — and answer — in order to determine optimal customization and implementation of PRO software at an individual clinic.

“Software must be implemented in order to collect electronic PROs through the electronic health record (EHR), however, in addition to installing programs, clinics must make many decisions that affect the workflow and function of that software,” says Nelson.

For example, one clinic might decide to collect PROs only in conjunction with patient visits to ensure that patients are always seen by a clinician quickly after completing PRO measures. Another clinic might decide to send requests to complete PROs on a regular schedule that does not correspond to patient visits. These clinics might both use the same software, but they would configure it differently. The researchers also highlighted the importance of human and organizational factors in effective implementation, including differing workflows, time constraints and cultures.

“It is often said that 95 percent of healthcare occurs outside of the healthcare system, but we have very little ability to measure that,” says Justin Starren, MD, PhD, FACMI, director of the NUCATS Center for Biomedical Informatics and Data Science and Chief of Health and Biomedical Informatics in the Department of Preventive Medicine. “Patient-reported outcomes give us a way to measure with precision the extent to which patients are actually getting better. We know that there is a huge placebo effect. We know that there is a huge social pressure to say that you are doing well, because you don’t want to disappoint the doctor. So, when you are on your own and the questions are well structured, that gives us an ability to know what’s happening with patients during their real lives, not just during the appointment.”

PROs can be envisioned as a new vital sign, offering clinical insight. Depending on how PROs are administered, they can entail a fundamental change in how medicine is practiced as care moves outside of the clinical visit.

“This paper is extremely relevant as interest in using PRO measures is increasing and in some cases is even mandated by particular specialties due to the ability of PROs to measure outcomes rather than process,” says Nelson. “When used together, the tools and lessons that we outlined lay the groundwork for future testing in additional real-world settings.”

The research collaboration included individuals supported by Clinical and Translational Science Awards at Northwestern (UL1TR001422), the University of Chicago (UL1TR002389), the University of Illinois at Chicago (UL1TR002003) and the University of Florida (UL1TR001427). This research was also supported, in part, by the EHR Access to Seamless Integration of PROMIS (EASI-PRO) grant U01TR001806.