Patient-Reported Data: Can Healthcare Execs Trust It?

Rosner

Patients are accurate self-reporters of certain healthcare utilization events such as readmissions and selected complications, according to a new study.

For the study, published in the Journal of Medical Internet Research, patients were surveyed about healthcare utilization and complications post-discharge through HealthLoop, an automated digital patient engagement (DPE) application already in use as part of routine remote guidance and telemonitoring care provided by their physicians.  Patient responses to readmission and complications questions were compared for accuracy to claims data from study partner, Anthem Blue Cross.

The study found that for certain utilization and complication events for which there may be more ambiguity, patients are moderately accurate. It further found that when such surveys are served in an automated DPE platform, patients are highly willing to participate (76.8% completion rate).Post-discharge benefits

Post-discharge benefits

The study was conducted to ascertain whether patients might be able to take a partnership role in healthcare quality improvement post-discharge by providing healthcare organizations a closed-loop feedback cycle to understand what happens to patients post-discharge.

“If it is determined that patients can be accurate and engaged self-reporters, then healthcare organizations bearing financial risk post-discharge might consider engaging the patient as a partner and solicit self-reports so that hospitals can identify gaps in quality and opportunities for quality improvement,” says lead study author Ben Rosner, MD, PhD, CMIO, HealthLoop Inc, Mountain View, California, and a practicing internal medicine doctor, Department of Hospital Medicine, Kaiser Permanente.

The data suggest that patients are highly accurate self-reporters for major events such as readmissions and significant complications such as pulmonary embolism. They are moderately accurate for other events such as emergency room/urgent care visits and complications such as deep vein thromboses and surgical-site infections. 

According to a task force led by Rosner, the Office of the National Coordinator defines patient-generated health data (PGHD) as “health-related data created, recorded, or gathered by or from patients (or family members or other caregivers) outside the clinical setting to help address a health concern.”

“This is a fairly broad umbrella that can include data such as health-related data from fitness trackers, glucometers, and other devices, but also includes data patients generate in surveys having to do with quality of life, patient satisfaction, patient-reported outcome metrics (PROMs), and other things pertaining to their health and healthcare,” says Rosner. “Increasingly, PGHD are used for engaging patients in their own health, and for learning about events of relevance to healthcare providers in the 99.9% of the time people are outside of the four walls of the clinical encounter.”

In an emerging era of value-based care, healthcare organizations are increasingly bearing financial risk for what happens to patients post-discharge, according to Rosner.

“By some estimates, for example, nearly 20% of Medicare beneficiaries who were hospitalized in the early 2000s were readmitted within 30 days,” he says. “And nearly 34% were readmitted within 90 days. These readmissions were associated with over $17 billion in potentially avoidable costs. A significant problem, however, is that the very organizations that are now bearing financial risk for readmissions and complications post-discharge have limited insight into these events.”

According to Rosner, this is due in part to leakage (e.g., presentation of patients to facilities other than the index facility for complications and readmissions) that is reported to occur in 31% to 65% of cases with some rates as high as 87.5%, and to lag (the time it takes for payment reconciliation with the index institution to occur, at which point the index institution might finally learn of a readmission or a complication that happened elsewhere).

“If we were to look at other industries that employ quality assurance, there is always a temporally proximate feedback cycle used-for example, testing of a product on the assembly line-that lets the stakeholder know in near-real time about potential quality concerns,” he says. “However, in the last mile of healthcare, there is no such closed loop, and when a failure occurs-such as a complication or readmission-the hospital does not necessarily know about it in a manner that permits a timely improvement in the system. Now imagine what could be accomplished if patients could report about readmissions and complications accurately, and in near real-time back to the index hospital. Suddenly the index facility would have actionable information to close the quality gap in the last mile of healthcare, the post-discharge period.”

This is the first study of its kind to rigorously evaluate patient self-report, including the ability to account for leakage across all practice locales, and the ability to account for bias (none was found) among patients who did not respond, according to Rosner.

“It also demonstrates that patients are very willing to participate, with completion rates of 76.8%,” he says. “These findings offer healthcare organizations a new way to close gaps in care for which they may bear risk, and to simultaneously engage the patient as a partner in his or her own care.”