Nonmedical workers and mobile technology can play a pivotal role in risk prediction.
An estimated $355 billion is wasted in the United States every year as a result of poor care coordination, failures in care delivery, and overtreatment, including as much as $44 billion spent on unplanned hospital readmissions.
A recent article in Perspectives in Health Management explored a possible solution to prevent readmissions—a quality improvement (QI) project that uses 2 readily available resources: mobile devices and nonmedical workers who provide home care and community support for the aging population.
The nonmedical workers leveraged by this project included “personal care attendants, home health aides, home meal delivery drivers, health coaches, community health workers, social worker case managers, and other providers of essential nonmedical functions.”
These workers have frequent interactions with at-risk patient populations and represent 8 of every 10 hours of paid services provided to the elderly and people with disabilities.
Authors studied a QI project implemented by an elder services agency in Massachusetts, designed to reduce 30-day hospital readmissions using mobile predictive analytics technology. Its strategy was to address what they had identified as a major driver of readmission: suboptimal communication between nonmedical and clinical staff.
The QI project stood to lower the cost of care transition and risk assessment by utilizing nonmedical workers at a pay rate about 70% to 90% lower than that of a nurse or doctor, respectively. Its success would hinge on the ability of a nonmedical worker-operated mobile app to stratify patients by readmission risk and determine when a patient’s care should be escalated to a clinician.
Authors performed a retrospective review of data from the Elder Services of Merrimack Valley (ESMV) QI project from July 10, 2013, to April 23, 2014. The study cohort included 2027 hospital-discharged patients enrolled in the ESMV care transition program with a mean age of 73 years, including Medicare fee-for-service patients.
The ESMV care transition program, called the mHealth Transitions Model, was adapted from the Care Transition Intervention (CTI). Whereas CTI uses transition coaches with nursing degrees, ESMV uses lay coaches with high school education and provides them with standard training consistent with CTI, another 1.5-hour training on their mobile technology, and centralized supervision by a nurse case manager.
The Care at Hand app asked a series of 15 questions to be answered by the nonmedical health coach during each patient encounter. Surveys used lay language and were designed to take no more than 2 to 5 minutes. A proprietary algorithm predicted likely risk factors for each patient and changed survey questions accordingly with each administration. If the system detected an elevated risk of readmission, it generated a real-time alert to a nurse care manager.
The nurse care manager used a different component of the software to triage the patient and assist the health coach with care coordination within 24 hours of the alert.
Results of Care Transition Quality Improvement
The study found that observations from nonmedical workers collected and analyzed by mobile app produced risk scores that were able to predict readmissions, and that the predictive nature of these scores was improved with nurse oversight and input. Observations from nonmedical workers significantly predicted readmission at 30 days and beyond 30 days after discharge.
Comparatively, risk scores from nonmedical worker observations associated with 30-day readmission rates with an odds ratio (OR) of 1.12 (95% CI; 1.09 — 1.15), whereas risk scores from nurse observations had an OR of 1.25 (95% CI; 1.19 – 1.32).
T-tests revealed that high risk scores had a significantly higher average 30-day readmission rate (32%) than baseline (14%), mild (20%), and moderate (19%) risk scores.
Fifty-five percent of interventions in response to a risk alert involved a nurse or physician, suggesting that nonmedical workers and mobile technology can play a pivotal role in risk prediction, but skilled clinicians are still needed to triage and address elevated risk, once detected.