Care Coordination for High-Need, High-Cost, Commercially Insured Patients
Key Takeaways
- A primary intention-to-treat analysis found no statistically significant differences in emergency department visits, inpatient hospitalizations, or total plan cost. The diabetes subsample and the instrumental-variable analysis also showed no statistically significant reductions.
- Most sensitivity analyses were consistent with the main findings, although some showed a statistically significant increase in inpatient hospitalizations.
Researchers conducted a national, pragmatic, 2-arm randomized trial in US adults aged 18 years or older with commercial insurance. Eligibility required high-need, high-cost status under a proprietary model that identified top-spend patients in claims and projected continued top-spend status. Patients were assigned monthly in a 60:40 ratio to telephonic care coordination or usual care from January 2018 through October 2019. The insurer randomized 157,102 patients, then excluded 63,723 after randomization, leaving 93,379 for analysis; mean age was 46 years, and 54% were female.
After an introductory mail or email message, staff made up to 4 telephone attempts to reach patients assigned to the intervention. A registered nurse conducted medication review, a clinical assessment to identify clinical and social risk factors, urgent coordination support, and a patient-centered case management plan with a defined outreach schedule. Nurses then continued follow-up outreach for 60 days, and some patients were referred to a higher level of case management when needs extended beyond that window. Engagement, defined by the insurer as an open dashboard, a completed barriers-to-care assessment, program enrollment, or a device reading, was 26% in treatment and 5% in control. The authors described the approach as relatively low intensity, with limited telephone contact over a short outreach window.
The coprimary outcomes were mean monthly emergency department visits, inpatient hospitalizations, and total plan cost over 12 months. Emergency department visits were 0.033 versus 0.033 per month, with a mean difference of 0, 95% CI -0.001 to 0.002, and P=.69. Inpatient hospitalizations were 0.009 versus 0.010, with a mean difference of 0.001, 95% CI 0 to 0.002, and P=.06. Total plan cost was $2507 versus $2568, with a mean difference of $60, 95% CI -$20 to $140, and P=.14. None of the coprimary endpoints was reduced in the primary analysis.
Investigators noted low intervention uptake and treatment contamination, which complicated interpretation of the intention-to-treat analyses and motivated the instrumental-variable analyses. Process outcomes and program fidelity could not be assessed, and the insurer did not provide detailed information on postrandomization exclusions. The authors framed the null findings as specific to this intervention, population, and timeframe, and linked them to a low-intensity telephonic model, fragmented care, and targeting challenges.