Care Coordination for High-Need, High-Cost Commercially Insured Patients: A Randomized Clinical Trial.

IMPORTANCE

High-need, high-cost (HNHC) patients account for 5% of the US population yet represent nearly half of health care spending.

OBJECTIVE

To evaluate whether national care coordination could reduce health care cost and utilization in a commercially insured HNHC population.

DESIGN, SETTING, AND PARTICIPANTS: This national, 2-arm randomized clinical trial with intention-to-treat and instrumental variable analyses included patients aged 18 years or older who were defined as HNHC according to a proprietary model (in the top 5% of spend within a rolling 12-month claims utilization window and projected to remain in the top 5% over the subsequent 12 months). Patients were randomized from January 2018 to October 2019. Data were analyzed from January 1 to December 31, 2024.

INTERVENTION

Participants were randomized monthly 60:40 to telephonic care coordination from a registered nurse, including medication review, a barriers-to-care survey, addressing urgent coordination needs (eg, patient unable to fill prescriptions), development of a case management plan addressing identified clinical risk factors, and establishing an outreach time frame, or to the control group with usual care. The nurse contacted patients over the 60 days after enrollment until all risk factors included in the management plan were addressed.

MAIN OUTCOMES AND MEASURES

The main outcomes were mean monthly emergency department visits, inpatient hospitalizations, and total plan cost (medical and pharmacy) over 12 months following the index date, defined as the enrollment date for intervention participants or the randomly generated synthetic enrollment date for nonparticipants. Outcomes were examined separately among patients with diabetes.

RESULTS

The analytic sample included 93 379 HNHC patients with a mean (SD) age of 46 (12) years (54% female). In intention-to-treat analyses, there were no differences between groups in mean (SE) monthly emergency department visits (0.033 [0.001] for control vs 0.033 [0.001] for treatment; mean difference [SE], 0 [0]; 95% CI, -0.001 to 0.002; P = .69), inpatient hospitalizations (0.009 [0] for control vs 0.010 [0] for treatment; mean difference [SE], 0.001 [0]; 95% CI, 0-0.002; P = .06), or cost (total: $2507 [$32] for control vs $2568 [$26] for treatment; mean difference [SE], $60 [$41]; 95% CI, -$20 to $140; P = .14). In the instrumental variable analyses and in the subsample with diabetes, no evidence of statistically significant reductions in these outcomes were found.

CONCLUSIONS AND RELEVANCE

In this randomized clinical trial of a national care coordination intervention, neither health care cost nor acute care utilization was reduced in the intervention group compared with the control group. The results emphasize the challenges of improving efficiency of care in a complex HNHC population with escalating health care costs.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT04415515.