Case 1 · Reducing hospital readmissions (Camden, NJ)

The goal. Reduce hospital readmissions among “super-utilizers” — patients with very high healthcare use.

The system. Identify the highest-risk, highest-cost patients from hospital data. For each one, a team of nurses, social workers and community health workers visits after discharge to coordinate outpatient care and connect them to social services.

The reception. Nationally celebrated. Profiled in the New Yorker. Expanded to cities across the country. Before-and-after numbers looked strong.

The result. A randomized controlled trial of ~800 patients (NEJM, 2020) found no difference in 180-day readmissions between the program group and usual care — about 62% in both.

Your task

  1. The model identified high-risk patients accurately. So what failed?
  2. Why did the before-and-after numbers look good?
  3. Was the highest-risk patient the one whose readmission was most preventable?
  4. What would you change — in the model, the intervention, or the evaluation?

Sources are on the main case studies page — read them after the session.