Why should 2020 be any different from earlier years? We start off with more debunking of hot initiatives. In this care, a study reported in the New England Journal of Medicine examines a “hotspotting” intervention for superutilizer patients. (NEJM Article) Since it is well-known that a small group of patients accounts for much of total spending, programs to identify and manage those patients are an obvious way to control spending. The program covered by this research was implemented in Camden, New Jersey and was developed by local providers to identify patients who were hospitalized and appeared to have high ongoing utilization. The program got highlighted in an article by Atul Gawande, which guaranteed that it probably was hype. And so the research to actually conduct a rigorous evaluation of its effects demonstrates. Patients included in the program had at least two hospital admissions in a 6 month period, at least two chronic conditions, and other characteristics like drug use, social problems or use of multiple medications. Once enrolled in the program, they were attended to by a team that included nurses, social workers, health coaches and community health workers. The team made home visits, coordinated care, managed medications and scheduled and made sure patients got to physician visits. All this hand-holding isn’t cheap. The control group consisted of similar patients who just got usual post-hospital discharge services.
Information was collected on the intervention and control groups regarding utilization and the primary outcome was hospital readmission within 180 days after the hospitalization that led to enrollment in the program. Hospital days, spending and mortality were also measured. The trial population averaged 1.8 hospital admissions in the 6 months before enrollment compared to less than .1 admission in the general Camden population. On average the intervention group had 7.6 home visits and 8.8 calls from program staff, and they had other indicators of intensive interaction. The 180 day readmission rate was 62.3% in the intervention group and 61.7% for the control group. Can you say “had no effect”. There was no effect on the secondary outcomes either. Now the interesting thing is that initially the intervention got a lot of positive press because readmission rates for the patients in the program appeared to decline significantly. They did, but when a real analysis was done, it turns out that readmission rates were declining in general for all patients, which isn’t surprising given all the attention paid to reducing readmissions. All the spending and effort put into the intervention didn’t create any further decline in readmissions, or improvement in other outcomes, beyond that which was already occurring in the general population. Now, as usual I will caution that this doesn’t mean the program isn’t worthwhile. It might make a significant difference in patients’ quality of life and longer term may ensure better health. But it is expensive, so a decision has to be made about how much money is worth spending on patients who aren’t likely to show big improvements in outcomes, and whose behavior is likely contributing significantly to their health problems.