Lead/Presenter: Jean Yoon, Resource Center - HERC
All Authors: Yoon J (HERC)
Chang E (COIN Los Angeles)
Park A (HERC)
Zulman D (COIN Palo Alto)
Stockdale S (COIN Los Angeles)
Ong M (COIN Los Angeles)
Asch S (COIN Palo Alto)
Schectman G (Patient Care Services)
Rubenstein L (COIN Los Angeles)
Objectives:
Targeting high-risk patients for intensive outpatient management may help reduce inefficient use of health care resources. We enrolled high-risk patients in an intensive outpatient program (PACT Intensive Management or PIM) by random allocation and examined cost and utilization impacts.
Methods:
Primary care patients in five Veterans Affairs (VA) medical centers were identified based on hospitalization risk ( >= 90th percentile of the Care Assessment Need Score) and a recent acute care episode and were randomized to PIM versus usual care in PACT from 9/1/2014 to 6/1/2015. In addition to routine primary care in PACT, PIM patients could opt to receive services including goals assessment, health coaching, medication reconciliation, care coordination with specialists, and home visits from a dedicated team staffed by primary care providers, social workers, psychologists, nurses, and other support staff. We measured patients' VA inpatient and outpatient utilization and costs 12 months prior to and following randomization from VA administrative data. We used differences-in-differences models to estimate the effect of PIM on utilization and costs.
Results:
A total of 1101 patients were allocated to PACT, and 1105 patients were allocated to PIM. There were no significant differences in patient demographic characteristics between the two groups, and patients had a mean age of 63 years, 90% were male, and 68% were not married. Mean costs per patient prior to randomization were $32,189 (SD = $43,717). PIM patients received a mean of 6 (SD = 11) PIM encounters over 12 months. Both PIM and PACT groups had decreased inpatient costs/patient from the pre to post-randomization periods, and differences for PIM were not significant relative to PACT (-$2933, P = 0.189). The PIM group had a significant increase in outpatient costs/patient post-randomization compared to PACT ($2302, P = 0.001) driven by higher utilization of primary care, home care, telephone care, and telehealth (all P < 0.05). Total health care costs per patient were similar pre and post randomization, and there were no significant differences associated with PIM (-$630, P = 0.791).
Implications:
High-risk patients with access to an intensive management program had increased utilization of outpatient care without an increase in total costs.
Impacts:
While intensive outpatient programs may shift costs, the impact on efficiency needs further examination.