In fiscal year 2009, 24% of the VA's health care budget was spent on patients with psychotic disorders. VA has invested heavily in evidence-based mental health practices to improve the quality of care for this population, but too often these services are not utilized by patients with mental illness. The HSR&D QUERI EQUIP study ("Implementing Effective, Collaborative Care for Schizophrenia") used evidence-based quality improvement implementation strategies to increase use of two evidence-based practices (weight services and Supported Employment) which targeted high priority outcomes in mental health.
EQUIP utilized routine assessment through a patient-facing kiosk which identified patients who needed services. Assertive care coordination and weight were implemented; existing work services were utilized. When comparing intervention to usual care sites, EQUIP resulted in 1) significantly increased utilization of weight and Supported Employment services; 2) significantly less weight gain, but not increased competitive employment; and 3) no difference in health related quality of life.
The purpose of this study was to (1) estimate the average VA cost-consequences of EQUIP during the 12 months following patients' enrollment, and (2) estimate the cost-effectiveness of the EQUIP intervention, compared to usual care, in achieving improvements in weight, employment, and health related quality adjusted life years, if applicable.
EQUIP was a clinic-level controlled trial. Within four VISNs, eight specialty mental health programs were assigned to intervention or control. 801 patients with schizophrenia were enrolled; 661 (82.5%) completed the 12-month follow-up.
We supplemented EQUIP records on dedicated staff time and other resources utilized for program implementation and operation with data on intervention-related healthcare utilization and costs, weight and employment outcomes, and health-related quality of life. The majority of these data were obtained in EQUIP study interviews or were extracted from the Decision Support System National Database Extracts.
Direct cost estimates included costs of personnel, kiosk equipment, weight group supplies, classroom space, and utility expenses. Personnel costs included time costs for the employment specialist, staff providing the weight services, and the site facilitator leading quality improvement. Personnel time was valued using site-specific average salaries plus fringe benefit and payroll taxes. Kiosk costs were drawn from study records and assumed to depreciate over a 5-year lifespan.
To develop estimates of staff effort we examined monthly implementation call minutes, care coordinator monthly activity logs, VISN coordinator weekly field notes, monthly quality improvement call minutes, quality improvement final reports, and study staff email correspondence and notes. After reviewing these records, two intervention sites that had comparatively more complete data were chosen as exemplary sites and used for the cost analysis.
In addition to the direct costs of EQUIP, we also examined indirect costs associated with use of other VA outpatient health care services during EQUIP. Indirect costs were examined by comparing use of outpatient mental health, primary care, and rehabilitation services between EQUIP and usual care sites, controlling for spending during the 6-month period prior to EQUIP. EQUIP was expected to increase spending for MOVE! and vocational services, as the program was designed to foster connections to these programs. We did not know how EQUIP would affect use of other services.
The one-time cost of the setting up EQUIP was $14,385 per site. These initial costs include the effort and salary of the staff who prepared the kiosks and weight groups, marketed the project, and engaged Supported Employment. These costs varied by site, from $11,624 to $17,147, depending on differences in staff effort.
The average annual cost of delivering EQUIP was $1,075 per patient. These delivery costs varied by site from $695 to $1,620 per patient, per year. The variation was largely attributable to site differences in the Supported Employment specialist's effort. Supported Employment was largely unavailable to study patients at one site, and as a result Supported Employment costs at that site were close to zero.
VA outpatient health care costs were greater, by $1,195 per person, during the 12 months of EQUIP compared to the 12 months prior to EQUIP ($11,629 per person during EQUIP compared with $10,434 per person before EQUIP). However, in the usual care group, 12-month post-baseline VA outpatient health care costs exceeded pre-baseline costs by $1,810 per person ($14,344 post-baseline versus $12,534 pre-baseline). As a result, on average EQUIP was associated with a savings of $615 per person compared with usual care. Most of these savings were the result of lower costs in EQUIP for intensive outpatient mental health services, such as partial hospitalization, Psychosocial Rehabilitation and Recovery Center (PRRC), and Mental Health Intensive Case Management (MHICM) services. Thus, our estimates suggest that more than half of the $1,075 per person cost of EQUIP was offset by lower outpatient costs for other services.
After offsetting savings are taken into account, EQUIP was associated with a net increase in VA health care costs of $460 per person during a one-year period. A substantial fraction (perhaps half) of these additional costs were attributable to weight and employment program services that patients utilized more frequently as a result of EQUIP, and the remainder was attributable to the costs of EQUIP's care coordination functions.
VA leaders are less inclined to invest in implementation of evidence-based practices without data on cost impacts. Within mental health, there have been a very limited number of cost-effectiveness analyses of evidence-based practices and those that have been conducted only include the cost of the evidence-based practice and not the effort to ensure utilization.
EQUIP results in higher utilization of weight and employment services, which are two high priority services. Both types of services are critical for the mental health recovery of individuals with schizophrenia but, until EQUIP, were inadequately utilized by the target population. The average costs of EQUIP, $1,075 per person per year, were modest by comparison to individuals' total expenses for outpatient health care services, which were over $14,000 per person per year. EQUIP also was associated with a reduction in use of expensive services, such as PRRC (approximately $13,000/per person annually) and MHICM (approximately $15,000/per person annually). On net, the cost of EQUIP was only $460 per person per year. This amount does not include the one-time expense of setting up an EQUIP service ($14,385 per site). EQUIP has shown that routine assessment, care coordination, and an investment in marketing and training of staff enables better outcomes for patients with schizophrenia at a price that is reasonable. The cost impact of EQUIP is on the low end of the range of costs of implementing evidence-based services for mental disorders.
- Cohen AN, Slade E, Hamilton AB, Young AS. Cost of an effective quality improvement intervention in specialty mental health. Poster session presented at: AcademyHealth Annual Research Meeting; 2015 Dec 14; Washington, DC.
- Cohen AN, Slade E, Hamilton AB, Young AS. Great, we've improved care quality, but how much did it cost? Paper presented at: VA HSR&D / QUERI National Meeting; 2015 Jul 8; Philadelphia, PA.
- Cohen AN, Slade E, Hamilton AB, Young AS. Assessing the cost of an effective quality improvement intervention in specialty mental health. Poster session presented at: AcademyHealth Annual Research Meeting; 2015 Jun 15; Minneapolis, MN.