Economic Grand Rounds: Can States Implement Involuntary Outpatient Commitment Within Existing State Budgets?

New York enacted Kendra’s Law in 1999, authorizing a form of outpatient commitment termed “assisted outpatient treatment” (AOT). A number of other states, including North Carolina, had previously adopted similar statutes with the aim of improving outcomes for individuals with severe mental illness. Research has shown that outpatient commitment must be sustained for at least six months and combined with frequent outpatient services to be effective (1). New York’s statute specified six months as the minimum duration for an initial court order. The law also specified that treatment must include an array of intensive services. These requirements might have been seen as a burdensome “unfunded mandate.” However, the legislature also authorized annual appropriations of $32 million to support the program and $125 million to expand statewide services for all consumers (2).

This funding was pivotal in the somewhat stormy acceptance of the AOT program. However, in the current era of constrained state budgets, appropriating new funds might have a chilling effect on other states’ interest in such a statute. For example, in 2010 New Jersey Governor Christie temporarily suspended implementation of a new outpatient commitment statute because of a lack of funding (3).

Contrary to the expectation of increased costs, recent evidence has demonstrated improved clinical outcomes and substantial net cost savings associated with New York’s AOT program (2,4). The cost impact study compared costs for selected participants in New York City for the year before and two years after AOT initiation and found that participation produced net cost savings of 50% in the first year and an additional 13% in the second year; in five other counties, savings of 62% in the first year and an additional 27% in the second year were noted (4). The New York study might persuade some states to implement outpatient commitment. However, it begs the question of whether a state could achieve positive outcomes without new funding.

A study of outpatient commitment in North Carolina permits estimation of cost impacts when a program is implemented within existing funding (1). The study was a randomized controlled trial of outpatient commitment conducted in the public mental health system in North Carolina between 1993 and 1996 (1). The county-based outpatient commitment programs involved were operated within existing state and county allocations and revenue sources. Below we estimate total program and treatment costs for persons randomly assigned to outpatient commitment, and we compare those costs to the costs for persons who received usual care.

The study population included consumers involuntarily hospitalized and scheduled to be discharged on an outpatient commitment order. At discharge, consenting consumers were randomly continued in outpatient commitment or randomly released from it. All participants received case management and were followed for a year. The initial outpatient commitment order was 90 days or less, but it could be renewed in subsequent court hearings.

The analysis presented here applied utilization rates derived from study data and combined them with estimated current unit prices for services to calculate and compare net costs in four groups: those randomly assigned to outpatient commitment, those randomly released from outpatient commitment (control group), those under outpatient commitment whose court orders were not renewed (nonrenewed group), and those under outpatient commitment whose court orders were renewed and who received on average more than six months of court-ordered outpatient treatment (renewed group). The renewed group most closely approximates patients under court order in New York.

Average utilization rates of participants were calculated for relevant inpatient and outpatient services. Criminal justice system costs were calculated for arrests and days of incarceration. Civil court costs and administrative overhead for outpatient commitment were also included, prorated to the number of days on outpatient commitment. All costs were adjusted to 2008 dollars to make estimates comparable to the cost impact study of Kendra’s Law (4).

Relevant program inputs and outcomes were monetized, including legal services, program administration, use of mental health services, hospitalizations, arrests, and incarcerations. For unit costs of mental health services, we used published data from the Medical Expenditure Panel Survey (5) to estimate average Medicaid rate payments for inpatient and outpatient services in 2008 dollars. For arrest unit costs, we used an estimate derived from a study of persons with severe mental illness and comorbid substance use disorders (6). For prison costs, we used data published by the North Carolina Department of Corrections (7). For jail costs, we used an average ratio of jail cost to prison cost derived from New York data (4) and applied this ratio to North Carolina prison costs. Court costs for outpatient commitment hearings were obtained from published estimates of court costs in the North Carolina Court System (8). Outpatient commitment program costs were estimated from the average salaries of the administrative staff positions assigned to oversee persons on outpatient commitment during the study.

Average per-person utilization rates and estimated unit costs were combined to produce annualized estimates of net costs, across all categories of cost, for the four groups. Monthly utilization events were multiplied by the estimated unit cost per event. Monthly costs were summed and annualized to obtain a total estimated cost per year.

Table 1 presents service costs for the four groups. Total annual costs for mental health services were $42,309 for the control group, $44,662 for the outpatient commitment group, $53,246 for the nonrenewed group, and $23,675 for the renewed group. Criminal justice costs were $1,207 for the control group, $754 for the outpatient commitment group, $787 for the nonrenewed group, and $596 for the renewed group. Costs for the outpatient commitment program were zero for the control group, $1,094 for the outpatient commitment group, $825 for the nonrenewed group, and $1,651 for the renewed group. Total annual costs were $43,516 for the control group, $46,510 for the outpatient commitment group, $54,858 for the nonrenewed group, and $25,922 for the renewed group who received sustained periods of outpatient commitment.

Total annual costs were roughly $3,000 higher for the outpatient commitment group than for the control group. However, when commitment was extended for six months or more (renewed group), total costs were $17,594 lower than for the control group, representing savings of approximately 40%.

Outpatient commitment in North Carolina appears to be relatively cost-neutral when relevant costs for persons on outpatient commitment are compared with costs for persons not on outpatient commitment, regardless of the duration of commitment. Outpatient commitment of six months or more, combined with frequent provision of outpatient services, appeared to result in a substantial cost offset, with decreases attributable mainly to reduced hospitalizations. The 40% savings for extended commitment is comparable to savings in New York’s AOT program (4).

These findings suggest that states seeking to implement outpatient commitment can do so without new funding for this specific purpose if they have adequate services to provide to these consumers. Further, if outpatient commitment is sustained for six months or more, substantial cost savings may be realized, although the savings may affect budgets other than that of the mental health system. Of note and concern, in the absence of new funding for outpatient commitment, some resources would have to be diverted from other priorities. As a result, this analysis should not be interpreted as an argument against new funding for such programs.

Several limitations suggest caution in interpreting these estimates. The data are based on a trial conducted in the mid-1990s and updated in 2008 dollars. Thus results may not generalize to current programs. It could also be argued that observation bias or experimental intervention may have influenced program performance and cost estimates. However, the study was a real-world effectiveness trial. The service systems of the mid-1990s are somewhat different than current systems. On the whole, participants in the North Carolina study probably received less intensive community services than they would today. In addition, secular trends toward reduced hospitalization could also have affected these findings, but even if hospitalizations were reduced by half, use of outpatient commitment for six months or longer would still reduce total costs.

Because savings are realized within a year, such outpatient commitment programs appear to be at least cost-neutral. These findings suggest that states with adequate services for consumers under outpatient commitment may implement a program without new funding for this specific purpose.

Partial funding for this project was provided by a research grant to the authors from the Stanley Research Foundation.

The authors report no competing interests.