Washington State Institute for Public Policy
Primary care in integrated settings (Veteran's Administration, Kaiser Permanente)
Adult Mental Health: Serious Mental Illness
Benefit-cost estimates updated December 2016.  Literature review updated May 2014.
Behavioral health settings (mental health and substance abuse treatment centers) provide primary care for patients on site or nearby. This collection of studies was conducted at Veterans Administration facilities or facilities of Kaiser Permanente where patients might have more ready access to primary care than community-based treatment centers.
BENEFIT-COST
META-ANALYSIS
CITATIONS
The estimates shown are present value, life cycle benefits and costs. All dollars are expressed in the base year chosen for this analysis (2015). The chance the benefits exceed the costs are derived from a Monte Carlo risk analysis. The details on this, as well as the economic discount rates and other relevant parameters are described in our Technical Documentation.
Benefit-Cost Summary Statistics Per Participant
Benefits to:
Taxpayers $467 Benefits minus costs $1,120
Participants $105 Benefit to cost ratio $5.90
Others $200 Chance the program will produce
Indirect $576 benefits greater than the costs 52 %
Total benefits $1,348
Net program cost ($229)
Benefits minus cost $1,120
1In addition to the outcomes measured in the meta-analysis table, WSIPP measures benefits and costs estimated from other outcomes associated with those reported in the evaluation literature. For example, empirical research demonstrates that high school graduation leads to reduced crime. These associated measures provide a more complete picture of the detailed costs and benefits of the program.

2“Others” includes benefits to people other than taxpayers and participants. Depending on the program, it could include reductions in crime victimization, the economic benefits from a more educated workforce, and the benefits from employer-paid health insurance.

3“Indirect benefits” includes estimates of the net changes in the value of a statistical life and net changes in the deadweight costs of taxation.
Detailed Monetary Benefit Estimates Per Participant
Benefits from changes to:1 Benefits to:
Taxpayers Participants Others2 Indirect3 Total
Crime $0 $0 $0 $0 $0
Property loss associated with alcohol abuse or dependence $0 $0 $1 $0 $1
Labor market earnings associated with illicit drug abuse or dependence $38 $85 $0 $474 $597
Health care associated with illicit drug abuse or dependence $29 $5 $25 $19 $78
Health care associated with general hospitalization $75 $4 $65 $38 $182
Health care associated with psychiatric hospitalization $286 $4 $64 $140 $495
Health care associated with emergency department visits $39 $7 $45 $19 $110
Adjustment for deadweight cost of program $0 $0 $0 ($115) ($115)
Totals $467 $105 $200 $576 $1,348
Detailed Annual Cost Estimates Per Participant
Annual cost Year dollars Summary
Program costs $228 2014 Present value of net program costs (in 2015 dollars) ($229)
Comparison costs $0 2014 Cost range (+ or -) 20 %
According to Saxon et al., (2006). Randomized trial of onsite versus referral primary medical care for veterans in addictions treatment. Medical Care, 44(4), 334-342, patients in the clinics with co-located primary care had an average of 1.1 more primary care visits than the comparison group in 12 months. We estimated additional cost of the program by multiplying 1.1 visits by the Medicaid enhanced payment rate for the longest primary care visit. See http://www.hca.wa.gov/medicaid/pages/aca_rates.aspx.
The figures shown are estimates of the costs to implement programs in Washington. The comparison group costs reflect either no treatment or treatment as usual, depending on how effect sizes were calculated in the meta-analysis. The cost range reported above reflects potential variation or uncertainty in the cost estimate; more detail can be found in our Technical Documentation.
Estimated Cumulative Net Benefits Over Time (Non-Discounted Dollars)
The graph above illustrates the estimated cumulative net benefits per-participant for the first fifty years beyond the initial investment in the program. We present these cash flows in non-discounted dollars to simplify the “break-even” point from a budgeting perspective. If the dollars are negative (bars below $0 line), the cumulative benefits do not outweigh the cost of the program up to that point in time. The program breaks even when the dollars reach $0. At this point, the total benefits to participants, taxpayers, and others, are equal to the cost of the program. If the dollars are above $0, the benefits of the program exceed the initial investment.

Meta-analysis is a statistical method to combine the results from separate studies on a program, policy, or topic in order to estimate its effect on an outcome. WSIPP systematically evaluates all credible evaluations we can locate on each topic. The outcomes measured are the types of program impacts that were measured in the research literature (for example, crime or educational attainment). Treatment N represents the total number of individuals or units in the treatment group across the included studies.

An effect size (ES) is a standard metric that summarizes the degree to which a program or policy affects a measured outcome. If the effect size is positive, the outcome increases. If the effect size is negative, the outcome decreases.

Adjusted effect sizes are used to calculate the benefits from our benefit cost model. WSIPP may adjust effect sizes based on methodological characteristics of the study. For example, we may adjust effect sizes when a study has a weak research design or when the program developer is involved in the research. The magnitude of these adjustments varies depending on the topic area.

WSIPP may also adjust the second ES measurement. Research shows the magnitude of some effect sizes decrease over time. For those effect sizes, we estimate outcome-based adjustments which we apply between the first time ES is estimated and the second time ES is estimated. We also report the unadjusted effect size to show the effect sizes before any adjustments have been made. More details about these adjustments can be found in our Technical Documentation.

Meta-Analysis of Program Effects
Outcomes measured Primary or secondary participant No. of effect sizes Treatment N Adjusted effect sizes (ES) and standard errors (SE) used in the benefit-cost analysis Unadjusted effect size (random effects model)
First time ES is estimated Second time ES is estimated
ES SE Age ES SE Age ES p-value
Alcohol abuse or dependence 3 684 -0.001 0.124 41 0.000 0.186 44 -0.001 0.995
Illicit drug abuse or dependence 2 643 -0.016 0.081 41 0.000 0.187 44 -0.016 0.845
Hospitalization 5 10449 -0.050 0.060 41 0.000 0.000 42 -0.050 0.403
Hospitalization (psychiatric) 1 59 -0.068 0.293 41 0.000 0.000 42 -0.068 0.818
Emergency department visits 3 753 -0.090 0.105 41 0.000 0.000 42 -0.090 0.388
Blood pressure 1 751 -0.168 0.071 41 n/a n/a n/a -0.168 0.019
Blood sugar 1 751 0.225 0.105 41 n/a n/a n/a 0.225 0.033
Death 2 98 -0.077 0.160 41 n/a n/a n/a -0.077 0.632
Cholesterol 1 751 0.071 0.122 41 n/a n/a n/a 0.071 0.562
Primary care visits 2 417 0.531 0.188 41 0.000 0.000 42 0.531 0.005
Citations Used in the Meta-Analysis

Druss, B.G., Rohrbaugh, R.M., Levinson, C.M., & Rosenheck, R.A. (2001). Integrated medical care for patients with serious psychiatric illness: a randomized trial. Archives of General Psychiatry, 58(9), 861-8.

Kilbourne, A.M., Pirraglia, P.A., Lai, Z., Bauer, M.S., Charns, M.P., Greenwald, D., . . . Yano, E.M. (2011). Quality of general medical care among patients with serious mental illness: does colocation of services matter?. Psychiatric Services, 62(8), 922-928.

Parthasarathy, S., Mertens, J., Moore, C., & Weisner, C. (2003). Utilization and Cost Impact of Integrating Substance Abuse Treatment and Primary Care. Medical Care, 41(3), 357-367.

Pirraglia, P.A., Kilbourne, A.M., Lai, Z., Friedmann, P.D., & O'Toole, T.P. (2011). Colocated general medical care and preventable hospital admissions for veterans with serious mental illness. Psychiatric Services, 62(5), 554-557.

Saxon, A.J., Malte, C.A., Sloan, K.L., Baer, J.S., Calsyn, D.A., Nichol, P., . . . Kivlahan, D.R. (2006). Randomized Trial of Onsite Versus Referral Primary Medical Care for Veterans in Addictions Treatment. Medical Care, 44(4), 334-342.

Weisner, C., Mertens, J., Parthasarathy, S., Moore, C., & Lu, Y. (2001). Integrating primary medical care with addiction treatment: A randomized controlled trial. JAMA : The Journal of the American Medical Association, 286(14), 1715-1723.

Willenbring, M.L., & Olson, D.H. (1999). A randomized trial of integrated outpatient treatment for medically ill alcoholic men. Archives of Internal Medicine, 159(16), 1946-1952.

Willenbring, M.L., Olson, D.H., & Bielinski, J. (1995). Integrated Outpatient Treatment for Medically Ill Alcoholic Men: Results from a Quasi-Experimental Study. Journal of Studies on Alcohol, 56(3), 337.

For more information on the methods
used please see our Technical Documentation.
360.664.9800
institute@wsipp.wa.gov