This Web site is a component of the SAMHSA Health Information Network

Enter Keywords Choose Website This Site Mental Health at HHS All Mental Health

Online Publications Order Publications National Library of Medicine National Academies Press Publications Homepage

printer friendly page e-mail this page bookmark this page shopping cart  current or new account

This Web site is a component of the SAMHSA Health Information Network.

ESTIMATING the COST of PREVENTIVE SERVICES
in Mental Health and Substance Abuse Under Managed Care

Introduction & Background

Dorfman's literature review (2000), commissioned by the Center for Mental Health Services, Substance Abuse and Mental Health Services Administration, screened more than 800 studies and yielded a large number (54) of published research studies documenting effective types of preventive interventions with a link to substance use (alcohol, tobacco, and illicit drugs) or mental health that have been published in peer-reviewed journals, including those addressing the cost impact of the interventions.1 Of the 54 studies outlined in the report, 21 were used to form the basis of the six recommended interventions, with a minimum of two supporting studies per service and as many as six studies supporting some services.

The following criteria were used for including an article in that literature review:

• The intervention fit within the definition of primary prevention, secondary prevention, or one of the three classifications (universal, selective, or indicated interventions) in the Institute of Medicine's Model of Prevention.
• The study evaluated or reviewed one or more interventions designed to prevent a substance abuse (i.e., alcohol, tobacco, or illicit drug) or mental health problem or a behavioral health intervention designed to prevent an associated health problem-such as a low-birthweight baby-resulting from use of tobacco.
• The intervention was implemented with human subjects; or the intervention model was applied to a hypothetical group of human subjects.
• The intervention was implemented in a medical care or referral setting.
• The intervention was shown to result in cost savings, cost offset, or neutral impact on the cost of care; or the intervention was shown to be effective, with the potential result of cost savings, cost offset, or neutral impact on the cost of care.
• The study was published from 1964 through 1999 in the English language, in a peer-reviewed journal. The use of these criteria yielded 54 published studies that were screened against the following additional criteria to be included in the recommendations for consideration by an MCO:
• The intervention's effectiveness has been demonstrated by two or more studies included in this review.
• The intervention's appropriateness for provision in a managed care or referral setting has been explicitly stated or is apparent.
• The intervention's feasibility for MCO coverage from a cost perspective has been documented or suggested (Dorfman, 2000, p. 19).

The application of these final criteria yielded six interventions supported by 21 articles. All the articles in Dorfman's review "document positive outcomes of preventive interventions in relation to mental health or substance abuse. Thirteen of the 54 articles address the cost of the intervention. . . . Overall, these studies represent the body of science-based evidence that interventions designed to prevent substance abuse and mental health problems [as well as some medical problems] have been proven effective and, in some cases, have produced net cost savings or have offset costs that would have been incurred absent the preventive intervention" (p. 13).

Based on these criteria and 21 studies (a minimum of two per intervention), Dorfman recommended the following six interventions, spanning the developmental stages from birth to old age (pp. 19-21):

1. Prenatal and Infancy Home Visits for High-Risk Mothers

These articles focused on women with high-risk pregnancies, who included pregnant women under 19 years of age, unmarried, or of low socioeconomic status; and low-birth-weight infants born prematurely. The timing of periodic home visits varied, ranging from the prenatal period to age 3. Home visits were made by nurses in one project and by a psychology graduate student teamed with a Comprehensive Education Training Act aide in another. One study focused home visits on maternal functioning and other studies on the training of mothers to stimulate their infants. Fewer subsequent pregnancies, greater spacing between births, less alcohol and drug impairment, and less child abuse and neglect were among the significant findings for mothers who received home visits. Higher weight, better scores on motor developmental tests, and reduced incidence of mental retardation were among the significant findings for infants whose mothers received intervention.

2. Targeted Cessation Education/Counseling for Smokers

Subjects in these articles included a "birth cohort" of women who smoked during pregnancy, pregnant smokers recruited through county maternity clinics, and a hypothetical group of male and female smokers receiving routine medical care. Interventions consisted of a 15-minute counseling session with a nurse or health educator supplemented by written materials and two followup telephone calls; a 15-minute counseling and skill development session with a trained health counselor supplemented by clinical patient reinforcement, social support, newsletter information, and mention in a prenatal education class; and 4 minutes of physician advice to quit smoking supplemented by a self-help booklet and a 1-year followup visit. The birth cohort model study estimated savings of $3.31 in the cost of caring for low-birthweight infants in a neonatal intensive care unit for every dollar spent on smoking cessation intervention. In the hypothetical patient group, brief physician advice was estimated to increase the cessation rate at 1 year by 2.7 percent. In the maternity clinics, the intervention produced a 14.3-percent quit rate, compared with an 8.5-percent quit rate in the control group.

3. Targeted Short-Term Mental Health Therapy

In a study of children up to age 15 who received one to six targeted behavioral therapy sessions with their parents from doctoral-level pediatric psychologists or predoctoral clinical psychology interns, those with behavioral problems (such as aggression, noncompliance, tantrums, excessive fears, or sleep or mealtime disturbances) reduced their medical encounters by almost one-third, while those with toileting problems reduced their medical encounters by almost one-half. In another group of individuals who sought short-term psychotherapy from a psychiatrist or other registered psychotherapist on an approved list of community practitioners, index cases significantly decreased days of medical hospitalization compared with matched controls.

4. Health Promotion Through Self-Care Education

Five of the six cited studies were conducted in managed care settings; the sixth was worksite-based. The interventions addressed health promotion and self-care issues that encompassed substance use and mental health. Interventions included group education workshops led by a nurse practitioner and supplemented by a self-care guide and videotapes; written materials, a telephone information service staffed by a nurse coordinator, and an individual health evaluation and planning conference with a trained nurse; computer-based, serial, personal health-risk reports supplemented by individualized recommendation letters and written materials; access to a self-care center; one-on-one education sessions with physicians; and slide-tape shows. The results were an estimated 28-percent savings in laboratory costs and 24-percent savings in x ray costs between experimental and control groups; and a 17-percent decrease in total medical visits and a 35-percent decrease in minor illness visits in experimental versus control groups. Also, significant improvements were noted: decreases in health-risk behaviors, including smoking, alcohol use, and reported stress; decreases ranging from 7.2 to 24 percent in ambulatory physician visits; and a decrease of 15 percent in total medical visits in the experimental group compared with controls. In one study, for every dollar expended on the program, an estimated $5 was saved in direct health care costs for physician visits and hospital days.

5. Presurgical Educational Intervention With Adults

In one of the cited studies, the intervention consisted of a workshop to enable staff nurses to provide psychoeducational care to adult surgical patients. Interventions described in the other two articles included giving patients information about what to expect; skills training to help patients prevent complications or reduce anxiety; psychosocial support with a health care provider to reduce anxiety or enhance ability to cope with hospitalization, supplemented with printed and taped materials; and visits to patients by an anesthetist before and after surgery to provide information and self-care guidance. Interventions were associated with less use of sedatives, antiemetics, hypnotics, and narcotics as well as earlier discharge from the hospital.

6. Brief Counseling/Advice to Reduce Alcohol Use

The four articles reviewed 47 studies conducted in the United States and internationally. Interventions included between 5 and 15 minutes of advice or counseling on reducing alcohol consumption provided by physicians, nurses, psychologists, or other professionals. In some studies, subjects also received a workbook or informational or self-help materials. Other intervention components included followup visits or telephone calls for reinforcement. Significant reductions in alcohol consumption were documented: among middle-aged men, 14 percent, and middle-aged women, 31 percent; among adults 65 and older, 40-percent reduction.

Cost Models of These Six Interventions

This publication reports on the results of cost simulations of these six types of interventions. The interventions were modeled specifically for their potential cost to an MCO when provided to enrolled members; the cost does not include cost offsets or potential savings to the plan. Based on a careful review of all the studies dealing with each category of intervention, financial spreadsheet models were created to reflect the variables that would drive the cost of each type of intervention.

The first part of this report summarizes the calculations and is intended to be useful to MCOs considering implementing the recommendations given in the previous report by Dorfman, Preventive Interventions Under Managed Care: Mental Health and Substance Abuse Services, published by the Center for Mental Health Services at SAMHSA. It should be useful to decisionmakers and purchasers of covered benefits packages, managed care organizations, employers, public payers, State health administrators, and mental health services researchers in financing.

Because of the multiple studies supporting each intervention and the fact that no one study provided all of the details in the design and operation of each kind of intervention within each overall category, a generic model was designed for each category using specifications from more than one study. For example, the category of "Prenatal and Infancy Home Visits for High-Risk Mothers" was supported by 11 different publications out of the overall 54 Dorfman reviewed. These 11 studies described a wide variety of interventions, ranging from home visits to classroom-based interventions to on-the-job training of young mothers within a child day-care center. Of these 11 studies, 7 did not provide any information specific to the cost of the intervention, 4 did not specify enough detail about what was done to establish a clear cost model, 3 formed the basis for the final recommendation, and only 2 with the most details were used to design the cost model for the home visit intervention.

General Method of Cost-Based Modeling

Each model was created based on the three primary variables that drive the cost of any intervention or service within a defined population:

Multiplying the value of each of these variables yields the "Total Cost" of the intervention. The total cost divided by the total number of eligible persons (covered lives, or members) yields the "Cost per Eligible Person." Because most managed care plans calculate their cost of health care benefits on a PMPM basis (in order to determine their monthly premium), the total cost of each type of intervention was calculated on a PMPM basis.

Figure 1 summarizes these three variables. Their multiplicative relationship can be used to calculate the total annual cost. PMPM costs are calculated based on total annual cost, divided by the average annual number of members, divided by 12 months.

Of course, the values of each of these primary variables may be the result of still other variables. For example, the cost of a unit of service delivered by an employee will depend on variables such as the following:

• The employee's salary and fringe-benefit costs
• The average hours in the year that the employee spends in "productive" service (i.e., the cost of a productive hour of service)
• The cost of direct expenses necessary to establish that particular service (e.g., equipment, supplies)
• The cost of G&A expenses necessary to support the general organization and that particular service (e.g., insurance, utilities, administrators' salaries)

Again, these variables may be broken down further and related to one another through formulas such as the following:

Cost of a Staff Person's Hour of Service = [(Salary) + (Fringe Benefits) + (Direct Expenses) + (G&A expenses) / Total Productive Hours]

Ongoing Service Costs on an Annualized Basis

Because managed care plans operate on a 1-year budget and members enroll through their employers annually, every model was designed as a "1-year" cost model and as if the MCO provides each intervention on an ongoing basis. Researchers commonly carry out their research on preventive interventions in a controlled environment designed for systematic data collection over a well-defined but limited time period. This method of operation understandably is somewhat different than the way a managed care plan would provide similar services.

Another major difference involves how the intervention is organized and delivered. Some of these interventions were offered as a group service to a well-defined but limited group of cases over a defined time period. Most clinical health services, however, are provided to individual patients on an ongoing basis as they occur intermittently throughout the year.

Furthermore, many of the interventions described in the literature were organized, carried out, and then terminated during a fixed time of less than 1 year. If a managed care plan were to undertake offering such an intervention, it would offer the service on an ongoing basis, probably starting up a new "class" or cohort of persons to receive the intervention at 6-month intervals. Still, other interventions do lend themselves to being offered on an ongoing, case-by-case basis, such as brief advice to reduce the use of alcohol.

In either example, a managed care plan is unlikely to want to incur one-time startup costs, establish some expectations among its enrolled members, and then terminate the program after a short time. Therefore, the model for the cost for a managed care plan assumes that the intervention will be offered throughout the year. At the same time, the cost models do include the startup costs that would have to be incurred during the first year of the service. After the first year, such costs could be omitted in the cost calculations so that the intervention would cost less in the ensuing year than it did in the initial year. The studies on which the models were based were all research with evaluations; obviously, the extra costs associated with carrying out the research, such as tracking outcomes on control group members, were not incorporated in the cost models.

Accounting for Variation

For a detailed discussion of how the model took into account variation due to chance (probability) and unknown values (uncertainty), refer to the Technical Appendix.

Accounting for Scenarios

A cost model of each of these preventive interventions was designed and then simulated under four different scenarios. At one extreme was the Least Expensive Scenario, defined as the scenario that assumed the least expensive values from among a reasonable range of values for each cost driver (e.g., prevalence, intensity of the intervention units/time period, staff salaries). At the other extreme was the Most Expensive Scenario, which assumed the most expensive values from among the same range of reasonable values.

Table 1 summarizes the relative values of the various cost drivers in each model.

The critical dependent variable for each model was the PMPM cost, which is the total cost divided by the total annual membership months (average membership per month over the year, multiplied by 12) of the managed care organization (MCO) that was assumed to sponsor the intervention.

Each model was simulated over 1,000 iterations using Monte Carlo simulation (Winston, 1996), resulting in a distribution of 1,000 possible PMPM cost values for each intervention.

The Technical Appendix presents the details of the four PMPM cost distributions for each intervention and the details of the assumed values of the cost drivers used as inputs to each model.

1  Dorfman reported that "the majority of the studies obtained for this review were located through Internet Grateful Med V2.3.2, which includes 11 databases: MEDLINE, HealthSTAR, PREMEDLINE, AIDSLINE, AIDSDRUGS, AIDSTRIALS, DIRLINE, HISTLINE, HSRPROJ, OLDMEDLINE, and SDILINE. The following search terms were used: cost behavior, cost-benefit analysis, cost-effectiveness, cost savings, evaluation studies, health education, health maintenance organizations (HMOs), health promotion, intervention studies, managed care programs, mental health, patient education, prevention, preventive health services, preventive medicine, primary prevention/economics, and substance abuse" (pp. 11-12).

Table of Contents | Previous | Next