Community Support

the Evaluation Center@HSRI Toolkit

Estimating Per Unit Treatment Costs for Mental Health and
Substance Abuse Programs

I. INTRODUCTION

This toolkit is designed for program evaluators and others who wish to understand how unit costs of specific mental health services are estimated, especially where data available to carry out estimations vary in both quantity and quality ¹. Per unit costs are the key to determination of per person expenditures. This toolkit has been developed to offer some practical methods for arriving at unit cost estimates when it is not possible to gain access to the cost accounts needed to calculate precisely the per unit cost of providing a specific service from a specific site. We also help the reader to gauge the precision of such estimates.

By "unit cost estimate," we mean an approximation of the monitized resources required to provide a specified amount of a particular service to a person with a mental disorder. Ultimately, estimates of unit costs can be linked to specific amounts of service provided a given individual and aggregated for different time periods, episodes of care, or groups of special populations. Unit cost estimates may be used for such purposes as:

Describe the mean per person costs of a set of specific services used within a network of programs or systems ². These per person costs might be compared to similar costs in other systems.
Develop cost effectiveness or cost-benefit ratios, using per person mean costs of a set of specific services used within a network of programs or systems as the denominator and the person's non-monitized or monitized outcomes as the numerator.
Simulate per person costs for purposes of planning under changing service system configurations.

Cost evaluation studies require four separate steps to arrive at per person costs. This toolkit is designed to take you through Steps 1 and 2 (the determination of per unit costs) so that you can then use your own data to complete steps 3 and 4. The toolkit is organized to be consistent with these steps and includes sections on each of the four steps.

Step 1. Identify and list all the different service types(e.g., inpatient hospitalization, individual therapy), by provider (specific hospital or agency) for the consumers in the study. This list becomes the starting point for knowing just what costs are to be determined.

Step 2. Determine the cost per unit of treatment for each of the service types, by provider (e.g., $x per diem or $x per hour, for each treatment/service for each provider from the list generated in Step 1).

Step 3. Specify the type and amount of service used in one year for each consumer (number of hospital days in Hospital A, number of hospital days in Hospital B, number of therapy sessions at Agency C) preferably in a spreadsheet format. Multiply the amount of service used for each service type by the cost per unit for that service for each consumer in the study. Repeat this step for every service used from every provider for each consumer in the study. The next step is to summarize group-level costs.

Step 4. Use these person-level data: (1) to determine cost-effectiveness ratios for specific programs or systems of care or (2) to be the foundation of simulations projecting cost differences under different conditions.

The toolkit has six sections:

1.  this introductory overview of the contents and intended audience,
2.  an examination of the tradeoffs in estimating per unit costs between level of precision attained and level of effort expended,
3.  a discussion of what services to include and how to define them,
4.  a discussion of how to construct per unit cost estimates,
5.  a discussion of how to develop a format for summarizing person and group level costs,
6.  a brief guide to economic concepts that are the foundation of cost evaluations.

The Appendix provides two articles and other materials for the toolkit user.

In the US, we cannot offer a national cost per unit for a set of commonly used services because of the wide variation across the country in treatment settings, cost of living, provider types and other resources used. Instead, we provide guidelines and case examples that illustrate how to estimate service-specific per unit costs. The emphasis is on how to estimate the costs for the particular programs used by clients participating in the evaluation. Almost all the examples we provide come from our work in Massachusetts and thus may need modification in other states.

One good example of the construction of per unit costs in countries where universal health care results in standardized expenditure and activity information is a series of reports from the Personal Social Services Research Unit at the University of Kent, Canterbury, United Kingdom. The most recent volume of Unit Costs of Health and Social Care was compiled by Ann Netten and Jane Dennett (1997). Two pages from this report are included in the Appendix. One page documents the per unit costs of a clinical psychologist and one page the per diem costs of acute NHS hospital services for people with mental illness. This approach does not assume that the services provided across the country are exactly the same, but for purposes of cost studies, national estimates are based on national expenditure data. In both cases, the estimates include a "London multiplier" in recognition that the capital city has expenditures higher than other areas of the country.

We assume the reader has some clinical, research or administrative experience in the mental health field (e.g., understands that "case management" is a generic term for service that can vary widely in intensity and function, depending on the program philosophy, the size of the case-load, the training and experience of the staff and the needs of the individuals served), and some knowledge of accounting and cost-finding (e.g., how administrative costs are stepped down). We do not expect the reader to have a background in economics or econometric analyses.

Table of Contents | Next