Title: 10 "Do's" and "Don'ts" for Cost Studies: Lessons learned so far from the COSP MultiSite
110 "Do's" and "Don'ts" for Cost StudiesLessons
learned (so far) from the COSP Multi-Site
- Brian Yates, Danyelle Mannix, Gazala Ansari,
Michael Freed - American University
- Washington, DC
2Cost Study Dos Donts 1 to 5
- Do treat costs and outcomes similarly
- Do apply the same scientific criteria to costs
and outcomes - Dont assume that everyone agrees
- Do make a distinction between Input Costs and
Output Costs - Do include monetary outcomes in the research plan
3Cost Study Dos Donts 6 to 10
- Dont assume program records will capture costs
- Do include all important costs
- Dont assume costs are like outcomes in all ways
- Do fund cost studies adequately!
- Do get support for the Cost Study
41. Do treat costs and outcomes similarly (in
research)
- Costs deserve the same attention as outcomes
- Like outcomes, costs require considerable time
and effort to measure well - Costs can be measured and analyzed
scientifically, too - ? Costs can be, and should be, included in
research and policy on evidence-based practices!
52. Do apply the same scientific criteria to costs
and outcomes
- Use experimental designs random assignment when
studying costs - Measure costs for each consumer separately
- Statistically test apparent differences in
- costs
- outcomes (effectiveness monetary benefits)
- cost-effectiveness cost-benefit ratios
63. Dont assume everyone agrees
- Some would ignore costs entirely
- Some would wait until outcomes are
well-researched - Some would simplify costs, ignoring some
- Some assume that all consumers cost the same
(only average or slot cost) - Some would only look at costs, and ignore outcomes
74. Do make a distinction between Input Costs
and Output Costs
- Input Costs
- the value of resources invested in the program,
e.g., - providers time
- consumers time!
- space, equipment, materials
- Output Costs
- costs avoided (cost offset)
- income produced
85. Do include monetary outcomes in research plan
- Cost savings
- Health care costs avoided
- Criminal justice costs avoided
- Direct benefits
- Income produced from increased employment
- Profit from consumer-involved enterprises
- Monetary outcomes can be derived
- from some measures of effectiveness
96. Dont assume program records will capture costs
- Just as you wouldnt assume that program records
can capture important outcomes - Accounting records dont usually capture cost
data with reliability and validity - Program records may record each use
- but not the identity of the consumer who used
the program! - Program records may under- or over-estimate
service use
107. Do include all important costs
- Omitting important outcomes is bad science, but
- omitting important costs (and benefits) can alter
the balance of costs to benefits - and can reverse policy and funding decisions!
11Critical Costs to Include
- Pharmacy
- Housing
- Administrative (not just direct costs)
- The time of
- Consumers! (waiting, as well as participating)
- Family and friends of consumers
- Providers (providing, waiting for no-shows)
- Transportation of consumer to site, other
12Include nonmonetary costs
- Volunteered and donated resources
- Critical resources in many COSs, TMHSs
- Calculate, present costs figures both ways
- Including volunteered donated resources
- Excluding volunteered donated resources
- This is specific case of entertaining multiple
perspectives on costs - just as there are multiple perspectives on
outcomes
138. Dont assume costs are like outcomes in all
ways
- Costs are not a subset of outcomes
- Dont assume that data collection systems for
other variables - can be used for Cost variables as well
- Costs occur as use occurs
- Use can be sampled, but perhaps should not be
- Dont think that research ethics always
generalize to costs
149. Do fund cost studies adequately!
- Insist that Cost Studies at each site are funded
adequately - e.g., health economist or accountant or
accounting contact at each site - Site contacts should communicate regularly with
Cost Study director - Allow adequate time for cost follow-up
- e.g., detecting service use following baseline
interviews may require years of post-baseline
inquiry with programs
1510. Get support for the Cost Study!
- Support from CC Director, POs, Consumers makes or
breaks the cost study - Jean Campbell, Crystal Blyler, Betsy McDonel
Herr, Zahira Duvall, Janine Elkanich, Bonnie
Schell, Nancy Erwin - Colleagues are essential
- Jeff Merrill, Matthew Johnsen, Sally Rogers,
David Lambert, Sarah Diwan - Associates are absolutely necessary
- Danyelle Mannix, Gazala Ansari, Mike Freed
16For more information
- Brian T. Yates, Ph.D.
- Department of Psychology
- American University
- 4400 Mass. Ave., N.W.
- Washington, DC 20016-8062
- briantyates_at_earthlink.net
- (please remember to type t between brian and
yates) Thanks! - 202-885-1727