Optimal allocation of testing dollars: the example of HIV counseling, testing, and referral

Med Decis Making. 2005 May-Jun;25(3):321-9. doi: 10.1177/0272989X05276955.


Background: Health screening programs can be represented as a pathway of sequential processes: offering a test, obtaining consent, conducting the test, providing results, and linking to appropriate care. Using the example of HIV testing, the authors explore the optimal targeting of funds within this pathway.

Methods: The authors develop a microsimulation of HIV testing services and decompose the likelihood that an unidentified HIV-infected person will receive care into the probability of testing [P(test)] and the probability of follow-up [P(follow)] defined as returning for results and linking to care. The authors examine the clinical impact and cost-effectiveness of alternative investments in these component probabilities.

Results: At 1% undiagnosed HIV prevalence, cost-effectiveness ratios for HIV testing cluster around $33,000/QALY (quality-adjusted life year) gained. A program with a yield of 0.16 via P(test)=0.20 and P(follow)=0.80 has a cost-effectiveness ratio of $32,900/QALY compared with $36,300/QALY for a program where P(test)=0.80 and P(follow)=0.20. Interventions that improve the probability of success in later stages in the testing pathway [P(follow)] are more cost-effective than investments devoted to earlier stages [P(test)].

Conclusions: Equivalent pathway outcomes in a screening program do not confer equal value. Limited screening resources are best targeted toward returning for results and linkage among those already identified with disease rather than offering testing to additional people.

Publication types

  • Research Support, N.I.H., Extramural
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • AIDS Serodiagnosis / economics
  • AIDS Serodiagnosis / statistics & numerical data*
  • Computer Simulation
  • Continuity of Patient Care / statistics & numerical data*
  • Cost-Benefit Analysis
  • Counseling / economics
  • Counseling / statistics & numerical data*
  • Decision Support Techniques
  • HIV Infections / diagnosis*
  • HIV Infections / epidemiology
  • HIV Infections / therapy*
  • Humans
  • Investments / statistics & numerical data
  • Mass Screening / economics
  • Mass Screening / statistics & numerical data*
  • Models, Statistical
  • Patient Acceptance of Health Care / statistics & numerical data*
  • Prevalence
  • Probability
  • Quality-Adjusted Life Years*
  • Referral and Consultation / economics
  • Referral and Consultation / statistics & numerical data*
  • Resource Allocation / economics*
  • Sensitivity and Specificity
  • United States / epidemiology