Statistics for Global Health

• Activity Number: 21
• Type: Contributed
• Date/Time: Sunday, August 8, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Section on Health Policy Statistics
• Author(s): David J. Fitch*+ and Roberto Molina Cruz and Patrick D. Baier
• Companies: Universidad del Valle de Guatemala and Universidad del Valle de Guatemala and George Washington University
• Keywords: LQAS ; developing countries ; decision-taking

Abstract

We take a second look at the use of Lot Quality Acceptance Sampling (LQAS) method for evaluating health projects success. As the method was designed for industrial quality control we go over the methodology to evaluate its applicability in assessing health programs in developing countries, especially because the great interest it has for important donor organizations like the United States Agency for International Development (USAID). We review its use, discuss how it is doing, and when it may not be efficiently providing information needed by program administrators. We also propose some other methods we believe may be more appropriate to decision-taking in health programs. We undertake a series of simulation studies to estimate the potential of LQAS and these other methods; e.g., our studies suggest there may be little gain in decision taking from sample sizes greater than 10 or 20. We observe the great potential for statistics for contributing to the health of the world's poor. We invite donor organizations like USAID and the statistical community, especially the members of the ASA and the IASS, to work together on better statistical methods to be used in the developing countries.