Back to Annual Meeting Page		American Public Health Association 133rd Annual Meeting & Exposition December 10-14, 2005 Philadelphia, PA

Abstract #106533

Magda A. Shaheen, MD, PhD, MS, MPH, RCMI, Charles R. Drew University, 1731 East 120th Street, Los Angeles, CA 90059, 323.563.5972, mashahee@cdrewu.edu and Ralph R. Frerichs, DrPH, DVM, Epidemiology Department, University of California Los Angeles, P.O.Box 951772, Los Angeles, CA 90095-1772.

Using a modified two-stage cluster survey method, we conducted three surveys in Los Angeles County between 1997 and 2003. Presented here will be the impact of field modifications on the variance estimate in the surveys. At the first stage, we sampled clusters with probability-proportionate-to-estimated-size of the child population. The second stage included enumeration of the sampled areas to develop a list from which we drew a constant-sized simple random sample. Data were collected through interview. Vaccine coverage data were abstracted from the immunization card or from clinic records. We trained community volunteers and students to conduct the enumeration and assist in the data collection. Participation was high in the surveys (81%-92%), which were completed in 4-6 months depending on the number of volunteers and the size of the area. The variance estimates from the surveys were close to that if the surveys had been conducted as simple random sample surveys (i.e., design effect close to 1). For the three surveys, the design effect range from 0.7 to 1.3. Examples of the design effect for the three surveys are as follows: complete immunization (0.7, 0.9, 0.7), immunization card at home (1.1, 1.1, 0.8) and varicella vaccine (1.0, 1.0, 1.1). We concluded that the modifications of the method resulted in reduced variance close to that of the random sample method. Surveys were completed in a short time period, making the method ideal for studying issues among inner-city populations. In addition, training of volunteers and students led to increased participation and reduced cost.

Learning Objectives: At the conclusion of the session, the participant (learner) in this session will be able to

• learn about modified cluster sampling to decrease the design effect (i.e., improve precision).
• learn methods to enhance participation in the survey and reduce cost of the survey

Presenting author's disclosure statement:

I wish to disclose that I have NO financial interests or other relationship with the manufactures of commercial products, suppliers of commercial services or commercial supporters.