Weighting Data from a Sample Survey of HIV Patients in a Community-Based Service Delivery System: The 2011 Los Angeles Coordinated Needs Assessment-Care (LACHNA-Care) Study

• Activity Number: 19
• Type: Topic Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Christopher H. Johnson*+ and Rhodri Dierst-Davies and Amy Rock Wohl
• Companies: CDC and Los Angeles County Department of Public Health and Los Angeles County Department of Public Health
• Keywords: population-based surveys ; health surveys ; multistage sampling ; PPS sampling ; administrative records

Abstract

U.S. jurisdictions receiving federal Ryan White (RW) funding for HIV care must justify needs and allocate resources using evidence-based data. For the 2011 LACHNA-Care study, we used a two-stage, stratified probability-proportional-to-size design to identify a cross-sectional sample of agencies and their clients in the RW system. Among the 97 agency sites providing services in 2009, 49 were sampled; 400 of their clients who presented during pre-determined sampling periods were interviewed. Respondents were asked about awareness, needs, utilization, and barriers for various RW services. The response rate was 94%.

To weight respondents' data, we adapted methods appropriate for clustered provider-patient surveys and for time-location samples. We calculated individual-level design weights for respondents based on known sampling probabilities at both the agency and client stages. Non-response adjustments yielded no significant changes to the design weights. Other adjustments accounted for clients' multiple opportunities to be sampled from the same or other agencies. The weighted sample size of 18,912 is similar to the 19,951 RW clients in the system according to administrative data.