Measuring quality in local public health emergency preparedness: The Connecticut H1N1 experience

Connecticut is a home rule State with a diverse local public health infrastructure. The Connecticut Practice Based Research Network determined a need to explore measuring the quality of the publc health services performed at the local health agency level. To standardize the quality of service performed, the 2009 local health agency H1N1 pre-vaccine and post-vaccine reponses were selected. A strategy was using available public report data released by the Conneticut Deaprtment of Public Health and self-reported survey responses from local health agencies provided measurement data to assess local health quality performance. Systematic testing of a new QI public health measures using advanced psychometric techniques based on Structural Equation Modeling (SEM) has been performed. We used in particular an appropriate index construction and validation procedures for causal indicators measures (known also as formative measures). The measure items (indicators) were identified in a qualitative phase of measure development which consisted of four focus group session with LHD representatives from four regions of CT. A list of activities performed before and after the H1N1 vaccine became available, was arrived at using the focus groups findings, the timeline of actual events, and the expertise of the research team. LHD participants were then asked in an online survey which activities they performed in what specific month during the two distinct periods. These variables representing activities were then analyzed with MIMIC (Multiple Cause Multiple Indicator) structural models which specified initially all indicators predicting a latent factor, which in turns predicted 2 presumed outcomes of QI/performance, self-reported overall performance (overall and during the pre-vaccine period), and vaccines administered by LHDs (as percent of all vaccines, which included the privately administered ones).