The development of a survey instrument for community health improvement

Objective. To develop a survey instrument that could be used both to guide and evaluate community health improvement efforts. Data Sources/Study Setting. A randomized telephone survey was administered to a sample of about 250 residents in two communities in Lehigh Valley, Pen nsylvania in the fall of 1997. Methods. The survey instrument was developed by health professionals repres enting diverse health care organizations. This group worked collaboratively over a period of two years to (1) select a conceptual model of health as a foundation for the survey; (2) review relevant literature to identify indi cators that adequately measured the health constructs within the chosen mod el; (3) develop new indicators where important constructs lacked specific m easures; and (4) pilot test the final survey to assess the reliability and validity of the instrument. Principal Findings. The Evans and Stoddart Field Model of the Determinants of Health and Well-Being was chosen as the conceptual model within which to develop the survey. The Field Model depicts nine domains important to the origins and production of health and provides a comprehensive framework fro m which to launch community health improvement efforts. From more than 500 potential indicators we identified 118 survey questions that reflected the multiple determinants of health as conceptualized by this model. Sources fr om which indicators were selected include the Behavior Risk Factor Surveill ance Survey, the National Health Interview Survey, the Consumer Assessment of Health Plans Survey, and the SF-12 Summary Scales. The work group develo ped 27 new survey questions for constructs for which we could not locate ad equate indicators. Twenty-five questions in the final instrument can be com pared to nationally published norms or benchmarks. The final instrument was pilot tested in 1997 in two communities. Administration time averaged 22 m inutes with a response rate of 66 percent. Reliability of new survey questi ons was adequate. Face validity was supported by previous findings from qua litative and quantitative studies. Conclusions. We developed, pilot tested, and validated a survey instrument designed to provide more comprehensive and timely data to communities for c ommunity health assessments. This instrument allows communities to identify and measure critical domains of health that have previously not been captu red in a single instrument.