Background: Identifying and intervening on health disparities requires representative community public health data. For cities with high vacancy and transient populations, traditional methods of population estimation for refining random samples are not feasible. The aim of this project was to develop a novel method for systematic observations to establish community epidemiologic samples. Results: We devised a four-step population randomization observation process for Flint, Michigan, USA: (1) Use recent total population data for community areas (i.e., neighborhoods) to establish the proportional sample size for each area, (2) Randomly select street segments of each community area, (3) Deploy raters to conduct observations about habitation for each randomly selected segment, and (4) Complete observations for second and third street segments, depending on vacancy levels. We implemented this systematic observation process on 400 randomly selected street segments. Of these, 130 (32.5%) required assessment of secondary segments due to high vacancy. Among the 130 primary segments, 28 (21.5%) required assessment of tertiary (or more) segments. For 71.5% of the 400 primary street segments, there was consensus among raters on whether the dwelling inhabited or uninhabited. Conclusion: Houses observed with this method could have easily been considered uninhabited via other methods. This could cause residents of ambiguous dwellings (likely to be the most marginalized residents with highest levels of unmet health needs) to be underrepresented in the resultant sample.
- Census block group
- Population sample
- Random sample
ASJC Scopus subject areas
- Computer Science(all)
- Business, Management and Accounting(all)
- Public Health, Environmental and Occupational Health