There are currently no national estimates of how many people die while unhoused in the US. Local jurisdictions have developed their own approaches for estimating homeless mortality. We aimed to examine these local approaches, document what is known about homeless mortality, and summarize local methodologies. We reviewed 17 publicly available homeless mortality reports (ie, gray literature). Reports were sought from government, Health Care for the Homeless, coalition to end homelessness, and other advocacy and social service websites. From each report, we extracted the number of homeless deaths, dates of observation, data source(s) used, determination of homeless status, manners and causes of death, and decedent demographics. Data collection and reporting on homeless mortality varied greatly across reports. This variation limits aggregation across reports. Medical examiner data was the most used data source. Manner of death was the most consistently collected field, with accidental deaths reported as the most prevalent manner of homeless deaths. Not all reports listed specific causes of death, but those that did reported toxicity (eg, overdose) and cardiovascular causes as most prevalent. The most granular age category of most homeless decedents was 40 to 60years. On average, 80% of decedents were of male sex. While over half of reports included race and ethnicity information, disparities could not be estimated without suitable denominators. Standardized data collection and reporting guidance is needed for homeless mortality. Health departments can work with local Health Care for the Homeless programs and Continuums of Care to establish data sharing processes. Matching vital statistics with homeless service utilization records may be one opportunity to improve these efforts. Until there is federal or national guidance on these standards, localities can consider adding housing or homelessness variables as optional or mandatory fields in electronic death reporting systems.
Read full abstract