Do official hospitalizations predict medical vulnerability among the homeless? A postdictive validity study of the vulnerability index

The current study tested the postdictive validity of the Vulnerability Index (VI), an instrument used to assess medical vulnerability among people who are homeless. It also examined the relationship between hospitalization records and self-reported health status. The VI is based on self-reports of hospital utilization and chronic health conditions. Data were collected over a one-year period from individuals receiving homeless services in a southwestern city (N = 97, 53.3% male, 57.7% African American). Vulnerability Index scores and three subcomponents of the measure (chronic health conditions, substance use, and mental health problems) were regressed on official reports of pastyear hospitalizations, controlling for gender and race, using four separate regression models. Official hospitalization records significantly predicted overall VI scores, but they did not predict the subcomponents of the measure. Results show that, within the current sample, official hospital records are predictive of overall VI scores and are correlated with self-reported hospitalization. The lack of relationship between hospital records and subcomponents of the VI may indicate an underutilization of health care for those with serious health conditions.