Sex-specific standardized mortality ratios for homeless youth: A meta-analysis

Several studies have compared sex-specific mortality rates in homeless youth to the general youth population. Due to low numbers of observed events, confidence intervals have been wide. We combined sex-specific standardized mortality ratios (SMRs) across studies to produce more precise estimates. Articles published from 1997 to 2012 were identified by using PubMed MeSH terms “mortality” and “homeless persons” or “homeless youth.” Articles were included if they reported sex-specific standardized mortality ratios and confidence intervals (or observed and expected deaths) in homeless youth. We also included data from our own, unpublished study. Summary SMRs were calculated using inverse variance weighting. Cochrane's Q statistic was used to test the assumption of fixed effects. Eight studies in four Western, industrialized countries were identified. All provided SMRs for male youth; five provided SMRs for female youth. Summary SMRs were 8.4 (95% confidence interval: 6.4 to 11.0) for males and 13.5 (95% CI: 7.4 to 24.7) for females. The p-value for the null hypothesis of equality of the gender specific SMRs was 0.16.Both male and female youth are at excess risk for mortality relative to the general youth population. While the difference between male and female SMRs was not significant at the Α = 0.05 level, female point estimates were higher for the summary SMR measure and in all studies that examined both genders except one. The summary estimate for female youth is still imprecise. Additional data are needed to improve estimation of relative mortality in females.