METHODS: Systematic review and meta-analysis guided by Preferred Reporting Items for Systematic Reviews and Meta-Analyses was conducted. Four electronic databases were searched to identify studies of any design that reported on the preferred and actual place of care and death of patients with cancer in LMICs. A random-effects meta-analysis estimated pooled prevalences, with 95% CI, with subgroup analyses for region and risk of bias.

RESULTS: Thirteen studies were included. Of 3,837 patients with cancer, 62% (95% CI, 49 to 75) preferred to die at home; however, the prevalence of actual home death was 37% (95% CI, 13 to 60). Subgroup analyses found that preferences for home as place of death varied from 55% (95% CI, 41 to 69) for Asia to 64% (95% CI, 57 to 71) for South America and 72% (95% CI, 48 to 97) for Africa. The concordance between the preferred and actual place of death was 48% (95% CI, 41 to 55) for South Africa and 92% (95% CI, 88 to 95) for Malaysia. Factors associated with an increased likelihood of preferred home death included performance status and patients with breast cancer.