Group 3 studies
Three high-quality adult studies, two from South Africa37,46 and one from Ethiopia49,50, reported the proportion of in-hospital deaths to which ADRs contributed (Table 5). All three studies were conducted among adults in secondary / tertiary hospitals, with one46 also including patients in the intensive care unit. In two studies46,49,50 the proportion of deaths attributed to ADRs was the primary outcome measure, while the third37 mentioned the proportion of deaths attributed to ADRs as an additional outcome. The Ethiopian survey49,50 only considered deaths from ADRs already present at time of admission, and not deaths from ADRs that developed during the hospital stay, and thus may have underestimated the deaths attributable to ADRs.