Lazarou · JAMA 1998 · meta-analysis of prospective studies · n=39 studies

Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies.

Cited 4837 times in the scientific literature.

Level 3 - non-randomized controlled study

Meta-analysis of prospective observational studies

PubMed 9555760 · doi:10.1001/jama.279.15.1200 · record verified 2026-08-28

What was done

A systematic search across four electronic databases (1966–1996) identified prospective studies evaluating adverse drug reactions (ADRs) in US hospitals. Out of 153 identified studies, 39 prospective studies were selected. Two investigators independently extracted data, which were pooled using a random-effects model. The analysis combined ADRs occurring during hospitalization with ADRs causing hospital admission. The authors excluded medication errors, noncompliance, overdose, drug abuse, therapeutic failures, and possible ADRs. Serious ADRs were defined as those resulting in hospitalization, permanent disability, or death.

What was found

The overall pooled incidence of serious ADRs was 6.7% (95% CI, 5.2% to 8.2%) and fatal ADRs was 0.32% (95% CI, 0.23% to 0.41%) among hospitalized patients. Extrapolated to 1994 US hospitalizations, an estimated 2,216,000 (95% CI, 1,721,000 to 2,711,000) patients experienced serious ADRs and 106,000 (95% CI, 76,000 to 137,000) experienced fatal ADRs, placing ADRs between the fourth and sixth leading cause of death in the United States.

Why it matters

This study provided the first comprehensive national estimate demonstrating that adverse drug reactions from standard medical therapy represent a major source of patient morbidity and mortality in US hospitals.

Limits

The findings are limited by heterogeneity among the included prospective studies and potential sampling biases. Total patient count across the 39 studies is not reported in the abstract. In addition, national fatal and serious ADR estimates depend on assumptions regarding how well the selected hospital cohorts represent all US hospital admissions.

Cited by