Objectives. This study examined the quality of recent reviews in epide
miology.Methods. All 1995 issues of 7 widely read epidemiology journal
s were searched to identify reviews. Results. Twenty-nine reviews were
identified. Methodology was not specified or incomplete for literatur
e searches in 79% of reviews; the same was true for inclusion criteria
in 83% and for combining studies in 62%. More than 60% of the reviews
were not methodologically systematic. Conclusions. There is a need to
improve the quality of review papers in epidemiology. If systematic m
ethodology were followed more frequently, epidemiologic science and it
s application could be improved.