Introduction: In 1972 Cochrane advocated the widespread use of RCTs in community health, but subjective opinion suggests they may yet be under-used and lack methodologic rigor.

Objective: To 1) identify all intervention studies in journals of community health and determine the number of RCTs and the proportion of the remainder which should have been done as RCTs; and 2) assess the internal validity and overall quality of the RCTs in comparison with community health RCTs in a general medical journal.

Methods: The 1992 issues of six journals of community medicine were manually searched. Studies assessing the effects of an intervention were counted and analyzed: Subjectively for those not employing randomization (for given or suspected rationale and opinion as to whether (using 35 items, 26 pertaining to internal validity). Community health RCTs from the New England Journal of Medicine were scored on the same criteria.

Results: Interventions comprised 14% of the 603 studies in community health journals; 4% were RCTs. Of those not employing randomization, 42% should have. Mean RCT scores were significantly lower for journals of community health than for the N Engl J Med: 48% and 50% versus 65% and 65% (internal validity and overall, respectively (p < 0.05).

Discussion: RCTs are under-utilized in community health. Those published in journals of community medicine lack methodologic rigor, thus calling into question the conclusions regarding the effectiveness of interventions.