Hand searching the Journal of Epidemiology and Community Health as part of the Cochrane Collaboration

Abstract

Study objective - To identify randomised controlled trials (RCTs) published in the Journal of Epidemiology and Community Health and to explore the contribution of these to the evaluation of public health issues. Design - Hand searching of the journal by both authors with independent assessment of topics of the reports and of their relevance to the Cochrane Collaboration. Agreement was assessed using κ scores. Setting - All papers and letters published in the journal from the first issue to the end of 1994. Subjects - Reports that might be RCTs were collected and classified into seven categories: definitely/probably/possibly RCTs or quasi-RCTs; or none of these. Main results - Eighty two definite RCTs were identified and a further 23 were probably/possibly RCTs or quasi-RCTs. Most reports dealt with health education, drug treatments, or 'other' health service interventions. Both authors failed to identify a number of trials on hand searching. Conclusions - The journal has published many trials of importance to the development of evidence-based public health policy. Hand searching may need to be done independently by more than one person.