Purpose: Diabetic retinopathy (DR) is a common cause of blindness among the working age group in the developed world. To identify potentially treatable DR, systematic screening is recommended. The aim of this study was to evaluate the quality of reporting of diagnostic accuracy studies in DR screening.
Methods: We conducted a systematic literature survey in which a highly sensitive MEDLINE search was performed. The aim was to identify reports of diagnostic accuracy for DR screening published during 1995-2006. Two reviewers appraised all selected papers independently. The Standards for Reporting of Diagnostic Accuracy (STARD) checklist was used to evaluate the quality of reporting of each publication. For every manuscript, each item was scored as completely reported (score = 2), partly reported (score = 1) or not reported (score = 0). We calculated the total number of reported STARD items per article, summary scores for each STARD item and the average number of reported STARD items per year.
Results: A total of 5887 citations were identified from Medline search. Seventy-six fulfilled the criteria for inclusion. The mean score +/- SD of articles included in our study was 19.8 +/- 6.5 out of a maximum of 50. The mean number +/- SD of fully reported items among included studies was 8 +/- 3.3. Only 9 (11.8%) manuscripts completely reported at least 50% of the STARD items. Descriptions of key aspects of the methodology were frequently missing. Only 8% (6/76) of the manuscripts had a flow chart describing flow of participants within a study.
Conclusion: The quality of reporting of diagnostic-accuracy studies for DR screening is suboptimal. Adherence to STARD guidelines may improve the quality of reporting of DR screening studies.