Methodological quality of the review: Medium confidence

Author: Zafar A, Khan GI, Siddiqui MAR.

Region: Details not provided

Sector: Diabetic retinopathy, screening, diagnostic accuracy studies (DAS), The Standards for Reporting of Diagnostic Accuracy (STARD).

Sub-sector: None specified

Equity focus: None specified

Review type: Effectiveness review

Quantitative synthesis method: Narrative analysis

Qualitative synthesis methods: Not applicable

Background

Diagnostic accuracy studies (DAS) are conducted to evaluate the efficacy of tests diagnosing a target disorder, such as diabetic retinopathy. DAS enables a clinician to make decisions regarding the potential utility of a test. However, if these accuracy studies are improperly conducted or incompletely reported, they may be prone to bias which may lead to ‘overly optimistic estimations of the diagnostic value of a test… exaggerated results may lead to premature adoption of new and more expensive tests’.

Tests used to screen for diabetic retinopathy are diverse and authors note that there is a lack of transparency in reports of DAS for diabetic retinopathy screening.

In 2003, the Standards for Reporting of Diagnostic Accuracy (STARD) statement was published to improve the quality of reporting of DAS and help authors clearly report the methodology.

Research objectives

‘To evaluate the quality of reporting of diagnostic accuracy studies in DR screening’.

Main findings

The authors identified 76 papers for inclusion in the review, all of which reported the diagnostic accuracy of tests for diabetic retinopathy screening (diagnostic accuracy studies).

Only 11.8% (nine out of 76) of manuscripts completely reported at least 50% of the STARD items. Descriptions of key aspects of the methodology were frequently missing. Only 8% (six out of 76) of the manuscripts had a flowchart describing the flow of participants within a study.

Based on the diagnostic accuracy studies included authors concluded that the quality of reporting of diagnostic accuracy studies for diabetic retinopathy screening was sub-optimal.

Authors suggested that the adherence to STARD guidelines may improve the quality of reporting of DR screening studies. Authors noted that only two of the top ten journals (British Medical Journal and British Journal of Ophthalmology) encouraged authors to use the STARD statement.

Methodology

Authors conducted a systematic literature survey in which a highly sensitive MEDLINE search was performed to identify reports of diagnostic accuracy or diabetic retinopathy screening studies published between 1995 and 2006. Only studies written in English were included in the review. Two reviewers independently assessed the titles and abstracts for potential inclusion. Selected papers were then independently assessed for inclusion by two investigators.

The STARD checklist was used to evaluate the quality of reporting of each publication. Authors noted that this study looked only at the quality of reporting and not the methodological quality of the included studies, which was outside the remit of this review.

Applicability/external validity

The authors did not specifically address the applicability of the results, but they did suggest that adherence to STARD guidelines would improve the quality of reporting of test accuracy studies in general.

Geographic focus

Authors do not report the geographical focus of the studies included in the review.

Quality assessment

Medium confidence was attributed to the conclusions about the effects of this review as limitations were identified. The authors used appropriate methods to reduce risk of bias in terms of independent screening and data extraction of included studies. However, this review had limitations in the search for literature. Authors did not report contacting authors/experts or reviewing reference lists in included studies as part of the search strategy.

The STARD checklist was used to evaluate the quality of reporting of each publication.  Authors noted that this study looked only at the quality of reporting and not the methodological quality of the included studies, which was outside the remit of this review.

The review was useful in highlighting the need for researchers to adhere to the STARD guidelines to improve the quality of reporting of diagnostic accuracy studies in diabetic retinopathy screening.