The quality of reporting of diagnostic accuracy studies in glaucoma using the Heidelberg retina tomograph

Methodological quality of review: Medium confidence

Author: Shunmugam M, Azuara-Blanco A

Geographical coverage: Not reported

Sector: Glaucoma

Sub-sector: Diagnostic accuracy

Equity focus: None specified

Review type: Other review

Quantitative synthesis method: Narrative synthesis

Qualitative synthesis method: Not applicable

Background: In the field of glaucoma practice, there has been a growth in the number of diagnostic tests designed to detect structural and functional damage at relatively early stages of the disease, such as the Heidelberg retina tomograph (HRT). Scanning laser tomography with the HRT has been proposed as a useful diagnostic test for glaucoma.

Objectives: To evaluate the quality of reporting of published studies using the HRT for diagnosing glaucoma.

Main findings: A total of 29 articles were included. Inter-observer rating agreement was observed in 83% of items (K= 0.76). The number of STARD items properly reported ranged from five to 18. Less than a third of studies (7/29) explicitly reported more than half of the STARD items. Descriptions of key aspects of the methodology were frequently missing. For example, the design of the study (prospective or retrospective) was reported in six of 29 studies, and details of participant sampling (e.g. consecutive or random selection) were described in five of 29 publications. The commonest description of diagnostic accuracy was sensitivity and specificity (25/29) followed by area under the ROC curve (13/29), with nine of 29 publications reporting both.

Based on these findings, authors conclude that the quality of reporting of diagnostic accuracy tests for glaucoma with HRT is sub-optimal. The STRAD initiative may be a useful tool for appraising the strengths and weaknesses of diagnostic accuracy studies.

Methodology: Authors included studies if they reported on measures of diagnostic accuracy of HRT for glaucoma. Authors conducted a search of Medline (1996-2005) using a validated search strategy, and PUBMED using medical subject headings and keywords such as specificity and false negative. The search was limited to publications written in English and studies focusing on human subjects.

Two authors independently assessed the quality of reporting using STRAD checklist. Checklist items are arranged under the following headings:

(a) title, abstract, and keywords, (b) introduction, (c) methods (11 items), (d) results (11 items), and (e) discussion. Each item was considered to be fully, partially or not reported. If the item was not applicable, it was marked as such. Inter-observer agreement (k statistic) and descriptive statistics were calculated.

Authors used a narrative approach to examine the standard of reporting of diagnostic accuracy publications using the HRT for detection of glaucoma.

Applicability/external validity: Authors note that the majority of included studies did not describe participant sampling, therefore authors were not able to discuss the generalizability of the results.

Geographic focus: Geographical location of included studies was not available.

Summary of quality assessment: Overall, there is medium confidence in conclusions about the effects of this study. Methods used to identify studies for inclusion in the review were not sufficiently comprehensive that we can be confident that relevant studies were not omitted. Review authors did not include studies written in languages other than English, and contacting authors/experts for further potentially relevant studies and searching for unpublished studies were not part of the search strategy.

Although authors did not extract data, authors used appropriate methods to critically appraise the quality of reporting of included studies. Taking into account the context and aim of the review, extracting data may not be required. Authors used appropriate methods to analyse findings of included studies and did not draw strong policy conclusions. However, limitations on the methods used to conduct this review were not reported.