The quality of diagnostic accuracy studies since the STARD statement: has it improved

OBJECTIVE: To assess whether the quality of reporting of diagnostic accuracy
studies has improved since the publication of the Standards for the Reporting of
Diagnostic Accuracy studies (STARD statement).
METHODS: The quality of reporting of diagnostic accuracy studies published in 12
medical journals in 2000 (pre-STARD) and 2004 (post-STARD) was evaluated by two
reviewers independently. For each article, the number of reported STARD items was
counted (range 0 to 25). Differences in completeness of reporting between
articles published in 2000 and 2004 were analyzed, using multilevel analyses.
RESULTS: We included 124 articles published in 2000 and 141 articles published in
2004. Mean number of reported STARD items was 11.9 (range 3.5 to 19.5) in 2000
and 13.6 (range 4.0 to 21.0) in 2004, an increase of 1.81 items (95% CI: 0.61 to
3.01). Articles published in 2004 reported the following significantly more
often: methods for calculating test reproducibility of the index test (16% vs
35%); distribution of the severity of disease and other diagnoses (23% vs 53%);
estimates of variability of diagnostic accuracy between subgroups (39% vs 60%);
and a flow diagram (2% vs 12%).
CONCLUSIONS: The quality of reporting of diagnostic accuracy studies has improved
slightly over time, without a more pronounced effect in journals that adopted the
STARD statement. As there is still room for improvement, editors should mention
the use of the STARD statement as a requirement in their guidelines for authors,
and instruct reviewers to check the STARD items. Authors should include a flow
diagram in their manuscript.