Skip to main content

Table 2 Summary of Wilcoxon–Mann–Whitney tests

From: Has the quality of reporting improved since it became mandatory to use the Standards for Reporting Diagnostic Accuracy?

Categorical variable

Summary of finding

Publication year

Studies published in 2019 (STARD mandatory) reported on more items compared to those in 2015 (STARD recommended) (19.5 [IQR 18.5–21.5] vs. 18.0 [IQR 15.5–19.5], p < 0.001, Vargha and Delaney’s A = 0.24)

Mode of data collection

No evidence of a difference in the total STARD score was found for data collection (retrospective, 18.8 [IQR 17.9–19.6] vs. prospective, 18.5 [IQR 16.8–20.0], p = 0.68, Vargha and Delaney’s A = 0.47)

Study design

No evidence of a difference in the total STARD score was found for study design (case control, 18.5 [IQR 17.5–19.0] vs. cohort, 18.5 [IQR 17.5–20.0], p = 0.81, Vargha and Delaney’s A = 0.53)

Citation rate (median split)a

No evidence of a difference in the total STARD score was found for citation rate (infrequently, 18.5 [IQR 18.0–20.5] vs. frequently, 18.5 [IQR 17.5–19.5], p = 0.54, Vargha and Delaney’s A = 0.54)

  1. STARD Standards for Reporting Diagnostic Accuracy
  2. aInfrequently < 0.56 citations/month; Frequently ≥ 0.56 citations/month