Table 2 |
||||||||||
|
Inter-assessment agreement: mean of first assessment and second assessment of the quality of reporting of diagnostic accuracy studies (n = 32), followed by mean differences between the two assessments, 95% limits of agreement, and smallest detectable difference (SDD). |
||||||||||
|
Outcome measure |
First assessment (A) |
Second assessment (B) |
Difference B – A MeandiffAB (SDdiffAB) |
95% Limits of Agreement* |
SDD† |
Systematic differences (95%CI)‡ |
||||
|
|
||||||||||
|
MeanA (SDA) |
RangeA |
MeanB (SDB) |
RangeB |
|||||||
|
Number of reported STARD items (0–25) |
12.08 (3.9) |
3.5 – 19 |
12.47 (3.4) |
7 – 19 |
0.39 (2.4) |
-4.27, 5.05 |
4.66 |
0.39 (-0.4, 1.2) |
||
|
|
||||||||||
|
SD = Standard Deviation, * 95% Limits of Agreement: mean(diffAB)± 1.96SDdiffAB, † SDD = Smallest Detectable Difference (SDD = 1.96 * SDdiffAB); ‡ systematic difference (bias) = MeandiffAB± 1.96 * SDdiffAB/√ n; 95%CI = 95% confidence intervals. |
||||||||||
|
Smidt et al. BMC Medical Research Methodology 2006 6:12 doi:10.1186/1471-2288-6-12 |
||||||||||
