Table 2.
CCTA Diagnostic Accuracy Study Quality and Internal Validity, as Assessed by Modified QUADAS Tool
QUADAS item | Response % (N = 42) | ||
---|---|---|---|
Yes | No | Unclear | |
Spectrum representative of disease risk | 29 | 66 | 5 |
Clear selection criteria | 73 | 19 | 8 |
Referent likely to classify CAD | 90 | 0 | 10 |
Limited time between diagnostic and referent | 81 | 2 | 17 |
Avoid partial verification bias | 83 | 10 | 7 |
Avoid differential verification bias | 94 | 4 | 2 |
Avoid incorporation bias | 98 | 0 | 2 |
Avoid test review bias | 71 | 2 | 27 |
Avoid diagnostic review bias | 54 | 8 | 38 |
Avoid clinical review bias | 67 | 29 | 4 |
Non-diagnostic results reported | 65 | 33 | 2 |
Withdrawals explained | 33 | 63 | 4 |
Established cut-off for significant stenosis | 96 | 2 | 2 |
Observer variation reported | 23 | 77 | 0 |
Data presented for appropriate subgroups | 90 | 2 | 8 |
True CAD prevalence reported | 79 | 21 | 0 |
“Response %” is based on the proportion of studies meeting criteria for each item
CCTA: Coronary computed tomographic angiography
QUADAS: Quality assessment of diagnostic accuracy studies
CAD: Coronary artery disease