Table 3.
Subgroup | Number of included studies | Sensitivity (95% CI) | Specificity (95% CI) | PLRa | NLRb | DORc | AUCd | ||||||||
Value of meta-analysis in all the included (bivariate and HSROCe method) | 4 | 0.81 (0.68-0.90) | 0.93 (0.82-0.98) | 12.3 (3.8-39.2) | 0.20 (0.11-0.38) | 61 (11-322) | 0.93 (0.90-0.95) | ||||||||
Value of meta-analysis in all the included (Moses-Shapiro-Littenberg method) |
|
0.90 (0.89-0.91) | 0.94 (0.93-0.95) | 11.1 (1.6-76.2) | 0.20 (0.08-0.52) | 56 (5-591) | 0.90 (0.71-0.99) | ||||||||
Methodological quality of included studies |
|
|
|
|
|
|
|
||||||||
|
High quality | 3 | 0.90 (0.87-0.91) | 0.94 (0.93-0.95) | 13.1 (1.4-124.5) | 0.22 (0.08-0.62) | 61 (4-919) | 0.87 (0.43-0.99) | |||||||
|
Low quality | 1 | Null | Null | Null | Null | Null | Null | |||||||
Total number of included patients |
|
|
|
|
|
|
|
||||||||
|
≤100 | 3 | 0.90 (0.87-0.91) | 0.94 (0.93-0.95) | 13.1 (1.4-124.5) | 0.22 (0.08-0.62) | 61 (4-919) | 0.87 (0.43-0.99) | |||||||
|
<100 | 1 | Null | Null | Null | Null | Null | Null | |||||||
Format of study |
|
|
|
|
|
|
|
||||||||
|
Retrospective | 3 | 0.90 (0.87-0.91) | 0.94 (0.93-0.95) | 13.1 (1.4-124.5) | 0.22 (0.08-0.62) | 61 (4-919) | 0.87 (0.43-0.99) | |||||||
|
Prospective | 1 | Null | Null | Null | Null | Null | Null | |||||||
Published year |
|
|
|
|
|
|
|
||||||||
|
After 2010 | 4 | 0.90 (0.89-0.91) | 0.94 (0.93-0.95) | 11.1 (1.6-76.2) | 0.20 (0.08-0.52) | 56 (5-591) | 0.90 (0.71-0.99) | |||||||
|
Before 2010 | 0 |
|
|
|
|
|
|
|||||||
Type of AIf |
|
|
|
|
|
|
|
||||||||
|
Neural network–based | 3 | 0.91 (0.90-0.92) | 0.97 (0.96-0.97) | 16.8 (2.0-141.7) | 0.17 (0.05-0.61) | 98 (6-1640) | 0.95 (0.75-0.99) | |||||||
|
SVMg–based | 1 | Null | Null | Null | Null | Null | Null | |||||||
Type of endoscopic image |
|
|
|
|
|
|
|
||||||||
|
WLIh | 3 | 0.91 (0.90-0.92) | 0.97 (0.96-0.97) | 16.8 (2.0-141.7) | 0.17 (0.05-0.61) | 98 (6-1640) | 0.95 (0.75-0.99) | |||||||
|
LCIi | 1 | Null | Null | Null | Null | Null | Null | |||||||
Classifying performance between Helicobacter pylori–positive vs H pylori–naïve images | 3 | 0.77 (0.71-0.82) | 0.96 (0.94-0.98) | 11.8 (3.7-38.3) | 0.26 (0.21-0.32) | 53 (17-161) | 0.88 (0.79-0.96) |
aPLR: positive likelihood ratio.
bNLR: negative likelihood ratio.
cDOR: diagnostic odds ratio.
dAUC: area under the curve.
eHSROC: hierarchical summary receiver operating characteristic.
fAI: artificial intelligence.
gSVM: support vector machine.
hWLI: white-light imaging.
iLCI: linked color imaging.