Table 1.
Source and size of included datasets.
Fig 1.
Flow chart on the systematic search and screening for articles to be included in this meta-analysis.
Table 2.
Characteristics of included G6PD quantitative PoC assays.
Table 3.
Study and participant details of included datasets.
Fig 2.
Forest plots of the performance of the SDB at the 30% threshold.
The plots were stratified by blood source: sensitivity (top left) and specificity (top right) for studies using capillary blood samples, and sensitivity (bottom left) and specificity (bottom right) for studies using venous blood samples. StudyId identified the first author, country, and year of publication. The sensitivity of the venous blood samples from Sadhewa et al [42] were excluded as no G6PD-deficient participants (positive results) were detected by the reference method.
Fig 3.
Forest plots of the performance of the SDB at the 70% threshold.
The plots were stratified by blood source: sensitivity (top left) and specificity (top right) for studies using capillary blood samples, and sensitivity (bottom left) and specificity (bottom right) for studies using venous blood samples. StudyId identified the first author, country, and year of publication.
Fig 4.
Bland-Altman plot of G6PD activity readings from SDB against reference spectrophotometry.
From studies using capillary (left) and venous (right) blood samples. Black dashed line indicates mean difference, grey shaded area indicates 95% limits of agreement (LoA).
Table 4.
Individual study and pooled performance of the CSA.
Fig 5.
Bland-Altman plot of G6PD activity readings from CSA against reference spectrophotometry.
From studies using capillary (left) and venous (right) blood samples. Black dashed line indicates mean difference, grey shaded area indicates 95% limits of agreement (LoA).
Table 5.
Individual study performances of the CSW and FBA.
Fig 6.
Bland-Altman plot of G6PD activity readings from CSW and FBA against reference spectrophotometry.
From studies evaluating the CSW (left) and FBA (right). Black dashed line indicates mean difference, grey shaded area indicates 95% limits of agreement (LoA).
Table 6.
Odds Ratio (OR) and the p-values of diagnostic accuracy comparisons from the mixed-effects logistic regression analysis.
Fig 7.
Qualitative assessment of articles and datasets included [
27–31,33,35–37,41,42] in this meta-analysis. Modified from the QUADAS-2 (quality assessment tool for diagnostic accuracy studies) tool [26].