Fig 1.
PRISMA flow chart.
Table 1.
Study characteristics of included studies in meta-analysis.
Table 2.
Test types, technologies used, and non-laboratory health care cadres in included studies.
Fig 2.
Forest plots of point-of-care CD4 diagnostic accuracy performance by non-laboratory staff.
(a) point-of-care CD4 testing by non-laboratory staff compared to conventional laboratory-based testing by professional laboratory staff; (b) point-of-care CD4 testing by non-laboratory staff compared to point-of-care CD4 testing by professional laboratory staff; the shaded box represents +/- 50 cells/ul in (a) and (b). (c) sensitivity of correctly classifying patients below a CD4 threshold of 350 cells/ul; (d) specificity of correctly classifying patients above a CD4 threshold of 350 cells/ul.
Fig 3.
Forest plots of point-of-care ALT and hemoglobin diagnostic accuracy performance by non-laboratory staff.
(a) point-of-care ALT testing by non-laboratory staff compared to conventional laboratory-based testing by professional laboratory staff; (b) point-of-care hemoglobin testing by non-laboratory staff compared to conventional laboratory-based testing by professional laboratory staff.
Fig 4.
Error rates for point-of-care CD4 technologies operated by non-laboratory staff.
Red forest plot and line indicate the overall pooled error rate. Blue forest plots indicate studies focused on program-wide, routine testing.