Table 2. Direct observational assessment of pulse oximetry use by health-care providers, Malawi, 2012–2014.
| Variable | All providers (n = 22) | HCB providers |
CHW (n = 6) | ||
|---|---|---|---|---|---|
| Medical assistants (n = 6) | Clinical officers (n = 5) | Nurses (n = 5) | |||
| SpO2 measurement quality, no. of measurements/total no. observed (%; 95% CI) | |||||
| Use of hallux if patient weighed < 10 kg | 733/790 (92.8; 90.8–94.5) | 199/219 (90.9; 86.2–94.3) | 162/179 (90.5; 85.2–94.4) | 165/173 (95.4; 91.1–98.0) | 207/219 (94.5; 90.6–97.1) |
| Patient calm when measurement made | 1287/1320 (97.5; 96.5–98.3) | 353/360 (98.1; 96.0–99.2) | 292/300 (97.3; 94.8–98.8) | 286/300 (95.3; 92.3–97.4) | 356/360 (98.9; 97.2–99.7) |
| Mean weighted kappaa | 0.41 | 0.51 | 0.40 | 0.36 | 0.37 |
CHW: community health workers; CI: confidence interval; HCB: health-centre-based; SpO2: peripheral arterial haemoglobin oxygen saturation.
a For level of provider–expert agreement on measured oxygen saturations.