Table 2.
Diagnostic test accuracy of the alert line for adverse neonatal outcomes
| Country (Year) references | Alert line status | Adverse neonatal outcome* | Percentage of alert line crossing | Prevalence of adverse neonatal outcome | Sensitivity (95% CI) | Specificity (95% CI) | Positive likelihood ratio (95% CI) | Negative likelihood ratio (95% CI) | Diagnostic odds ratio (95% CI) | J statistic (95% CI) | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Present | Absent | ||||||||||
| Apgar Score at 5 min <7 | |||||||||||
| South Africa (2006)26 | Crossed | 28 | 433 | 75.9% | 7.6% | 60.9% | 22.8% | 0.79 | 1.72 | 0.46 | −16.3% (−30.8 to −1.8) |
| Not crossed | 18 | 128 | (46.5–73.6) | (19.5–26.5) | (0.6–1.0) | (1.2–2.5) | (0.3–0.9) | ||||
| Ecuador (2008)32 | Crossed | 3 | 289 | 58.4% | 0.6% | 100.0% (43.9–100) | 41.9% (37.6–46.2) | 1.72 (1.6–1.9) | NA | NA | |
| Not crossed | 0 | 208 | |||||||||
| Brazil (2009)34 | Crossed | 2 | 98 | 42.9% | 1.3% | 66.7% (20.8–93.6) | 57.4% (50.9–63.6) | 1.56 (0.7–3.5) | 0.58 (0.1–2.9) | 2.69 (0.2–30.1) | 24.1% (−29.6 to 77.8) |
| Not crossed | 1 | 132 | |||||||||
| India(2014)27 | Crossed | 43 | 53 | 19.2% | 17.2% | 50.0% (39.7–60.3) | 87.2% (83.6–90.0) | 3.91 (2.8–5.4) | 0.57 (0.5–0.7) | 6.81 (4.1–11.4) | 37.20% (26.2–48.2%) |
| Not crossed | 43 | 361 | |||||||||
| India (2016)31 | Crossed | 4 | 109 | 56.5% | 2.0% | 100.0% (51.0–100) | 44.4% (37.6–51.4) | 1.80 (1.6–2.0) | NA | NA | 44.4% (37.4–51.3) |
| Not crossed | 0 | 87 | |||||||||
| Nigeria and Uganda (2018)33 | Crossed | 143 | 3990 | 49.0% | 2.8% | 60.9% (54.5–66.9) | 51.4% (50.3–52.5) | 1.25 (1.1–1.4) | 0.76 (0.7–0.9) | 1.64 (1.3–2.1) | 12.2% (5.9–18.6) |
| Not crossed | 92 | 4215 | |||||||||
| Overall | Crossed | 223 | 4972 | 49.6% | 3.6% | 59.2% (54.1–64.0) | 50.8% (49.8–51.8) | 1.20 (1.1–1.3) | 0.80 (0.7–0.9) | 1.49 (1.2–1.8) | 9.9% (4.9–15.0) |
| Not crossed | 154 | 5131 | |||||||||
| Birth asphyxia | |||||||||||
| Nigeria (2008)28 | Crossed | 22 | 186 | 45.4% | 10.3% | 46.8% (33.3–60.8) | 54.7% (49.9–59.5) | 1.03 (0.8–1.4) | 0.97 (0.7–1.3) | 1.06 (0.6–2.0) | 1.6% (−13.5 to 16.6) |
| Not crossed | 25 | 225 | |||||||||
| India (2016)31 | Crossed | 7 | 106 | 56.5% | 4.5% | 77.8% (45.3–93.7) | 44.5% (37.6–51.6) | 1.4 (0.97–2.0) | 0.5 (0.2–1.7) | 2.81 (0.6–13.9) | 22.3% (−5.8 to 50.3%) |
| Not crossed | 2 | 85 | |||||||||
| India (2016)30 | Crossed | 7 | 79 | 32.8% | 5.7% | 46.7% (24.8–69.9) | 68.0% (62.0–73.5) | 1.46 (0.8–2.6) | 0.78 (0.5–1.3) | 1.86 (0.7–5.3) | 14.7% (−11.2 to 40.6) |
| Not crossed | 8 | 168 | |||||||||
| Overall | Crossed | 29 | 292 | 48.8% | 8.5% | 51.8% (39.0–64.3) | 51.5% (47.5–55.5) | 1.07 (0.8–1.4) | 0.94 (0.7–1.2) | 1.14 (0.7–2.0) | 3.3% (−10.4 to 17.0) |
| Not crossed | 27 | 310 | |||||||||
| Early neonatal mortality | |||||||||||
| South Africa (2006)26, ** | Crossed | 8 | 453 | 75.9% | 3.0% | 44.4% (24.6–66.3) | 23.1% (19.9–26.7) | 0.58 (0.3–0.97) | 2.41 (1.6–3.7) | 0.24 (0.1–0.6) | −32.5% (–55.7 to −9.3) |
| Not crossed | 10 | 136 | |||||||||
NA, not applicable.
*Stillbirths and unknown outcomes excluded from the denominator as follows. Dujardin24: outcome unknown for 33 neonates and 28 stillbirths excluded; WHO12: three stillbirths excluded; Van Bogaert26: three stillbirths excluded Orji28: five stillbirths excluded; Souza33: 49 stillbirths excluded).
**Deaths resulted from low Apgar scores and unsuccessful resuscitation.