Table 2.
Pooled Sensitivity and Specificity of Cord Blood Bilirubin in Predicting the Need for Phototherapy.
Risk group | Cord bilirubin cut-off | No of studies and no of participants | Sensitivity (95% confidence interval) | Specificity (95% confidence interval) | Diagnostic odds ratio (95% confidence interval) | Post-test probability for an abnormal resulta | Post-test probability for a normal resulta |
---|---|---|---|---|---|---|---|
All neonatesb | <1.5 mg/dl | 4; 13,070 | 0.96 (0.7–0.996) | 0.32 (0.13–0.59) | 11.3 (9.63–37.2) | 14% (13%–14%) | 1% (1%–4%) |
All neonatesb | 1.51–2.0 mg/dl | 13; 15,005 | 0.8 (0.72–0.86) | 0.73 (0.65–0.79) | 10.8 (9.66–11.4) | 25% (22%–28%) | 3% (2%–4%) |
All neonatesb | 2.01–2.5 mg/dl | 7; 2999 | 0.55 (0.34–0.75) | 0.89 (0.76–0.95) | 9.84 (9.8–9.92) | 36% (30%–42%) | 5% (4%–7%) |
All neonatesb | 2.51–3.0 mg/dl | 5; 13,420 | 0.31 (0.18–0.47) | 0.98 (0.96–0.99) | 22.5 (21.1–22.9) | 62% (50%–75%) | 7% (6%–8%) |
All neonatesb | >3.0 mg/dl | 4; 11,860 | 0.11 (0.03–0.34) | 0.999 (0.95–1) | 124 (31–∞) | 92% (59%–99%) | 9% (8%–10%) |
High riskc | Below 1.5 mg/dl | 3; 1159 | 0.97 (0.83–1) | 0.29 (0.08–0.66) | 13.2 (9.48–∞) | 13% (13%–14%) | 1% (0–3%) |
High riskc | 1.51–2.0 mg/dl | 8; 3091 | 0.9 (0.81–0.95) | 0.55 (0.33–0.75) | 10.98 (9.46–12.91) | 18% (17%–20%) | 2% (1%–4%) |
High riskc | 2.01–2.5 mg/dl | 7; 1210 | 0.84 (0.67–0.93) | 0.92 (0.82–0.96) | 60.3 (48.8–60.9) | 55% (48%–59%) | 2% (1%–3%) |
High riskc | 2.51–3.0 mg/dl | 4; 924 | 0.8 (0.39–0.96) | 0.95 (0.78–0.99) | 75.5 (63–85.7) | 64% (57%–71%) | 2% (2%–3%) |
Low riskd | 1.51–2.0 mg/dl | 11; 3441 | 0.89 (0.77–0.95) | 0.77 (0.58–0.89) | 26.97 (26.24–27) | 30% (27%–33%) | 2% (1%–3%) |
Low riskd | 2.01–2.5 mg/dl | 13; 3996 | 0.79 (0.64–0.88) | 0.89 (0.74–0.95) | 30.3 (20.94–33.58) | 44% (39%–50%) | 3% (2%–4%) |
Low riskd | 2.51–3.0 mg/dl | 10; 2886 | 0.74 (0.39–0.93) | 0.97 (0.91–0.99) | 91.9 (64–134.14) | 74% (65%–80%) | 3% (2%–4%) |
The probabilities are calculated at a prevalence of 10%.
Studies including all neonates without stratifying their risk for jaundice.
Studies including only neonates at risk for blood group incompatibility.
Studies including only low risk neonates (excluded those with blood group incompatibility).