Table 2. Performance of LMP and Ballard to determine gestational age at birth compared to ultrasound.
| Gestational age at birth | Preterm birth | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean in weeks (SD) | Convergent validity | Bland-Altman statistics | Proportion preterm | kappa | Sensitivity | Specificity | Diagnostic Odds Ratio (95% CI) |
Positive Predictive Value | Negative Predictive Value | ||||
| Pearson’s correlation coefficient | Bias Correction Factor | Lin’s concordance correlation (95% CI) |
95% limits of agreement, weeks | Mean difference, weeks (CI) | |||||||||
| Ultrasound (n = 188) | 38.3 (1.6) | (Ref.) | (Ref.) | (Ref.) | (Ref.) | (Ref.) | 16.0% | (Ref.) | (Ref.) | (Ref.) | (Ref.) | (Ref.) | (Ref.) |
| New Ballard (n = 156) | 39.0 (1.4) | 0.42 | 0.92 | 0.39 (0.26, 0.51) |
-2.51 to 3.73 | 0.61 (0.36, 0.86) |
6.4% | 0.1090 | 13.6% | 94.8% | 2.61 (0.73, 9.34) |
30.0% | 87.0% |
| LMP (n = 185) | 38.4 (2.3) | 0.52 | 0.92 | 0.48 (0.38, 0.58) |
-3.93 to 4.20 | 0.13 (-0.16, 0.43) |
17.8% | 0.2925 | 43.3% | 87.1% | 5.16 (2.21, 12.09) |
39.4% | 88.8% |