Table 4.
Observed and Bayesian posterior means, 95% credible intervals and posterior probabilities of a dose effect for early preterm birth rate by quartile baseline DHA status and adherence (postpartum RBC-PL-DHA ≥5.5%) with the high dose. Posterior means and Bayesian 95% credible intervals and posterior probabilities of a dose effect.
| Observed early preterm birth <34 weeks, n/N (%) | Posterior mean% of EPTB (95% Bayesian credible interval)a | Bayesian posterior probabilities | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 200mg | 1000- mg | 1000+ mg | 200mg | 1000- mg | 1000+ mg | 1000+ mg better than 200 mg | 1000+ mg better 1000- mg | 1000- mg better than 200 mg | |
| DHA Enrollment (%) | |||||||||
| Quartile 1 (<5.1] | 5/112 (4.5) | 3/63 (4.8) | 2/79 (2.5) | 3.4 (1.2,8.0) | 3.1 (1.1,7.7) | 1.3 (0.4,3.6) | 0.92 | 0.91 | 0.51 |
| Quartile 2 (5.1–6.2] | 6/132 (4.6) | 0/29 (0.0) | 0/97 (0.0) | 3.5 (1.5,7.7) | 2.7 (0.0,6.4) | 1.1 (0.0,2.4) | 0.97 | 0.85 | 0.63 |
| Quartile 3 (6.2–7.4] | 1/110 (0.9) | 1/22 (4.5) | 2/116 (1.7) | 1.3 (0.0,3.8) | 3.0 (0.6,8.3) | 1.3 (0.4,2.8) | 0.50 | 0.89 | 0.16 |
| Quartile 4 (>7.4) | 0/136 (0.0) | 0/10 (0.0) | 1/122 (0.8) | 0.6 (0.0,3.3) | 2.9 (0.0,7.2) | 1.1 (0.2, 2.6) | 0.34 | 0.86 | 0.12 |
a
Bayesian hierarchical model was fitted using the approach in Berry and Berry,8 except mu~N(−3.1,10^2). This approach avoids false discoveries in subgroups (quartiles). We burned-in 10,000 draws and used 40,000 draws for inference. High adherence, 1000+ mg, postpartum RBC-PL-DHA ≥ 5.5%; low adherence, 1000- mg, postpartum RBC-PL-DHA <5.5%.