7 Oral azithromycin ± amoxicillin to prevent peripartum infection in laboring high-risk women: A 3-Arm RCT


      In low-income countries (LICs), maternal peripartum infections are a major cause of maternal mortality and mostly occur after vaginal birth. We compared the effectiveness of single dose azithromycin (AZI) ± amoxicillin (AMOX) vs placebo to prevent peripartum infection in laboring women.

      Study Design

      Multicenter 3-arm double-blind RCT (NCT3248297) of women with viable term non-anomalous gestations with prolonged labor >18 hrs or ROM >8 hrs in Cameroon, Africa (2018-2020). Women with chorioamnionitis, study drug contraindications, or planned cesareans were excluded from enrollment. Women were randomized to: Arm 1) AZI 1g PO/Placebo, Arm 2) AZI 1g PO/AMOX 2g PO or Arm 3) Placebo/Placebo. All groups received “usual care” including antibiotics (abx) given at provider discretion. The primary outcome was composite maternal peripartum infection/death up to 6 wks postpartum. Two comparisons (Arm 1 vs 3; Arm 2 vs 3) were planned. We estimated 250 women/group (750 total) would provide 80% power at two-sided α=0.05 (0.025 per comparison) to detect a 50% effect size with 20% baseline composite infection rate.


      Of 6531 women screened, 756 were randomized (Figure). Baseline characteristics (BMI, duration of ROM, labor, etc) were balanced except women were younger in the placebo arm (p=0.04). Over 60% of women in each arm received “usual care” abx: >90% penicillin and ∼50% for prolonged ROM >12-18 hrs. Composite outcome incidences were lower in each arm 1 and 2 vs placebo (Table), but differences were not significant (RR 0.6 [0.3-1.2]; 0.7 [0.4-1.3] respectively). Of note, chorioamnionitis and wound infection were significantly lower in Arm 2. There were no differences in other maternal or neonatal outcomes such as sepsis (Table).


      Single oral dose AZI ± AMOX for prolonged labor/ROM at term did not significantly reduce maternal peripartum or neonatal infection/death, partly due to lower than expected infection rates and frequent abx use as usual care. Given the direction of our results, larger trials should test for smaller, clinically important reductions in maternal infections in LICs.
      Figure thumbnail fx1
      Figure thumbnail fx2