The Effect of Antibiotics on Latency When Given at the Time of Membrane Rupture Prior to Viability [ID: 1357946

Abstract

INTRODUCTION: Antibiotics administered when membranes rupture (ROM) after viability increase latency to delivery. This may also be true in previable prelabor preterm rupture of membranes (pPPROM). This study assesses the effect of prophylactic antibiotics on latency in individuals with pPPROM. METHODS: Retrospective cohort of pregnancies with pPPROM less than 23 weeks 0 days in a single health system (2013–2022). Patients opting for termination or with contraindication to expectant management were excluded. Prophylactic antibiotic administration (48 hours IV azithromycin/ampicillin followed by 5 days oral amoxicillin) was at clinician discretion. The primary outcome was latency (weeks) from diagnosed pPPROM to delivery. Secondary outcomes included maternal and neonatal morbidity and mortality. Bivariate statistics compared patients who did and did not receive antibiotics. Kaplan-Meier/Cox proportional hazards ratio using significant covariates (P<.1) in bivariate analysis models antibiotic effect on latency. RESULTS: Ninety-three patients had pPPROM; 46 (49%) met inclusion criteria. Thirty-four (74%) received prophylactic antibiotics. Median gestational age (GA) at ROM trended later among those who received antibiotics (22.0 weeks [20.6, 22.4] versus 20.9 weeks [19.6, 21.7], P=.09). Median latency (interquartile range) did not differ with antibiotic receipt (1 week [0.4, 2.6] versus 0.6 weeks [0.3, 0.9], P=.27). When adjusted for GA at ROM, antibiotics were not associated with longer latency (hazard ratio 1.33 [0.91, 1.93]). Antibiotic receipt was associated with lower rates of previable delivery (23.0 weeks [22.7, 24.0] versus 21.3 weeks [20.5, 23.1], P=.006). After controlling for GA at ROM, adjusted odds of previable delivery remained lower with receipt of antibiotics (adjusted odds ratio 0.20 [0.04, 0.90]). CONCLUSION: Antibiotics at the time of pPPROM were not associated with longer latency but, after controlling for confounders, did increase the odds of delivering after viability. Further study should address optimal antibiotics strategies for this unique population.