Ambulation for latency during expectant management of preterm prelabor rupture of membranes: a randomized controlled trial (AMBLE)

Background: Individuals hospitalized with preterm prelabor rupture of membranes are often advised to limit their activity or adhere to bed rest. Some evidence suggests that greater activity is associated with longer latency and improved outcomes, but no high-quality evidence from a randomized controlled trial exists.

Objective: This study aimed to evaluate whether encouragement to ambulate at least 2000 steps daily affects latency among individuals with preterm prelabor rupture of membranes compared with usual care.

Study design: This was a multisite unblinded, 2-arm randomized trial of individuals at 23 0/7 to 35 0/7 weeks of gestation undergoing inpatient expectant management of preterm prelabor rupture of membranes with planned delivery at least 7 days away. Each participant wore a Fitbit Inspire that tracked steps. The intervention arm was encouraged (verbal and Fitbit-based reminders) to reach a goal of 2000 steps per day. The usual-care arm was allowed ad libitum activity with no step goal or reminders. The primary outcome was latency (days) from randomization to delivery. Secondary analyses included composite neonatal and maternal clinical outcomes and maternal mental health survey results. Statistical analyses were conducted with an intent-to-treat approach under a Bayesian framework using neutral priors (a priori assumed 50:50 likelihood of longer latency in either arm). A total of 100 participants were required to have 80% power to demonstrate a 4-day difference in latency with 75% certainty (Bayesian probability).

Results: Among 163 eligible individuals, 100 (61%) were randomized, and after loss to follow-up, 95 were analyzed. Gestational age at randomization was 29 3/7 weeks (interquartile range, 26 2/7 to 31 5/7) in the intervention arm and 27 4/7 weeks (interquartile range, 25 4/7 to 29 6/7) in the usual-care arm. Median step counts were 1690 per day in the intervention arm (interquartile range, 1031-2641) and 1338 per day in the usual-care arm (interquartile range, 784-1913). Median days of latency were 9 days in the intervention arm (interquartile range, 4-17) and 6 days in the usual-care arm (interquartile range, 2-14). The primary analysis indicated a 65% posterior probability that the intervention increased latency relative to usual care (posterior relative risk, 1.09; 95% credible interval, 0.70-1.71). The relative risk was 0.55 (95% credible interval, 0.32-0.82) for the composite neonatal adverse outcome, with 99% posterior probability of intervention benefit, and was 0.94 (95% credible interval, 0.72-1.20) for the composite maternal adverse outcome, with 70% posterior probability of intervention benefit. There was a 94% posterior probability of the intervention arm having a greater decrease in maternal stress score from baseline to before delivery compared with the usual-care arm (mean arm difference, 3.24 points [95% credible interval, -7.23 to 0.79]). Adjustment for gestational age at randomization had minimal impact on secondary outcome results.

Conclusion: Individuals with preterm prelabor rupture of membranes randomized to encouragement to ambulate had a longer latency to delivery and improved neonatal and mental health outcomes, with similar maternal clinical outcomes compared with usual care.