The Effect of Calcium Carbonate on Labor Induction: A Pilot Study

Purpose: Since 1976, cesarean deliveries in the United States have increased approximately 30%. While sometimes lifesaving, cesarean deliveries are associated with greater maternal risks than vaginal deliveries. The most common indication for cesarean delivery is labor dystocia. Calcium carbonate (CaCO3), brand name TUMS, has gained popularity for off-label use to prevent labor dystocia based only on anecdotal evidence, as no human trials have been conducted. Pharmacologically, calcium may improve uterine muscle contractility, and carbonate may decrease lactic acid buildup. This led us to investigate the potential for CaCO3 to prevent labor dystocia caused by poor uterine contractility during labor inductions and decrease cesarean delivery rates. We conducted a pilot study to explore the use of CaCO3 to augment labor induction by assessing (1) induction duration with oxytocin administration, (2) rate of labor dystocia, (3) rate of cesarean delivery, (4) maternal/neonatal outcomes, and (5) GI side effects. Methods: We conducted a quasi-experimental pilot study, inclusive of a prospective treatment group and a retrospective control group, among English and Spanish-speaking pregnant adults who presented for labor induction at an urban, teaching hospital in southeastern Wisconsin. The prospective treatment group (CaCO3 group, n=50) was consented between June 2024 and September 2024. After cervical ripening, the prospective treatment group received CaCO3 per standard treatment protocol (500mg orally every 4 hours), along with standard-dose oxytocin. After birth, participants completed a validated gastrointestinal (GI) effect survey. A randomly selected retrospective control group (control group, n=200) who presented for induction and received standard-dose oxytocin without CaCO3 between 2020-2022 was also identified. The primary outcomes assessed were (1) induction duration with oxytocin administration, excluding cervical ripening, and (2) rate of labor dystocia. Secondary outcomes assessed were the rate of cesarean delivery, amount of oxytocin used, blood loss, postpartum hemorrhage rate, neonatal outcomes, and GI side effects. All data were collected from the electronic medical record or GI side effects survey (scale 1-5 no problem-very severe problem) and were stored in REDCap. Means, medians, or percentages were reported as appropriate. Given that the goal of this pilot study was to inform sample size and explore any side effects or protocol concerns for a future randomized control trial, sample size estimates were not completed. Differences between groups were assessed using Pearson chi-squared test of independence, Fisher’s exact test, or Wilcoxon’s rank sum test, as appropriate. Multivariable Gamma and logistic regression models were used to carry out adjusted analysis. Two tailed p < 0.05 was considered statistically significant. Results: Baseline characteristics, including maternal age, body mass index, gestational age, parity, and starting Bishop score, were similar between groups. However, the groups were significantly different on race (e.g., African Americans were 38% in CaCO3 group vs. 62.5% in control group, p<0.01). After adjusting for this difference, mean induction duration with oxytocin was longer but not significantly between groups (879.3 minutes CaCO3vs. 759.5 minutes control group, β = 0.142, 95% CI: –0.05 to 0.33, p = 0.14). Despite this, there was a non-statistically significant but clinically relevant decrease in the rate of labor dystocia in CaCO3 group (4% vs. 11%, p=0.12). For our secondary outcomes, there was a clinically relevant, but non-statistically significant lower cesarean delivery rate in the CaCO3 group (14% vs. 22%, p=0.34). In a subgroup analysis of those who had cesarean delivery (n=50), dystocia as an indication was less likely, although not statistically significant, in the CaCO3 group (14% vs 44%, p=0.21). Use of CaCO3, when compared to the control group, did not lead to a significant difference in the total amount of oxytocin used (median 4799mU vs. 3707mU, p=0.76), blood loss (150 mL vs. 200mL, p=0.69), or postpartum hemorrhage ≥ 1000mL (4% vs. 5.5%, p=0.91). Favorably, there were no identifiable differences in neonatal outcomes such as 5-minute Apgar (median 9 vs. 9 control group, p=0.36), respiratory support (4% vs. 14% control group, p=0.19), and Neonatal Intensive Care Unit admission (6% vs. 7.5% control group, p=0.61). The GI side effects survey revealed that most patients reported few negative GI symptoms, reporting “no” for: burping/belching (84%), heartburn (72%), bloating (96%), passing gas (70%), sour taste (96%), nausea (56%), bad breath (98%), diarrhea (100%), constipation (96%), and vomiting (82%). Conclusion: Although there was a non-statistically significant longer induction time within the CaCO3 group, there were potentially clinically relevant decreases in labor dystocia and cesarean delivery rate. We did not identify differences in maternal and neonatal adverse outcomes, and our GI survey revealed overall maternal tolerability. As this was a pilot, we were not powered to detect statistically significant differences. However, our study will be useful for estimating sample size for future studies, as further information is needed to detect meaningful differences regarding CaCO3 use in labor inductions.