Abstract
Introduction: Immediate postpartum long-acting reversible contraception (IPP-LARC) after delivery may decrease rapid repeat pregnancy (RRP). RRP is associated with preterm birth, small for gestational age, low birth weight, and infant/maternal mortality. In April 2017, the Ohio legislature passed SB 332, mandating immediate postpartum LARC coverage. Later that year, our hospital made the IPP-LARC Nexplanon (and not intrauterine devices [IUDs]) available to all delivering patients. This study aimed to see if SB 332 is associated with decreased RRP.
Methods: Using electronic health record and state registry data, we examined sociodemographic and perinatal care variables for 726 pregnant adolescents and adults delivering at our hospital during the 12 months before and after IPP-LARC became available. We then determined the association between RRP and these variables.
Results: After LARC became available, 62/368 patients chose Nexplanon. Gravidity (OR: 1.128) and vaginal delivery (OR: 1.682) were negatively associated with RRP while age at delivery was inversely associated (OR: 0.902, p < .001). The adjusted mean probability of RRP was 6% for those who had IPP-LARC placement, compared with 20% for those who did not. LARC availability was positively associated with RRP but not significantly (OR: 1.287, p = .195).
Conclusions: After SB 332’s passage, IPP-LARC availability was positively associated with RRP, but the percentage of patients who chose IPP-LARC was low. However, RRP occurred less often among patients who chose IPP-LARC placement. Hospital systems must actively invest in patient education and provider support to ensure interventions to reduce rapid repeat pregnancy are optimally utilized by vulnerable populations.
- Contraceptives
- Counseling
- Desogestrel
- Drug Implants
- Electronic Health Records
- Etonogestrel
- Health Education
- Health Disparities
- Health Policy
- Long-Acting Reversible Contraception
- Maternal Health Services
- Medicaid
- Obstetrics
- Ohio
- Perinatal Care
- Population Health
- Postpartum Period
- Preconception Care
- Pregnancy
- Premature Birth
- Preventive Medicine
- Public Health
- Reproductive Health
- Women's Health
Introduction
The World Health Organization and the American College of Obstetrics and Gynecology recommend birth spacing between 18 and 24 months to avoid a variety of adverse maternal and child outcomes.1,2 However, nearly one-third of women in the United States have had a rapid repeat pregnancy (RRP), an interpregnancy interval of less than 18 months.3 RRP has been associated with preterm birth,4–6 small for gestational age,4,5 low birth weight,7 and infant mortality.5,8 RRP occurs more often among younger women,9 among women covered by Medicaid,9 and among non-Hispanic Black mothers compared with non-Hispanic White mothers and Hispanic mothers.10 Because unintended pregnancy and RRP are modifiable risk factors associated with adverse birth outcomes, reducing unintended pregnancy and RRP may reduce adverse health outcomes, mitigate health disparities, and improve population health.
Prior research has shown that the rate of RRP differs depending on the contraceptive method, if any, used between pregnancies. For instance, one study found that RRP rates were 6% with long-acting reversible contraception (LARC) and 12% with depot medroxyprogesterone (DMPA), compared with 21% with the pill, patch, or ring and 23% with no contraception.11 LARCs, which have the lowest difference between prescribed and ideal use among reversible contraceptives, are therefore promoted as the most effective method for reducing the rate of unintended pregnancy and RRP.12
While evidence demonstrates that increasing access to LARCs can significantly decrease both unintended and rapid repeat pregnancy, LARCs’ effectiveness are limited by the setting and circumstances in which they are placed. The immediate postpartum period, in particular, may present a critical time period to initiate LARC therapy. Placing a LARC during the immediate postpartum visit during hospitalization (IPP-LARC) significantly reduces RRP compared with placing a LARC during an outpatient follow-up visit.13–16
As of November 2023, most states have mandated Medicaid coverage for IPP-LARC.17 Reimbursing LARC insertion and LARC devices separately from global obstetric payments makes IPP-LARC more financially attractive to hospitals. Although most states have mandated separate Medicaid payment for both the LARC device itself and LARC insertion, some states, like Ohio, have decided to separately cover just one and not both aspects of LARC placement.17 Starting in April 2017, Ohio Senate Bill 332 (SB 332), reflecting recommendations of the Ohio Commission on Infant Mortality, required hospitals with maternity units and birthing centers to give delivering patients the option of having a LARC placed before discharge.18 SB 332 mandated Medicaid payment for the physical LARC device separate from the Global Maternity Fee that covers other inpatient services, including LARC insertion. OhioHealth Grant Medical Center, a teaching hospital located in central Ohio, subsequently made IPP-LARCs available in July 2017.17
It is unknown whether the availability of IPP-LARC after the passage of SB 332 is associated with rates of utilization of IPP-LARC and RRP in Ohio. We therefore performed a retrospective chart review to examine the rates of RRP before and after IPP-LARC became available in our hospital. We also studied how patient socio-demographic, prenatal care, and perinatal care variables were associated with RRP.
Methods
Design
This quasi-experimental observation study used a retrospective cohort design to compare rates of RRP before and after availability of IPP-LARC in an urban hospital that primarily serves a low-income, Medicaid population, and to examine differences in socio-demographic, prenatal, and perinatal variables associated with the presence or absence of RRP.
Study Sample
The cohort studied was identified and organized using REDCap.19 Inclusion criteria included having an index birth within the 1 year before IPP-LARC became available and the 1 year after IPP-LARC became available at Grant Medical Center. For adolescent patients, IPP-LARC became available on January 2016 and for adult patients IPP-LARC became available on July 2017. We followed patients for 18 months after the date of delivery to assess for RRP.
Subjects were also required to be persons with uteri aged 13 to 44 years, receive prenatal care from an OhioHealth Grant Medical Center medical education-affiliated clinic, and deliver during their second or third trimester at Grant Medical Center. We excluded subjects who died within 18 months of their delivery date; who delivered through spontaneous abortion or whose pregnancies resulted in an in utero fetal demise; who received a LARC postpartum due to an oncologic diagnosis; who delivered at Grant Medical Center but followed with a non-OhioHealth clinic for their prenatal care; who underwent a hysterectomy or bilateral salpingectomy; or who relied on their partners’ vasectomies for family planning. We excluded patients who underwent permanent sterilization in our analysis because prior randomized controlled trials of IPP- versus delayed-LARC included patients who had already decided to use LARC after delivery.20–23
Independent Variables
We collected patient demographic, prenatal care, and perinatal care variables using our Electronic Health Record, CareConnect 3.0, EpicCare Inc.24 Patient demographic data included gravidity and parity on the date of index birth, race (black vs other), language (English vs other), and insurance type (public & self-pay vs private). Prenatal care variables included gestational age at first visit and the total number of prenatal visits. Perinatal care variables included type of delivery (caesarian section vs vaginal), intent to breastfeed (yes or no), gestational age at delivery, patient age at delivery, the type of postpartum contraception chosen between delivery and hospital discharge (ie, Etonogestrel Implant [Nexplanon®], DMPA, combined oral contraceptive pill [OCP], progestin-only pill, condom, patch, vaginal ring [NuvaRing®], or none), and whether LARC was available immediately postpartum or not (pre-IPP-LARC or post-IPP-LARC). We dichotomized postpartum contraception into LARCs and non-LARCs. The only LARC we had available at our hospital was Nexplanon. The non-LARCS included DMPA, OCP, progestin-only pill, condom, patch, NuvaRing, or none.
Dependent Variable
Our outcome of interest was RRP, defined as pregnancy within 18 months of the patient’s delivery date.25 We obtained information about whether RRP occurred by examining our electronic health record system, supplementing this information using the Care Everywhere feature of EPIC and birth certificate data from the Ohio Department of Health.
Statistical Analysis
We performed descriptive analysis to characterize patient demographic, prenatal care, and perinatal care information for the pre-IPP-LARC and post-IPP-LARC cohorts. LARC (ie, Nexplanon) became available for adolescents in January 2016 and for adults in July 2017 at our hospital. The pre-IPP-LARC cohort included adolescents who delivered from January 2015 to December 2015 and adults who delivered from July 2016 to June 2017. The post-IPP-LARC cohort included adolescents with a birth from January 2016 to December 2016 and adults with a birth from July 2017 to June 2018. We followed patients for 18 months after their index births to determine whether RRP occurred. We then described patient demographic, prenatal care, and perinatal care variables (including whether LARC was available or not) according to presence or absence of RRP, assessing differences in the independent variables based on RRP status using χ2 and Student’s t test.
We used logistic regression to determine the adjusted relationship between RRP and our patient demographic, prenatal care, and perinatal care variables. The variables we included in our initial model were based on our study aims and previous research: gravidity, self-identified race, language, insurance, gestational age at first visit, total number of prenatal visits, LARC availability, delivery type, intention to breastfeed, gestational age at delivery, age at delivery, and LARC vs non-LARC.9,10 Variables were eliminated from the full model through likelihood ratio tests. To aid in interpretation, we used STATA’s margins command to calculate the adjusted probability of a rapid repeat pregnancy by each of the variables in our full and final (reduced) model.
All analyses were performed using STATA software, Version 14.2 Seconds.E. (StataCorp, College Station, TX).
Results
Our study sample comprised 726 subjects. Mean gravidity and parity were 3.2 and 2.6, respectively, and were similar for both the pre-IPP-LARC and post-IPP-LARC time periods (Table 1). Roughly 63% of subjects identified as Black in the total study sample, with a greater proportion of Black subjects in the pre-IPP-LARC group compared with the post-IPP-LARC group. Most subjects spoke English and were covered by public insurance or not insured.
Patient Demographic, Prenatal Care, and Perinatal Care Variables by LARC Availability
The mean gestational age at the first visit was higher in the pre-IPP-LARC group (15.5 weeks) than the post-IPP-LARC group (14.4 weeks), but the total number of prenatal visits was similar in both groups.
When we combined both the pre-IPP-LARC and post-IPP-LARC time periods, roughly three-fourths of deliveries were vaginal deliveries. Fewer deliveries in the pre-IPP-LARC period were vaginal deliveries (72%) than in the post-IPP-LARC period (77%). There were fewer subjects who intended to breastfeed in the pre-IPP-LARC period (64%) compared with the post-IPP-LARC period (70%). Gestational age at delivery and age at delivery were roughly the same during both periods.
Zero percent of patients (0/358) in the pre-IPP-LARC period and 17% of patients (62/368) in the post-IPP-LARC period chose a LARC (Nexplanon). Among patients who chose a non-LARC, most chose DMPA or no contraception during both the pre-IPP-LARC and post-IPP-LARC time periods.
When we examined our independent variables by presence or absence of RRP, we found that the group who had RRP (yes-RRP group) had a greater proportion of patients who identified as Black than the group who did not (no-RRP group) but not to a statistically significant degree (69% vs 61%; p = .083) (Table 2). Ninety-five percent of the yes-RRP group were uninsured or covered by public insurance, compared with 89% in the no-RRP group (p = .053). The yes-RRP group also had a lower mean number of total prenatal visits than the no-RRP group (8.3 vs 8.9; p < .05).
Patient Demographic, Prenatal Care, and Perinatal Care Variables by RRP
Interestingly, the yes-RRP and no-RRP groups did not differ significantly by LARC availability in the immediate postpartum period (52% vs 50%, p = .646). On the other hand, the type of delivery differed significantly between the yes-RRP and no-RRP groups, with 84% of the yes-RRP group and 72% of the no-RRP group having vaginal deliveries (p = .004). Patients in the yes-RRP group were younger at delivery compared with the no-RRP group (24.7 years old vs 26.9 years old, p < .0001). Three percent of patients in the yes-RRP group compared with 10% of patients in the no-RRP group (p = .013) chose IPP-LARC for their immediate postpartum contraception.
The reduced model to calculate the odds of RRP included gravidity, LARC availability, type of delivery, patient age at delivery, and whether a LARC was used or not (LARC vs non-LARC) immediately postpartum (Table 3). Gravidity was positively associated with RRP; the adjusted mean probability of RRP was 0.133 with a gravidity of 1 compared with 0.175 with gravidity of 5 (Appendix Figure 1). Patients who were younger at delivery had a greater probability of RRP compared with those who were older; the adjusted mean probability of RRP was 0.447 for subjects aged 14 and 0.052 for subjects aged 40 (Appendix Figure 2). C-section delivery was negatively associated with RRP; the adjusted mean probability of RRP was 0.203 for patients who delivered vaginally and 0.132 for patients who delivered through c-section.
Adjusted Associations Between Rapid Repeat Pregnancy and Patient Demographics, Prenatal Care Variables, and Perinatal Care Variables
Availability of LARC immediately postpartum was significantly associated with RRP in neither the full model that adjusted for patient demographic, prenatal care, and perinatal care variables nor the final, reduced-form model (Table 3). Patients who chose LARC immediately postpartum had an adjusted mean probability of 6% for RRP. Patients who did not choose LARC had a probability of 20% of RRP.
Study Limitations
Because this is an observational study, we are unable to make claims of causality about the availability of IPP-LARC and RRP. Moreover, since only Nexplanon and not intrauterine devices (IUDs) were offered as IPP-LARC, our results are not generalizable to hospital systems that offer both types of contraception in the immediate postpartum period. Our study exclusively looked at patients who received prenatal care at OhioHealth Grant Medical Center medical education-affiliated clinics. Studies of patients following with private physicians, not associated with a medical education-affiliated clinic, or with higher rates of private insurance coverage may yield different associations between IPP-LARC availability and RRP. Although we supplemented data on RRP from our electronic health record with state birth data, we may have missed RRP for patients who moved out of state or transitioned care to a health system with electronic health records that are not connected to our EHR. Since we excluded patients who relied on bilateral salpingectomy or vasectomy, underwent a hysterectomy, or received a LARC postpartum because of an oncologic diagnosis, our results are not generalizable to these populations. Lastly, we did not collect data on contraception discontinuation, changes, and initiation during postdelivery outpatient follow-up visits, so we are unable to determine the associations between these variables and RRP; data from these variables may have explained the lack of association between IPP-LARC availability and RRP.
Discussion
In our urban, tertiary-care hospital located in Ohio that treats a primarily underserved, publicly insured obstetrics population, state legislation mandating Medicaid coverage for IPP-LARC devices was positively associated with RRP. Congruent with previous research,14,16 patients who chose a LARC after IPP-LARCs became available had lower odds of RRP than those who did not chose LARC (OR 0.254; 95% CI 0.097-0.665). In addition, consistent with previous research, only 3% of those who chose Nexplanon in the immediate postpartum period had RRP.15
Our finding that the availability of IPP-LARC is associated with increased RRP in our final model is not consistent with previous randomized controlled studies that show that IPP-LARC results in less RRP compared with delayed LARC placement.16 However, in contrast to our observational study, where patients were offered Nexplanon and several non-LARC methods immediately after delivery, in these randomized controlled trials participants were included only after they had agreed to IPP-LARC placement.20–23
The positive association between IPP-LARC availability and RRP may have also resulted from inadequate statistical power since only about 17% of patients chose IPP-LARC after it became available. Future studies that use a longer time period before and after IPP-LARC availability would result in a larger sample size and this larger sample size, in turn, may yield a negative association between IPP-LARC availability and RRP. Moreover, since our hospital did not offer IUDs as a form of IPP-LARC, patients who preferred an IUD may have chosen a less effective form of contraception over an IPP-LARC, thus diminishing a potential association between IPP-LARC availability and RRP. In another observational study, 66% of patients who initiated LARC before discharge chose a contraceptive implant and 34% chose an IUD.26 Additionally, since Ohio offers separate reimbursement for LARC devices but not placement,17 the financial incentive to place an IPP-LARC may have been insufficient for providers to promote IPP-LARC placement. Our hospital also did not offer a structured program to promote and educate hospital providers about the availability of IPP-LARC, which may have adversely affected uptake of IPP-LARC.27,28
The 17% of patients who chose Nexplanon placement during the post-IPP-LARC period is similar to the 2 to 11% of patients who underwent IPP-LARC placement among states that were early adopters of mandating IPP-LARC coverage for Medicaid patients.29 Interestingly, in this state-level analysis, only 10% (38/366) of hospitals provided IPP-LARC for more than 1% of deliveries. In general, these 38 hospitals were nonrural, not affiliated or owned by Catholic organizations, and teaching institutions–all characteristics of our hospital.
Patients in our sample with higher gravidity at delivery, younger age, and vaginal deliveries had higher odds of RRP compared with their counterparts in our final model. Interestingly, while mean gravidity was similar between patients that experienced RRP compared with those who did not in descriptive bivariate analysis, the odds of RRP were greater among patients with higher gravidity in our final model. The direction of the association between RRP and age at delivery and delivery type (vaginal vs c-section) were roughly consistent between our descriptive analysis and our reduced form model. We hypothesize that age at delivery and delivery type may have confounded the association between gravidity and RRP since older patients may have higher gravidity and may be more likely to undergo c-section.30 Gemmil et al. also found a positive association between gravidity and RRP and between younger age and RRP.9 In addition, our finding that the mean probability of RRP is higher among patients who delivered vaginally compared with c-section may be due to the increased education provided to patients who underwent c-section on the risks of RRP and increased likelihood for a repeat c-section.
Given the positive association between IPP-LARC availability and RRP in our final model, tailoring interventions to patient preferences, life stage (age and gravidity), and type of delivery may be helpful in decreasing the “time lag” between IPP-LARC availability and widespread IPP-LARC uptake.31 One possible intervention could be a structured program designed to increase provider awareness and buy-in about IPP-LARC, thus fostering more robust discussions about IPP-LARC. This intervention should take advantage of the 24 to 72 hours during which patients who have recently delivered are still under the care of their treatment teams. These teams would use the time that they have available to provide contraception counseling and effective contraception. Ideally, this intervention would also offer both forms of IPP-LARC (IUDs and hormonal implants), and payment for both the LARC itself and insertion to incentivize providers.
Our findings suggest that it is not enough for health systems to institute policies making IPP-LARC available to reduce RRP. To bridge the gap between the decreased RRP seen in trials of IPP-LARC and real-world settings,32 health systems must actively invest in patient education and provider support,33 and offer all long-term contraception options to ensure innovations like IPP-LARC are optimally utilized by vulnerable populations.
Acknowledgments
We appreciate Robin Devine’s support on this project.
Appendix
Figure shows the adjusted mean probability of RRP by gravidity. We adjusted for time period (i.e. immediate postpartum LARC available vs not), type of delivery (vaginal vs c-section), patient age at delivery, and whether LARC was placed or not (LARC vs non LARC). The adjusted mean probability of rapid repeat pregnancy was 0.133 for those with a gravidity of 1 compared to 0.175 for those with gravidity of 5.
Figure shows the adjusted mean probability of RRP by age at delivery. We adjusted for time period (i.e. immediate postpartum LARC available vs not), type of delivery (vaginal vs c-section), gravidity, and whether LARC was placed or not (LARC vs non LARC). The adjusted mean probability of rapid repeat pregnancy was 0.447 for subjects aged 14 and 0.052 for subjects aged 40.
Notes
This article was externally peer reviewed.
Funding: None.
Conflict of interest: None.
Author contribution statement: RA conceived the study and had full access to all study data and takes responsibility for the integrity of the data and the accuracy of the data analysis. AH, BC, RA, CF, AC, and JY analyzed the data, drafted the manuscript, and contributed substantially to its revision.
- Received for publication October 10, 2024.
- Revision received June 17, 2025.
- Accepted for publication June 30, 2025.
References
Jump to section
Related Articles
Cited By...
More in this TOC Section
Similar Articles
Keywords
- Contraceptives
- Counseling
- Desogestrel
- Drug Implants
- Electronic Health Records
- Etonogestrel
- Health Education
- Health Disparities
- Health Policy
- Long-Acting Reversible Contraception
- Maternal Health Services
- Medicaid
- Obstetrics
- Ohio
- Perinatal Care
- Population Health
- Postpartum Period
- Preconception Care
- Pregnancy
- Premature Birth
- Preventive Medicine
- Public Health
- Reproductive Health
- Women's Health