Pregnancy Rates Among Women Treated with Medication for Opioid Use Disorder

BACKGROUND: Treatment-seeking people with opi - oid use disorder (OUD) who are capable of pregnancy need accurate information about the potential impact of medication to treat OUD (MOUD) on fertility to make informed choices about treatment that are consistent with their reproductive wishes. There is a dearth of research on fertility associated with MOUD receipt in birthing people with OUD.

To estimate the association between treat- ment with MOUD and odds of conception among birth- ing people using national administrative claims. DESIGN: Retrospective case-crossover study using multi-state US administrative data (2006–2016). Dates of conception were estimated from delivery dates and served as “case” days for which MOUD exposures were compared to those on all other (“control”) days of insur- ance enrollment. PARTICIPANTS: Treatment-seeking people with OUD with a delivery during the observation period. MAIN MEASURES: Odds ratios for conception from within-person fixed effects models were modeled as a function of exposure to MOUD (buprenorphine, metha- done, extended-release depot naltrexone, or oral nal - trexone) using conditional logistic regression. KEY RESULTS: A total of 21,928 births among 19,133 people with OUD were identified. In the sample, 5873 people received buprenorphine, 1825 methadone, 486 extended-release naltrexone, and 714 oral naltrex- one. Participants could receive more than one type of MOUD. Mean age was 28.2 years (SD = 2.2; range = 16–45), with 76.2% having Medicaid. vs. commercial insurance. Compared to no MOUD, periods of metha- done (aOR = 0.55 [95% CI = 0.48–0.63]) or buprenor - phine receipt (aOR = 0.84 [0.77–0.91]) were associ- ated with fewer conceptions. Treatment periods with extended-release depot naltrexone compared to no medication were associated with higher odds of con- ception (aOR = 1.75 [1.22–2.50]) and there was no sig- nificant difference in conception with oral naltrexone (aOR = 1.02 [0.67–1.54]).

The association between MOUD and odds of conception among birthing people varied by type of MOUD, with extended-release naltrexone associated with higher odds of conceiving compared to no treat- ment. Clinical studies are urgently needed to investigate these findings further. KEY WORDS: fertility; pregnancy; opioid use disorder; buprenorphine; naltrexone; methadone J Gen Intern Med DOI: 10.1007/s11606-024-08689-8 © The Author(s), under exclusive licence to Society of General Internal Medicine 2024

Estimates of unintended pregnancy among people capable of pregnancy with opioid use disorder (OUD) are higher than the general population.1–5 Yet, simultaneously, chronic opi- oid use is hypothesized to reduce fertility by suppressing the hypothalamic-pituitary axis (HPA).6 Among women taking opioids for chronic non-malignant pain, more than half aged 30–50 years experienced amenorrhea compared to 20% in the control group.7 Other studies have shown lower levels of estradiol, luteinizing hormone, follicle-stimulating hormone, free and total testosterone, and dehydroepiandrosterone sul- fate (DHEAs), concerning for impaired endocrine function and sex steroid balance. 7–11 Amid this complex interplay, many women with OUD may be unsure if they can conceive at the time they enter treatment.12 The association between medication to treat opioid use disorder (MOUD) and fertility is understudied. Estimates of fertility include pregnancies that end in live births, in contrast to fecundity which is the capability to produce offspring,13 Although it can be posited that the partial and complete opioid agonists, i.e., buprenorphine and metha- done, may be associated with lower rates of pregnancy than the opioid antagonist naltrexone, this has not been inves- tigated using human clinical or real-world data. Low-dose naltrexone is posited to induce ovulation through poten- tial immunomodulating effects14 specifically by stimulat- ing endorphin production. 15,16 While naltrexone is not an approved treatment for infertility in the US, some studies have suggested naltrexone may facilitate menstruation in people with hypothalamic amenorrhea17 and improve odds of ovulation in people with polycystic ovarian syndrome18 Jennifer K. Bello and Kevin Y. Xu have equal contribution.Received October 27, 2023 Accepted February 16, 2024 1343Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM and endometriosis. 19 However, the association of naltrex- one with fertility overall is poorly characterized. To gener - ate essential preliminary data and catalyze future investiga- tion, we use multi-state insurance claims data to estimate the association between treatment with MOUD and odds of conception among people with delivery between 2006 and 2016. We hypothesized that people treated with a full opi- oid agonist may have a greater reduction in odds of concep- tion that ends in live birth, relative to no MOUD than those treated with an opioid partial agonist, and that treatment with an opioid antagonist would increase the odds of conception.

Summary of Analytic Design Our retrospective analysis employed a case-crossover design using multi-state insurance claims data to investi- gate whether days characterized by conception events, ascer- tained via procedure codes on claims, were associated with days characterized by MOUD receipt. The within-person case-crossover design was previously described in detail20–22 and is illustrated in eFigure 1. In brief, the case-crossover

characterizes each person who experiences concep- tion by the presence or absence of treatment with medica - tions.23,24 Case days correspond to estimated dates of con- ception; control days correspond to all other days. Only people with at least one conception can be analyzed because the design requires within-person variation on the outcome variable. Using a conditional logistic regression approach and conditioning on a person, each person serves as their own control and associations can be estimated by comparing exposures at the time of conception (outcome variable) to exposure during control periods, with the exposure of inter- est being types of MOUD. Overview of the Dataset and Cohort We used the Merative™ MarketScan® Commercial and Multi-State Medicaid Databases (2006–2016), 22 which include claims for inpatient, outpatient, and prescription drugs from over 200 million unique people in the US with employee-sponsored commercial health insurance or Medic- aid. We extracted data on clinical information (i.e., delivery, sterilization, and contraception insertion procedures), enroll- ment, and pharmacy claims. Because no identifiable private data were used, this study was deemed exempt from IRB review by the authors’ Research Protection Office. Our cohort of birthing reproductive-age people (ages 16–45 years), with at least one insurance claim for live birth and conception estimated to have happened during insurance enrollment, was derived from a larger OUD treatment cohort previously described. Briefly, that cohort comprised 304,676 individuals aged 12–64 years with OUD, commercial insur- ance (2006–2016) or Medicaid (2011–2016), prescription drug coverage, and at least one OUD treatment claim. 22 After excluding 155,618 men and people outside the age range of 16–45, we derived 123,311 reproductive-age peo- ple, of whom 29,919 had at least one delivery during insur - ance enrollment. See Figure  1 for flow diagram of inclusion and exclusion criteria. OUD was operationalized by ICD9/10 diagnoses for opioid “abuse or dependence.” (These terms are still used for International Classification of Disease (ICD) diagnostic codes despite the stigmatizing nature of the term “abuse”25,26). OUD treatment was defined by receipt of at least one of the following services: prescription fills for buprenorphine, oral naltrexone or extended-release naltrex- one, procedures for methadone or extended-release naltrex- one, or procedure codes for psychosocial treatment with an OUD diagnosis. Outcome Variable After identifying all people with live births from the OUD treatment cohort, we generated a dataset comprising days nested within persons with each day a person was enrolled in insurance corresponding to one observation, i.e., a “per - son-days” level data set. Conception dates were estimated for each delivery as delivery date—270 days. We identified dates of deliveries using CPT and ICD-9/10 procedure codes and preterm deliveries using ICD9/10 diagnosis codes. In cases where the delivery was accompanied by a diagnostic code for premature delivery, the conception date was esti- mated as delivery date—240 days. 27 Then the outcome vari- able (conception) was coded as 1 on estimated conception date and 0 on all other dates. Individuals ( n = 3015) who received sterilization procedures were censored following the procedure. (See eTable 1 for CPT codes for steriliza- tion procedures.28 ) In addition, periods between conception and delivery were censored. Because conception dates are estimated from deliveries, some conceptions were excluded because they occurred on dates that fell outside of the range of insurance enrollment (thus precluding the observation of medication exposures at time of conception). Our final cohort thus consisted of 21,922 deliveries among 19,133 people. In this analysis, we examine fertility—conceptions that end in a live birth, rather than fecundity—the biological capacity to reproduce; thus, conception ending in pregnancy loss (known or unknown) or termination was not classified as conception days. We include a supplemental discussion in eMethods 2 describing the effect on the estimates if misclas- sification error is present. Exposures The primary predictor variables were MOUD exposure (buprenorphine vs methadone vs naltrexone-PO vs naltrex- one-extended release) coded at a day level with all days of insurance eligibility without MOUD serving as the reference condition. For each person-day, we recorded all variables 1344 Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM of interest from the relevant pharmacy claims or procedure codes. For oral naltrexone and buprenorphine, we identified all prescription fills using NDC codes with the assumption that each date of medication possession corresponded to a date of medication consumption. Days of medication pos- session were imputed from the prescription fill date and the “days’ supply” field. Extended-release naltrexone utiliza- tion was identified using NDCs and HCPCS code J2315; we assumed that the medication was active for 30 days after injection. Methadone dispensing was identified using Healthcare Common Procedure Coding System procedure code H0020 with days’ supply imputed from the number of units (days) or, in cases where units were missing or zero, the service end date. Because buprenorphine monproduct was used much less frequently than buprenorphine-naloxone (i.e., monpoproduct comprised only 5.1% of all use), our analyses combined buprenorphine monoproduct and combi- nation product. Because conception dates are approximate, we include a sensitivity analysis to simulate the effect of additional random error in estimating the conception date (see eTable 2 and accompanying text for further rationale). Covariates As a within-person analysis, all covariates that do not vary over time are implicitly controlled in the conditional logistic regression analysis (i.e., by conditioning on person). Birth control methods other than sterilization were included as time-varying covariates in the analysis because they are reversible and exhibit various failure rates even with perfect use. Moreover, since contraceptive medications are expected to markedly reduce but not eliminate the possibility of con- ception, they can serve as “positive control” variables, i.e., variables for which we expect to see strong and significant (negative) effect sizes. A sensitivity analysis examined the impact of excluding days on which birth control methods were used (and their associated conceptions) from the analy- sis (rather than including contraceptive methods as covari- ates; eTable 3). This resulted in the removal of 969 con- ceptions from the cohort described above. See eMethods 1 for details on contraception coding. Additional sensitivity analysis added a covariate for opioid use (eTable 4). Because days of heterogeneity in the reference condition (i.e., days on which MOUD was not received could reflect pre-treatment, relapse, or successful recovery), two additional sensitivity analyses were conducted. One excluded all days on which 123,311 reproduc/g415ve-age women, 16-45 years 304,676 people aged 12-64 with a diagnosis of opioid "abuse or dependence" in the MarketScan Commercial (2006-2016) and Mul/g415-State Medicaid (2011-2016) Database who have prescrip/g415on drug coverage and at least one treatment claim for opioid "abuse or dependence" during enrollment 149,058 women aged 12-64 years 19,133 reproduc/g415ve-age women, 16-45 years, with at least one delivery falling during enrollment 29,919 reproduc/g415ve-age women, 16-45 years, with at least one delivery EXCLUDE 155,618 men, aged 12-64 EXCLUDE 25,747 women 46 years EXCLUDE 93,392 reproduc/g415ve-age women without at least one delivery EXCLUDE 10,786 people without deliveries falling during enrollment Figure 1 Derivation of the analytic sample. 1345Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM no MOUD was received (eTable 5). A second coded all days prior to the first observed initiation of treatment as the refer- ence condition, while all subsequent days without MOUD were coded as “post-treatment” days (eTable 6). Statistical Analysis First, to describe the clinical and demographic character - istics of the cohort, we computed demographics and the prevalence of co-occurring psychiatric and substance use disorders in the 6 months preceding the start of treatment, as well as medical comorbidity ascertained via the Charlson Comorbidity Index. Next, we conducted a case-crossover analysis, with each individual serving as their own control using conditional logistic regression to model the likelihood of conception as a function of MOUD exposure while con- ditioning on person. As a within-person analysis, the case- crossover design implicitly controls time-invariant confound- ing; although time-varying confounding may still occur,23,24 this was mitigated using the date as a covariate (restricted cubic spline). Analyses were conducted using SAS 9.4. P values were 2-sided, and statistical significance was set at P = .05. Data were analyzed from February 1, 2023, through January 26, 2024. There are no additional data available. Authors intend to provide relevant code on written reason- able request.

Our sample consisted of 19,133 reproductive-age people initiat- ing treatment for OUD, age 16–45, who had at least 1 delivery during insurance eligibility, encompassing 21,928 instances of conception. Demographic and clinical characteristics are denoted in Table 1. The mean age at treatment initiation is 27.2 (SD 2.2, Mdn 27). Treatment was as follows: 5873 (30.7%) received buprenorphine (1,309,383 person-days), 1825 (9.5%) methadone (504,399 person-days), 714 (3.7%) oral naltrexone (35,731 person-days), 486 (2.5%) extended-release naltrexone (32,280 person-days). Because the analysis is at the level of person-days, participants could have used more than one type of MOUD during the observation period. In the conditional logistic regression (case-crossover) analysis (Table  2), person-days of buprenorphine and methadone were associated with lower odds of conception (aOR = 0.84 [0.77–0.91] for buprenorphine, aOR = 0.55 [0.48–0.63] for methadone). In contrast, extended-release naltrexone was associated with 75% increased odds of conception (aOR = 1.75, 95% CI 1.22–2.50). No relation- ship was observed between person-days of oral naltrexone exposure and conception. As expected, days of contracep- tion exposure (positive control) correlated strongly with decreased odds of conception. Sensitivity analyses examining the robustness of results to artificially increased error in the estimation of concep- tion dates (eTable 2) yielded slightly wider confidence intervals but no significant changes in OR estimates. Anal- yses eliminating days on which contraception was utilized (vs. inclusion of contraception variables as covariates; eTable 3) similarly showed no significant differences from the primary model. The inclusion of prescription opioids as a covariate (eTable 4) did not impact results. Elimina- tion of days without MOUD exposure from the analyses (eTable 5) showed the same order of OR estimates for MOUDs (Naltrexone-ER > buprenorphine > methadone). Analyses differentiating pre-treatment from post-treatment days without MOUD exposure suggested that odds of con- ception were lower on all days following initiation of treat- ment (eTable 6), but otherwise showed similar associations between MOUD and conception.

In a cohort of reproductive-age people with OUD who expe- rienced a delivery between 2006 and 2016, we observed an association between MOUD type and odds of concep- tion, with the magnitude and direction of the association Table 1 Demographic and Clinical Characteristics (n = 19,133) * Naltrexone PO: oral naltrexone † Naltrexone ER: long-acting injectable Naltrexone Characteristic N (percentage) Demographics   Age >30 years 5805 (30.3)   Mean age (sd) 27.2 (2.2)   Medicaid 14,582 (76.2)   Race (among Medicaid only)     Non-Hispanic White 12,412 (87.1)     Non-Hispanic Black 948 (6.7)     Hispanic 140 (1.0)     Other 753 (5.3) Co-occurring disorders   Mood disorder 7435 (38.9)   Anxiety disorder 6142 (32.1)   Psychotic disorder 331 (1.7)   Alcohol use disorder 1753 (9.2)   Cocaine use disorder 1583 (8.3)   Amphetamine use disorder 1207 (6.3)   Sedative use disorder 1276 (6.7)   Cannabis use disorder 2949 (15.4)   Charlson comorbidity index (CCI)     CCI = 0 17,568 (91.8)     CCI = 1 1347 (7.0)     CCI = 2 156 (0.8)     CCI = 3 62 (0.3) MOUD types   Buprenorphine 5873 (30.7)   Methadone 1825 (9.5)   Naltrexone PO * 714 (3.7)   Naltrexone ER † 486 (2.5) 1346 Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM varying by MOUD type. The associations identified between MOUD and delivery are consistent with past research on the physiological mechanisms of MOUDs and how they impact fertility. While deliveries do not represent all conceptions that may have occurred during the observation period, they provide a preliminary estimate of how MOUD may impact fertility, defined as pregnancies that end in delivery.13 Con- sistent with the suppression of the HPA axis found with chronic opioid use, 6 we observed that full opioid agonists (methadone) were associated with lower odds of concep- tion relative to MOUD-free days, while partial agonists (buprenorphine) were associated with a smaller reduction in the odds of conception relative to methadone. Those who received the opioid antagonist naltrexone in the extended- release depot formulation had higher odds of conceiving, while treatment with oral naltrexone, compared to no treat- ment, had no association. The lack of association between oral naltrexone and odds of conception may be attributable to high treatment discontinuation rates associated with oral naltrexone,29 relative to extended-release formulations. Despite potential improvements in the odds of ovulation associated with low-dose naltrexone in preliminary stud- ies,14–16,18 the real-world impact of naltrexone at doses used to treat OUD in both oral and long-acting depot formulations is not typically included in prescribing information for these medications. Moreover, women have been underrepresented in studies of naltrexone in opioid use disorder,30 representing a significant gap in our understanding of the effects of these medications on fertility. Our findings are strengthened by the use of active controls. Consistent with estimates of contraceptive method effective- ness, person-days while using a contraceptive method were associated with strong reductions in odds of conception. In addition, the highest reduction in conception was found with the use of methods that do not require daily medication con- sumption, such as LARC methods and the medroxyproges- terone injection.31 Another important strength of our study was its use of a within-person case-crossover approach, mitigating selection and sampling bias that would result from recruitment of different case and control persons. Research has largely focused on helping people with OUD avoid or delay pregnancy as well as treating OUD after pregnancy is established, leading to a missed opportunity to consider that people with OUD may desire pregnancy or be unsure how they feel about conceiving. 32 A qualitative analysis of sexually active, non-pregnant people revealed that people with OUD may believe they are infertile amid a history of unprotected sex without conceiving33; partici- pants’ belief that they were infertile was cited as a reason to not use contraception. In another study, pregnant people who used substances were twice as likely to believe they could not get pregnant as people who did not use substances. 34 These factors underscore the need to use a reproductive jus- tice lens to address the whole person when treating people with OUD capable of pregnancy.35 Providing education about the impact of medical and non- medical drugs on reproductive function is a core component of a reproductive justice approach. Namely, clinicians need the knowledge and tools to engage in discussion with peo- ple about their perceptions of fertility and the impact that stopping opioids may have on their likelihood of pregnancy, as their HPA axis function returns to normal, and the ways that the type of MOUD they choose may also impact fertil- ity. Filling this gap in knowledge is critical to allow clini- cians treating patients with OUD to provide patient-centered counseling that will help people make choices about MOUD that are in line with their reproductive wishes. For example, if a person with OUD does not desire pregnancy, knowing treatment with long-acting depot naltrexone may increase the odds they will conceive, a clinician should provide accu- rate information about all forms of contraception including LARCs, specifically their safety and efficacy compared to other contraception methods.36 If pregnancy is desired, and long-acting depot naltrexone is the preferred MOUD, a dis- cussion weighing the risks and benefits of becoming preg- nant while on this medication is needed. Table 2 Association of Person-Days of Medication for Opioid Use Disorder (MOUD) Receipt and Conception * Naltrexone ER: long-acting injectable Naltrexone † Naltrexone PO: oral naltrexone ‡ LARC: long-acting reversal contraception § IUD: intrauterine device (progesterone and copper) Effect OR (95% CI) Buprenorphine 0.84 (0.77–0.91) Methadone 0.55 (0.48–0.63) Naltrexone-ER * 1.75 (1.22–2.50) Naltrexone-PO † 1.06 (0.72–1.55) Oral contraception (positive control, n = 5268) 0.39 (0.35–0.43) Patch (positive control, n = 366) 0.38 (0.23–0.63) Ring (positive control, n = 1016) 0.42 (0.32–0.55) LARC ‡ (IUD, § Implant) (positive control, n = 3465) 0.06 (0.06–0.07) Medroxyprogesterone Injection (positive control, n = 1913) 0.02 (0.01–0.04) 1347Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM There are several limitations of this analysis. First, phar - macy claims do not always reflect actual consumption of medication, and misclassification bias cannot be ruled out. However, for MOUD in the MarketScan Databases, our group has found that the proportion of days covered in a treatment episode was high, suggesting few gaps between refills and providing some evidence of likely consumption.37 Second, the MarketScan data are limited to services billed through insurance claims, which will not capture persons receiving medications with out-of-pocket payment; because many OUD treatment programs do not accept insurance, our study may be underestimating the number of people who receive MOUD, especially methadone dispensing. 38 Simi- larly, our analysis does not account for contraception self- discontinuation or use of non-provider-dependent methods, such as condoms. Third, because conception was estimated via delivery date, our analysis does not account for termi- nations and pregnancy losses. Fourth, pregnancy intention is not captured in claims data, meaning it is not known if during the study period someone was actively trying to con- ceive or prevent pregnancy which may impact their like - lihood of conception regardless of MOUD. We attempt to address this limitation by removing days that contraceptive

were being used, presumably to avoid pregnancy. Fifth, despite the use of positive controls as well as a case- crossover design, we cannot rule out residual confounding by indication. For example, estimates may reflect associations of specific MOUDs with unmeasured exposures, such as illicit substance use, pregnancy intentions, sexual behaviors, and pregnancy loss or termination, as well as impacting the likelihood of conception. Finally, non-treatment days could reflect a combination of potential illicit opioid use, pre-OUD days, and post-treatment days, as well as abstinence from opioids. However, in a sensitivity analysis limited to only person-days when individuals were receiving MOUD, the association between MOUD in overall magnitude and direc- tion remains robust. In summary, treatment-seeking people with a delivery during the observation period who received MOUD experi- enced increased or decreased odds of conception that were associated with MOUD treatment days and differed by type of MOUD. These findings do not support a call to initiate naltrexone for individuals of reproductive age. Rather the findings support that data on the influence of MOUD on fertility is an important part of preserving patients’ auton- omy and meeting them where they are. These associations warrant further investigation because people capable of pregnancy lack the information needed to make informed choices about MOUD treatment considering their reproduc- tive wishes. Supplementary Information The online version contains supplementary

available at https:// doi. org/ 10. 1007/ s11606- 024- 08689-8. Corresponding Author: Jennifer K. Bello, MD, MS; Department of Family and Community Medicine, Saint Louis University School of Medicine, St. Louis, MO, USA (e-mail: jennifer.bellokottenstette@ health.slu.edu). Author Contribution: Patricia Cavazos-Rehg PhD and Laura Bierut MD of the NIDA K12 Program of Washington University for obtaining funding to support effort for personnel (Dr. Xu); Matthew Keller MS, John Sahrmann MS, Dustin Stwalley MA from the Center for Administrative Data Research (CADR) of Washington University for technical support of the MarketScan Databases. Merative and MarketScan are trademarks of Merative Corporation in the US, other countries, or both. Funding This project was funded by R21 DA044744 (PI: Richard Grucza/Laura Bierut). Effort for some personnel was supported by grants K23 DA053433 (Jennifer K. Bello) and T32 DA015035 (Kevin Xu, PI: Kathleen Bucholz, Jeremy Goldbach) and St. Louis University Research Institute Fellowship (Richard Grucza), but these grants did not fund the analyses of the Merative™ MarketScan® Commercial and Multi-State Medicaid Database data performed by Dr. Xu. In addition, we acknowledge Matt Keller MS, John Sahrmann MS, Dustin Stwalley MA, and the Center for Administrative Data Research (CADR) at Wash- ington University for assistance with data acquisition, management, and storage. Merative and MarketScan are trademarks of Merative Corporation in the US, other countries, or both. CADR is supported in part by the Washington University Institute of Clinical and Transla- tional Sciences via grants UL1 TR002345 (from the National Center for Advancing Translational Sciences of the National Institutes of Health). Data Availability No additional data available. We intend to pro- vide relevant code on written reasonable request. Declarations: Conflict of Interest: Dr Grucza reported receiving grants from the NIH and Arnold Ventures LLC during the conduct of the study, con- sulting for Janssen Pharmaceuticals, and receiving personal fees for grant reviews from the NIH and Washington University outside the submitted work. Dr. Bello reported receiving personal fees for grant reviews from the NIH outside of the submitted work.

1. Heil SH, Jones HE, Arria A, et al. Unintended pregnancy in opioid- abusing women. J Subst Abuse Treat. 2011;40(2):199-202. 2. Rossen LM, Hamilton BE, Abma JC, et al. Updated methodology to estimate overall and unintended pregnancy rates in the United States. Vital and health statistics. Series 2, Data evaluation and methods research. 2023; 201. [Internet]. https:// stacks. cdc. gov/ view/ cdc/ 124395. Accessed 24 Aug 2023 3. Fischbein RL, Lanese BG, Falletta L, Hamilton K, King JA, Kenne DR. Pregnant or recently pregnant opioid users: contraception deci- sions, perceptions and preferences. Contracept Reprod Med. 2018;3:4. 4. Collier KW, MacAfee LK, Kenny BM, Meyer MC. Does co-location of medi- cation assisted treatment and prenatal care for women with opioid use disorder increase pregnancy planning, length of interpregnancy interval, and postpartum contraceptive uptake? J Subst Abuse Treat. 2019;98:73-7. 5. Smith C, Morse E, Busby S. Barriers to Reproductive Healthcare for Women With Opioid Use Disorder. J Perinat Neonatal Nurs. 2019;33(2):E3-e11. 6. Corsi DJ, Murphy MSQ. The Effects of opioids on female fertility, preg- nancy and the breastfeeding mother-infant dyad: A Review. Basic Clin Pharmacol. 2021;128(5):635-41. 7. Daniell HW. Opioid endocrinopathy in women consuming prescribed sustained-action opioids for control of nonmalignant pain. J Pain. 2008;9(1):28-36. 1348 Bello et al.: Pregnancy Rates and MOUD (2024) 39:1342–1348 JGIM 8. Rhodin A, Stridsberg M, Gordh T. Opioid Endocrinopathy: A Clini- cal Problem in Patients With Chronic Pain and Long-term Oral Opioid Treatment. Clin J Pain. 2010;26(5):374-80. 9. Abs R, Verhelst J, Maeyaert J, et al. Endocrine consequences of long- term intrathecal administration of opioids. J Clin Endocrinol Metab. 2000;85(6):2215-22. 10. Kim CH, Garcia R, Stover J, Ritchie K, Whealton T, Ata MA. Andro- gen deficiency in long-term intrathecal opioid administration. Pain Physician. 2014;17(4):E543-8. 11. Reddy RG, Aung T, Karavitaki N, Wass JA. Opioid induced hypog- onadism. Bmj. 2010;341:c4462. 12. Stowell MA, Thomas-Gale T, Jones HE, Binswanger I, Rinehart DJ. Perspectives among women receiving medications for opioid use disorder: Implications for development of a peer navigation inter - vention to improve access to family planning services. Subst Abus. 2022;43(1):722-32. 13. Leridon H. Studies of fertility and fecundity: comparative approaches from demography and epidemiology. C R Biol. 2007;330(4):339-46. 14. Li Z, You Y, Griffin N, Feng J, Shan F. Low-dose naltrexone (LDN): A promising treatment in immune-related diseases and cancer therapy. Int Immunopharmacol. 2018;61:178-84. 15. Wildt L, Leyendecker G, Sir-Petermann T, Waibel-Treber S. Treat- ment with naltrexone in hypothalamic ovarian failure: induction of ovulation and pregnancy. Hum Reprod. 1993;8(3):350-8. 16. Roozenburg BJ, van Dessel HJ, Evers JL, Bots RS. Successful induc- tion of ovulation in normogonadotrophic clomiphene resistant anovula- tory women by combined naltrexone and clomiphene citrate treatment. Hum Reprod. 1997;12(8):1720-2. 17. Genazzani AD, Petraglia F, Gastaldi M, Volpogni C, Gamba O, Genaz- zani AR. Naltrexone treatment restores menstrual cycles in patients with weight loss-related amenorrhea. Fertil Steril. 1995;64(5):951-6. 18. Ahmed MI, Duleba AJ, El Shahat O, Ibrahim ME, Salem A. Naltrex- one treatment in clomiphene resistant women with polycystic ovary syndrome. Hum Reprod. 2008;23(11):2564-9. 19. Maksym RB, Hoffmann-Młodzianowska M, Skibińska M, Rabijew- ski M, Mackiewicz A, Kieda C. Immunology and Immunotherapy of Endometriosis. J Clin Med. 2021;10(24):5879. 20. Xu KY, Presnall N, Mintz CM, et al. Association of Opioid Use Disor - der Treatment With Alcohol-Related Acute Events. Jama Netw Open. 2021;4(2):e210061. 21. Xu KY, Mintz CM, Presnall N, Bierut LJ, Grucza RA. Association of Bupropion, Naltrexone, and Opioid Agonist Treatment With Stimulant- Related Admissions Among People With Opioid Use Disorder: A Case- Crossover Analysis. J Clin Psychiat. 2022;83(4):21m14112. 22. Xu KY, Borodovsky JT, Presnall N, et al. Association Between Benzo- diazepine or Z-Drug Prescriptions and Drug-Related Poisonings Among Patients Receiving Buprenorphine Maintenance: A Case-Crossover Analysis. Am J Psychiat. 2021;178(7):651-9. 23. Allison P, Christakis NA. Fixed-effects methods for the analysis of nonrepeated events. Sociol Methodol. 2006;36:155-72. 24. Allison P, editor. Fixed Effects Regression Methods in SAS. SUGI 31 Proceedings; Paper 184-31; 2006; San Francisco, CA. Accessed August 25, 2023. Available at https:// suppo rt. sas. com/ resou rces/ papers/ proce edings/ proce edings/ sugi31/ 184- 31. pdf. 25. Saitz R, Miller SC, Fiellin DA, Rosenthal RN. Recommended Use of Terminology in Addiction Medicine. J Addict Med. 2021;15(1):3-7. 26. Volkow ND, Gordon JA, Koob GF. Choosing appropriate language to reduce the stigma around mental illness and substance use disorders. Neuropsychopharmacology. 2021;46(13):2230-2. 27. Xu KY, Jones HE, Schiff DM, et al. Initiation and treatment discon- tinuation of medications for opioid use disorder in pregnant compared with nonpregnant people. Obstet Gynecol. 2023;141(4):845-853. 28. Yoder M, Boudreaux M. The effect of contraceptive access reform on privately insured patients: evidence from delaware contraceptive access now. Plos One. 2023;18(1):e0280588. 29. Johansson BA, Berglund M, Lindgren A. Efficacy of maintenance treatment with naltrexone for opioid dependence: a meta-analytical review. Addiction (Abingdon, England). 2006;101(4):491-503. 30. Krupitsky E, Nunes EV , Ling W, Illeperuma A, Gastfriend DR, Silver- man BL. Injectable extended-release naltrexone for opioid dependence: a double-blind, placebo-controlled, multicentre randomised trial. Lan- cet. 2011;377(9776):1506-13. 31. World Health Organization Department of Reproductive Health and Research (WHO/RHR) and Johns Hopkins Bloomberg School of Pub- lic Health/Center for Communication Programs (CCP). Knowledge for Health Project. Family Planning: A Global Handbook for Providers (2018 update). Baltimore and Geneva: CCP and WHO; 2018. Available from: https:// apps. who. int/ iris/ bitst ream/ handle/ 10665/ 260156/ 97809 99203 705- eng. pdf? seque nce=1. Accessed 25 Aug 2023 32. Bello JK, Baxley C, Weinstock J. Preconception Health Services for Women with Opioid Use Disorder (OUD): A Review and Best Practice Recommendation. Transl Issues Psychol Sci. 2021;7(2):154-65. 33. Bornstein M, Gipson JD, Bleck R, Sridhar A, Berger A. Perceptions of Pregnancy and Contraceptive Use: An In-Depth Study of Women in Los Angeles Methadone Clinics. Womens Health Issues. 2019;29(2):176-81. 34. MacAfee LK, Dalton V , Terplan M. Pregnancy Intention, Risk Percep- tion, and Contraceptive Use in Pregnant Women Who Use Drugs. J Addict Med. 2019;13(3):177-81. 35. Gilliam ML, Neustadt A, Gordon R. A call to incorporate a reproduc- tive justice agenda into reproductive health clinical practice and policy. Contraception. 2009;79(4):243-6. 36. Matusiewicz AK, Melbostad HS, Heil SH. Knowledge of and concerns about long-acting reversible contraception among women in medication- assisted treatment for opioid use disorder. Contraception. 2017;96(5):365-9. 37. Mintz CM, Xu KY, Presnall NJ, et al. Analysis of Stimulant Prescrip- tions and Drug-Related Poisoning Risk Among Persons Receiving Buprenorphine Treatment for Opioid Use Disorder. JAMA Netw Open. 2022;5(5):e2211634. 38. Schiff DM, Work EC, Foley B, et al. Perinatal opioid use dis- order research, race, and racism: a scoping review. Pediatrics. 2022;149(3):e2021052368. Prior Presentations Bello JK, Xu K, Grucza R. (May 2023) Association between Medications for Opioid Use Disorder (MOUD) and Pregnancy. 5th Annual Institute of Clinical and Translational Sciences Symposium and Poster Display. Washington University in Saint Louis, MO. Poster presentation. Bello, JK, Xu K, Grucza R. (September 2023) Using Big Data to Optimize Reproductive Health in Pregnant Women with Opioid Use Disorder. Institute of Clinical and Translational Sciences Big Data Research Symposium. The Advanced Health Data Institute, Saint Louis University and the Administrative Data Core Services, Washington University in Saint. Louis, MO. Oral presentation. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.