Abstract
This pharmacoepidemiology study uses Medicaid data to estimate associations between first-trimester use of intravenous ondansetron and risk of cardiac malformations and oral cleft in children of exposed mothers.
A population-based cohort study previously evaluated the association between first-trimester exposure to the oral formulations of ondansetron and congenital malformations. After accounting for potential confounding variables, there was no significant association with congenital malformations overall or cardiac malformations, but a small increased risk of oral clefts could not be excluded.1 Subsequent research has suggested that intravenous administration of ondansetron may be associated with greater risks of cardiac malformations and oral clefts.2 Potential explanations for the apparent difference in teratogenic effect of intravenous compared with oral formulations include (1) residual confounding by indication and associated factors (eg, nutritional deficiencies) in women with hyperemesis severe enough to require intravenous medication, (2) higher doses administered intravenously vs orally, and (3) higher specificity for exposure because studies of oral use in health care utilization data are based on dispensed medication, and not all women may consume the medication as prescribed. We therefore conducted a follow-up study to examine the association between intravenous ondansetron and congenital malformations.
Methods
The data source, study design, and analytic methods were the same as described in our previously published evaluation of oral ondansetron.1 Briefly, we used a mother-infant–linked cohort nested in the 2000 to 2014 Medicaid Analytic eXtract.3 The research was approved by the institutional review board of Brigham and Women’s Hospital, with waiver of informed consent.
Women were considered exposed if a claim for Healthcare Common Procedure Coding System code J2405 indicating ondansetron injection was recorded during the first trimester of pregnancy. The reference group consisted of women without exposure to either oral or intravenous ondansetron from 3 months before the start of pregnancy through the end of the first trimester. Outcomes included cardiac malformations, oral clefts, and congenital malformations overall. Potential confounders and proxies for confounders considered are shown in the Table.
Table. Selected Cohort Characteristics of Women With and Without Exposure to Intravenous Ondansetron During the First Trimester After Propensity Score Stratificationa,b.
| Characteristics | No. (%) | ||
|---|---|---|---|
| Ondansetron | Unexposed | Standardized Differencec | |
| No. of pregnancies | 23 866 | 1 762 018 | |
| No. of women | 23 309 | 1 437 378 | |
| Age, mean (SD), y | 25 (5.16) | 25 (5.15) | 0.03 |
| Year of delivery | |||
| 2000-2005 | 1311 (5.49) | 104 436 (5.93) | −0.02 |
| 2006-2010 | 9339 (39.13) | 707 899 (40.18) | −0.02 |
| 2011-2013 | 13 216 (55.38) | 949 683 (53.90) | 0.03 |
| Race/ethnicityd | |||
| White | 9456 (39.62) | 705 822 (40.06) | −0.01 |
| Black | 8735 (36.60) | 639 205 (36.28) | 0.01 |
| Hispanic | 3215 (13.47) | 238 643 (13.54) | 0.00 |
| Other or unknown | 2460 (10.31) | 178 348 (10.12) | 0.01 |
| Multiple gestation | 484 (2.03) | 36 138 (2.05) | 0.00 |
| Maternal conditions | |||
| Anxiety | 2224 (9.32) | 174 627 (9.91) | −0.02 |
| Depression | 2858 (11.98) | 221 793 (12.59) | −0.02 |
| Migraine/headache | 4326 (18.13) | 345 101 (19.59) | −0.04 |
| Diabetes | 593 (2.48) | 50 713 (2.88) | −0.02 |
| Hypertension | 982 (4.11) | 76 765 (4.36) | −0.01 |
| Illicit drug abuse or dependence | 908 (3.80) | 69 749 (3.96) | −0.01 |
| Alcohol abuse or dependence | 291 (1.22) | 23 729 (1.35) | −0.01 |
| Smoking | 2434 (10.20) | 188 525 (10.70) | −0.02 |
| Obstetric comorbidity index, mean (SD) | 1.16 (1.59) | 1.19 (1.63) | −0.02 |
| Concomitant medications | |||
| Antidepressants | 3932 (16.48) | 307 949 (17.48) | −0.03 |
| Benzodiazepines | 1908 (7.99) | 150 350 (8.53) | −0.02 |
| Oral hypoglycemics | 257 (1.08) | 20 967 (1.19) | −0.01 |
| Insulin | 302 (1.27) | 26 361 (1.50) | −0.02 |
| Antihypertensives | 1073 (4.50) | 85 601 (4.86) | −0.02 |
| Suspected teratogense | 2062 (8.64) | 158 417 (8.99) | −0.01 |
| Metoclopramide | 4007 (16.79) | 294 274 (16.70) | 0.00 |
| Promethazine | 7780 (32.60) | 598 253 (33.95) | −0.03 |
| Pyridoxine | 927 (3.88) | 65 348 (3.71) | 0.01 |
In addition to the characteristics listed in the table, the following additional potential confounding variables were included in the propensity score (PS): treatment indication (nausea and vomiting in pregnancy, hyperemesis gravidarum) and associated conditions (weight loss, electrolyte and laboratory abnormalities, dehydration, gastroesophageal reflux), state of residence, maternal chronic conditions (kidney disease, Crohn disease, irritable bowel syndrome, ulcerative colitis, overweight or obesity, underweight), concomitant medications (anticonvulsants, triptans, progestins, corticosteroids), and general markers of comorbid illness and disease severity (number of prescriptions for medications other than antiemetics, number of distinct diagnoses, number of outpatient visits, hospitalizations, and emergency department visits). Maternal conditions and concomitant medication use were measured from 3 months before the start of pregnancy through the end of the first trimester. General markers of the burden of illness were measured during the 3 months before pregnancy because these measures may be affected by early pregnancy complications.
Observations in the nonoverlapping regions of the PS were trimmed to focus the estimation of the treatment effect in a population with clinical equipoise. We defined 50 strata based on the PS distribution in the exposed. All ondansetron-exposed women received a weight of 1, and all unexposed women were weighted in proportion to the distribution of the exposed in the stratum into which they fell. The unexposed group weights were scaled to sum to the number of unique unexposed women included in the analysis: (No. exposed women included in a given stratum/No. exposed women)/(No. unexposed women included in a given stratum/No. unexposed women).
An absolute standardized difference <0.1 was considered evidence of covariate balance.
Race or ethnic group was considered to represent a potentially important confounding factor, and was categorized based on information submitted to the Centers for Medicare & Medicaid Services by individual states, which was based on information that had been collected and coded from Medicaid applications.
Suspected teratogens were fluconazole, methimazole, danazol, propylthiouracil, aminoglycosides, folic acid antagonists, potassium iodide, and tetracycline. Women exposed to known teratogenic mediations were excluded from the cohort (ie, warfarin, antineoplastic agents, lithium, isotretinoin, misoprostol, thalidomide).
Relative risks (RRs) and risk differences (RDs) with their 95% CIs were estimated using generalized linear models. Propensity score (PS) fine stratification was used for confounding adjustment (50 strata based on the PS distribution in the exposed); all covariates listed in the Table were used to estimate the PS. Weighted regression models were used to derive the adjusted associations.4 All analyses were conducted using SAS version 9.4 (SAS Institute). Precision around the measures of association is provided using 2-sided 95% CIs.
Results
The cohort consisted of 1 880 594 pregnancies, of which 23 877 (1.3%) received at least one ondansetron injection during the first trimester. Characteristics of exposed and unexposed women were well balanced after PS stratification (Table).
The adjusted RR for cardiac malformations was 0.97 (95% CI, 0.86-1.10) and the adjusted RD was −2.9 (95% CI, −15.7 to 9.8) per 10 000 births. For oral clefts, the adjusted RR was 0.95 (95% CI, 0.63-1.43) and the adjusted RD was −0.5 (95% CI, −4.5 to 3.5) per 10 000 births. For malformations overall, the RR was 1.02 (95% CI, 0.96-1.08) and the RD was 7.1 (95% CI, −17.9 to 32.2) per 10 000 births (Figure).
Figure. Risk of Congenital Malformations in Infants Following Exposure to Intravenous Ondansetron During the First Trimester.
Relative risks and 95% CIs are presented to show the risk of any congenital malformation, cardiac malformations, and oral clefts among infants born to women with exposure to intravenous ondansetron during the first trimester compared with infants born to women without such exposure.
Discussion
Evidence from this national cohort of publicly insured pregnant women with confounding adjustment suggests that intravenously administered ondansetron was not associated with an increase in the risk of cardiac malformations, oral clefts, or congenital malformations overall. While the point estimate for oral clefts was lower for intravenous ondansetron compared with oral ondansetron (RR, 1.24 [95% CI, 1.03-1.48]), the CI was wide with an upper limit similar to that for oral ondansetron.1
The strengths and limitations of using large health care databases to evaluate drug safety in pregnancy have been described.5 Limitations of this study include the potential for residual confounding due to unmeasured or poorly measured characteristics, restriction of the cohort to live births and individuals with Medicaid, and availability of data through 2014 only.1 An advantage of studying intravenous administration of ondansetron is the high specificity for exposure because a claim for reimbursement is generated. Overall, these findings suggest that ondansetron is not a major teratogen.1
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Huybrechts KF, Hernández-Díaz S, Straub L, et al. Association of maternal first-trimester ondansetron use with cardiac malformations and oral clefts in offspring. JAMA. 2018;320(23):2429-2437. doi: 10.1001/jama.2018.18307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zambelli-Weiner A, Via C, Yuen M, Weiner DJ, Kirby RS. First trimester ondansetron exposure and risk of structural birth defects. Reprod Toxicol. 2019;83:14-20. doi: 10.1016/j.reprotox.2018.10.010 [DOI] [PubMed] [Google Scholar]
- 3.Palmsten K, Huybrechts KF, Mogun H, et al. Harnessing the Medicaid Analytic Extract (MAX) to evaluate medications in pregnancy: design considerations. PLoS One. 2013;8(6):e67405. doi: 10.1371/journal.pone.0067405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Desai RJ, Rothman KJ, Bateman BT, Hernandez-Diaz S, Huybrechts KF. A propensity-score-based fine stratification approach for confounding adjustment when exposure is infrequent. Epidemiology. 2017;28(2):249-257. doi: 10.1097/EDE.0000000000000595 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Huybrechts KF, Bateman BT, Hernández-Díaz S. Use of real-world evidence from healthcare utilization data to evaluate drug safety during pregnancy. Pharmacoepidemiol Drug Saf. 2019;28(7):906-922. doi: 10.1002/pds.4789 [DOI] [PMC free article] [PubMed] [Google Scholar]
