The study’s primary outcome was a composite of adverse pregnancy outcomes, including stillbirth, defined as Apgar score of 0; any hypertensive disorder of pregnancy, including gestational hypertension, preeclampsia, eclampsia, and HELLP syndrome (hemolysis, elevated liver enzymes, and low platelet count); preterm birth, defined as spontaneous delivery before 37 weeks gestation; and infants born small for gestational age, defined as a birth weight below the 10th percentile after adjustment for gestational age and sex.
Secondary outcomes included pregnancy outcomes including placental abruption, mode of delivery, and gestational age at delivery. Neonatal outcomes included weight, length, and head circumference at birth, and neonatal intensive care unit admission. An Apgar score less than 7 at 5 minutes was considered an adverse neonatal outcome.
The sample size was determined by an estimate drawn from previous chart abstraction that the composite outcome would be seen in 16% of the clinic’s non–marijuana exposed patients, and 24% of the marijuana-exposed patients. The investigators also factored in that 18% of adolescents in the clinic database were marijuana users.
Dr. Rodriguez and her collaborators used a variety of models for statistical analysis, some of which included self-report alone or in conjunction with urine toxicology. In the end, they found that significant associations between their composite endpoint and marijuana use were seen when patients were dichotomized into those who had at least one positive urine toxicology test, versus those who had no positive toxicology results.
One of the study limitations was that the study didn’t permit investigators to get accurate information about the quantity, timing, or route of marijuana dosing. Also, this methodology may primarily identify heavier marijuana users, said Dr. Rodriguez.
Tobacco use was determined only by self-report, and outcomes were followed over a relatively short period of time.
Still, said Dr. Rodriguez, the study had many strengths, including the use of biological sampling to determine exposure and the near-universal participant urine toxicology testing. The investigators were able to capture and account for many important factors that could confound the results, she said. “Uncertainty regarding the impact of [marijuana] on pregnancy outcomes in the literature may result from incomplete ascertainment of exposure,” she and her coinvestigators wrote in the abstract accompanying the presentation.
SOURCE: Rodriguez C et al. Am J Obstet Gynecol. 2018 Jan;218:S37.