Menu Close

Comparing Clinical Research Participation and Retention Between Pregnant Cannabis-Exposed and Unexposed Participants

Nayana Sojin
University of Florida

Co-Authors: Rhea Parimoo1, Lauren Agliano1, Amie J Goodin1, Deepthi S Varma1, Bruce A Goldberger1, Kay Roussos-Ross1
1University of Florida

Background: While there is currently limited data on the effects of perinatal cannabis use, in-utero exposure has been linked to adverse neonatal health outcomes. As perinatal cannabis use rates reportedly increase, it is necessary to obtain conclusive, pregnancy-specific safety data through well-designed clinical research studies.

Objective: To compare clinical research participation and retention trends in a pilot study among cannabis-exposed and unexposed participants in the perinatal period.

Methods: Pregnant patients self-reporting cannabis exposure during prenatal visits at an academic health-system were recruited, along with control patients with no self-reported exposure.

Biospecimens and imaging were collected throughout the perinatal period: maternal urine and fetal ultrasound each trimester; fetal MRI in the third trimester; maternal urine, placenta, umbilical cord, and neonatal meconium at delivery; and postpartum maternal urine and breastmilk. Participants received compensation for each biospecimen/imaging completion. Participation trends were identified by calculating proportions of missed/completed biospecimens/imaging, study visit reschedules/cancellations, and losses-to-follow-up (LTFUs).

Results: 25 participants were recruited over 18 months: 80% (n=20) self-reported as cannabis-exposed and 20% (n=5) reported as unexposed (control).

Of 37 possible fetal ultrasounds, 30 (81%) were performed (82%, n=23 cannabis-exposed; 78%, n=7 control). 68% of participants (60%, n=12 cannabis-exposed; 100%, n=5 control) completed a fetal MRI. Of 239 biospecimen collections, 170 (71%) were completed (66%, n=126 cannabis-exposed; 90%, n=44 control). 44% of participants (45%, n=9 cannabis-exposed; 40%, n=2 control) missed at least one delivery sample. 24% of participants (30%, n=6 cannabis-exposed; 0%, n=0 control) did not complete one or more postpartum breastmilk samples due to not breastfeeding or no longer lactating.

40% (n=8) of cannabis-exposed participants were loss-to-follow-up (LTFU), with the majority (88%, n=7) occurring postpartum and 13% (n=1) after the third trimester. Notably, no LTFUs were observed in the control group. 45% of cannabis-exposed participants either canceled or required rescheduling of at least one appointment, compared to 0% (n=0) of controls. Additionally, 40% (n=8) of cannabis-exposed participants canceled, did not attend, or did not schedule their postpartum appointment, compared to 0% (n=0) of controls.

Conclusions: Participants with self-reported perinatal cannabis exposure were more likely to miss postpartum visits, become lost-to-follow-up, and require rescheduling of study visits compared to controls. Despite this trend, cannabis-exposed participants completed 68% of study visits, suggesting that it is feasible to retain cannabis-exposed pregnant women for long-term study. To improve retention/compliance rates among cannabis-exposed participants, future research could consider implementing a staggered increase in incentives and home visits for sample collections.