Chaput, K.H., McMorris, C.A., Metcalfe, A. et al. Development and validation of the Cannabis Exposure in Pregnancy Tool (CEPT): a mixed methods study. BMC Pregnancy Childbirth 24, 280 (2024). https://doi.org/10.1186/s12884-024-06485-0
Abstract
Background
Evidence of associations between prenatal cannabis use (PCU) and maternal and infant health outcomes remains conflicting amid broad legalization of cannabis across Canada and 40 American states. A critical limitation of existing evidence lies in the non-standardized and crude measurement of prenatal cannabis use (PCU), resulting in high risk of misclassification bias. We developed a standardized tool to comprehensively measure prenatal cannabis use in pregnant populations for research purposes.
Methods
We conducted a mixed-methods, patient-oriented tool development and validation study, using a bias-minimizing process. Following an environmental scan and critical appraisal of existing prenatal substance use tools, we recruited pregnant participants via targeted social media advertising and obstetric clinics in Alberta, Canada. We conducted individual in-depth interviews and cognitive interviewing in separate sub-samples, to develop and refine our tool. We assessed convergent and discriminant validity internal consistency and 3-month test–retest reliability, and validated the tool externally against urine-THC bioassays.
Results
Two hundred fifty four pregnant women participated. The 9-item Cannabis Exposure in Pregnancy Tool (CEPT) had excellent discriminant (Cohen’s kappa = -0.27–0.15) and convergent (Cohen’s kappa = 0.72–1.0) validity; as well as high internal consistency (Chronbach’s alpha = 0.92), and very good test–retest reliability (weighted Kappa = 0.92, 95% C.I. [0.86–0.97]). The CEPT is valid against urine THC bioassay (sensitivity = 100%, specificity = 82%).
Conclusion
The CEPT is a novel, valid and reliable measure of frequency, timing, dose, and mode of PCU, in a contemporary sample of pregnant women. Using CEPT (compared to non-standardized tools) can improve measurement accuracy, and thus the quality of research examining PCU and maternal and child health outcomes.