The overall objective of this study is to determine whether the use of domperidone is associated with an increased risk of serious cardiac events among postpartum women in the six months following delivery. More specifically, the investigators will assess whether domperidone is associated with the risk of ventricular tachyarrhythmia (VT) and sudden cardiac death (SCD).
A distributed, common-protocol approach will be used to conduct retrospective cohort studies using administrative health care data from five Canadian provinces (Alberta, British Columbia, Manitoba, Ontario and Saskatchewan). Briefly, the Canadian databases include population-level data on physician billing, diagnoses and procedures from hospital discharge abstracts, and dispensations for prescription drugs. The Ontario data will be restricted to social assistance recipients.
To assess this potential drug safety issue, the investigators will use a prevalent new-user cohort design as described by Suissa et al., 2017. In each jurisdiction, the investigators will first assemble a base cohort that includes all women aged between 15 and 55 years old with live births between April 1, 1997 and December 31, 2016 (or the latest available data at each site). Base cohort entry will be defined by the hospital discharge date after delivery. From this base cohort, a study cohort will be formed including all women who received a prescription for domperidone during the six months following delivery and matched women who have a physician visit at the same follow-up time and do not receive domperidone. Study cohort entry will be defined by the prescription date of domperidone for exposed women or the corresponding date of the physician visit for unexposed women. Women will be followed from the study cohort entry date until an event (defined below) or end of follow up (six months after delivery). Women will be permitted to contribute more than one observation to the study provided that each observation meets all the inclusion criteria.
Exposure to domperidone will be defined as a prescription for domperidone during the six months following delivery. Women not exposed to domperidone will serve as the reference group. Analyses will be conducted using an approach analogous to an intention-to-treat. Unexposed women who are subsequently exposed to domperidone will be censored at the time of the domperidone prescription. The primary outcome will be a composite endpoint of VT or SCD. Secondary outcomes will be the individual endpoints of VT, SCD, and all-cause mortality; separate follow-up durations will be estimated for each outcome.
Using a prevalent new-user design with time-based exposure sets, each user of domperidone will be matched to 10 unexposed women with a physician visit within ±30 days of the first domperidone dispensing of the exposed woman. Matching will be done on propensity score and calendar date of base cohort entry (i.e., hospital discharge from delivery hospitalization). Cox proportional hazards models will be used to estimate site-specific adjusted hazards ratios (HR) and corresponding 95% confidential intervals (CI) for each outcome of interest among postpartum women. Robust sandwich variance estimators will be used to account for potential within-subject clustering. As secondary analyses, the composite endpoint of VT or SCD will be stratified by domperidone dose and duration if feasible. Sensitivity analyses will be performed to assess the robustness of study results. Meta-analyses of the site-specific results will be performed using random-effects models.
In additional analyses, the demographic and clinical characteristics (including perinatal and peripartum characteristics) of women prescribed domperidone in the six months postpartum with those of women not prescribed it during this period will be compared. Furthermore, the impact of Health Canada safety advisories on prescribing patterns will be assessed via interrupted time series analyses.
