Decreasing Opioid Misuse and Habit-forming Behaviors Following Prescription in Patients Undergoing Cesarean Section

Overview

Background: Cesarean delivery is one of the most common obstetric procedures experienced among women who are pregnant.1 Women with Cesarean deliveries have a higher rate of peripartum opioid prescriptions and persistent opioid use compared to those with vaginal deliveries.2 Since 2002, prescription opioid use and misuse has significantly increased among women, including those who are pregnant, showing over 31% increase in past-month heroin use among women of childbearing age.3 This indicates the importance of focusing on maternal population for prescribed opioid medication management during the immediate postpartum period to prevent long-term persistent opioid misuse. Few evidence-based approaches are available to remotely manage prescription opioid use post-discharge.4 Recent advances in mobile technology have made it possible to monitor behavior and maintain communication in near real-time, long after patients are discharged from their surgical procedures.5-7 Using a virtual platform via use of mobile technology offers potential for sustainable implementation of a behavioral intervention and patient-provider communication even during the COVID-19 pandemic.8 Continuous Precision Medicine (CPM™; Research Triangle Park, NC) has developed a mobile app to overcome these barriers for tracking pain and pain medication use among post-surgery patients and tested the logistical and technological feasibility in postpartum patients at Temple OB/GYN. Collectively, our team brings expertise and collaborations between Temple University Hospital, RTI International, and CPM for the following Specific Aims: Aim 1: To examine the preliminary impact of the CPM mobile app to reduce the use of opioids among women post-Cesarean surgery Hypothesis 1: Patients using the CPM application will use fewer Morphine Milligram Equivalents (MME) compared to the blister package group. Aim 2: To establish correlates of pain medication use among women post-Cesarean surgery to estimate the appropriate recommendable dosages per model. Hypothesis 2: Structural and intermediary social determinants such as younger age, lower socioeconomic status, violence and trauma exposure, substance use disorder, and mental and physical health issues will be associated with more opioid medication use.

Experimental Design: A randomized pilot trial will be conducted with 100 females at Temple University Hospital immediately post-Cesarean delivery. The women will be randomized (1:1) to either: using the CPM mobile app and electronic monitoring blister packs (experimental group) or to blister packs only (control group). The primary outcome is Morphine Milligram Equivalents (MME). Secondary outcomes will be delay to dosages and pain scores. We will examine whether a mobile application found feasible by post-surgery patients and healthcare professionals has a preliminary impact on opioid medication use among women who are immediately postpartum and had Cesarean surgery. The CPM application provides patients with a detailed prescription regimen, allows a patient to report dose-by-dose pain scores, and helps the patient consider whether a dose is needed via gamification, and creates usage traceability for clinicians to reduce over-prescription. Linear regression, including ANOVA and repeated mixed models will be used to compare groups on the primary continuous outcomes of daily MME during the 10-day period. The initial models will only include a term for study group. Variables that are statistically significant (\<0.05) by univariate analyses, as well as theoretically important covariates will be considered for inclusion in adjusted analyses to improve their precision. Pain scores as a secondary continuous outcome will be evaluated using regression models for linear and non-linear mixed effect models. If no significant differences in these outcomes were detected, resampling Bootstrap method will be used to estimate effect sizes (i.e., Cohen's d) and its 95% confidence intervals, considering potential non-normal distribution of data. Further, we will utilize structural equation modeling (SEM) for its greater flexibility and capacity of handling the structural relationship between measured variables and latent constructs.