PERIOPERATIVE OPIOID EXPOSURE PATTERNS IN PEDIATRIC ACL PATIENTS: A TEN YEAR ADMINISTRATIVE DATABASE STUDY

摘要

Background: Pediatric patients sustaining anterior cruciate ligament (ACL) or related injuries are at high risk for opioid exposure in the acute, perioperative and postoperative phases of injury. Early and repeated exposure to these medications may increase the risk of future misuse. While variation in opioid prescribing practices has been documented in the outpatient setting and other realms of pediatric care, perioperative opioid exposure in this procedural cohort has not been previously examined on a national level. Purpose: To assess for demographic, temporal, regional, and hospital-level variability in perioperative opioid exposure in pediatric ACL patients. Methods: The Pediatric Health Information Systems Database (PHIS) was used to identify pediatric patients (≤18 years old) undergoing surgical treatment for ACL injury between January 2008 and December 2017. Perioperative opioids were converted to a morphine equivalent dose (MED) and summed for each patient. A hierarchical bayesian regression was performed to identify demographic factors that predicted opioid exposure while adjusting for the effect of hospital. Results: The study cohort included 23,071 patients across 52 hospitals. We report model estimates in Table 1 and mean MME by hospital in Figure 1. Compared to older adolescents (15-18yo), younger adolescents (11-14yo; b=-0.13 95% Credible Interval[-0.19, -0.07]) and children &10yo (b=-1.62 [-1.72, -1.52]) received less MED. Patients located in an observation unit (b=-1.15 [-1.25, -1.06]) or an inpatient unit (b=-1.31;[-1.42, -1.20]) received less MED than patients in an ambulatory surgical setting. Patients with commercial insurance also were dispensed more MED compared to those with other payers (b=0.10 [0.04, 0.16]). Female patients received less opioids than male patients (b=-0.08 [-0.13, -0.02]). Of the hospital random effects, the 95% credible intervals of 24 (46%) intercepts and 21 (40%) slopes did not include zero. Conclusions: This administrative database study identified hospital and patient-level characteristics predictive of perioperative narcotic exposure among pediatric ACL patients. Those who were older, in an ambulatory surgery setting, or had commercial insurance received more opioids. Cumulative perioperative opioid exposure has not declined on a national level in recent years and significant variability in opioid exposure exists between hospitals. Future work should seek to identify and utilize opioid-minimizing practices that appear present in some clinical settings. Table 1. Estimates from the Hierarchical Bayesian Regression Model Predicting Log Opioid Exposure During ACL-Related Repair Figure 1. ACL Opioid Prescribing by hospital