Predictors of narcotic prescriptions in pediatric patients with symptomatic ureteral stone
Anja Zann, MD, Christina Ching, MD, Yuri Sebastiao, PhD, MPH, Tran Bourgeois, MPH, Lauren Nicassio, BS, Daryl McLeod, MD.
Nationwide Children's Hospital, Columbus, OH, USA.
BACKGROUND
Pediatric stone disease is a growing problem, as is the use and misuse of opioid medication in children. Currently there are no guidelines for prescribing opioids to children with symptomatic ureteral stones. The aim of this study was to identify factors that lead to increased likelihood of narcotic prescription upon discharge from the Emergency Room (ER).
METHODS
A retrospective review of patients discharged with a diagnosis of ureteral stone from a single pediatric center ER and affiliated urgent cares from 2013 to 2018 was performed. All patients with a primary diagnosis of stone in the urinary tract were identified using ICD-10 N20, and ICD-9 592 codes. Individual chart review identified patients treated during that encounter for a stone specifically in the ureter, excluding kidney and bladder stones. Patients without confirmed stone on imaging, but treated for a presumed ureteral stone were included. This diagnosis and treatment decision was at the ER provider's discretion. Demographics, insurance status, imaging results, medications administered, medications prescribed, follow-up plan, and outcomes were recorded when available.
Chi-square or T-test was used to analyze parametric data. Non-parametric data was compared using Wilcox two-sample test analyses. Predictors of opioid prescribing were analyzed with logistic regression. Variables with p 0.10 in the univariate model were further evaluated by multivariable analysis with a p<0.05 considered significant.
RESULTS
559 patients were identified with 242 patients meeting inclusion criteria for a symptomatic ureteral stone. The median age was 15.8 years (Interquartile Range (IQR) 13.3-17.4), 55% of which were female. The majority of patients were white (81%), and had private insurance (55.8%). Imaging was obtained in 88.4% of patients. The remaining 11.6% were treated for a presumed ureteral stone based on history and symptoms. Of those imaged, 47.5% did not have a discrete stone identified, and were treated for a "presumed stone".
The rate of narcotic prescription upon discharge from the ER was 62.4%. Multivariable analysis identified three factors that significantly predicted whether a narcotic medication was prescribed upon discharge: 1) increasing age (Odds Ratio (OR) 1.09, 95% Confidence Interval (CI): 1.01 - 1.19); 2) presence of stone on imaging (OR 3.06, 95% CI: 1.70 - 5.51); 3) administration of narcotic while in the ER (OR 5.99, 95% CI: 3.03 – 11.86). Race, gender, stone size, stone location, urologic consultation and insurance status were not significant predictors.
The number of narcotic doses prescribed at ER discharge was highly variable, ranging from 3 to 102. The two most commonly prescribed number of doses were 10 and 20. The majority of patients received 6-10 doses (39%), with near equal numbers receiving 11-15 doses (22.6%) and 16-20 doses (21.9%).
CONCLUSIONS
Older patients, those with imaging demonstrating a stone, and those who received narcotics in the ER are more likely to get a home-going prescription for narcotic medication. In patients who did receive a narcotic upon discharge, the number of doses provided was highly variable. Future studies are needed to determine appropriate justification for narcotic prescription as well as establish guidelines for dosing practices.
Back to 2019 Abstracts
