Importance: Practice guidelines recommend nonopioid medications in children after tonsillectomy, but to date, studies have not used recent national data to assess perioperative opioid prescribing patterns or the factors associated with these patterns in this population. Closing this knowledge gap may help in assessing whether such prescribing and prescription duration could be safely reduced.
Objective: To assess national perioperative opioid prescribing patterns, clinical and demographic factors associated with these patterns, and association between these patterns and complications in children after tonsillectomy compared with children not using opioids.
Design, setting, and participants: This cohort analysis used the 2016 to 2017 claims data from the database of a large national private insurer in the United States. Opioid-naive children aged 1 to 18 years with a claims code for tonsillectomy with or without adenoidectomy between April 1, 2016, and December 15, 2017, were identified (n = 22 567) and screened against the exclusion criteria. The final sample included 15 793 children.
Main outcomes and measures: The percentage of children with 1 or more perioperative fills (prescription drug claims for opioids between 7 days before to 1 day after tonsillectomy) was calculated, along with the duration of perioperative prescriptions (days supplied). Linear regression was used to identify the demographic and clinical factors associated with the duration of perioperative opioid prescriptions. Logistic regression was used to assess the association between having 1 or more perioperative fills and their duration and the risk of return visits 2 to 14 days after tonsillectomy for pain or dehydration, secondary hemorrhage, and constipation compared with children not using opioids.
Results: Among 15 793 children, the mean (SD) age was 7.8 (4.2) years, 12 807 (81.1%) were younger than 12 years, 2986 (18.9%) were between 12 and 18 years of age, and 8289 (52.6%) were female. In total, 9411 (59.6%) children had 1 or more perioperative fills, and the median (25th-75th percentile) duration was 8 (6-10) days; 6382 had no perioperative fills. The probability of having 1 or more perioperative fills and the duration of prescription varied across US census divisions. Having 1 or more perioperative fills was not associated with return visits for pain or dehydration (adjusted odds ratio [AOR], 1.13; 95% CI, 0.95-1.34) or secondary hemorrhage (AOR, 0.90; 95% CI, 0.73-1.10) compared with children not using opioids, but it was associated with increased risk of return visits for constipation (AOR, 2.02; 95% CI, 1.24-3.28). Duration was not associated with return visits for complications.
Conclusions and relevance: These findings suggest that reducing perioperative opioid prescribing and the duration of perioperative opioid prescriptions may be possible without increasing the risk of these complications.