Inpatient Opioid Consumption Variability following Total Knee Arthroplasty: Analysis of 4,038 Procedures

This study examined an early iteration of an inpatient opioid administration-reporting tool, which standardized patient opioid consumption as an average daily morphine milligram equivalence per surgical encounter (MME/day/encounter) among total knee arthroplasty (TKA) recipients. The objective was to assess the variability of inpatient opioid administration rates among surgeons after implementation of a multimodal opioid sparing pain protocol. We queried the electronic medical record at our institution for patients undergoing elective primary TKA between January 1, 2016 and June 30, 2018. Patient demographics, inpatient and surgical factors, and inpatient opioid administration were retrieved. Opioid consumption was converted into average MME for each postoperative day. These MME/day/encounter values were used to determine mean and variance of opioids prescribed by individual surgeons. A secondary analysis of regional inpatient opioid consumption was determined by patient zip codes. In total, 23 surgeons performed 4,038 primary TKA. The institutional average opioid dose was 46.24 ± 0.75 MME/day/encounter. Average intersurgeon (IS) opioid prescribing ranged from 17.67 to 59.15 MME/day/encounter. Intrasurgeon variability ranged between ± 1.01 and ± 7.51 MME/day/encounter. After adjusting for patient factors, the average institutional MME/day/encounter was 38.43 ± 0.42, with average IS variability ranging from 18.29 to 42.84 MME/day/encounter, and intrasurgeon variability ranging between ± 1.05 and ± 2.82 MME/day/encounter. Our results suggest that there is intrainstitutional variability in opioid administration following primary TKA even after controlling for potential patient risk factors. TKA candidates may benefit from the implementation of a more rigid standardization of multimodal pain management protocols that can control pain while minimizing the opioid burden. This is a level of evidence III, retrospective observational analysis.