Notable Variability in Opioid-Prescribing Practices After Common Orthopaedic Procedures

Sophia A. Traven, MD; Daniel L. Brinton, PhD; Shane K. Woolf, MD; Lee R. Leddy, MD; Michael B. Gottschalk, MD; Harris S. Slone, MD


J Am Acad Orthop Surg. 2021;29(5):219-226. 

In This Article


Initial Oral Morphine Equivalent

A total of 73,921 opioid-naive patients undergoing the abovementioned procedures were identified. The regional variability differed notably between procedure types (Table 1). Although the median OME was relatively consistent between regions for patients undergoing procedures such as single-level ACDF, arthroscopic meniscectomy, and total hip arthroplasty (THA), wide IQRs were noted for all of the procedures (Table 1). For example, although the median OME for patients undergoing total shoulder arthroplasty (TSA) in the Northeast was 400, the IQR had a wide range of 260 OME to 750 OME.

The widest IQRs were seen in the initial prescriptions for patients undergoing total knee arthroplasty (TKA). Although the national median OME was 495, the IQR ranged from 320 in the first quartile to 750 in the third quartile—a difference of 430 OMEs (Table 1). Similar variations were seen within regions as well. For patients undergoing TKA in the South, the range in IQR was from 300 to 800, a difference of 500 OME. In other words, one in four patients are being discharged with at least 2.67 times more opioids compared with other opioid-naive patients after undergoing the same procedure.

In addition, the West and South consistently prescribed more opioids than the Midwest and Northeast for many procedures including ACL reconstruction, bimalleolar ankle fracture ORIF, carpal tunnel release, distal radius fracture ORIF, TKA, and TSA (Table 1).

90-day Total Oral Morphine Equivalent

Inter-regional variability and IQRs increased when examining the 90-day total OME prescribed for a given procedure (Table 2). Notable interregional variability in prescriptions was observed for all procedures excluding total joint arthroplasty (P ≤ 0.007, Table 2). Patients residing in the West and the South consistently had prescriptions at or greater than national medians (Figure 1). Notable intraregional variability was also observed. For example, although the median prescription for patients undergoing arthroscopic RCR in the West was 625 OME, the IQR ranged from 425 to 1,200—a difference of 775 OME.

Figure 1.

Figure demonstrating the variation in the 90-day OME by procedure and region. This figure demonstrates the difference from the national median that each region prescribes for each procedure. The error bars represent the IQRs within each region. For example, while the West prescribes 75 OME more than the national median for ACDF (675 OME versus 600 OME), the IQR within the West ranges from 225 OME below the median to 450 OME above the median. ACDF = anterior cervical discectomy and fusion, ACL = anterior cruciate ligament, IQR = interquartile range; OME = oral morphine equivalent, ORIF = open reduction and internal fixation

By far, the greatest number of opioids prescribed consistently across all regions was for patients undergoing TKA. The median 90-day OME across the United States for this procedure was 1,200, which was more than twice that of the median OME for the initial prescription (495 OME, Table 1). In addition, a wide range in IQR was noted for patients undergoing TKA, with the most intraregional variability in prescriptions seen in the West. For patients undergoing TKA in this region, one in four opioid-naive patients were given prescriptions for more than 3 times the amount of opioids than others undergoing the exact same procedure (IQR 700 to 2,350, Table 2). However, no interregional variability in 90-day prescriptions was observed in patients undergoing primary total knee, hip, or shoulder arthroplasties (P ≥ 0.162, Table 2). Subsequently, a post hoc power analysis was conducted using Power and Sample Size version 15.0.3 NCSS LLC. With an alpha level of 0.05 and running 500 simulations, this study achieved power to find an interregional difference using the Kruskal-Wallis test for 90-day OME among those undergoing TSA of 23.6%, TKA 7.0%, and THA 17.6%. In other words, these data are likely underpowered to detect any statistical differences, if those exist, between regions in patients undergoing total joint arthroplasty.

Less disparate median OMEs between initial and 90-day total prescriptions were seen in procedures such as carpal tunnel release (200 OME median initial prescription compared with 200 OME median 90-day prescription), arthroscopic meniscectomy (300 OME median initial prescription compared with 300 OME median 90-day prescription), and distal radius fracture ORIF (300 OME median initial prescription compared with 380 OME median 90-day prescription), suggesting that most patients undergoing these procedures did not require refills.

Prescribed Refills

Variability in the percentage of patients who were prescribed refills was also observed between procedure type and geographic region (Table 3). Across all regions, patients undergoing carpal tunnel release were prescribed the fewest number of refills compared with patients undergoing TKA, who received the greatest number of refills (≤9.7% versus ≥54.1%, Table 3). Patients in the Midwest and South also were prescribed the greatest number of refills, whereas patients in the Northeast consistently received fewer refills than the national median for all procedures. For patients who did receive refills, the median number of refills was one for all procedure types.