Prospective Evaluation of Opioid Consumption Following Cubital Tunnel Decompression Surgery

Introduction

The opioid epidemic in the United States has become a national priority.^1-3 Prescription opioid abuse has been rising steadily for the past 20 years in the United States and abroad.^4-6 With increasing opioid abuse, overdose-related deaths have been rising as well. ⁷ Recent evidence shows that orthopedic surgeons prescribe more opioid pain medication than any other surgical specialty. ⁸ Although this is partly due to the painful nature of musculoskeletal procedures, ⁹ orthopedic surgeons need to be even more responsible when prescribing pain medications. However, there remains uncertainty as to the appropriate number of opioids to prescribed postoperatively after various surgeries, and inadvertent overprescribing is common.^10,11 This inadvertent overprescribing of opioids puts patients at risk of abuse and leaves a significant number of extra prescription pills in the community for potential diversion.^3,12,13 Recently, there have been efforts to establish opioid prescription guidelines for upper extremity surgeons to address this problem.^12,14 In that spirit, the goal of our study was to analyze opioid use in patients undergoing cubital tunnel decompression surgeries to better determine typical opioid consumption and guide more optimal opioid prescribing postoperatively.

Methods

After obtaining institutional review board approval, 9 hand surgery fellowship-trained, board-certified orthopedic surgeons practicing in a single private academic group prospectively collected postoperative opioid consumption data for 16 consecutive months (in February 2016 to June 2017) for patients undergoing surgery for cubital tunnel syndrome. Data were collected via a standardized intake form at the preoperative and first postoperative appointments. The surgeons were not blinded, were asked to continue their normal prescribing patterns, and were aware that their patients would be asked about their opioid consumption postoperatively. Nicotine use information was not collected. Data on patient age, gender, handedness, and insurance type were collected. Preoperatively, patients completed the 11-item Quick Disabilities of the Arm, Shoulder and Hand (QuickDASH) questionnaire. ¹⁵ Patient McGowan Stage was determined at this preoperative visit. ¹⁶ All patients also had a preoperative electrodiagnostic study (electromyography [EMG]) performed by a board-certified physiatrist or neurologist. EMG was graded negative, mild, mild-moderate, moderate, moderate-severe, or severe.

On the day of the surgical procedure, the surgeon recorded the following variables on the intake form: the procedure type (cubital tunnel release [CuTR], endoscopic CuTR, subcutaneous ulnar nerve transposition [UNT], or submuscular UNT), the anesthesia type (local, sedation, general, and/or regional), the opioid type, and the quantity of the opioid prescribed. Additional procedures performed during the surgery were recorded as well. Patients were instructed to bring unused pills to their first postoperative appointment for counting and disposal.

At the first postoperative visit, the following data were solicited directly from the patient by a member of the research team and were added to the standardized intake form: the quantity of the prescribed opioid used, the total days of opioid use, the reason for discontinuation (the opioids no longer being necessary or the side effects associated with the opioids), and the side effects (if any). Finally, patients were asked whether opioid disposal instructions were given to them at any point in time. A research team member collected the intake form, and the data were subsequently entered into a central database. Patients completed the QuickDASH questionnaire at this visit as well.

The opioids prescribed and evaluated in this study included Percocet (5 mg oxycodone and 325 mg acetaminophen; Endo Pharmaceuticals Inc, Malvern, Pennsylvania) or an oxycodone 5-mg equivalent, Vicodin (5 mg hydrocodone and 325 mg acetaminophen; AbbVie Inc, Chicago, Illinois) or a hydrocodone 5-mg equivalent, and Tylenol #3 (30 mg codeine and 325 mg acetaminophen; Janssen Pharmaceuticals, Inc, Raritan, New Jersey). For the purposes of this study, amounts prescribed and used were converted to morphine equivalent units (MEUs) for more accurate comparisons.

For the statistical analysis, t tests were used for comparing two independent groups (gender, surgery type, concomitant surgery, arm dominance), one-way analysis of variance (ANOVA) for comparing more than two groups (McGowan Stage), Kruskal-Wallis nonparametric test was used to compare three or more groups with small sample sizes (EMG severity), while Mann-Whitney U nonparametric test was used as post hoc tests to make all pairwise comparisons following a significant Kruskal-Wallis test (surgery type, anesthesia type, insurance type). Finally, linear regression was used to examine the relationship between opioid intake and continuous predictors (age, QuickDASH scores).

The main goal of the current study was exploratory and aimed to simply estimate the average postoperative opioid intake level for this particular procedure. Sample size was set at 150 patients. This provides 95% power to detect whether the sample’s mean opioid level significantly differs from 0 (some constant). It also provides just over 85% to detect a medium difference between 2 groups of patients who undergo different variants of this procedure and 95% power to detect even a small correlation between opioid use and other continuous variables like age.

Results

Initially, 154 patients were enrolled in the study. Of these, 30 patients did not attend their postoperative visit and were lost to follow-up. Of the remaining 124 patients, 21 were either unwilling or unable to accurately report their pain medicine consumption and 3 had additional surgeries during the immediate postoperative period prior to their first postoperative visit, yielding a total of 100 patients for study analysis.

The first postoperative visit occurred at an average of 11.8 days after the procedure. During this time, the 100 patients consumed an average of 50 morphine equivalent units (MEUs) postoperatively (range, 0-300). This is equivalent to 7 Percocet or oxycodone 5-mg pills, 11 Tylenol #3 pills, 12.5 Vicodin pills, 12.5 Morphine (1 mg) pills, 10 Dilaudid (hydromorphone 1 mg) pills, or 10 Tramadol (50 mg) pills.

Overall, older patients had significantly lower opioid use than younger patients (beta = −0.03, P = .03) (Figure 1). Postoperative QuickDASH score was positively related to opioid intake when controlling for baseline scores (P = .009) (Figure 2), meaning patients who had more disability postoperatively tended to consume more opioids.

OPEN IN VIEWER

Figure 1.

Morphine equivalent units (mg) vs age (years) (beta = −0.03, P = .03).

OPEN IN VIEWER

Figure 2.

Morphine equivalent units (mg) vs postoperative QuickDASH score (beta = 0.02, P = .009).

Cubital tunnel release patients had marginally lower opioid use compared with UNT patients (40.4 vs 62.5 MEU, 5.4 vs 8.3 oxycodone 5-mg pills, P = .08) (Table 1). In post hoc testing using Mann-Whitney U nonparametric test to compare two groups, patients undergoing submuscular UNT consumed more opioids than patients undergoing CuTR (115.0 vs 40.4 MEU, 15.3 vs 5.4 pills, P = .003) and patients undergoing subcutaneous UNT (115.0 MEU vs 37.8 MEU, 15.3 vs 5.0 pills, P = .03) (Figure 3).

OPEN IN VIEWER

Table 1.

MEUs vs Type of Procedure.

Surgery	Number of patients	MEUs (95% confidence interval)	Oxycodone 5-mg tablets
Total	100	50.4 (38.4-62.4)	6.7
Cubital tunnel release	76	40.4 (16.5-64.4)	5.4
Ulnar nerve transposition	24	62.5 (34.2-90.7)	8.3
Subcutaneous ulnar nerve transposition	16	37.8 (19.6-55.9)	5.0
Submuscular ulnar nerve transposition	8	115.0 (46.1-183.9)	15.3

Note. Cubital tunnel release patients had marginally lower opioid use compared with ulnar nerve transposition patients (P = .08). MEUs = morphine equivalent units.

OPEN IN VIEWER

Figure 3.

Patients undergoing submuscular ulnar nerve transposition consumed more opioids than patients undergoing cubital tunnel release (P = .003) and patients undergoing subcutaneous ulnar nerve transposition (P = .03).

Opioid consumption was not significantly different among insurance types, but post hoc testing showed there was lower opioid use in Medicare patients compared with privately insured (42.7 vs 54.1 MEU, 5.7 vs 7.2 pills, P = 0.02) and workers’ compensation patients (42.7 vs 76.8 MEU, 5.7 vs 10.2 pills, P = 0.04) (Table 2, Figure 4).

OPEN IN VIEWER

Table 2.

MEUs vs Insurance Type.

Insurance type	Number of patients	MEUs (95% confidence interval)	Oxycodone 5-mg tablets
Medicare	14	42.7 (0-92.0)	5.7
Private	52	54.1 (39.8-68.5)	7.2
Workers’ compensation	14	76.8 (24.1-129.5)	10.2

Note. There was lower opioid use in Medicare patients compared with privately insured (P = .02) and workers’ compensation patients (P = .04). MEUs = morphine equivalent units.

OPEN IN VIEWER

Figure 4.

Graph showing lower opioid use in Medicare patients compared with privately insured (P = .02) and workers’ compensation patients (P = .04).

Patients who had procedures on their dominant hand used similar amounts of opioids compared with nondominant-hand procedures (39.5 vs 56.0 MEU, 5.3 vs 7.5 pills, P = .09) (Supplemental Table 1, Supplemental Figure 1).

There were similar amounts of opioid use with higher preoperative EMG severity (P = .14) (Supplemental Table 2, Supplemental Figure 2).

McGowan Stage did not predict opioid use, although McGowan Stage 2 was marginally higher than 3 (P = .096) (Supplemental Table 3, Supplemental Figure 3).

Anesthesia type, concomitant surgeries, and gender did not have an association with opioid use.

Discussions

The opioid crisis in the United States is becoming a top priority for medicine and for the country in general. As prescribers of opioid medications, orthopedic surgeons have a difficult task in controlling postoperative pain in patients undergoing painful procedures.^8,9 As we attempt to control the amounts of opioids we prescribe, there has been a push to understand how much pain medication is appropriate for certain procedures. Kim et al and Rodgers et al have reported that upper extremity surgeons tend to overprescribe opioids.^10,12 Rodgers et al found that 53% of opioid medications were consumed by postoperative day 2, and they recommended decreasing prescriptions to 15 pills with an option of one refill. O’Neill et al also found older patients to consume fewer pain medications. ¹⁷ Their recommendation for distal radius fractures was to prescribe 15 to 20 opioid pills. Stanek et al reported a significant decrease in opioid prescriptions after implementing a limit to their opioid prescriptions for upper extremity patients. ¹⁴ They had a limit of 40 pills for their larger upper extremity procedures such as wrist fusions, with fewer pills recommended for smaller soft tissue procedures.

In our study, patients consumed an average of 7 (range, 0-40) opioid pills, meaning 50 (range, 0-300) MEUs. Younger patients, patients with worse functioning scores postoperatively, privately insured patients, workers’ compensation patients, and patients undergoing submuscular UNT procedures tended to consume more opioids. Overall, our patients consumed 44.5% of the pills they were prescribed.

Using 7 pills as a baseline, orthopedic surgeons should be able to assess how many more pills patients may need in the immediate postoperative period. Patients undergoing submuscular UNT required significantly more opioids postoperatively. This intuitively makes sense considering the greater soft tissue dissection involved compared with the other techniques. Therefore, surgeons can adjust the number of pain pills these patients may need postoperatively if a submuscular nerve transposition was performed.

Strengths of this study include its prospective nature and standard protocol in collecting the data. Patients brought their pain medications to the office to count the number of pills consumed, which eliminated potential recall bias.

Although these data were collected prospectively, patients still may have misplaced or miscounted pills in the time period between surgery and their first postoperative visit. In addition, recall bias may affect a patient’s ability to accurately represent their opioid pill consumption. Moreover, psychological and psychiatric factors can also impact a patients’ consumption pattern of opioids, which were outside the scope of this study and therefore not controlled for nor measured.

This information can be useful for both surgeons and patients. Surgeons can take this opioid consumption pattern after CuTR surgery to better inform opioid prescribing to effectively manage their patients’ postoperative pain while avoiding inadvertent overprescribing after CuTR surgery.

Conclusions

Patients consumed an average of 7 oxycodone 5-mg pills after cubital tunnel decompression surgery. Younger patients, patients with worse functional scores, privately insured patients, workers’ compensation patients, and patients undergoing submuscular UNT procedures consumed more opioids.

Supplemental Material