Lindsey Wurster and Sarah Brandt are Physician Assistants, Patricia Mecum is a Family Nurse Practitioner, Kenneth Gundle and Lucas Anissian are Attending Orthopedic Surgeons, all at US Department of Veterans Affairs Portland Health Care System in Oregon. Erik Woelber is an Orthopedic Surgery Resident, and Kenneth Gundle is an Attending Physician, both in the Orthopedic Department at Oregon Health and Sciences University in Portland. Correspondence: Lindsey Wurster (lindsey.wurster@va.gov)
Author disclosures The authors report no actual or potential conflicts of interest with regard to this article.
Disclaimer The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the US Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review the complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.
Electronic health records were reviewed and data were abstracted. The data included demographic information (age, gender, body mass index [BMI], diagnosis, and procedure), surgical factors (American Society of Anesthesiologists score, Risk Assessment and Predictive Tool score, operative time, tourniquet time, estimated blood loss), hospital factors (length of stay [LOS], discharge location), postoperative pain scores (measured on postoperative day 1 and on day of discharge), and postdischarge events (90-day complications, telephone calls reporting pain, reoperations, returns to the ED, 90-day readmissions).
The primary outcome was the mean postoperative daily MED during the inpatient stay. Secondary outcomes included pain on postoperative day 1, pain at the time of discharge, LOS, hospital readmissions, and ED visits within 90 days of surgery. Because different opioid pain medications were used by patients postoperatively, all opioids were converted to MED prior to the final analysis. Collected patient data were de-identified prior to analysis.
Power analysis was conducted to determine whether the study had sufficient population size to reject the null hypothesis for the primary outcome measure. Because practitioners controlled postoperative opioid use, a Cohen’s d of 1.0 was used so that a very large effect size was needed to reach clinical significance. Statistical significance was set to 0.05, and patient groups were set at 20 patients each. This yielded an appropriate power of 0.87. Population characteristics were compared between groups using t tests and χ2 tests as appropriate. To analyze the primary outcome, comparisons were made between the 2 cohorts using 2-tailed t tests. Secondary outcomes were compared between groups using t tests or χ2 tests. All statistics were performed using R version 3.5.2. Power analysis was conducted using the package pwr.11 Statistical significance was set at P < .05.
Results
Forty patients met the inclusion criteria, evenly divided between those undergoing TKA before and after instituting the MOJO protocol (Table 2). A single patient in the MOJO group died and was excluded. A patient who underwent bilateral TKA also was excluded. Both groups reflected the male predominance of the VA patient population. MOJO patients tended to have lower BMIs (34 vs 30, P < .01). All patients indicated for surgery with preoperative opioid use were able to titrate down to their preoperative goal as verified by prescriptions filled at VA pharmacies. Twelve of the patients in the MOJO group received adductor canal blocks.
Results of t tests and χ2 tests comparing primary and secondary endpoints are listed in Table 3. Differences between the daily MEDs given in the historical and MOJO groups are shown. There were significant differences between the pre-MOJO and MOJO groups with regard to daily inpatient MEDs (82 mg vs 29 mg, P < .01) and total inpatient MEDs (306 mg vs 32 mg, P < .01). There was less self-reported pain on postoperative day 1 in the MOJO group (5.5 vs 3.9, P < .01), decreased LOS (4.4 days vs 1.2 days, P < .01), a trend toward fewer total ED visits (6 vs 2, P = .24), and fewer discharges to skilled nursing facilities (12 vs 0, P < .01). There were no blood transfusions in either group.
There were no readmissions due to uncontrolled pain. There was 1 readmission for shortness of breath in the MOJO group. The patient was discharged home the following day after ruling out thromboembolic and cardiovascular events. One patient from the control group was readmitted after missing a step on a staircase and falling. The patient sustained a quadriceps tendon rupture and underwent primary suture repair.
Discussion
Our results demonstrate that a multimodal approach to significantly reduce postoperative opioid use in patients with TKA is possible without increasing readmissions or ED visits for pain control. The patients in the MOJO group had a faster recovery, earlier discharge, and less use of postoperative opioid medication. Our approach to postoperative pain management was divided into 2 main categories: patient optimization and surgical optimization.
Patient Selection
Besides the standard evaluation and optimization of patients’ medical conditions, identifying and optimizing at-risk patients before surgery was a critical component of our protocol. Managing postoperative pain in patients with prior opioid use is an intractable challenge in orthopedic surgery. Patients with a history of chronic pain and preoperative use of opioid medications remain at higher risk of postoperative chronic pain and persistent use of opioid medication despite no obvious surgical complications.8 In a sample of > 6,000 veterans who underwent TKA at VA hospitals in 2014, 57% of the patients with daily use of opioids in the 90 days before surgery remained on opioids 1 year after surgery (vs 2 % in patients not on long-term opioids).8 This relationship between pre- and postoperative opioid use also was dose dependent.12
