Postoperative Opioid Reduction Using a Multimodal Pain Protocol for Outpatient Orthopaedic Sports Medicine Surgery

Abstract

Background:

Orthopaedic surgery regularly employs opioids for postoperative pain management. Multimodal pain protocols have been shown to reduce opioid prescriptions in orthopaedic surgery.

Purpose:

To analyze the impact of a division-level multimodal pain protocol for orthopaedic sports medicine surgery on opioid prescription reduction and pain level postoperatively.

Study Design:

Cohort study; Level of evidence, 3.

Methods:

All orthopaedic sports medicine procedures at a military treatment facility were categorized into 1 of 3 pain tiers. A pain protocol emphasizing multimodal pain control was implemented for each tier. A retrospective cohort study compared pre- and postprotocol pain groups for each respective tier (n = 40 in each) for visual analog scale (VAS) for pain scores 2 weeks postoperatively, postoperative opioid prescription, and time to discontinuation of postoperative opioids.

Results:

The mean number of opioid pills prescribed for all tiers decreased by 43.6% (preprotocol 35.7 ± 3.1; postprotocol 20.1 ± 1.5; P < .0001) after pain protocol implementation. Of the total opioids prescribed in the postprotocol cohort, a mean of 64.1% were consumed. There was no significant difference in overall visual analog scale for pain scores at 2 weeks postoperatively (preprotocol 2.72 ± 0.41; postprotocol 2.99 ± 0.43; P = .40). At 2 weeks postoperatively, only 1 patient continued opioids in the postprotocol group compared with 20 patients with continued opioid use in the preprotocol group (P < .001).

Conclusion:

A division-level multimodal pain protocol applied to orthopaedic sports medicine procedures led to decreased opioid prescription postoperatively with no significant difference in 2-week postoperative pain scores compared with more opioid reliant and variable protocols in a cohort of military service members. Despite the reduced prescription, patients consumed a mean of 64.1% of pills, indicating continued overprescription.

Pain control after orthopaedic sports medicine surgery requires tailored protocols for the range of postoperative pain from relatively minor minimally invasive arthroscopic procedures to major multiligamentous reconstructions and osteotomies. Historically, opioids have been a mainstay of postoperative pain management. However, we now have a greater understanding of the complications and addiction risk associated with liberal opioid use.^3,36 Without judicious prescription, opioids have the potential for severe consequences to morbidity and mortality on a societal level.^18-20,29,30 Continued evidence of variable prescription patterns can contribute to excessive opioid use after the acute postoperative period, which is associated with an increased risk of addiction.^25,31,34,38 Clinical teams aim to provide adequate perioperative pain management for comfortable recovery and successful rehabilitation while maintaining judicious opioid prescription. To accomplish this goal, some institutions have turned to standardized multimodal pain protocols.^5,10,13,33

Recent clinical practice guidelines on postoperative pain from the American Pain Society describe multimodal analgesia as “the use of a variety of analgesic medication and techniques that target different mechanisms of action in the peripheral and/or central nervous system, which can also be combined with nonpharmacologic interventions and might have additive or synergistic effects and more effective pain relief compared with single-modality interventions.” ⁶ In a meta-analysis of multimodal analgesia, Elia et al ⁹ demonstrated superior pain relief and decreased opioid requirements compared to isolated opioid therapy.

Multimodal pain management for orthopaedic surgery has incorporated a multitude of medications and prescription quantities over time with the goal of improving postoperative pain control and decreasing opioid prescription. Postoperative pain protocols can be variable among surgeons within the same department and across institutions.^31,38 Given no standard protocol for various levels of surgery, the purpose of our study was to analyze a tiered multimodal pain protocol applied to orthopaedic sports medicine procedures to decrease opioid prescriptions while maintaining appropriate pain management. Our hypothesis is that many surgeons may be unknowingly overprescribing opioids and that standardized multimodal pain protocols designed with surgical tiers for respective surgical procedures will reduce opioid prescription while maintaining adequate perioperative pain control.

Methods

This retrospective cohort study was approved by the institutional review board at the military treatment facility. An orthopaedic sports medicine division protocol leveraging multimodal perioperative pain techniques was developed in 2019 by board-certified sports medicine surgeons. Three tiers of sports medicine procedures were developed based on perceived pain and opioid demands (Table 1). Given the limited available literature regarding orthopaedic opioid-prescribing guidelines, medication protocols were developed by consensus.^32,36,37

Table 1.

Procedures Grouped in Tiers ^a

Basic knee arthroscopy: meniscal debridement, loose body removal, chondroplasty

Moderate knee arthroscopy: meniscus repair, meniscus root repair, microfracture, allograft ACLR Basic open knee: medial patellofemoral ligament reconstruction Basic shoulder arthroscopy: subacromial decompression, distal clavicle excision (open or arthroscopic), labral repair, biceps tenotomy/tenodesis, subscapularis repair

Complex knee: autograft ACLR, posterior cruciate ligament reconstruction, osteochondral allograft transplant, meniscal transplant, high tibial osteotomy, distal femoral osteotomy, tibial tubercle osteotomy Complex shoulder: rotator cuff repair, Laterjet, shoulder arthroplasty

^aProcedures grouped by tier 1, tier 2, and tier 3 based on complexity and surgical pain. ACLR, anterior cruciate ligament reconstruction.

A literature review was performed to assess other best practices for opioid-prescribing methodologies for surgical procedures in orthopaedic sports medicine and nonsports medicine. The literature review also investigated other potential concerns regarding nonopioid modalities, such as nonsteroidal anti-inflammatory drug (NSAID) use after revision anterior cruciate ligament surgery, healing rates after osteotomies, and rotator cuff retear rates.^11,24,27,35 The orthopaedic sports medicine division was queried on its current opioid-prescribing practice. Using the most recent literature at the time by Stepan et al, ³⁶ the division elected to decrease prescribed opioid quantities by >50% of Stepan's recommended levels and introduced nonopioid modalities. The goal of the protocol was to have as many patients as possible weaned off opioid and gabapentinoids by the end of postoperative day 3. The protocol was then reviewed and approved by board-certified pain management anesthesiologists in the pain management service and department of anesthesia.

Order sets were created in the electronic medical record with instructions provided to all department residents and staff on determining tier status and correctly finding order sets. Simple go-by forms with pain protocol tiers and surgical procedures were placed in all physician workspaces. Orders were reviewed periodically to ensure that staff and residents followed new ordering protocols. Orthopaedic sports medicine providers were ultimately given discretion to order other medications and/or quantities based on their clinical judgment but were encouraged to follow the protocol. All 6 surgeons in the orthopaedic sports medicine division agreed with this protocol before initiating.

All patients reviewed and signed a pain protocol instruction (Figure 1) explaining the importance of the aim to wean off opioids as soon as able and to maximize nonopioid medication use. Patients were also informed that, if needed, a refill could be provided during business hours. However, patients were not informed that the protocol allowed patients to receive 1 refill of 50% of their initial opioid prescription quantity so as to not create a bias by letting patients think they were supposed to obtain a refill.

Figure 1.

Naval Medical Center San Diego Orthopaedic Sports Medicine pain protocol instruction.

Protocol

The multimodal pain protocol (Table 2) was developed with a consistent base of nonopioid analgesic agents throughout all 3 tiers. Celecoxib 200 mg was prescribed 1 to 2 hours preoperatively, followed by 200 mg daily for 2 weeks postoperatively. Acetaminophen 975 mg was prescribed once preoperatively, followed by 975 mg 3 times daily postoperatively for as long as necessary. For patients in the tier 2 and tier 3 groups, gabapentin 600 mg was given 1 to 2 hours preoperatively, followed by 300 mg 3 times daily for 3 days postoperatively. All patients received a peri- or intra-articular injection of weight-based 0.25% bupivacaine local anesthetic intraoperatively. The opioid prescription was oxycodone 5 mg immediate release for all patients. No long-acting opioids were prescribed. Oxycodone 5 mg immediate release was prescribed as 10 tablets for patients in tier 1, 20 tablets for patients in tier 2, and 30 tablets for patients in tier 3. Aspirin 325 mg daily for 3 weeks was provided for deep vein thrombosis prophylaxis for all patients in all tiers unless a different anticoagulant was indicated. In addition, docusate sodium and ondansetron were prescribed to manage adverse effects as needed. All prescriptions were ordered at the preoperative appointment, given that many procedures occurred at outpatient surgery centers without a pharmacy.

Table 2.

Multimodal Pain Protocols per Surgical Tier ^a

Surgical Tier	Medication	Dosage/Frequency	No. of Tabs	Refill
1	Preop
	Celebrex	200 mg/1-2 h preoperatively	1	—
	Tylenol	325 mg/1-2 h preoperatively	3	—
	Intraop
	Peri-/intra-articular injection (0.25% bupivacaine)	Max weight-based/×1	—	—
	Postop
	Tylenol	325 mg/3 tabs TID	60	OTC
	Celebrex	200 mg/1 tab daily	14	OTC NSAIDs
	Oxycodone IR	5 mg/1 tab q4-6 h PRN	10	5
2	Preop
	Celebrex	200 mg/1-2 h preoperatively	1	—
	Tylenol	325 mg/1-2 h preoperatively	3	—
	Gabapentin	300 mg/1-2 h preoperatively	2	—
	Intraop
	Peri-/intra-articular injection (0.25% bupivacaine)	Max weight-based/×1	—	—
	Postop
	Tylenol	325 mg/3 tabs TID	90	OTC
	Celebrex	200 mg/1 tab daily	14	OTC NSAIDs
	Oxycodone IR	5 mg/1 tab q4-6h PRN	20	10
	Gabapentin	300 mg/1 tab TID	9	—
3	Preop
	Celebrex	200 mg/1-2 h preoperatively	1	—
	Tylenol	325 mg/1-2 h preoperatively	3	—
	Gabapentin	300 mg/1-2 h preoperatively	2	—
	Intraop
	Peri-/intra-articular injection (0.25% bupivacaine)	Max weight-based/×1	—	—
	Postop
	Tylenol	325 mg/3 tabs TID	90	OTC
	Celebrex	200 mg/1 tab daily	14	OTC NSAIDs
	Oxycodone IR	5 mg/1 tab q4-6 h PRN	30	15
	Gabapentin	300 mg/1 tab TID	9	—

^aIntraop, intraoperative; IR, immediate release; NSAIDs, nonsteroidal anti-inflammatory drugs; OTC, over the counter; postop, postoperative; preop, preoperative; PRN, as needed; TID, 3 times daily.

The electronic medical record was used to collect patient self-reported demographics, number of opioid pills prescribed, type of opioid in the preprotocol group (oxycodone 5 mg or oxycodone-acetaminophen 5-325 mg), opioid tablets consumed at 2 weeks in the postprotocol group, and number of opioid refills. Visual analog scale for pain (VAS-P) scores were the primary outcome and are recorded as a standard of practice within the orthopaedic sports medicine division at postoperative follow-up by orthopaedic staff in clinic. The standard of practice also includes recording which patients remain reliant on opioids for pain relief at all postoperative follow-ups. Patients in the postprotocol group were instructed to bring any unused opioid tablets to their first postoperative visit at 2 weeks for documentation of total pills remaining and referred to the pharmacy for safe opioid disposal; this is not a standard of practice and was not performed for any patients in the preprotocol group.

Cohort Selection

In total, 40 patients per pre- and postprotocol tiers were retrospectively collected (Figure 2). Patients in the preprotocol group were collected consecutively in reverse chronologic order between January 2019 and November 2018. The surgical schedule registry was filtered for all cases completed by orthopaedic sports medicine division surgeons. Once 40 patients were identified for a tier based on the list of procedures outlined in Table 1, that tier no longer continued recruitment for qualifying cases. This method of recruitment was performed to not dilute tier 3 effects in the overall combined analysis. All patients from this range were opioid naïve.

Figure 2.

Recruitment flowchart.

Patients in the postprotocol group were collected consecutively in chronologic order between September and November 2019. At this time, pain protocols were fully integrated and consistent across the orthopaedic sports medicine division. The surgical schedule registry was filtered for all cases completed by orthopaedic sports medicine division surgeons. As in preprotocol recruitment, once 40 patients were identified for a tier based on procedures outlined in Table 1, that tier no longer continued recruitment for qualifying cases. One patient from this range was not opioid naïve and was excluded. Patients were excluded from the postprotocol group if they deviated from protocol by not having the correct level of procedure for the corresponding pain protocol according to tier. For instance, a surgeon may have prescribed a tier 2 regimen with 20 oxycodone tablets for someone they believed would require a meniscus repair but ultimately performed a meniscus debridement, which is a tier 1 procedure necessitating 10 oxycodone tablets. Eight patients were excluded for this reason from tier 1, 4 patients from tier 2, and 6 patients from tier 3.

Statistical Analysis

Analyses were performed for 40 patients per tier from the pre- and postprotocol groups. A sample of 40 patients per tier was estimated based on VAS-P scores as the primary outcome using a Student t test with alpha .05, 80% power, and medium-large effect size 0.63 using G*Power 3.1.7. VAS-P scores were compared between the pre- and postprotocol groups for each tier and overall. Total pills prescribed were compared between the pre- and postprotocol groups using a Student t test for each tier and overall. The distribution of baseline characteristics was compared between the pre- and postprotocol groups using a Student t test, Fisher exact test, and chi-square test. Statistical significance was set at P < .05. All statistical analyses were performed using SAS University Edition (SAS Institute Inc).

Results

In total, 40 patients per tier from the pre- and postprotocol groups were included in the analyses. There was no statistically significant difference in age, race, or sex between the pre- and postprotocol groups (Table 3).

Table 3.

Patient Demographic Data ^a

	Preprotocol	Postprotocol	P
Race			.55
White	63 (52.5)	71 (59.17)
Black	18 (15)	18 (15)
Asian	8 (6.7)	4 (3.3)
Other	31 (25.8)	27 (22.5)
Male	101 (84.2)	109 (90.8)	.17
Age, y	31.89 ± 9.16	32.06 ± 9.34	.95

^aData are reported as n (%) or mean ± SD.

There was an overall significant decrease of 43.6% (preprotocol 35.7 ± 3.1; postprotocol 20.1 ± 1.5; P < .0001) in total opioid pills prescribed across all 3 tiers after implementation of the pain protocol, and statistically significant decreases in all 3 tiers. The total number of opioid pills prescribed decreased by 64.1% in tier 1, by 40.6% in tier 2, and by 33.2% in tier 3 (Table 4).

Table 4.

Opioid Pills Prescribed Pre- and Postprotocol ^a

Tier	Preprotocol	Postprotocol	P
Pills prescribed overall	35.7 (32.6-38.7)	20.1 (18.6-21.7)	<.0001
Pills prescribed in tier 1	27.9 (23.0-32.8)	10 (10-10)	<.0001
Pills prescribed in tier 2	33.7 (29.4-38.0)	20 (20-20)	<.0001
Pills prescribed in tier 3	45.5 (40.1-50.9)	30.4 (29.6-31.1)	<.0001

^aData are reported as mean (95% CI).

In the postprotocol group, a mean of 6.8 ± 1.3 opioid pills were consumed postoperatively in tier 1, a mean of 11.4 ± 2.4 opioid pills were consumed in tier 2, and a mean of 20.6 ± 3.3 opioid pills were consumed in tier 3. The overall postprotocol cohort consumed a mean of 12.9 ± 11.7 opioid pills (Table 5). Opioid consumption was 64.1% of the total pills prescribed in the postprotocol group.

Table 5.

Opioid Pills Consumed by Tier ^a

Tier	Postprotocol
Opioid pills consumed overall	12.9 (11.2-14.7)
Opioid pills consumed in tier 1	6.8 (5.5-8.1)
Opioid pills consumed in tier 2	11.4 (9.0-13.9)
Opioid pills consumed in tier 3	20.6 (17.3-24.0)

^aData are reported as mean (95% CI).

There were 5 refills among 4 patients in the tier 3 preprotocol group, 1 of whom received 2 refills. There was 1 refill for 1 patient in the tier 3 postprotocol group who underwent rotator cuff repair and biceps tenodesis. The other 2 tiers, both pre- and postprotocol, had no refills. There was no statistically significant difference between the rate of refills in patients in tier 3 (P = .20).

At 2 weeks postoperatively, 1 patient in the postprotocol group continued opioids compared with 20 patients in the preprotocol group with continued opioid use (P < .001). At 6 weeks postoperatively, all patients in the postprotocol group had discontinued opioids compared with 9 patients in the preprotocol group with continued opioid use (P < .001). All patients in the pre- and postprotocol groups discontinued opioids by 12 weeks postoperatively.

There was no statistically significant difference between overall pre- and postprotocol VAS-P scores at 2 weeks postoperatively (preprotocol 2.72 ± 0.41; postprotocol 2.99 ± 0.43; P = .40) (Table 6). Tier 1 demonstrated a statistically significant increase in VAS-P scores at 2 weeks (preprotocol 1.83 ± 0.82; postprotocol 3.18 ± 0.76; P = .017). Tier 2 had no statistically significant difference in VAS-P scores (preprotocol 2.74 ± 0.89; postprotocol 1.96 ± 0.72; P = .17). Tier 3 had no statistically significant difference in VAS-P scores (preprotocol 3.33 ± 0.82; postprotocol 3.72 ± 0.73; P = .47).

Table 6.

Mean VAS-P Scores Postoperatively ^a

Tier	Preprotocol	Postprotocol	P
Overall VAS-P at 2 weeks	2.72 (2.31-3.21)	2.99 (2.56-3.43)	.40
Tier 1 VAS-P at 2 weeks	1.83 (1.01-2.66)	3.18 (2.42-3.93)	.017
Tier 2 VAS-P at 2 weeks	2.74 (1.85-3.63)	1.96 (1.24-2.68)	.17
Tier 3 VAS-P at 2 weeks	3.33 (2.51-4.15)	3.72 (2.99-4.46)	.47

^aData are reported as mean (95% CI). VAS-P, visual analog scale for pain.

Discussion

This study analyzed the impact of a division-level multimodal pain protocol with 3 tiers designed to address the breadth of orthopaedic sports medicine procedures and drive judicious postoperative opioid prescriptions. We found an overall mean decrease of 43.6% in opioid pills prescribed after its implementation. This is an improvement on the study by Stepan et al, ³⁶ which demonstrated a 21.2% decrease in mean oral morphine equivalents for ambulatory sports surgery with their pain protocol. Patients consumed a mean of 64.1% of the total opioid pills prescribed despite the 43.6% decrease in opioid prescriptions. This supports the hypothesis that surgeons may be unknowingly overprescribing opioids.

Although there was a significant reduction in overall opioid prescription, we did not find a statistically significant increase in overall VAS-P scores or opioid refills comparing the pre- and postprotocol groups. Tier 1 VAS-P scores were statistically significant and have debatable clinical significance, which previous literature supports as 13 mm on a 0- to 100- mm VAS-P scale for minimal clinically significant difference while other orthopaedic specific studies challenge this threshold.^4,7,12 The lack of statistical significance in the overall VAS-P score indicates adequate postoperative pain control in the setting of a significant opioid reduction. The efficacy of multimodal pain protocols on decreasing pain has been mixed.^9,27 We hypothesized that we did not find a statistically significant reduction in VAS-P scores at 2 weeks given that scores were relatively low at that postoperative juncture in both the pre- and postprotocol cohorts. It is possible that a more profound difference would be identified in the acute postoperative period, which requires further study.

Orthopaedic surgeons have been identified to be contributors to inadvertent opioid overprescribing as the third-highest opioid prescribers among all specialties and the highest among surgical specialties.^17,23 The use of opioids after the acute perioperative period is associated with addiction. Prior literature demonstrates that the duration, rather than the dosage, of opioid prescriptions has a greater association with misuse. ¹¹ Johnson et al ¹⁶ found that 13% of opioid-naïve patients continue to fill opioid prescriptions more than 90 days after hand surgery. In a review of military patients undergoing hip arthroscopy, Rhon et al ²⁸ found that 45.9% of their 1642 participants remained on opioids 1 year postoperatively. They also found the following predictors of chronic opioid use: socioeconomic status; prior opioid use; prior nonopioid pain medication; high health-seeking behavior; insomnia; mental health disorder; and substance abuse. Shah and Martin ³⁴ assessed risks of chronic opioid use based on postoperative prescriptions and found that the incidence of long-term usage is approximately 15% in patients whose first episode of consumption is >8 days and increases to 30% if the initial opioid prescription lasts 31 days or more. Therefore, attempts at reducing postoperative opioid prescriptions via multimodal pain management plays a critical role in reducing the risk of chronic opioid use. This study's multimodal pain protocol observes these principles, as demonstrated in tier 3 of our protocol; if patients were taking the maximum dose of oxycodone, they would have completed their prescription by 2.5 days postoperatively.

Multimodal pain management in orthopaedic surgery has been complicated by concerns regarding the use of NSAIDs on fracture healing, impaired muscle regeneration, and weakened tendon-bone healing. ²⁶ NSAID use in orthopaedic sports medicine has had mixed results regarding its effects on revision rates and graft survival for anterior cruciate ligament reconstruction and rotator cuff repair specifically.^15,24,35 Although controversial, the trend across orthopaedic surgery has been an increase in NSAID use in order to decrease opioid consumption.

Other components of multimodal pain protocols include gabapentin, acetaminophen, peri-incisional local anesthesia, and regional blocks. The addition of gabapentin to postoperative pain protocols has demonstrated mixed effects across a multitude of literature. Gabapentin has been found to have heterogeneous outcomes, with Bang et al ² identifying a statistically significant decrease in VAS scores but no difference in opioid consumption in arthroscopic rotator cuff repair. However, Mardani-Kivi et al ²¹ found a statistically significant decrease in opioid consumption but no difference in VAS-P scores in the first 24 hours postoperatively. The combination of preoperative gabapentin and celecoxib in patients who had undergone elective spine fusion was found to have a statistically significant decrease in mean VAS-P scores, morphine equivalent consumption, and earlier mobility compared with the placebo group. In arthroscopic ACL reconstruction, a single preoperative dose of gabapentin 1200 mg was found to have a significant decrease in morphine consumption and improved passive and active range of motion at 24 and 48 hours after surgery. ²²

Pain protocols in other orthopaedic subspecialties have seen similar decreases in opioid prescriptions. Dwyer et al ⁸ found a 35% to 55% reduction in opioid prescriptions for mini-open carpal tunnel and volar plating of distal radius fractures when surgeons were provided written guidelines and patients were given educational handouts. Stepan et al ³⁶ found a 50% decrease in opioid pills prescribed for carpal tunnel and distal radius fixation with prescribed education and an institutional protocol. Patient education may have also been a factor that contributed to a decrease in opioids in that study as the pain contract contained information about decreasing opioid use as well as a renewed focus on addressing postoperative pain management at the preoperative visit. Further research should build upon combining the most promising features of each protocol to better educate the orthopaedic surgeon and patient on adequate postoperative pain management.

Limitations

The limitations of this study include its retrospective design and possible patient diversion of opioids ahead of postoperative visits, which would contribute to inaccurate opioid pill counts. The findings do not provide evidence regarding the comparison of pre- and postprotocol opioid consumption, only prescription. Due to the simultaneous introduction of multiple pain medications within the multimodal protocol, the quantified effect of each individual component is unable to be isolated. Preprotocol opioid prescription also included occasional combination opioids such as oxycodone-acetaminophen compared with only oxycodone in the postprotocol group; however, the oxycodone 5-mg dose was consistent across patients in the pre- and postprotocol groups. Lastly, this sample of opioid-naïve, active-duty military and their dependents may not be generalizable to other populations. ²⁵

The military health care system does provide an advantage in which all patients receive their prescriptions in a single system, enabling comprehensive review and assurance that no additional refills or opioids were prescribed by other providers. Paramount to any future opioid guideline is the role of patient counseling. Prior literature has identified a 13% decrease in discontinued opioid use in patients with orthopaedic trauma at 6 weeks postoperatively and a 66% decrease in opioid tablets for patients with carpal tunnel after undergoing perioperative opioid counseling.^1,14 As this study protocol iterates and refines, it will expand to include standardization of this preoperative counseling.

Discussion

In conclusion, our study found a 43.6% decrease in opioid pills prescribed with no significant increase in VAS-P scores or opioid refills after the implementation of a multimodal pain protocol applied in an orthopaedic sports medicine division. Patients consumed 64.1% of their total prescribed opioids in the postprotocol group, indicating adequate pain control and evidence of continued overprescription. Further research should be performed to identify mean opioid consumption for specific surgical procedures to further tailor opioid-prescribing guidelines.