Translate this page into:
Effect of post-operative NSAID use on rotator cuff repair outcomes
, Fernando A. Huyke-Hernandeza,b, Rebekah M. Kleinsmitha,b, Stephen A. Doxeya,b, Adam Schweitzerc, Jacob Ristowd, Brian P. Cunninghama,b, Jonathan Bramane
⁎Corresponding author: Adam Hadro. adam.hadro@tria.com
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
The impact of non-steroidal anti-inflammatory drugs (NSAIDs) on rotator cuff repair is an ongoing area of study within orthopedics, with conflicting results in current literature. Despite concerns over the deleterious effects of NSAIDs on rotator cuff healing, they are becoming an integral part of a multimodal post-operative pain control regiment. The purpose of this study was to compare post-operative patient-reported outcomes (PROs), complications rates, and retear rates of arthroscopic rotator cuff repairs in patients using ibuprofen post-operatively to those who abstained from NSAIDs for six weeks after surgery. It was hypothesized that a short course of ibuprofen post-operatively would not lead to inferior PRO scores, increased retear rates, nor increased complication rates after arthroscopic rotator cuff repair.
Patients of the primary surgeon who underwent arthroscopic rotator cuff repair between 2012 and 2022 were evaluated by retrospective chart review. In May 2017 the primary surgeon changed his protocol from avoiding NSAIDs for six weeks after surgery to routinely prescribing two weeks of Ibuprofen 800 mg TID post-operatively. Patients who avoided NSAIDs for six weeks were compared to patients who were prescribed NSAIDs post-operatively. Patient demographic data, pre-operative MRI results, pre-operative and post-operative PROs were collected from the EMR. Additionally, post-operative complications and repair failures requiring reoperation within one year were evaluated.
125 patients met inclusion criteria for this study with 36 patients in the NSAID group and 89 in the no NSAID group. When comparing improvement in PROs, the NSAID group reached MCID at one year in 83.8 % of patients and the no NSAID group reached MCID at one year in 73.9 % of patients. There was no significant difference between the groups in reaching MCID improvement at one year (p = 0.471). Five post-operative complications were reported in the no NSAID group and two in the NSAID group (5.7 % vs 5.4 %, respectively, p = 0.827). Finally, there was no significant difference in the percentage of post-operative rotator cuff repair failures requiring revision in the first year between the groups (2.3 % vs 2.7 %, p = 1.000).
There was no difference in percent of patients improving their PRO by the MCID between the groups that used ibuprofen and the group that did not. There was also no difference in post-operative complication rates and rates of symptomatic retear requiring reoperation between the groups. This supports that a short course of NSAIDs post-operatively, specifically ibuprofen, after rotator cuff repair does not increase reoperation rates nor lead to a clinically significant decrease in PROs at one year.
Keywords
NSAID
Rotator cuff
Shoulder
Ibuprofen
Arthroscopy
1 Introduction
Non-steroidal anti-inflammatory drugs (NSAIDs) are increasingly being used as an integral part of multimodal post-operative pain control to reduce the use of opioid medications after orthopedic surgery. Mandava et al. demonstrated that NSAIDs are effective in decreasing post-operative opioid consumption after rotator cuff repair surgery.1 Despite this benefit, concerns remain that NSAIDs can inhibit soft tissue healing after rotator cuff repair. In multiple animal studies on rotator cuff repair, NSAID administration resulted in lower tendon healing rates, lower tendon tension to failure, and decreased collagen organization and maturation.2–5 As a result, NSAID usage after rotator cuff repair has historically been discouraged despite the benefit of lowering narcotic consumption.6
Current literature regarding the post-operative effect of NSAIDs on rotator cuff repair outcomes are limited with conflicting results. Tangtiphaiboontana et al. randomized 101 patients undergoing rotator cuff repair to receive ibuprofen or placebo for two weeks after surgery. They reported no difference in healing rate on ultrasound at one year. However, they did report significantly greater forward flexion and higher ASES scores at 6 months in the ibuprofen group. These differences were no longer significant at 12 months.7 Lim et al. performed a metanalysis including 916 patients undergoing rotator cuff repair. They reported no significant difference in retear rates between the 443 patients using post-operative NSAIDs and the 473 in the no NSAID control group.8 In contrast, Oh et al. performed a double-blinded randomized controlled trial of 180 patient undergoing rotator cuff repair who were randomly allocated to receive ibuprofen, celecoxib, or tramadol after surgery. At a minimum two-year follow-up, the celecoxib patients had a significantly higher rate (11/30) of tendon healing failure on MRI than the ibuprofen (2/27) and tramadol groups (1/25). Even with these differences in healing, there was no difference in functional scores between the three groups.9
1.1 Study goals
Further studies are needed to clearly define the effects of specific NSAIDs on rotator cuff repair outcomes. The purpose of this study was to compare post-operative outcomes of rotator cuff repair in patients prescribed a short course of ibuprofen after surgery to those who abstained from NSAID usage for 6 weeks after surgery. Specifically, PROs improvement at one-year, post-operative complication rates, and failure of tendon healing requiring revision surgery. This study is one of only a few that directly addresses outcomes related to ibuprofen consumption after rotator cuff repair. It is hypothesized that a short course of ibuprofen would not lead to a clinically significant decrease in PROs nor a greater number of post-operative complications at one year.
2 Materials and methods
2.1 Patient selection
This retrospective cohort study was deemed exempt from full Institutional Review Board review in 2023. All patients who underwent primary arthroscopic rotator cuff repair between 2012 and 2022 by the primary surgeon were identified by CPT code and reviewed for inclusion in this study. Exclusion criteria included: patients younger than 18 years old, those with irreparable cuff tears or partial cuff repairs, history of prior surgery on the operative shoulder, those requiring open rotator cuff repair, less than one year of clinical follow-up, and those with incomplete PROs pre-operatively and one year postoperatively. The primary surgeon made a change in practice in May 2017 from instructing patients to abstain from NSAIDs for 6 weeks following rotator cuff repair, to routinely prescribing Ibuprofen 800 mg TID for post-operative pain control. Patients were divided into two groups: those prescribed ibuprofen postoperatively (all rotator cuff repairs performed by primary surgeon after May 2017), and those instructed not to take NSAIDs for six weeks postoperatively (prior to May 2017).
2.2 Outcomes of interest
Patient demographics such as age, sex, body mass index (BMI), smoking status, and workers compensation status were collected from the electronic medical record along with pertinent medical and surgical history. Magnetic resonance imaging was blindly reviewed by a fellowship trained orthopedic surgeon to determine injury characteristics including level of retraction and Goutallier Classification. Number of tendons involved and amount of retraction was confirmed in the operative note as well as number of anchors used for repair, and concomitant procedures. Preoperative and postoperative range of motion (ROM) was obtained from clinic notes.
Primary study outcomes were PROs that were prospectively collected and stored in a secure electronic data capture system (REDCap) until retrospective review for the purposes of this study. The preoperative and one-year post-operative Western Ontario Rotator Cuff (WORC) index or American Shoulder and Elbow Surgeons (ASES) scores were used. In 2018, the primary surgeon changed from WORC to ASES for PROs data. Therefore, both were collected based on availability but WORC was primarily available for the no NSAID group and ASES for the NSAID group. These were compared based on an improvement of greater than the minimum clinically important difference (MCID) from the pre-operative to one-year post-operative PRO testing. Achievement of MCID was based on previously reported values.10 Secondary study outcomes evaluated were rate of post-operative complications occurring in the first year and repair failures requiring surgery.
2.3 Statistical analysis
Statistical analysis was performed using Intellectus (Clearwater, FL) statistical software. Descriptive statistics are reported as mean ± standard deviation for scale variables and n (%) for categorical variables. Statistical tests included Fischer Exact tests, t-tests, Mann-Whitney U tests, One-Way Analysis of Variance (ANOVA) and Pearson correlations as appropriate. Statistical significance was set at an alpha value of p ≤ 0.05.
3 Results
3.1 Overall cohort analysis
In total, 272 patients underwent rotator cuff repair by the primary surgeon in the study timeframe and were chart reviewed for inclusion in the study. Of these, 147 were excluded based on the afore mentioned criteria with the most common reason for exclusion being lack of one year follow up. 125 patients met inclusion criteria, 36 patients in the no NSAID group and 89 in the NSAID group.
3.2 Subgroup analyses
The two groups showed no significant difference in average age, sex, BMI, smoking status, diabetes, nor workers compensation status (Table 1). There was also no significant difference found in average number of tendons involved, tendon retraction, or Goutallier classification on MRI between the two groups (Table 2). Comparing intra-operative characteristics, the NSAID group had significantly fewer average number of tendons involved (p = 0.022) and number of anchors required for repair (p = 0.039). The groups had no significant difference in amount of tendon retraction noted intra-operatively (Table 3). Pre-operatively, the NSAID group had significantly greater passive and active shoulder abduction than the non-NSAID group (163.3 vs. 157.4, p = 0.040, 163.6 vs 157.4, p = 0.033) (Table 4). At one-year post-op, the no NSAID group had greater passive and active shoulder external rotation (59.1 vs 50.5, p = 0.036, 57.1 vs 48.5, p = 0.037) (Table 5). The remainder of pre-operative and post-operative ROM results were not significantly different between the groups.
| No NSAID | NSAID | P-valuea | |
| Age (Years) | 58.1 | 60.0 | 0.630 |
| Sex (Female) | 21 (23.9 %) | 14 (37.8 %) | 0.092 |
| BMI | 28.7 | 29.2 | 0.984 |
| Currently Smoking | 3 (3.4 %) | 1 (2.7 %) | 1.000 |
| Workers Compensation | 7 (8.0 %) | 0 (0.0 %) | 0.102 |
| Diabetes | 8 (9.1 %) | 6 (16.2 %) | 0.349 |
| No NSAID | NSAID | P-valuea | |
| Supraspinatus Tendon retraction (cm) | 0.840 | ||
| <1 cm | 28 (31.8 %) | 12 (32.4 %) | |
| 1–2 cm | 16 (18.2 %) | 6 (16.2 %) | |
| >2 cm | 40 (45.5 %) | 13 (35.1 %) | |
| Goutallier Classification | 0.310 | ||
| 0 | 58 (65.9 %) | 19 (51.4 %) | |
| 1 | 18 (20.5 %) | 12 (32.4 %) | |
| 2 | 4 (4.6 %) | 3 (8.1 %) | |
| 3 | 3 (3.4 %) | 0 (0.0 %) | |
| 4 | 1 (1.1 % | 0 (0.0 %) |
| No NSAID | NSAID | P-valuea | |
| # of Tendons involved | 0.022 | ||
| 1 | 21 (23.9 %) | 18 (48.7 %) | |
| 2 | 35 (39.8 %) | 11 (29.7 %) | |
| 3 | 29 (33.0 %) | 6 (16.2 %) | |
| 4 | 1 (1.1 %) | 1 (2.7 %) | |
| Supraspinatus Tendon retraction (cm) | 0.437 | ||
| <1 cm | 22 (25.0 %) | 3 (8.1 %) | |
| 1–2 cm | 11 (12.5 %) | 2 (5.4 %) | |
| >2 cm | 44 (50.0 %) | 14 (37.8 %) | |
| # of anchors required for repair | 0.039 | ||
| 1 | 13 (14.8 %) | 10 (27.0 %) | |
| 2 | 14 (15.9 %) | 11 (29.7 %) | |
| 3 | 22 (25.0 %) | 7 (18.9 %) | |
| 4 | 18 (20.5 %) | 6 (16.2 %) | |
| 5 | 17 (19.3 %) | 1 (2.7 %) | |
| >6 | 2 (2.3 %) | 0 (0.0 %) | |
| Concomitant Procedures | |||
| GH Debridement | 56 (63.6 %) | 24 (64.9 %) | 0.896 |
| LHB Tenotomy | 18 (20.5 %) | 3 (8.1 %) | 0.092 |
| LHB Tenodesis | 49 (55.7 %) | 24 (64.9 %) | 0.342 |
| DCE | 16 (18.2 %) | 9 (24.3 %) | 0.433 |
| SAD | 67 (76.1 %) | 27 (73.0 %) | 0.709 |
| No NSAID | NSAID | P-valuea | |
| Shoulder flexion (°) | 158.7 ± 27.6 | 160.9 ± 18.2 | 0.700 |
| Shoulder abduction (°) | 145.6 ± 36.6 | 148.3 ± 35.5 | 0.861 |
| Shoulder external rotation (°) | 59.1 ± 19.1 | 50.5 ± 17.8 | 0.036 |
| No NSAID | NSAID | P-valuea | |
| Shoulder flexion (°) | 162.8 ± 16.9 | 163.7 ± 21.7 | 0.497 |
| Shoulder abduction (°) | 157.4 ± 22.1 | 163.6 ± 29.5 | 0.033 |
| Shoulder external rotation (°) | 53.6 ± 17.2 | 58.5 ± 16.9 | 0.143 |
The average ASES score in the NSAID group improved from an average pre-operative score of 34.3–84.6 at one-year post-operative follow-up, with 83.8 % of patients improving by at least the MCID. Utilizing the WORC score, the patients in the no NSAID group improved from a pre-operative average of 37.0–82.8 at one year follow-up, with 73.9 % of patients improving by at least the MCID. The percent of patients improving by the MCID at one year did not show significant difference between the groups (p = 0.471)
Five post-operative complications were documented in the no NSAID group, including three pulmonary emboli (PE) requiring anti-coagulation, one transient ulnar nerve palsy, and one shoulder requiring manipulation to improve range of motion. Two post-operative complications were reported in the NSAID group; one superficial femoral vein DVT requiring anti-coagulation and one shoulder requiring manipulation to improve range of motion. The rate of post-operative complications was not significantly different between the no NSAID and NSAID groups (5.7 % vs 5.4 % respectively, p = 0.827).
In the no NSAID group, two patients had failure of rotator cuff healing that required revision rotator cuff repair in the first year. One patient in the NSAID group had a failed rotator cuff repair and required revision rotator cuff repair in the first year. There was no significant difference in the percentage of post-operative rotator cuff repair failures requiring revision in the first year after surgery between the no NSAID and NSAID groups (2.3 % vs 2.7 %, p = 1.000).
4 Discussion
Concerns remain in the orthopedic community over the deleterious effects of NSAIDs on soft tissue healing in the acute post-operative period. The purpose of this study was to compare post-operative outcomes of rotator cuff repair in patients using a short course of ibuprofen post-operatively to those that avoided NSAID usage for 6 weeks after surgery. The literature contains several studies on NSAIDs effect on post-operative rotator cuff repair outcomes, but this is one of the few studies on ibuprofen specifically. It was hypothesized that in this single center retrospective cohort study, a short course of ibuprofen in the immediate post-operative period would not lead to decreased PROs, nor greater post-operative complication rates or re-operation rates at one year.
This study demonstrated no statistically significant difference in percent of patients reaching greater than MCID improvement in PROs from pre-operatively to one year follow up post-operatively between the ibuprofen group and the control group. At one year post-operatively there were also not statistically significant differences between groups in post-operative complication rates nor rate of failure of rotator cuff healing requiring re-operation. These results are in line with Tangtiphaiboontana et al.’s randomized study of rotator cuff repairs as stated previously.7 Similarly, the meta-analysis by Lin et al. found no difference in retear rates between patients using NSAIDs for post-operative pain control and those who did not.11 In contrast, Oh et al.’s double blinded randomized study reported increased retear rates in patients using Celebrex for post-operative pain control, but not ibuprofen or tramadol.9 The result of this study further supports no difference in healing rates or clinical outcomes after a short course of ibuprofen in rotator cuff repairs. It further contributes to the current literature by providing a larger study population than previously reported randomized studies on NSAIDS effect on post-operative rotator cuff repair outcomes.
Concern over the deleterious effects of NSAIDs on surgical repairs are present in all areas of orthopedics including spinal fusion, open reduction internal fixation of long bones, and many other areas. Current studies in these areas have largely proven the efficacy and safety of a short course of NSAIDs in the acute post-operative period. A meta-analysis from 2021 demonstrated that a short course of NSAIDs (less than two weeks) did not increase the risk of fracture non-union.12 Another review found that a short course of NSAIDs (less than two weeks) did not decrease the rates of spinal fusion.13 The current body of literature supports a short course of NSAIDs in the acute post-operative period after rotator cuff repair. As such, in solidarity with other subspecialties it is recommended for shoulder surgeons to consider the use of NSAIDs for post-operative pain control.
The strengths of this study are that it is a single center study with one primary surgeon, thus limiting confounding operative techniques. Additionally, the change from instructing patients to avoid NSAIDs after surgery to prescribing NSAIDs was a hard change and was made at a specific time point, which limits selection bias. Additionally, Ibuprofen was the NSAID used in all patients. Thus, standardizing the medication used and eliminating the confounding factors of various NSAID medications. Multiple limitations are also present. This is a retrospective study without true randomization. Additionally, a high rate of patient exclusion was noted. The greatest limitation is that the study had no way of confirming whether patients in each group did or did not consume NSAIDs, nor the frequency and quantity they consumed. The groups were defined based on NSAID prescription, not patient reported consumption of the medication.
5 Conclusion
There was no difference in reaching MCID between the group that used ibuprofen after rotator cuff repair and the group that did not. There was also no difference in post-operative complications nor re-operations at one year due to failure of rotator cuff healing between the groups. This supports other current literature in that the use of NSAIDs, specifically ibuprofen, after rotator cuff repair does not increase reoperation rates nor decrease PROs at one year. Further studies should be completed in a prospective randomized fashion with regular blood draws to confirm patient compliance in the use or disuse of NSAIDs post-operatively. Studies are also needed that include multiple cohorts utilizing different NSAIDs to determine if various NSAIDs have different effects.
Disclaimer
BPC's spouse is the founder and CEO of CODE Technology. JPB is a paid consultant for Zimmer Biomet. No other authors have any financial disclosures nor conflicts of interest to report.
IRB
This retrospective cohort study was deemed exempt from full Institutional Review Board review by the Health Partners IRB in February 2023.
Author contribution
Adam Hadro: contributed to formal analysis, data curation, writing of the original draft and reviewing/editing and visualization. Jacob Ristow: contributed to conceptualization, methodology, validation, formal analysis, data curation, writing of the original draft and reviewing/editing and visualization. Brian Cunningham: contributed to data curation, writing of the original draft and reviewing/editing and visualization, supervision and project administration. Rebekah Kleinsmith: contributed to data curation, writing of the original draft and reviewing/editing and visualization. Stephen Doxey: contributed to data curation, writing of the original draft and reviewing/editing and visualization. Adam Schweitzer: contributed to conceptualization, methodology, validation, formal analysis, data curation, writing of the original draft and reviewing/editing. Jonathan Braman: contributed to conceptualization, methodology, validation, formal analysis, data curation, writing of the original draft and reviewing/editing, supervision and project administration. All authors including on this manuscript have mad significant contributions to the manuscript.
Ethical statement
This work performed in the construction of this study and manuscript was prepared in accordance with The Declaration of Helenski.
Conflicts of interest and source of funding
The BPC's spouse is the CEO and founder of a patient-reported outcome measure company. For the remaining authors, none were declared. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
References
- Opioid requirement after rotator cuff repair is low with a multimodal approach to pain. J Shoulder Elbow Surg. 2021;30(7):e399-e408.
- [Google Scholar]
- The effect of diclofenac on matrix metalloproteinase levels in the rotator cuff. Arch Orthop Trauma Surg. 2014;134:1739-1744.
- [Google Scholar]
- Timing matters: NSAIDs interfere with the late proliferation stage of a repaired rotator cuff tendon healing in rats. Arch Orthop Trauma Surg. 2014;134:515-520.
- [Google Scholar]
- Indomethacin and celecoxib impair rotator cuff tendon-to-bone healing. Am J Sports Med. 2006;34(3):362-369.
- [Google Scholar]
- Do different cyclooxygenase inhibitors impair rotator cuff healing in a rabbit model? Chinese Med J. 2015;128(17):2354-2359.
- [Google Scholar]
- Nonsteroidal anti-inflammatory drugs (NSAIDs) and their effect on musculoskeletal soft-tissue healing: a scoping review. JBJS reviews. 2019;7(12):e4.
- [Google Scholar]
- The effects of nonsteroidal anti-inflammatory medications after rotator cuff surgery: a randomized, double-blind, placebo-controlled trial. J Shoulder Elbow Surg. 2021;30(9):1990-1997.
- [Google Scholar]
- Non-selective NSAIDs do not increase retear rates post-arthroscopic rotator cuff repair: a meta-analysis. J ISAKOS. 2023;8(4):216-226.
- [Google Scholar]
- Do selective COX-2 inhibitors affect pain control and healing after arthroscopic rotator cuff repair? A preliminary study. Am J Sports Med. 2018;46(3):679-686.
- [Google Scholar]
- Minimal clinically important difference (MCID) for patient-reported shoulder outcomes. J Shoulder Elbow Surg. 2020;29(7):1484-1492.
- [Google Scholar]
- Post-operative non-steroidal anti-inflammatory drugs do not affect clinical outcomes of rotator cuff repair. J Orthop. 2020;17:113-115.
- [Google Scholar]
- The effect of NSAIDs on postfracture bone healing: a meta-analysis of randomized controlled trials. OTA International. 2021;4(2)
- [Google Scholar]
- The effect of NSAIDs on spinal fusion: a cross-disciplinary review of biochemical, animal, and human studies. Eur Spine J. 2017;26:2719-2728.
- [Google Scholar]