Study finds increased risk of additional surgeries after rotator cuff repair

Corticosteroid injections (CSI) are frequently used in the nonsurgical management of rotator cuff tears. According to the authors of Scientific Paper 838 displayed at the AAOS 2018 Annual Meeting, however, individuals who undergo rotator cuff repair (RCR) surgery within a year of receiving a CSI are at much higher risk of needing a revision RCR than those who had not been treated with CSI.

The risks generally associated with CSI include tendon rupture, articular cartilage degeneration, and septic arthritis. Evidence supports a temporal relationship between CSI and infection following shoulder arthroscopy, shoulder arthroplasty, and total knee arthroplasty. The authors hypothesized that the rate of a revision RCR would be higher for patients who had a preoperative CSI within three months of undergoing RCR and that this correlation would weaken as the length of time between injection and surgical intervention increased.

A cohort study

To identify patients for the study, the researchers conducted a retrospective analysis of claims data from an administrative claims database for the years 2011 to 2014. They identified patients between 18 and 64 years old with a diagnosis of rotator cuff tear (International Statistical Classification of Diseases and Related Health Problems, 9th revision, Clinical Modification [ICD-9-CM] 726.10, 726.13, 727.61, 840.4) and RCR (ICD-9-CM 83.63 or CPT 29827) in any inpatient or outpatient service claim between Jan. 1, 2011, and Dec. 31, 2011, who were continually insured for 12 months pre- and post-RCR. The first claim date for RCR served as the index date and patients were followed for three years or until the end of their continuous insurance enrollment.

Only patients with verified laterality were included (n = 4,959). More than half (53.6 percent) were male; the mean age of patients was 49.2 years. Treatment records for the 12 months prior to surgery were examined to identify injections associated with an ICD-9 shoulder code. Only those injections with confirmed laterality were included. Based on these data, researchers identified two patient cohorts—patients with no history of injection and those with a history of one or more injections.

They used multivariable logistic regression models to compare the odds of reoperation between groups and covariate adjustment to control for differences in failure rates that might be attributable to other factors. The primary outcome was revision RCR within three years of the index date.

Identifying the impact of injections

Of the 4,959 patients who met the inclusion and exclusion criteria, 553 required reoperation within the following three years. The researchers noted that baseline characteristics for the two groups differed in several ways. The group that did not require reoperation was generally younger (48.8 versus 52.7 years, P < 0.0001) and nonsmoking (3.6 percent versus 7.4 percent, P < 0.0001). “Older patients, males, diabetics, and smokers were more likely to undergo reoperation following primary RCR, which is consistent with previously published literature,” noted the researchers. All potential confounders were controlled for in the statistical analysis. Their results showed that patients who had an injection within the year prior to surgery had 43.5 percent higher odds (P = 0.0002) of reoperation when compared to individuals who did not have an injection in the year prior to RCR.

A subgroup analysis revealed that patients who had an injection within six months preceding the RCR were at a much higher risk of requiring a revision RCR, with an adjusted odds ratio of 1.528 (95 percent confidence index [CI]: 1.202–1.942, P = 0.0005). Patients who had their injection more than six months prior to the index RCR did not have an elevated risk of revision RCR when compared to the cohort that did not have any injections.

“The risk of revision RCR significantly declines if there is more than six months between injection and RCR, becoming comparable to the group which did not have an injection,” noted the researchers. “As the number of RCRs continues to rise, it is imperative that orthopaedic surgeons critically evaluate the work-up and management of rotator cuff tears as well as the role of injections in this setting.”

Although previous animal models revealed what appeared to be a “dose-dependent effect of preoperative CSI on the quality of the cuff tissue,” this has yet to be shown in humans. Determining how many injections to provide and when to offer them remains an issue, which this study helps to address.

Noting the limitations of the study, the researchers pointed to potential errors in reporting and inability to capture all patients and selection bias inherent to database utilization. The database did not include information on presenting symptoms, chronicity, or the size of the tear. As a result, they acknowledged that “we are unable to comment on the type of injection, number of injections, or location (subacromial, acromioclavicular, glenohumeral, or biceps tendon sheath).”

Nonetheless, they concluded, “As more patients elect to undergo RCR, orthopaedic surgeons must critically evaluate use of preoperative shoulder injections. Undergoing an RCR within six months of a shoulder injection is associated with an increased likelihood of subsequent revision RCR. Therefore, it may be reasonable to consider minimizing preoperative injections or delaying surgery following these injections in patients with rotator cuff tears.”

The authors of Scientific Paper 838, “Shoulder Injection Prior to Rotator Cuff Repair Is Associated with Increased Risk of Subsequent Surgery” are Sophia A. Traven, MD; Daniel Brinton, MHA, MAR; Kit N. Simpson, DrPH; Zachary Adkins, MD; Alyssa Althoff, BS; John Palsis, MD; William Ashford, MD; and Harris S. Slone, MD.

The study was partially supported by the South Carolina Clinical & Translational Research Institute, through an NIH-NCATS Grant.