Abstract
Background:
Rotator cuff disease is one of the most common causes of shoulder disability. Pathologic conditions of the shoulder are often complicated by concomitant conditions such as cervical spine (C-spine) disease.
Purpose/Hypothesis:
The purpose of this article is to assess whether patients with concomitant C-spine disease undergoing arthroscopic rotator cuff repair (ARCR) demonstrate lower absolute values in Patient-Reported Outcomes Measurement Information System (PROMIS) scores compared with those without C-spine disease and whether it affects the achievement of the minimal clinically important difference (MCID). It was hypothesized that patients with C-spine disease undergoing ARCR demonstrate lower absolute values in PROMIS scores compared with those without concomitant C-spine disease, but those patients experience a similar improvement and achievement of MCID after ARCR surgery.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
A total of 1387 patients who underwent ARCR at a single institution were identified. A retrospective chart review of a prospectively collected database was conducted to capture PROMIS Pain Interference (PI), Physical Function (PF), and Depression (Dep) scores. Patients were stratified into 2 groups based on the presence of concomitant C-spine disease at the time of surgery. The MCID was calculated using a distribution-based methodology.
Results:
Of the total, 528 patients met inclusion criteria for final analysis, 120 of whom had concomitant C-spine disease. Unadjusted analysis of descriptive variables revealed no significant differences between the C-spine and control groups. Both cohorts showed significant improvement at final follow-up compared with their preoperative PROMIS values for PF, PI, and Dep (P < .001). Compared with patients without it, those with concomitant C-spine disease had statistically worse preoperative PROMIS scores for PI (62.6 ± 6.8 vs 60.1 ± 6.8; P = .001), Dep (50.4 ± 9.8 vs 47.6 ± 9.8; P = .004), and PF scores (39.5 ± 7.3 vs 42.6 ± 8.4; P = .001), as well as worse postoperative scores for PI (53.9 ± 7.9 vs 50.5 ± 8.1; P = .001), Dep (43.4 ± 9.4 vs 41.1 ± 7.8; P = .003), and PF (45.6 ± 7.6 vs 48.6 ± 7.9; P = .001). Notably, these differences while statistically significant were not clinically significant due to the overlap between the C-spine and control group PROMIS scores in each domain. No significant difference was found in the achievement of MCID between the 2 groups for PF, PI, or Dep. Regression analysis revealed that C-spine disease was not a significant predictor of achievement of MCID in any of the 3 PROMIS domains.
Conclusion:
This study underscores the broad applicability of ARCR for patients with complex musculoskeletal pathology such as concomitant C-spine disease. The findings highlight the importance of setting expectations for patients regardless of their concomitant C-spine pathology before surgery.
Chronic shoulder pain affects up to 8% of adults in the United States, with rotator cuff disease being the most common cause of disability in the shoulder. Rotator cuff injuries and methods of repair have been studied extensively. The increasing incidence of patients being evaluated for chronic shoulder pain treatment has prompted increased interest in optimizing diagnosis, treatment, and cost-effective care for this aging population.1,2,4,5,8,12,13,15
Cervical spine (C-spine) conditions are difficult to distinguish from shoulder conditions since both can cause overlapping areas of pain in the neck and the shoulder (Figure 1). Cervical myelopathy and radiculopathy are common pathologies in older adults being evaluated for pain, weakness, or numbness of the upper extremity. While some patients may be diagnosed with either shoulder or C-spine disease, other patients have concomitant involvement of both the neck and the shoulder creating an even more complex clinical picture.12,13,15 One study by Dunn et al 3 found that the coincidence of C-spinal stenosis and rotator cuff tears increased from 9% to 13% from 2005 to 2014. This overlap creates challenges in diagnosis and management, including the decision to operate on either the C-spine or the shoulder. 6 Studies have been conducted to assess the effect of specific cervical pathologies such as cervical spinal stenosis 12 or cervical radiculopathy 7 on outcomes after open or arthroscopic rotator cuff repair (ARCR). 13 These studies showed comparable functional outcomes in patients with radiculopathy 7 but increased risk of reoperation on the ipsilateral side in patients with cervical degenerative disease. 16
Cervical (C) dermatomes and brachial plexus pathology. Anterior and posterior views of the shoulder with C4-C7 dermatomes 14 illustrated in corresponding colors and rotator cuff pathology as documented in previous literature 10 overlayed in gray stripes. This image highlights the overlapping pain distributions of cervical spine pathology and rotator cuff injury.
Patient-reported outcomes (PROs), especially using the Patient-Reported Outcomes Measurement Information System (PROMIS), is a validated and relatively time-efficient metric in evaluating health outcomes. PROMIS is a useful tool to assess the overall patient experience in the presence of comorbid conditions, such as C-spine and rotator cuff disease, as historical legacy measures for each condition alone would not fully capture the patient's status.
The purpose of this study is to assess whether patients with concomitant C-spine disease undergoing ARCR demonstrate lower absolute values in PROMIS scores compared with those without C-spine disease and whether it affects the achievement of a minimal clinically important difference (MCID). C-spine disease is defined as patients with symptomatic degenerative disease, radiculopathy, or myelopathy of the C-spine. The hypothesis is that patients with C-spine disease undergoing ARCR demonstrate lower absolute values in PROMIS scores compared with those without concomitant C-spine disease, but those patients experience a similar improvement and achievement of MCID after ARCR surgery.
Methods
After obtaining institutional review board approval, a retrospective observational study of a prospectively collected outcome database was conducted, encompassing patients who underwent elective single- or double-row ARCR between January 1, 2015, and December 31, 2018. This cohort was utilized due to the established data availability and the consistent group of surgeons performing the procedures at the institution during this period. The surgical procedures were performed by 1 of 6 fellowship-trained, board-certified orthopaedic surgeons at an academic institution. In total, 1387 patients were identified, of which 750 were excluded as they did not have a preoperative PROMIS score ≤6 months of their surgical period, and 109 were excluded for not having >6-month postoperative PROMIS scores. The patient cohort was stratified into 2 groups based on the presence of a concomitant symptomatic cervical condition at the time of ARCR versus those without existing C spine–related disease. C spine–related conditions that were included in the study were degenerative disc disease (n = 36), radiculopathy (n = 35), myelopathy (n = 9), herniated disc (n = 2), facet arthritis (n = 2), spondylosis (n = 14), and spinal stenosis (n = 22).
PROMIS Pain Interference (PI), Depression (Dep), and Physical Function (PF) questionnaires were administered on Apple iPads (Apple) as part of routine clinical care at outpatient orthopaedic clinics, through the electronic medical record, physical therapy sessions, and surgery centers. Results were promptly accessible in the patient's electronic medical record and stored in a prospectively maintained research database. Computer adaptive testing technology based upon item response theory was employed to tailor each questionnaire, considering each patient's previous responses for efficient and consistent results. The PROMIS domains followed a normal distribution, with a mean T-score of 50 and an SD of 10 for the general population. For the respective PROMIS domains, higher PI and Dep scores indicate worse symptomatology, while lower scores in PF represent poorer function.
A retrospective chart review captured demographic and surgical data, including patient age at the time of surgery, body mass index (BMI kg/m2), sex, insurance status, race, smoking history, and comorbidities assessed through the Charlson Comorbidity Index. Additionally, the number of anchors used intraoperatively and rotator cuff tear size (small to medium, large, massive) were also collected. Patients with an active diagnosis related to the C-spine (by, eg, cervical myelopathy or cervical radiculopathy) were identified through International Classification of Diseases, Version 10, codes. Symptomatic status of the condition was confirmed using the search terms “cervicalgia” and “neck pain” in each patient chart. Pearson chi-square analysis or Fisher exact test assessed descriptive features and MCID achievement among the 2 cohorts. Bivariate Student t test analysis evaluated descriptive differences, such as age, BMI, and PROMIS outcomes. Multivariate logistic regression was applied to determine the role of concomitant C-spine conditions as a predictive feature for MCID achievement. Statistical significance was set at a P value <.05.
Results
In total, 528 patients were included in the final analysis. The mean time to final follow-up was 2.42 ± 0.43 years. Of the 528 patients, 120 were identified as having symptomatic C-spine disease. The remaining 408 patients were stratified into a separate group to serve as a control. Overall, it was found that there was no significant difference in age, sex, BMI, race, insurance, smoking history, mean number of anchors, and rotator cuff tear size between patients with concomitant cervical conditions and those without any cervical condition who underwent ARCR (Table 1).
Covariate Unadjusted Analysis of Demographic Variables a
Data are presented as mean ± SD or n (%). CCI, Charlson Comorbidity Index; C-spine, cervical spine disease.
Both cohorts demonstrated significant improvement at the final follow-up compared with their preoperative PROMIS values for PF, PI, and Dep (Figure 2). Nevertheless, patients with concomitant cervical diagnoses had significantly worse preoperative PI (62.6 ± 6.8 vs 60.1 ± 6.6; P = .001), Dep (50.4 ± 9.8 vs 47.6 ± 9.8; P = .004) and PF scores (39.5 ± 7.3 vs 42.6 ± 8.4; P = .001) when compared with the control group (Table 2). Moreover, there remained a significant difference in the improvement at the final follow-up for PI (53.9 ± 7.9 vs 50.5 ± 8.1; P = .001), Dep (43.3 ± 9.4 vs 41.1 ± 7.8; P = .003), and PF (45.6 ± 7.6 vs 48.6 ± 7.9; P = .001). Nevertheless, it was observed that there was no significant difference between the 2 groups for change in PROMIS score from preoperative to final follow-up for PROMIS PI (P = .31), Dep (P = .55), or PF (P = .82). Importantly, all comparisons between the C-spine and control groups had overlap in the ranges of PROMIS score across all 3 domains. These results, while statistically significant, are not clinically significant due to the overlapping ranges between groups.
(A) Unadjusted mean PROMIS PF scores comparing pre- with postoperative scores for patients in the C-spine and no C-spine groups and between C-spine and no C-spine patients for pre- and postoperative scores showing significantly higher (improved) scores for patients in the no C-spine group at pre- and postoperative time points and higher (better) scores postoperatively for both groups. (B) Unadjusted mean PROMIS PI scores showing significantly lower (improved) postoperative scores compared with preoperative scores in each group, and significantly higher (worse) scores at both time points for patients in the C-spine cohort. (C) Unadjusted mean PROMIS Dep scores showing significantly lower (better) scores for patients in the no C-spine group both pre- and postoperatively compared with the C-spine group, and improved scores postoperatively for both cohorts. C-spine, cervical spine; Dep, Depression; PF, Physical Function; PI, Pain Interference; Post-op, postoperative; Pre-op, preoperative; PROMIS, Patient-Reported Outcomes Measurement Information System.
Unadjusted PROMIS Outcomes a
Unadjusted mean PROMIS PF, PI, and Dep scores comparing the C-spine cohort to the no C-spine cohort for each domain at both pre- and postoperative time points showing significantly higher (better) scores for the no C-spine cohort in the PF domain, and lower (better) scores in the PI and Dep domains at all time points. C-spine, cervical spine; Dep, Depression; PF, Physical Function; PI, Pain Interference; Post-op, postoperative; Pre-op, preoperative; PROMIS, Patient-Reported Outcomes Measurement Information System.
The differences in PROMIS score required to achieve MCID were found to be −3.34 for PI, 4.14 for PF, and −4.95 for Dep. In total, 78.7% of patients achieved MCID for PI, 51.5% for Dep, and 55.4% for PF. It was observed that no difference was found in the achievement of MCID between patients who had concomitant C-spine conditions compared with those with no C-spine conditions that underwent ARCR for PI (80.5% vs 78.3%; P = .77), Dep (57.7% vs 50.3%; P = .27), and PF (51.2% vs 56.3%; P = .47). Logistic regression analysis found that concomitant cervical condition was a nonsignificant predictor for the achievement of MCID PI (P = .38), Dep (P = .92), or PF (P = .09) (Table 3). Moreover, patients with worse preoperative PI (odds ratio [OR], 1.17; 95% CI, 1.11-1.22; P = .001), Dep (OR, 1.13; 95% CI, 1.10-1.16; P = .001), and PF (OR, 0.95; 95% CI, 0.93-0.9; P = .04) scores had a significantly greater odds of achieving MCID. Finally, patients who had higher BMI (OR, 0.95; 95% CI, 0.93-0.98; P = .04) and those who had government insurance (OR, 0.56; 95% CI, 0.34-0.78; P = .02) had a decrease in odds of achieving MCID PF.
Multivariate Logistic Regression MCID a
BMI, body mass index; CCI, Charlson Comorbidity Index; C-spine, cervical spine; MCID, minimal clinically important difference; OR, odds ratio; PROMIS, PROMIS, Patient-Reported Outcomes Measurement Information System.
Discussion
This study revealed that despite both groups showing significant improvement at their final follow-up compared with preoperative scores, patients with concomitant cervical conditions had significantly worse PROMIS PI, Dep, and PF at their preoperative visit and final follow-up than patients with no history of previous C-spine conditions. While these differences were statistically significant, there were overlapping ranges in all PROMIS domains between the control cohort and the cohort with C-spine disease. Additionally, there was no significant difference in the degree of improvement (change in PROMIS scores from preoperative to final follow-up) or achievement of MCID between the 2 groups for all 3 PROMIS domains.
A shift from traditional fee-for-service models to value-based care signifies a fundamental change, prioritizing high-quality, cost-effective health care over sheer service volume. This transformative landscape poses a challenge for health care providers, requiring them to optimize patient outcomes within the complexities of the health care system. Given the prevalent and often ambiguous nature of pain related to the neck/spine versus the shoulder in C-spine conditions, it is crucial to examine the effect of cervical spinal issues on patients with shoulder pain. Despite inferior scores in patients with concomitant C-spine pathology, the small differences in these outcomes are unlikely to represent clinically relevant differences for patients undergoing ARCR.
The complex nature of C-spine conditions adds an extra layer of difficulty in addressing rotator cuff disease, particularly when pain presents ambiguously and overlaps between the neck, spine, and shoulder. Understanding the interplay between these conditions is crucial for health care providers aiming to enhance the effectiveness of ARCR procedures. Comparing patients with cervical radiculopathy to those without C-spine conditions, Moorthy et al 7 did not find differences in functional outcomes 1 year after ARCR. These results are supported by the findings in this study demonstrating clinically insignificant differences in both preoperative and postoperative PROs for patients with cervical pathology undergoing ARCR compared with those without coincident cervical pathology. The authors also found that patients achieved comparable MCID in both of their patient groups, which is consistent with the results in the present study. 7
Previous studies evaluating the effect of C-spine conditions on rotator cuff outcomes relied on a shoulder-specific legacy metrics, potentially limiting accuracy and sensitivity in measuring improvement and offering an incomplete perspective. This study uses PROs to measure outcomes after ARCR, which have been validated compared with shoulder-specific legacy metrics such as the American Shoulder and Elbow Surgeons (ASES) shoulder assessment form. A study by Nicholson et al 9 examining patients specifically with rotator cuff disease found an excellent-good correlation score (r = 0.62; P < .0001) between ASES and physical PROMIS scores. A systematic review evaluating operative shoulder outcomes compared PROMIS with legacy metrics and found the closest correlation between PROMIS upper extremity scores and ASES (r = 0.70). 11
Limitations
This study is not without limitations. The retrospective nature of this research prevented data from being collected from all patients who received ARCR surgeries and met inclusion criteria. The data were also collected in 2018 and not updated to preserve consistency among the surgeon cohort included in the study; however, the lapsed time since data collection may decrease the relevance of these findings. This study was limited to including only patients who completed PROMIS surveys and was unable to include those with missing data; thus, transfer bias was present in the cohorts with significant losses to follow-up, which may limit the validity of these findings. Another limitation is the potential for undiagnosed C-spine disease as well as the notable variability in symptomatology among patients with the C-spine diseases included in this study. These limitations, while inclusive of varying C-spine pathology, diminish the utility of applying the findings to patients with specific pathologies, as there may be differential findings based on each disease process. Patients are diagnosed through routine testing at clinic visits with sports medicine physicians and may not be diagnosed for a range of reasons including limited access to care, cost-prohibitive diagnostic testing, and subclinical levels of pain or dysfunction. However, the center at which this study took place, as part of a large health care system with a diverse patient population, enabled the identification and monitoring of a range of concomitant conditions, as patients typically remain within the system for their health care needs. Despite this advantage, the ability to detect all potential concomitant conditions may still be limited in this study. Finally, a limitation of PROMIS is that while it provides a more holistic view of the patient, it cannot distinguish between overlapping conditions. It captures only a snapshot of the patient's health on a given day, which may not offer the granularity needed to reflect the progress of individual diseases and could affect result interpretation.
Conclusion
Patients undergoing ARCR with concomitant C-spine conditions exhibited worse preoperative PI, Dep, and PF scores compared with their counterparts without C spine–related conditions. Despite these initial differences, both groups demonstrated significant improvements at the final follow-up, without notable disparities in the degree of improvement or achievement of the MCID in PROMIS scores. Along with the primary outcomes, the findings in this study demonstrate the practical challenges associated with incorporating specific legacy metrics emphasizes the efficiency of PRO's like PROMIS in enabling real-time tracking and intervention. This study underscores the importance of managing shoulder conditions despite the presence of concomitant C-spine disease. Despite possible differences in PROMIS scores, these findings indicate that patients should be counseled that they are expected to clinically improve equally to those without C-spine pathology.
Footnotes
Final revision submitted May 15, 2025; accepted June 16, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: B.D.G. is a paid consultant for and receives royalties from Arthrex. I.V. is a paid consultant for Arthrex and FH Orthopedics Inc; has received nonconsulting fees from Arthrex, Zimmer Biomet, Arthrosurface Inc; and has received royalties from Smith and Nephew. M.D.M. is a paid consultant for Arthrex. S.M. is a paid consultant for Arthrex. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was waived by University of Rochester (STUDY00005121).
