Impact of osteoporosis on post-operative outcomes following rotator cuff repair

Abstract

Background

Osteoporosis, a systemic skeletal disorder marked by reduced bone mineral density and increased fracture risk, is especially prevalent in postmenopausal women and older adults, populations that commonly undergo rotator cuff repair (RCR) surgery. However, large-scale clinical evidence evaluating the association between osteoporosis and post-operative outcomes following RCR remains limited. This study evaluated post-operative complication rates and health care utilization following RCR in patients with and without osteoporosis.

Methods

Using the PearlDiver Mariner170 database, patients with and without a diagnosis of osteoporosis who also underwent RCR surgery were propensity matched 1:1. Primary study outcomes included 2-year RCR revision surgery, conversion to reverse total shoulder arthroplasty (rTSA), and pathological fractures of the humerus. Secondary outcomes included 90-day wound complications, shoulder stiffness, thromboembolic events, infections, and health care utilization. Analyses were conducted for the full RCR cohort as well as separately for open and arthroscopic RCR subgroups. Within the cohort of osteoporotic patients, outcomes were compared between patients who underwent open vs. arthroscopic RCR. Welch t-test and chi-square tests were used for continuous and categorical variables, respectively, with a Bonferroni-corrected significance ($\alpha$ = 0.002).

Results

After matching, 65,118 patients of similar age (66.4 years), gender distribution (87.1% female), and comorbidity burdens (Elixhauser Comorbidity Index: 5.3) were included in the osteoporosis and no-osteoporosis cohorts. At 2 years following RCR, osteoporotic patients had significantly higher rates of rTSA and pathological humeral fractures, but no difference in revision RCR. Within 90 days, osteoporotic patients experienced a higher incidence of general surgical complications, anemia, shoulder stiffness, and medical visits. Subgroup analyses between open and arthroscopic RCR demonstrated similar trends between the 2 procedure types. Among osteoporotic patients, those who underwent open RCR had higher rates of rTSA, revision RCR, and pathological humeral fractures than their arthroscopic RCR counterparts.

Conclusion

In the setting of RCR, patients with osteoporosis demonstrated higher rates of conversion to rTSA and pathological humeral fractures within 2 years of tendon repair compared to patients without a diagnosis of osteoporosis, regardless of surgical approach. In addition, patients with osteoporosis may experience a higher incidence of medical complications within the 90-day global period than nonosteoporotic patient. In the subgroup of patients with osteoporosis, open RCR was associated with higher 2-year rates of rTSA, revision RCR, and pathological fractures when compared to arthroscopic RCR. These findings suggest that bone health evaluation prior to RCR may be prudent to facilitate pre-operative optimization, surgical planning, and post-operative protocols to reduce complications and improve outcomes.

Osteoporosis is a systemic skeletal disorder characterized by reduced bone mineral density (BMD) that results in an increased susceptibility to fracture.² It affects over 10 million individuals in the United States and is more prevalent in postmenopausal women and older adults, populations that commonly present with degenerative shoulder conditions requiring surgical intervention.⁶

Rotator cuff tears are one of the most common causes of shoulder pain and dysfunction in this subset of the population and increase in prevalence with age.³⁶ With the activity levels in the adult population now remaining high later into life, the incidence of rotator cuff repair (RCR) has risen in recent decades. However, RCR outcomes vary widely and retear rates range from 5% to 34%, particularly in elderly patients and those with biologic or structural factors that impair healing.¹³ Suboptimal BMD has been associated with impaired tendon-to-bone healing and increased risk of repair failure following RCR, as successful tendon-to-bone healing relies on the integrity of the underlying bone for adequate tendon fixation and biologic incorporation.^10,11,14,29 In patients with osteoporosis, decreased BMD may increase the risk of suture anchor pullout, fixation failure, retear, and overall suboptimal functional recovery.¹¹

Despite the known relationship between osteoporosis and rotator cuff healing, the broader impact of osteoporosis on post-operative complications and health care utilization following RCR remains incompletely defined.⁵ Some studies have reported higher retear rates in osteoporotic patients, while others have found no significant differences in patient-reported outcomes such as pain, strength, or patient satisfaction, suggesting that functional recovery may be preserved even in the setting of compromised bone quality.^7,30 Furthermore, limited data exist regarding whether surgical approach modifies these associations, particularly in the context of open vs. arthroscopic techniques.

The comparison of open vs. arthroscopic RCR in osteoporotic patients is particularly relevant given the distinct surgical characteristics of each approach. Open RCR may involve detachment and subsequent repair of the deltoid muscle to access the rotator cuff, whereas arthroscopic repair uses minimally invasive portal-based techniques that preserve deltoid integrity and the surrounding soft tissue envelope.^24,28,36 In addition, open approaches may place greater mechanical stress on compromised bone and disrupt soft tissue, potentially increasing the risk of repair failure, complications, or subsequent procedures.¹⁰ While prior studies have shown similar long-term outcomes and retear rates between arthroscopic and open/mini-open RCR in general populations, the interaction between surgical approach and osteoporosis has not been well studied, representing a critical knowledge gap.^24,28,36

In this retrospective review of a large national insurance database, our objectives were twofold. We sought to determine if, at a population level, a diagnosis of osteoporosis is associated with post-operative outcomes and health care utilization following RCR. In addition, among patients with osteoporosis, we sought to determine if open versus arthroscopic RCR is associated with a difference in post-operative outcomes. We hypothesized that among patients undergoing RCR, those with a diagnosis of osteoporosis would demonstrate suboptimal post-operative outcomes and increased health care utilization compared to those without a diagnosis of osteoporosis. Among osteoporotic patients, we hypothesized that open RCR would be associated with worse post-operative outcomes compared to arthroscopic RCR due to the potential for increased mechanical stress on osteoporotic bone and greater soft tissue disruption with open approaches.

Material and methods

This study was a retrospective analysis using the PearlDiver Mariner170 database. This database contains medical and prescription claims for approximately 170 million patients between January 1, 2010, and April 30, 2023. The database aggregates data across all payer types and provider networks, incorporating both International Classification of Diseases codes (Ninth and Tenth Revisions [ICD-9 and ICD-10]) and Current Procedural Terminology codes to capture diagnoses, surgical procedures, reimbursements, and prescription records. Post-operative outcomes and health care utilization were assessed using PearlDiver's timeline filters to identify events occurring within the 90-day and 2-year windows following RCR. This study was conducted in accordance with the REporting of studies Conducted using Observational Routinely-collected health Data statement.⁴ Data within PearlDiver are deidentified and compliant with the Health Insurance Portability and Accountability Act; therefore, institutional review board approval was not required.

To build our study populations, we first identified patients diagnosed with osteoporosis through a predefined cohort available in PearlDiver. Osteoporosis diagnosis was defined based on the presence of ICD diagnosis codes within the administrative record. The database does not provide BMD measurements or diagnostic confirmation using dual-energy X-ray absorptiometry. Therefore, osteoporosis severity and diagnostic confirmation could not be assessed. Patients with a diagnosis of osteoporosis were then filtered to include those who underwent a RCR procedure after the diagnosis of osteoporosis, identified by Current Procedural Terminology codes 29827, 23410, 23412, and ICD-9 procedure code 8,363. To address potential confounding and selection bias inherent to administrative database studies, propensity score (1:1) matching was employed to create a control group of patients without osteoporosis who underwent RCR, accounting for age at the time of surgery, sex, and Elixhauser Comorbidity Index (ECI). Primary outcomes included revision RCR surgery, reverse total shoulder arthroplasty (rTSA), and pathological humeral fractures within 2 years of index RCR. Secondary outcomes included general surgical complications, Surgical Site Infections, wound complications, venous thromboembolism, pulmonary embolism, deep vein thrombosis, myocardial infarction, sepsis, methicillin-resistant Staphylococcus aureus infection, anemia, pneumonia, effusion, shoulder stiffness, arthrocentesis/steroid injection, medical visits, emergency room visits, and readmissions within 90 days of RCR. Primary and secondary outcomes were compared between patients with and without osteoporosis for the full RCR cohort. Primary outcomes were also compared between patients with and without osteoporosis within the open and arthroscopic RCR subgroups. A secondary analysis within the cohort of patients with osteoporosis was performed to compare primary outcomes between patients undergoing open vs. arthroscopic RCR. The specific ICD-9 and 10 codes used to identify complications and the comorbidities used in ECI calculations can be found in Supplementary File 1. General surgical complications were identified using a predefined coding bucket within PearlDiver designed to encompass a wide spectrum of ICD-9 and ICD-10 codes reflecting post-operative adverse events outside of orthopedic-specific outcomes. Conditions commonly included in this category are post-operative bleeding, anesthesia-related complications, and various other perioperative medical or surgical problems.

All statistical analyses were performed using tools built into the PearlDiver platform. Chi-square tests were used to compare categorical variables, including sex distribution and post-operative complication rates. Welch two-sample t-tests were employed for continuous variables, such as age and ECI. To account for multiple comparisons among our 23 outcomes of interest, we applied the Bonferroni correction. By dividing the overall significance level ($\alpha$ = 0.05) by the number of comparisons, we set a corrected alpha level of 0.002. Thus, only P values $\le$.002 were deemed statistically significant.

Results

A total of 65,121 patients with a diagnosis of osteoporosis and 1,026,847 patients without a diagnosis of osteoporosis underwent RCR. Osteoporotic patients were more likely to be female (osteoporosis: 87.1% female, no osteoporosis: 44.9% female; P $\le$ .001), older (osteoporosis: 64.6 ± 7.79 years, no osteoporosis: 59.3 ± 10.3 years; P $\le$ .001), and had higher a comorbidity burden as measured by ECI scores (osteoporosis: 5.33 ± 3.67, no osteoporosis: 3.19 ± 2.93; P $\le$ .001) (Table I). Following matching, all demographic disparities were nonsignificant.

Table I.

Demographic information between patients with osteoporosis who underwent RCR and those without osteoporosis who underwent RCR before and after propensity matching.

Demographics	Unmatched cohort					Matched cohort
	Osteoporosis (N = 65,121)		No osteoporosis (N = 1,026,847)		P value	Osteoporosis (N = 65,118)		No osteoporosis (N = 65,118)		P value
	Value	%	Value	%		Value	%	Value	%
Full cohort
Average age (yr) (mean ± SD)	64.6 ± 7.79	-	59.3 ± 10.3	-	<.001	66.4 ± 7.86	-	66.4 ± 7.85	-	.94
Female	56,727	87.1	460,624	44.9	<.001	56,724	87.1	56,728	87.1	.98
ECI	5.33 ± 3.67	-	3.19 ± 2.93	-	<.001	5.33 ± 3.67	-	5.33 ± 3.65	-	.79

ECI, Elixhauser Comorbidity Index; SD, standard deviation.

Across the entire cohort, 2-year rates for revision RCR and conversion to rTSA were 2.59% and 1.02%, respectively. Over the study period, the proportion of open RCR decreased significantly with time (slope = −1.46, R² = 0.99), while the proportion of arthroscopic RCR increased with time (slope = 1.68, R² = 0.98) (Fig. 1). Within 2 years following RCR, patients with a diagnosis of osteoporosis had significantly higher rates of conversion to rTSA (1.12% vs. 0.91%; odds ratio [OR]: 1.23, 95% confidence interval [CI]: 1.10-1.37, P $\le$ .001) and pathological humeral fractures (0.11% vs. 0.06%; OR: 1.89, 95% CI: 1.27-2.82, P = .002) compared to those without osteoporosis (Table II). No significant differences were found in the incidence of revision RCR between groups (2.55% vs. 2.63%; OR: 0.97, 95% CI: 0.90-1.04, P = .37) (Table II).

Figure 1.

Trends in the percentage of open and arthroscopic RCR among total cohort of patients from 2011 to 2022. Percentages were calculated as the number of procedures performed using each approach divided by the total number of RCR in total cohort for each year. RCR, rotator cuff repair.

Table II.

Primary 2-year post-operative outcomes following RCR surgery.

Complications	Osteoporosis (N = 65,118)		No osteoporosis (N = 65,118)		OR	95% CI	P value
	Value	%	Value	%
Full cohort
RCR revision	1,659	2.55	1,711	2.63	0.97	0.90, 1.04	.37
rTSA	729	1.12	595	0.91	1.23	1.10, 1.37	<.001
Pathological humeral fracture	70	0.11	37	0.06	1.89	1.27, 2.82	.002

Open RCR
	Osteoporosis (N = 14,390)		No osteoporosis (N = 14,390)		OR	95% CI	P value
	Value	%	Value	%
RCR revision	375	2.61	417	2.90	0.90	0.78, 1.03	.14
rTSA	213	1.48	151	1.05	1.42	1.15, 1.75	.001
Pathological humeral fracture	45	0.31	25	0.17	1.80	1.10, 2.94	<.001

Arthroscopic RCR
	Osteoporosis (N = 51,751)		No osteoporosis (N = 51,751)		OR	95% CI	P value
	Value	%	Value	%
RCR revision	1,263	2.44	1,249	2.41	1.01	0.93, 1.09	.79
rTSA	537	1.04	458	0.89	1.17	1.04, 1.33	.01
Pathological humeral fracture	23	0.04	8	0.02	2.75	1.22, 6.18	<.001

RCR, rotator cuff repair; rTSA, reverse total shoulder arthroplasty; OR, odds ratio; CI, confidence interval.

OR is defined as likelihood of event to occur in the osteoporosis population compared to the no osteoporosis population, CI of 2.5% and 97.5%.

Among patients undergoing open RCR, those with osteoporosis had significantly higher rates of conversion to rTSA (1.48% vs. 1.05%; OR: 1.42, 95% CI: 1.15-1.75, P $\le$ .001) and pathological humeral fractures (0.31% vs. 0.17%; OR: 1.80, 95% CI: 1.10-2.94, P$\le$ .001) compared to patients without osteoporosis (Table II). Among patients undergoing arthroscopic RCR, those with osteoporosis experienced increased rates of pathological humeral fractures (0.04% vs. 0.02%; OR: 2.75, 95% CI: 1.22-6.18, P $\le$ .001), but there was no significant difference in rates of rTSA (1.04% vs. 0.89%; OR: 1.17, 95% CI: 1.04-1.33, P = .01). There was no significant difference in rates of revision RCR between patients with and without osteoporosis in either the open or arthroscopic RCR groups.

In the subanalysis of the osteoporosis group, patients who underwent open RCR had significantly higher 2-year rates of revision RCR (3.78% vs. 2.68%; OR: 1.43, 95% CI: 1.29-1.58, P $\le$ .001), conversion to rTSA (1.04% vs. 0.89%; OR: 1.53, 95% CI: 1.31-1.79, P $\le$ .001), and pathological humeral fractures (0.31% vs. 0.04%; OR: 7.06, 95% CI: 4.27-11.7, P$\le$ .001) compared to those who underwent arthroscopic RCR (Table III). During the study period, temporal trends in utilization of open vs. arthroscopic RCR were similar to the full cohort, with the proportion of open RCR decreasing significantly with time (slope = −0.96, R² = 0.98) and the proportion of arthroscopic RCR increasing with time (slope = 1.20, R² = 0.98).

Table III.

Primary 2-year post-operative outcomes for patients with a diagnosis of osteoporosis who underwent open RCR compared to those who underwent arthroscopic RCR.

Complications	Open RCR (N = 14,390)		Arthroscopic RCR (N = 51,751)		OR	95% CI	P value
	Value	%	Value	%
RCR revision	544	3.78	1,388	2.68	1.43	1.29, 1.58	<.001
rTSA	229	1.59	541	1.05	1.53	1.31, 1.79	<.001
Pathological humeral fracture	45	0.31	23	0.04	7.06	4.27, 11.7	<.001

RCR, rotator cuff repair; rTSA, reverse total shoulder arthroplasty; OR, odds ratio; CI, confidence interval.

OR is defined as likelihood of event to occur in the open RCR osteoporosis population compared to the arthroscopic RCR osteoporosis population, CI of 2.5% and 97.5%.

In the full matched cohort of patients undergoing RCR, those with osteoporosis experienced a higher general surgical complication rate within 90 days post-operatively compared to patients without osteoporosis (1.59% vs. 1.35%; OR: 1.18, 95% CI: 1.08-1.30, P $\le$ .001) (Table IV). Patients with osteoporosis also experienced higher rates of post-operative adverse events, including anemia (5.18% vs. 4.56%; OR: 1.14, 95% CI: 1.09-1.20; P $\le$ .001), shoulder stiffness (2.72% vs. 2.43%; OR: 1.12, 95% CI: 1.05-1.20; P$\le$ .001), and medical visits (8.17% vs. 6.72%; OR: 1.23, 95% CI: 1.18-1.29; P$\le$ .001) (Table IV).

Table IV.

Secondary 90-day post-operative outcomes following RCR surgery for the full cohort.

Complications	Osteoporosis (N = 65,118)		No osteoporosis (N = 65,118)		OR	95% CI	P value
	Value	%	Value	%
General surgical complication	1,038	1.59	879	1.35	1.18	1.08, 1.30	<.001
Surgical Site Infection	23	0.04	19	0.03	1.21	0.66, 2.22	.64
Wound complications	141	0.22	111	0.17	1.27	0.99, 1.63	.07
Venous thromboembolism	375	0.58	357	0.55	1.05	0.91, 1.22	.53
Pulmonary embolism	200	0.31	237	0.36	0.84	0.70, 1.02	.08
Deep vein thrombosis	215	0.33	164	0.25	1.31	1.07, 1.61	.01
Myocardial infarction	109	0.17	113	0.17	0.96	0.74, 1.26	.84
Sepsis	231	0.35	233	0.36	0.99	0.83, 1.19	.96
MRSA	21	0.03	17	0.03	1.24	0.65, 2.34	.63
Anemia	3,377	5.18	2,971	4.56	1.14	1.09, 1.20	<.001
Pneumonia	674	1.04	703	1.08	0.96	0.86, 1.07	.45
Effusion	425	0.65	341	0.52	1.25	1.08, 1.44	.003
Shoulder stiffness	1,768	2.72	1,582	2.43	1.12	1.05, 1.20	.001
Arthrocentesis/steroid injection	3,504	5.38	3,453	5.30	1.02	0.97, 1.07	.54
Medical visits	5,320	8.17	4,377	6.72	1.23	1.18, 1.29	<.001
ER visits	109	0.17	75	0.12	1.45	1.08, 1.95	.01
Readmission	964	1.48	1,022	1.57	0.94	0.86, 1.03	.20

RCR, rotator cuff repair; OR, odds ratio; CI, confidence interval; ER, emergency room; MRSA, methicillin-resistant Stphylococcus aureus.

OR is defined as likelihood of event to occur in the osteoporosis population compared to the no osteoporosis population, CI of 2.5% and 97.5%.

Discussion

This study evaluated the relationship between a diagnosis of osteoporosis and post-operative outcomes following RCR using a large national insurance claims database. We found that patients with osteoporosis experienced higher rates of conversion to rTSA and pathological fracture within 2 years of RCR, and a higher rate of post-operative complications within 90 days following RCR, compared to a propensity-matched control group without osteoporosis. Subgroup analysis revealed that among osteoporotic patients, those undergoing open RCR had higher rates of revision RCR, conversion to rTSA, and pathological humeral fractures within 2 years of the index surgery compared to those undergoing arthroscopic repair.

The influence of osteoporosis on post-operative outcomes has traditionally been discussed in the context of fracture fixation or spinal fusion. More recently, the potential adverse impact of poor bone quality on elective shoulder procedures has become increasingly recognized. Similar to prior studies on shoulder arthroplasty linking osteoporosis to increased rates of periprosthetic fracture, implant loosening, and revision surgery, our results indicate that poor bone quality may alter the risk profile for RCR.^7,8,19 Understanding this relationship requires consideration of the fundamental biology of rotator cuff healing. Successful repair depends on regeneration of the native fibrocartilaginous enthesis at the tendon–bone interface, a specialized transitional tissue with distinct zones progressing from tendon to unmineralized fibrocartilage, mineralized fibrocartilage, and bone.^18,32 Following repair, healing occurs through one of 2 pathways: regenerative healing, which recapitulates the native enthesis structure with organized fibrocartilage formation, or reparative healing, which produces disorganized fibrovascular scar tissue prone to mechanical failure.^3,9,10,14 In osteoporotic patients, compromised bone quality impairs healing both mechanically, by reducing suture anchor pullout strength, and biologically, by altering bone formation and mineralization at the repair interface.¹⁸ This dual impact explains why osteoporosis is an independent risk factor for failure of tendon healing after RCR, as demonstrated in the Rotator Cuff Healing Index scoring system.¹⁸ Patients identified as high risk by Rotator Cuff Healing Index may benefit from alternative repair strategies, including structural or biologic augmentation of the repair construct, as well as pharmacologic optimization of bone quality, such as bisphosphonates or anabolic agents, to enhance tendon-to-bone healing.^20,27,37

Following primary RCR, reported 2-year rates of revision RCR range from 2.0% to 12.6%, while conversion to rTSA ranges from 0.9% to 3.3%.^15,16,21 Our data are consistent with these reports, with 2-year rates for revision RCR of 2.59% and rTSA of 1.02% across the entire cohort. In this large propensity-matched cohort of patients undergoing RCR, osteoporotic patients had significantly higher rates of conversion to rTSA within 2 years, while revision RCR rates were not significantly different between groups. As the PearlDiver database does not provide sufficient granularity to identify all cases of failure after RCR surgery, revision RCR and conversion to rTSA are the best available proxies. Thus, it is possible that patients may have experienced repair failure—radiographically or clinically—but did not undergo a subsequent surgery within 2 years. The significantly higher rate of conversion to rTSA in the group of patients with osteoporosis indicates that these patients may be predisposed to failure after index RCR and may be less suitable to undergo a revision soft-tissue procedure.

Across the full RCR cohort, as well as among both the open and arthroscopic repair subgroups, 2-year rates of pathological fracture were significantly higher in patients with osteoporosis compared to those without. This serves as a reminder of the essential role of the orthopedic surgeon promoting bone health and fracture prevention, even when performing soft-tissue procedures. A directive by the United States Surgeon General in 2004 urged health care providers to allocate attention and resources to the evaluation and management of osteoporosis.²⁵ Despite this, broad underscreening and undertreatment continues to perpetuate a gap between best-practice and the reality of clinical care.^22,35 Orthopedic surgeons are often the first contact point for these patients and must be advocates for bone health when treating patients in at-risk populations, even when performing soft-tissue procedures.^26,31 This can be facilitated by using prepopulated information from programs such as the American Orthopedic Association's “Own the Bone” initiative regarding adequate calcium and vitamin D intake, weight-bearing exercise, smoking cessation, alcohol reduction, and fall prevention, as well as initiating referrals for BMD testing and pharmacologic therapy when indicated.^5,33 The relevance of these measures is further underscored by the higher rate of conversion to rTSA, a procedure for which osteoporosis is known to be linked with adverse events and suboptimal outcomes.^1,19

Analysis within the cohort of patients with osteoporosis demonstrated that those who underwent open RCR demonstrated significantly higher rates of revision RCR, conversion to rTSA, and pathological fracture compared to those undergoing arthroscopic repair. Over the study period extending from 2010 to 2023, the proportion of open RCR decreased substantially with time (slope = −0.96, R² = 0.98). As a result, a higher number of open RCR cases may have been performed with older repair methods, such as drilling bone tunnels, the outcomes of which may have been disproportionately impacted by poor bone quality of the proximal humerus.¹⁷ In addition, open RCR may be associated with higher intraoperative mechanical stress on already compromised bone.^28,34 Conversely, arthroscopic repair, with greater preservation of the soft-tissue envelope, less mechanical stress, and modern bone-sparing fixation techniques, may mitigate these risks in the osteoporotic population.^12,30

In the 90-day post-operative period, osteoporotic patients had significantly higher rates of general surgical complications, anemia, shoulder stiffness, and increased medical visits. Anemia may be more common in osteoporotic patients following RCR due to chronic inflammation, nutritional deficiencies, or concomitant comorbidities such as chronic kidney disease or malabsorption, which are more prevalent in this population.²³ Increased shoulder stiffness following RCR in osteoporotic patients may be attributed to delayed mobilization, either due to surgeon-directed restrictions intended to protect the repair or to reduced functional capacity limiting patient participation in post-operative rehabilitation.¹⁰ In addition, osteoporosis is associated with systemic frailty, which may lead to more medical encounters during the early post-operative period.^7,23

Although this study provides valuable insights into the association between osteoporosis and outcomes following RCR, several notable limitations must be considered. The primary limitation is the retrospective design using a large insurance claims database, which introduces inherent risks of selection bias and coding inaccuracies. Osteoporosis diagnosis was based on billing codes and could not be confirmed using BMD measurements or stratified by severity, and underdiagnosis may have resulted in some patients with poor bone quality being included in the control cohort. The use of predefined code groupings, such as those for general surgical complications, lacks specificity and may encompass heterogeneous outcomes not directly relevant to RCR. Because our only proxy for repair failure was subsequent surgery, it is likely that true failure rates were underestimated. In addition, important orthopedic-specific clinical data, such as tendon quality, tear size, repair technique, fixation method, and patient-reported outcome measures, were not available, limiting the ability to assess repair integrity and functional recovery. The database also does not capture radiographic or intraoperative findings, which are key in evaluating tendon-to-bone healing and anchor fixation in osteoporotic patients. While we employed propensity score matching to control for baseline demographic and comorbidity differences, residual confounding from unmeasured variables remains possible, and causality cannot be established due to the observational, retrospective design. It is also important to note that although several post-operative variables reached statistical significance at a P $\le$ .002 threshold, many of these differences represented relatively small absolute differences between the osteoporosis and nonosteoporosis cohorts (often less than 0.5%). This suggests that while the large sample size provides strong statistical power to detect even modest differences, the clinical significance of these findings should be interpreted cautiously. These small effect sizes may indicate that the observed associations, although real, contribute only modestly to short-term post-operative risk at the individual level. Finally, osteoporosis severity, treatment status, and adherence to pharmacologic therapy were not available in this dataset, which may have influenced outcomes and limited the precision of group comparisons. Further research incorporating detailed parameters of bone health, radiographic findings, intraoperative evaluations, and patient-reported measures is needed to further clarify the associations between osteoporosis and RCR outcomes.

Conclusion

In this propensity-matched cohort analysis using a large national insurance database, a diagnosis of osteoporosis was associated with higher rates of conversion to rTSA, pathological humeral fractures, and select 90-day post-operative adverse events following RCR. Among osteoporotic patients, open RCR was associated with higher rates of revision RCR, conversion to rTSA, and pathological humeral fracture compared to arthroscopic repair. These findings suggest that a bone health evaluation in at-risk patients prior to RCR may be prudent to facilitate pre-operative optimization, surgical planning, and post-operative protocols in order to improve outcomes and decrease the incidence of adverse events.

Disclaimers:

Funding: No funding was received for the completion of this study.

Conflicts of interest: The authors, their immediate family, and any research foundation with which they are affiliated did not receive any financial payments or other benefits from any commercial entity related to the subject of this article.