The Affordable Care Act changed the landscape of health care in the United States by reducing the population of uninsured patients.1,2 The Affordable Care Act expanded the availability of Medicaid to include all persons with income less than 138% of the poverty level.1,2 This change promised to expand on what was already the largest insurer in the country, Medicaid, which included 70 million patients in 2012.1,2 States were given the option to proceed with Medicaid expansion, and as of 2016, a total of 31 of the 50 states had elected to move forward, increasing the Medicaid population by approximately 9.7 million patients.1–3 Despite this increase, 29 million people remained uninsured as of 2015.3 Additionally, it remains unclear how increased insurance coverage equates to health care access and outcomes.
Total shoulder arthroplasty (TSA) has become an effective treatment for the growing elderly population of patients with multiple types of shoulder pathology.4,5 The number of TSA and shoulder hemiarthroplasty procedures increased more than 400% from 2000 to 2008 and continues to increase at annual rates similar to those for total knee arthroplasty.4,6 Success rates with TSA in the 1990s were up to 90% with intermediate- and long-term follow-up.4 Reverse total shoulder arthroplasty (rTSA) was introduced as a new surgical technique to treat conditions such as rotator cuff arthropathy, proximal humerus fracture, rheumatoid arthritis, malunion, nonunion, revision, malignancy, dislocation, and glenoid bone deficiency.4,7 Multiple studies have shown that this procedure is as effective as anatomic TSA in terms of pain and American Shoulder and Elbow Surgeons (ASES) scores.5,6,8,9 The ASES score is a validated patient-reported outcome measure that is commonly used for patients with shoulder arthroplasty.7,10,11 The score has a reported minimal clinically important difference of 9 to 21 for shoulder arthroplasty.10,11
The disparity in musculoskeletal health and access to care is an evolving area of study. Disparity in care is correlated with insurance status for several reasons. In many cases, insurance status is somewhat confounded by age and is a proxy for patient socioeconomic status. Uninsured patients and those with Medicare or Medicaid are at risk for increased mortality after major surgery.12,13 Additionally, compared with private insurance, Medicaid has been associated with longer length of stay, higher total cost, increased risk of wound complications, infectious complications, pulmonary complications, and systemic complications after major surgery.12,14–16 The hip and knee arthroplasty literature showed that patient outcomes differ profoundly by insurance status.12,17–19 However, the relationship between insurance status and functional outcome after shoulder arthroplasty is not as clear. The primary goal of this study was to compare functional outcome in terms of ASES scores in patients after TSA and rTSA with respect to payer type. The authors hypothesized that the differences in functional outcome after TSA and rTSA will be consistent with the data on outcomes after hip and knee arthroplasty. These findings will provide both patients and surgeons with valuable information on outcomes after shoulder replacement.
Materials and Methods
A retrospective review was performed of all patients who underwent total shoulder arthroplasty (Current Procedural Terminology code 23472) at one institution between January 1, 2013, and January 1, 2016. Institutional review board approval was obtained. Inclusion criteria included patient age older than 18 years and primary anatomic TSA or rTSA. Exclusion criteria included lack of initial ASES score, inadequate follow-up (<1 year), arthroplasty in the setting of proximal humerus fracture, hemiarthroplasty, bone grafting or tendon transfer, and/or revision procedures. The final study population included 91 shoulders (84 patients) after exclusion for lacking initial ASES score (n=18) and/or inadequate follow-up (n=19).
All study patients had an anatomic shoulder prosthesis (Aequalis Ascend Flex; Wright Medical/Tornier, Memphis, Tennessee) or a reverse shoulder prosthesis (AltiVate Reverse; DJO, Lewisville, Texas). The postoperative protocol did not vary among patients. Those who underwent anatomic TSA were placed in a sling for 4 weeks. Between 10 and 14 days postoperatively, patients were prescribed physical therapy to include early passive and active range of motion. Patients remained non–weight bearing in the operative shoulder for 4 weeks. Those who underwent rTSA were placed in a sling for 2 to 3 weeks. No therapy was used for rTSA.
Adequate clinical follow-up was defined as meeting one of the following criteria: (1) documentation by the treating surgeon that the patient had optimized his or her outcome and that follow-up was no longer required and (2) diagnosis of failed arthroplasty requiring revision surgery. Clinic notes were reviewed to determine when the treating surgeon declared the arthroplasty stable and function optimized through correlation with documentation of each patient's subjective report of pain and function, the clinician's physical examination, and the clinician's interpretation of serial radiographs. Patients who did not have postoperative follow-up for at least 1 year were excluded. Typical follow-up intervals were 2 and 6 weeks, 3 and 6 months, and 1 year. Radiographs were obtained immediately postoperatively and at 6 months and 1 year unless symptoms or previous imaging indicated the need for more frequent monitoring. Typical series included anteroposterior, true anteroposterior, axillary, and scapular Y views of the shoulder. All patient records were analyzed in their entirety to determine whether late healing complications occurred among patients who were previously declared healed.
Each patient was assessed for pre- and postoperative visual analog scale (VAS) score and ASES score 1 year postoperatively. Further, the electronic health record was reviewed to obtain demographic information on age, sex, body mass index, history of smoking, and history of diabetes.
Patient characteristics were presented as mean and standard deviation or as count and percentage. For patients who underwent more than one procedure, age was calculated at the time of the first procedure. Patient characteristics were compared across insurance types with analysis of variance and Fisher's exact test for continuous and nominal variables, respectively. Least squares mean preoperative ASES scores were estimated from a generalized linear mixed model that included procedure type and insurance type as independent variables. A second generalized linear mixed model was specified to predict ASES score at 1 year and included preoperative ASES score as an adjustment variable. These models also included random intercepts for patients to account for within-individual correlation. Similar generalized linear mixed models were used to calculate least squares mean VAS scores at each time point. Finally, a generalized linear mixed model with a binomial distribution and logit link was used to compare reoperation rates. P values for omnibus tests for insurance were reported, and Sidak adjustments were applied to P values for pairwise comparisons among insurance types. Analyses were performed with SAS version 9.4 software (SAS Institute, Cary, North Carolina).
A total of 84 patients (91 shoulders) met the inclusion criteria and were followed clinically for a minimum of 1 year after surgery. Of the 84 patients, 50 were women (59.5%) and 34 were men (40.5%). Mean age was 65.3±10.4 years. Most of the patients had a primary diagnosis of primary osteoarthritis (59.3%). The second most common diagnosis was rotator cuff arthropathy (28.6%). A minority of patients had a preoperative diagnosis of avascular necrosis (6.59%), fracture (4.40%), or rheumatoid arthritis (1.10%). Six patients (7.1%) had documented tobacco use. Of the patients, 20 (23.8%) had a history of diabetes mellitus (Table 1).
Patient Characteristics by Type of Insurance
Medicare patients were older than patients with all other insurance types (P<.05 for each pairwise comparison). Patients who received workers' compensation were more likely to be smokers than those with other insurance types (40.0% vs 5.1%, respectively), although this difference was not statistically significant after adjustment for multiple comparisons. No difference was found between the groups for sex or the prevalence of diabetes.
In the TSA group (N=53 shoulders), 20 shoulders were associated with Medicare (37.7%) and 19 were associated with private insurance (35.8%). Of the remaining patients, 13 had Medicaid (24.5%) and 1 had workers' compensation (1.9%). In the rTSA group (N=38), most of the shoulders (n=29) were associated with Medicare (76.3%). Additionally, 2 patients had Medicaid (5.3%), 3 had private insurance (7.9%), and 4 had workers' compensation (10.5%). No uninsured patients were included in either group.
Before surgery, ASES score differed by insurance type (P=.014), with lower scores among Medicaid patients compared with those with private insurance (mean [SD], 20.4 [4.5] vs 38.8 [3.9], P=.009). One year postoperatively and controlling for baseline ASES, postoperative ASES scores differed by insurance type as well (P<.001), and patients with private insurance had better mean ASES scores (85.6 [4.6]) compared with those with Medicaid (55.2 [5.4], P<.001) and workers' compensation (57.1 [8.5], P=.028). Patients with Medicare (80.6 [2.8]) also has better ASES scores at follow-up compared with those with Medicaid (P<.001). No other pairwise significant differences were found in ASES scores at either time point (Table 2).
Least Squares Mean American Shoulder and Elbow Surgeons Scores by Type of Insurance
The greatest change in mean preoperative and postoperative ASES scores was noted in the Medicare cohort. For this group, the mean preoperative ASES score was 31.4±2.4 and the mean postoperative ASES score was 80.6±2.8 (delta, 48.1). The improvement observed in the private insurance group was also greater than 40 points on the ASES score. The Medicaid group averaged a much smaller change in ASES score, with mean preoperative and postoperative ASES scores of 20.4±4.5 and 55.2±5.4, respectively (delta, 34.8) (Table 2).
The VAS pain scores showed similar trends. Preoperative VAS scores did not differ significantly by insurance type (P=.18). However, postoperative VAS scores differed by insurance type (P<.001), with patients with Medicaid (mean [SD], 4.31 [0.64]) having significantly higher VAS scores compared with those with both private insurance (1.05 [0.55], P<.001) and Medicare (1.23 [0.34], P<.001) (Table 3). Changes in mean VAS scores mirrored the results for ASES scores, with the greatest change seen in patients with Medicare and private insurance. Reoperation rates for each insurance type demonstrated no statistically significant difference (Table 4).
Least Squares Mean Visual Analog Scale Scores by Type of Insurance
Reoperation Rate by Type of Insurance
In the ever-evolving insurance landscape, payer status has had implications for comorbidities, patient satisfaction, and surgical outcomes.4,16 Medicare is the most common form of payer for shoulder arthroplasty. Although Medicaid accounts for 1.5% to 1.8% of payers for this procedure, Medicaid expansion promises to increase the incidence of this population seeking shoulder arthroplasty.1,4
Norris and Iannotti13 showed significant improvement in ASES scores at 1-year follow-up after primary TSA for primary osteoarthritis, with a mean increase of 52.3 points. Long-term follow-up showed lasting improvement in motion and strength as well as high patient-reported satisfaction.14 Wall et al15 showed an average 37-point increase in Constant score for patients with various etiologies treated with rTSA.
In this series of 84 patients undergoing primary anatomic TSA or rTSA, the mean preoperative ASES score was 32.3 and the mean postoperative ASES score was 76.8, falling within the range of previous studies.17,19 The mean magnitude of change between pre- and postoperative scores was 44.5, which is consistent with previous literature.5,6
The current analysis showed variable improvement between pre- and postoperative ASES scores that correlated with insurance status. Preoperative ASES scores were lower in the Medicaid cohort, and postoperative ASES scores remained lower in this cohort at 1-year follow-up. The magnitude of improvement in ASES scores was greater in the private insurance and Medicare cohorts compared with the Medicaid and workers' compensation groups.
These changes are seemingly isolated to payer status, and demographic analysis showed that comorbidities, such as smoking and diabetes, which may be linked to increased pain scores and dissatisfaction, remained relatively similar between insurance cohorts. The current study included a statistically significantly younger cohort of Medicaid patients compared with Medicare patients. Previous studies pointed out that this may indicate earlier physical deterioration in this population compared with patients with private insurance and Medicare.18 Medicaid patients are younger on average and have demonstrably lower preoperative self-reported pain and functional scores, as evidenced by VAS and ASES scores. Many social and economic factors likely contribute to this finding.
The current results indicate that previously reported functional outcomes from the hip and knee literature may not be generalizable to outcomes after upper extremity joint replacement. The discrepancy in the magnitude of change is in contrast to previous studies of functional outcomes in TKA and THA in which pre- and postoperative functional scores were lower for the Medicaid cohort, although the magnitude of improvement was similar between groups.18,19 In addition, VAS scores were significantly higher in the Medicaid cohort compared with the private insurance and Medicare groups. Although previous studies showed increased complication rates in patients with Medicaid, this study did not show a significant difference in re-operation rates between patients with different insurance types.
In a cohort of patients who underwent total hip or knee arthroplasty, patients with private insurance had a greater change in Short Form-36 physical health scores on average compared with patients with Medicare and Medicaid.17 Rosenthal et al18 compared a cohort of Medicaid patients with non-Medicaid patients undergoing total knee arthroplasty and found that Medicaid patients had lower pre- and postoperative Knee Society Scores than those with Medicare or private insurance. The magnitude of change in pre- to postoperative Knee Society Scores was similar for patients with Medicaid, Medicare, and private insurance.18
Waldrop et al16 found disparities in patient-reported outcomes, including ASES scores, between patients with socioeconomically disadvantaged insurance and private insurance. Pre- and postoperative ASES scores were greater in the private insurance group; however, the magnitude of change in ASES scores was similar, contrary to the current findings.16 In their study, Waldrop et al16 grouped Medicaid and Medicare patients younger than 65 years in the socioeconomically disadvantaged category and did not include any patients older than 65 years. The current study adds to the generalizability of these findings because mean patient age was 65.3 years.
Limitations of the current study included its retrospective design and the patient population that was drawn from a single urban area. The results may not be generalizable to other settings. Additionally, it is possible that patients with arthroplasty that appeared clinically and radiographically stable at final follow-up may have presented to another institution with late failure. A large number of patients had follow-up of 1 year or less, which may correspond to a satisfactory outcome or failure with follow-up with a different surgeon. Although the authors examined aggregate numbers from 2 different types of implant (TSA and rTSA), multiple studies have shown similar magnitudes of change within the time frame of this study as well as similar complication rates and other patient-reported outcomes. Simovitch et al5 compared functional improvement after anatomic TSA and rTSA with multiple outcome scores, including ASES scores. No statistically significant difference was found in the magnitude of change in ASES scores between anatomic TSA and rTSA, 51.0 and 50.6, respectively.5 The current study also noted that most of the improvement in outcome scores had occurred by 6 months postoperatively.5 Flurin et al8 showed similar magnitude of improvement in ASES scores between anatomic TSA and rTSA, 51.7 and 51.5, respectively. Additional studies showed similar complication rates, patient-reported outcomes, and pain scores between the 2 procedures.9 Finally, although the current results showed that insurance status was not a significant prognostic indicator for reoperation, this finding is beyond the scope and power of this study. Future studies may investigate the specific factors that contribute to the discrepancies between Medicaid and other insurance groups.
Medicaid patients are younger, have more debilitating disease, and obtain less benefit after TSA compared with those with private insurance and Medicare. This finding is important as the population of public aid patients continues to grow and becomes a larger proportion of the practice of many surgeons. This study provided objective data to aid in preoperative discussion and management of expectations for TSA and adds to the findings of Waldrop et al16 in showing the possibility of diminished benefit in certain populations. Future studies may continue to investigate the specific factors that contribute to discrepancies between patients with Medicaid and other types of insurance and may seek to determine which of these factors may be modifiable before shoulder arthroplasty.
- Hahn JA, Sheingold BH. Medicaid expansion: the dynamic health care policy landscape. Nurs Econ. 2013;31(6):267–272, 297. PMID:24592530
- Iglehart JK, Sommers BD. Medicaid at 50: from welfare program to nation's largest health insurer. N Engl J Med. 2015;372(22):2152–2159. doi:10.1056/NEJMhpr1500791 [CrossRef] PMID:26017826
- Clarke TC, Norris T, Schiller JS. Early release of selected estimates based on data from 2016 National Health Interview Survey. https://www.cdc.gov/nchs/data/nhis/earlyrelease/earlyrelease201705.pdf. Accessed May 21, 2019.
- Jain NB, Higgins LD, Guller U, Pietrobon R, Katz JN. Trends in the epidemiology of total shoulder arthroplasty in the United States from 1990–2000. Arthritis Rheum. 2006;55(4):591–597. doi:10.1002/art.22102 [CrossRef] PMID:16874781
- Simovitch RW, Friedman RJ, Cheung EV, et al. Rate of improvement in clinical outcomes with anatomic and reverse total shoulder arthroplasty. J Bone Joint Surg Am. 2017;99(21):1801–1811. doi:10.2106/JBJS.16.01387 [CrossRef] PMID: 29088034
- Flurin PH, Roche CP, Wright TW, Marczuk Y, Zuckerman JD. A comparison and correlation of clinical outcome metrics in anatomic and reverse total shoulder arthroplasty. Bull Hosp Joint Dis. 2015;73(suppl 1):S118–S123.
- Richards RR, An KN, Bigliani LU, et al. A standardized method for the assessment of shoulder function. J Shoulder Elbow Surg. 1994;3(6):347–352. doi:10.1016/S1058-2746(09)80019-0 [CrossRef] PMID:22958838
- Flurin PH, Marczuk Y, Janout M, Wright TW, Zuckerman J, Roche CP. Comparison of outcomes using anatomic and reverse total shoulder arthroplasty. Bull Hosp Joint Dis. 2013;71(suppl 2):S101–S107.
- Kiet TK, Feeley BT, Naimark M, et al. Outcomes after shoulder replacement: comparison between reverse and anatomic total shoulder arthroplasty. J Shoulder Elbow Surg. 2015;24(2):179–185. doi:10.1016/j.jse.2014.06.039 [CrossRef] PMID:25213827
- Werner BC, Chang B, Nguyen JT, Dines DM, Gulotta LV. What change in American Shoulder and Elbow Surgeons score represents a clinically important change after shoulder arthroplasty?Clin Orthop Relat Res. 2016;474(12):2672–2681. doi:10.1007/s11999-016-4968-z [CrossRef] PMID:27392769
- Tashjian RZ, Hung M, Keener JD, et al. Determining the minimal clinically important difference for the American Shoulder and Elbow Surgeons score, Simple Shoulder Test, and visual analog scale (VAS) measuring pain after shoulder arthroplasty. J Shoulder Elbow Surg. 2017;26(1):144–148. doi:10.1016/j.jse.2016.06.007 [CrossRef] PMID:27545048
- Fisher DA, Dierckman B, Watts MR, Davis K. Looks good but feels bad: factors that contribute to poor results after total knee arthroplasty. J Arthroplasty. 2007;22(6)(suppl 2):39–42. doi:10.1016/j.arth.2007.04.011 [CrossRef] PMID:17823013
- Norris TR, Iannotti JP. Functional outcome after shoulder arthroplasty for primary osteoarthritis: a multicenter study. J Shoulder Elbow Surg. 2002;11(2):130–135. doi:10.1067/mse.2002.121146 [CrossRef] PMID:11988723
- Deshmukh AV, Koris M, Zurakowski D, Thornhill TS. Total shoulder arthroplasty: long-term survivorship, functional outcome, and quality of life. J Shoulder Elbow Surg. 2005;14(5):471–479. doi:10.1016/j.jse.2005.02.009 [CrossRef] PMID:16194737
- Wall B, Nové-Josserand L, O'Connor DP, Edwards TB, Walch G. Reverse total shoulder arthroplasty: a review of results according to etiology. J Bone Joint Surg Am. 2007;89(7):1476–1485. doi:10.2106/JBJS.F.00666 [CrossRef] PMID: 17606786
- Waldrop LD II, King JJ III, Mayfield J, et al. The effect of lower socioeconomic status insurance on outcomes after primary shoulder arthroplasty. J Shoulder Elbow Surg. 2018;27(6S):S35–S42. doi:10.1016/j.jse.2018.01.002 [CrossRef] PMID:29519585
- Martin CT, Callaghan JJ, Liu SS, Gao Y, Warth LC, Johnston RC. Disparity in total joint arthroplasty patient comorbidities, demographics, and postoperative outcomes based on insurance payer type. J Arthroplasty. 2012;27(10):1761–1765. doi:10.1016/j.arth.2012.06.007 [CrossRef]
- Rosenthal BD, Hulst JB, Moric M, Levine BR, Sporer SM. The effect of payer type on clinical outcomes in total knee arthroplasty. J Arthroplasty. 2014;29(2):295–298. doi:10.1016/j.arth.2013.06.010 [CrossRef] PMID:23927908
- Hinman A, Bozic KJ. Impact of payer type on resource utilization, outcomes and access to care in total hip arthroplasty. J Arthroplasty. 2008; 23(6)(suppl 1):9–14. doi:10.1016/j.arth.2008.05.010 [CrossRef] PMID:18722298
Patient Characteristics by Type of Insurance
|Characteristic||Overall (N=84)||Private Insurance (n=19; 22.6%)||Medicare (n=46; 54.8%)||Medicaid (n=14; 16.7%)||Workers' Compensation (n=5; 6.0%)||P|
|Age, mean (SD), y||65.3 (10.4)||59.8 (9.3)||71.6 (7.6)||53.6 (5.5)||60.8 (5.3)||<.01|
|Male, No. (%)||34 (40.5)||6 (31.6)||21 (45.7)||5 (35.7)||2 (40.0)||.74|
|Smoker, No. (%)||6 (7.1)||0 (0)||3 (6.5)||1 (7.1)||2 (40.0)||.05|
|Diabetes, No. (%)||20 (23.8)||4 (21.1)||12 (26.1)||3 (21.4)||1 (20.0)||.97|
Least Squares Mean American Shoulder and Elbow Surgeons Scores by Type of Insurance
|Type of Insurance||No. (%)||Mean (SD) American Shoulder and Elbow Surgeons Score|
|Private||22 (24.2)||38.8 (3.9)||85.6 (4.6)|
|Medicare||49 (53.8)||31.4 (2.4)||80.6 (2.8)|
|Medicaid||15 (16.5)||20.4 (4.5)||55.2 (5.4)|
|Workers' compensation||5 (5.5)||35.5 (7.2)||57.1 (8.5)|
Least Squares Mean Visual Analog Scale Scores by Type of Insurance
|Type of Insurance||No. (%)||Mean (SD) Visual Analog Scale Score|
|Private||22 (24.2)||7.81 (0.51)||1.05 (0.55)|
|Medicare||49 (53.8)||7.40 (0.31)||1.23 (0.34)|
|Medicaid||15 (16.5)||8.49 (0.59)||4.31 (0.64)|
|Workers' compensation||5 (5.5)||9.14 (0.96)||4.01 (1.04)|
Reoperation Rate by Type of Insurance
|Type of Insurance||No.(%)|
|Private||22 (24.2)||2 (9.1)|
|Medicare||49 (53.8)||7 (14.3)|
|Medicaid||15 (16.5)||3 (20.0)|
|Workers' compensation||5 (5.5)||0 (0)|