Isometric strength, range of motion, and impairment before and after total and reverse shoulder arthroplasty

J Shoulder Elbow Surg (2013) 22, 869-876

www.elsevier.com/locate/ymse

SHOULDER

Isometric strength, range of motion, and impairment before and after total and reverse shoulder arthroplasty Brian Puskas, MDa, Kevin Harreld, MDa, Rachel Clark, BAb, Katheryne Downes, MPHc, Nazeem A. Virani, MD, MPHb, Mark Frankle, MDa,* a

Shoulder & Elbow Division, Florida Orthopaedic Institute, Tampa, FL, USA Clinical Research Department, Foundation for Orthopaedic Research and Education, Tampa, FL, USA c Office of Research, University of South Florida, Tampa, FL, USA b

Background: Medicare Part A provides similar resources for coverage of inpatient hospitalization costs for patients treated with total shoulder arthroplasty (TSA) and reverse shoulder arthroplasty (RSA). This is based on an assumption that TSA and RSA are used to treat similar patient populations with comparable disease severity. However, no objective clinical information is available to support this resource allocation. The purpose of this study is to quantify the disease severity and subsequent improvement from primary TSA, primary RSA, and revision arthroplasty (TSA and RSA). Methods: From March 2004 through May 2006, 174 shoulders (87 primary TSA, 55 primary RSA, and 32 revision cases) were prospectively studied using Biodex (Biodex Medical Systems, Shirley, NY, USA) isometric strength and standardized video range of motion measurements performed by an independent third-party observer at 1 week before surgery and at an average of 49 months (range, 32-69 months) postoperatively. Patient impairment ratings were calculated using the Florida Impairment Guidelines. Results: Primary TSA had the lowest average preoperative impairment (21%), and revision arthroplasty had the highest (28%). All patients demonstrated improvement in the parameters tested. At an average 49 months, all 3 groups demonstrated a similar reduction in impairment ratings (TSA: 21% to 10%; RSA: 25% to 15%; revision arthroplasties: 28% to 20%). Conclusion: There are distinct differences in preoperative disease severity among patients undergoing primary TSA, primary RSA, and revision arthroplasty. Greater impairment is evident in patients undergoing a revision arthroplasty. However, all groups may be expected to achieve improvements and maintain these improvements 4 years postoperatively. Level of evidence: Level II, Prospective Cohort Design, Treatment Study. Ó 2013 Journal of Shoulder and Elbow Surgery Board of Trustees. Keywords: Shoulder arthroplasty; cost-effectiveness; impairment; strength; range of motion; economic evaluation

This study was approved by the Western Institutional Review Board (Study # 1098441). *Reprint requests: Mark Frankle, MD, Shoulder & Elbow Division, Florida Orthopaedic Institute, 13020 N Telecom Pkwy, Tampa, FL 33637, USA. E-mail address: [email protected] (M. Frankle).

When nonoperative treatments are unsuccessful, total shoulder arthroplasty (TSA) and reverse shoulder arthroplasty (RSA) effectively decrease pain and improve function in patients with end-stage degenerative shoulder disease.1,3,6,9,12,15-18 The indications for these 2 distinct procedures vary greatly, reflecting differences in the

1058-2746/$ - see front matter Ó 2013 Journal of Shoulder and Elbow Surgery Board of Trustees. http://dx.doi.org/10.1016/j.jse.2012.09.004

870 underlying patient populations. However, these 2 patient populations are considered equivalent for purposes of hospital resource allocation, as directed in Section 1886(d) of the Social Security Act. This provision sets forth a system of hospital payment under Medicare Part A, which covers the costs of inpatient hospitalization in which reimbursement rates are the same for patients in a given homogenous cluster, called a diagnosis-related group (DRG). Each DRG has a payment weight assigned to it based on the average resources used to treat Medicare patients. When the amount of resources used is not known, as was the case when RSA was introduced in the United States in 2004 and 2005, payment was based on the assumption that patients within a given category are clinically similar, have similar severity of disease, and are therefore expected to use the same level of hospital resources. The assignment of a DRG code is therefore linked to the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) code. This ICD-9-CM code (81.80) was the same for RSA and TSA from 2004 to 2006. As a result, TSA and RSA are both currently assigned the same resources from Medicare Part A. However, little objective information exists to provide for a rational allocation of resources based on the severity of the underlying shoulder condition. Patients managed with RSA are generally felt to represent a different patient population from those undergoing TSA. Patients treated with RSA suffer from muscular insufficiency of the rotator cuff and also variable amounts of glenohumeral cartilage loss, whereas, patients undergoing TSA suffer primarily from articular cartilage disease. Patients requiring a revision from a previous arthroplasty may also have additional pathology. These differences are not captured in the current allocation scheme. The purpose of the current study was to characterize disease severity preoperatively and postoperatively in 3 patient groupsdprimary TSAs, primary RSAs, and revision shoulder arthroplastiesdusing objective outcome measures. Our main hypothesis is that patients undergoing primary TSA would be less impaired than those undergoing primary RSA. Our secondary hypothesis is that patients undergoing a primary arthroplasty would have lower postoperative impairment ratings than the revision shoulder arthroplasties. Lastly, we hypothesize that identifiable patient factors will have an effect on postoperative disease severity as measured by impairment.

Materials and methods All patients provided informed consent before participating in the study. To assess the difference in disease severity between TSA and RSA patients, we used subjective and objective measures to perform a prospective cohort study of a single surgeon’s shoulder arthroplasty practice. Objective measures of isometric strength and range of motion (ROM) were obtained, as described below. Data were collected 1 week before surgery and at a minimum of 2 years postoperatively.

B. Puskas et al. Table I

Reasons for nonparticipation

Reason for nonparticipation Deceased Unable to locate Missed scheduled appointment Moved out of state In assisted living facility No longer able to travel Poor health Unsatisfied with surgery No longer desire to participate Bilateral shoulder) Totals )

Patients

Shoulders

No. (%)

No. (%)

41 (10) 47 (12) 19 (5) 20 (5) 3 (<1) 33 (8) 24 (6) 10 (3) 19 (5) 0 (0) 216

42 (11) 50 (13) 19 (5) 20 (5) 3 (<1) 33 (8) 24 (6) 10 (3) 19 (5) 2 (<1) 222

Only most recent shoulder included.

Patient population Between March 2004 and May 2006, 396 shoulders in 390 patients underwent TSA (n ¼ 204) or RSA (n ¼ 192). All surgical procedures were performed by the senior author (M.A.F.) after failure of a reasonable trial of nonoperative management. The type of prosthesis used was consistent throughout this interval (DJO TSA and DJO Reverse Shoulder Prosthesis, Encore Medical, Austin, TX, USA). Inclusion criteria for the current study included all patients undergoing shoulder arthroplasty within the specified timeframe regardless of diagnosis or surgical history; however, patients had to complete preoperative and postoperative isometric Biodex strength testing (Biodex Medical Systems, Shirley, NY, USA) in the method described below. If the patient had bilateral shoulder arthroplasties, only the most recent shoulder was included to maintain statistical assumptions of independence of observations. There were no other exclusion criteria. Table I summarizes reasons for nonparticipation in the current study. Preoperative and postoperative data at our follow-up visits were available for 174 patients, comprising 87 with primary TSA, 55 with primary RSA, and 32 with revision arthroplasties.

Primary TSA group A primary TSA was performed in those patients with radiographic evidence of glenohumeral arthritis on preoperative imaging (radiographs and computed tomography scan) and an intact rotator cuff by physical examination and intraoperative inspection. The TSA group consisted of 87 shoulders in 87 patients (38 women, 49 men) who were a mean age of 66 years (range, 35-89 years). Mean follow-up was 49 months (range, 32-69 months). In this group, none of the 87 primary TSA patients had undergone any previous shoulder surgeries.

Primary RSA group A primary RSA was performed in those patients with rotator cuff deficiency of the shoulder along with glenohumeral subluxation, glenohumeral arthritis, or pseudoparesis (<90 of elevation) of the shoulder. A combination of findings from the physical examination (rotator cuff weakness, dynamic instability, and limited ROM)

11-170) 28-180) 40-75) 0-135) 3-35) (95, (90, (20, (15, (20, 96 90 19 25 20 (53, 3-135) (48, 15-160) (20, 35-70) (0, 0-75) (28, 17-36) 56 52 19 16 28 <.001y <.001y .041 <.001y <.001y (135, 40-180) (120, 40-190) (30, 55-90) (30, 0-105) (14, 4-34) 128 121 30 39 15 20-160) 20-157) 35-70) 0-60) 13-37) 74 68 20 18 25 (90, (80, (25, (15, (20, 90 81 23 14 21

15-160) 30-165) 25-63) 0-75) 8-35)

139 127 51 37 10

(140, 45-180) (130, 40-180) (50, 10-90) (30, 0-105) (8, 1-32)

<.001y <.001y <.001y <.001y <.001y

(68, (65, (20, (15, (23,

(18, 0-68) (6, 0-53) (0, 0-48) (0, 0-34) (20, 0-51) (14, 0-49) 21 11 5 4 21 16 (10, 0-50) (0, 0-36) (0, 0-31) (0, 0-21) (15, 2-36) (8, 1-27) 13 5 3 1 16 9 <.001y .023 .009 .011 .034 <.001y (20, 0-55) (9, 0-87) (3, 0-24) (1, 0-27) (22, 0-47) (17, 0-40) 22 13 6 4 22 17 (15, 0-45) (6, 0-30) (0, 0-22) (0, 0-16) (18, 1-54) (8, 0-30) 17 8 3 2 20 9 .001y .008 .006 .006 <.001y <.001y (33, 2-92) (20, 0-64) (12, 0-65) (9, 0-50) (32, 4-99) (22, 2-68) 35 23 15 11 35 25 (23, 4-95) (14, 0-70) (6, 0-58) (2, 0-56) (22, 0-75) (13, 0-49) 29 18 10 8 25 14

Post-op Pre-op Post-op

P

Primary RSA

ER, external rotation; FF, forward flexion; IR, internal rotation; N-m, Newton-meters; ROM, range of motion; RSA, reverse shoulder arthroplasty; TSA, total shoulder arthroplasty. ) All values are displayed as mean (median, range). y Statistically significant (P < .001).

To better represent the collective objective strength and ROM data sets, we used the Florida Impairment Guidelines5 to develop one numeric score representative of objective shoulder function. The Florida Impairment Guidelines was selected because of its distinct and purposeful omission of any scoring component for pain. This

Strength (N-m) FF 0 30 60 90 IR neutral ER neutral ROM FF, Abduction, ER, IR, Impairment rating

Impairment rating

Pre-op

ROM assessment was performed using a digital goniometer on a videorecorded physical examination according to a previously published ROM protocol.6 Patients were asked not to go beyond the point of pain or discomfort. Measurements were performed by an independent observer blinded to study design and purpose. Videotaped preoperative ROM data were not available for 23 patients. As a result, a patient questionnaire completed by every patient, which includes self-assessed ROM indicated by marking the highest attainable motion on a picture, was substituted.11,13

Primary TSA

ROM measurements

Variable

Isometric shoulder strength measurements were performed by an independent physical therapist blinded to the purpose of the study, using a Biodex System II dynamometer. Measurements were obtained in 6 shoulder positions: forward flexion (FF) at 0 , 30 , 60 , and 90 , internal rotation (IR), and external rotation (ER) at 0 of abduction. Patients performed three 5-second trials, advancing to maximal isometric exertion at each position. The highest score obtained was recorded.

Preoperative vs postoperative strength, range of motion, and impairment ratings for all three groups)

Strength measurements

Table II

Revision shoulder arthroplasty was performed in 32 patients. Of these, 26 underwent revision RSA, including 19 patients with failed hemiarthroplasties, 3 failed RSAs, and 4 failed TSAs. An additional 6 patients underwent revision TSA (5 failed hemiarthroplasties, 1 failed TSA). For the purposes of evaluating the 3 groups, revision RSA and TSAwere merged for comparison with the primary groups. The rationale for including all revisions procedures in one group is that the evolution of the default procedure for a failed arthroplasty is more frequently consisting of a conversion to a reversed prosthesis, but intraoperative decision making can factor into this and therefore may include revision to a standard anatomic TSA.

P

Revision arthroplasty group

Pre-op

All revisions

Post-op

P

and radiographs (decreased joint space and abnormal joint position) were used to determine if RSA was indicated. The RSA group comprised 55 shoulders in 55 patients (35 women, 20 men) who were a mean age of 71 years (range, 46-88 years). Mean follow-up was 48 months (range, 31-72 months). The group consisted of 29 patients with rotator cuff surgeries that had failed, 23 with primary rotator cuff arthropathy, and 3 with proximal humeral malunions. For this study we did not consider a prior rotator cuff repair failure as equivalent to having had a previous arthroplasty surgery, and thus, these patients were evaluated in the primary RSA group (Tables II and III). Prior studies by Cuff et al2 and Mulieri et al.11 found no difference between primary shoulder arthroplasty patients who did or did not have prior rotator cuff surgery.

<.001y <.001y .909 .154 <.001y

871 <.001y .003 .047 .041 .005 .001y

Shoulder arthroplasty strength, motion, and impairment

872

B. Puskas et al.

Table III All revisions vs primary total shoulder arthroplasty and primary reverse shoulder arthroplasty: preoperative, postoperative, and preoperative to postoperative comparison of impairment rating Variable Pre-op Post-op D (Pre-Post)

All revisions

Primary TSA

Mean (median, range)

Mean (median, range)

P

Mean (median, range)

P

21 (20, 8-35) 10 (8, 1-32) 11 (12, 29 to 10)

<.001) <.001) .088

25 (23, 13-37) 15 (14, 4-34) 10 (10, 22 to 13)

.014 .001) .209

28 (28, 17-36) 20 (20, 3-35) 8 (8, 21 to 6)

Primary RSA

D, change; RSA, reverse shoulder arthroplasty; TSA, total shoulder arthroplasty. ) Statistically significant (P < .001).

system allows one to establish a whole-body impairment rating based solely on strength and motion deficits in a given extremity. Preoperative and postoperative strength and ROM measurements were taken for the 174 patients, as described above. We next calculated independent strength and ROM deficits for each plane of motion using the charts in the Florida Impairment Guidelines.5 For the combined strength deficit measurement, we converted all Biodex isometric strength measurements to Newton-meters (originally measured in foot-pounds) to enable a comparison with previously published normative data stratified according to age, handedness, and sex,7 and added the strength deficit percentages of 3 individual planes of motion measurements (FF at 30 þ IR at 0 þ ER at 0 ). This is a modification of the strength deficit calculation described in the Florida Impairment Guidelines. Finally, the Florida Impairment Guidelines describe how to merge the combined strength deficit calculation with the combined ROM deficit to produce an impairment rating of the entire upper extremity while adjusting for hand dominance (Section 15: Combined Values Chart).

Statistical methods Statistical analysis was conducted using SPSS 20.0 software (IBM SPSS, Armonk, NY, USA). Normality of continuous variables was evaluated using the Kolmogorov-Smirnov test. Statistical significance was assessed using the Wilcoxon signed rank test for paired data, the Mann-Whitney U and t test (where appropriate) for unpaired data, and the Spearman r correlation. A Bonferroni correction for multiple comparisons was performed, yielding an adjusted P < .001 for statistical significance. A regression analysis was used to identify independent predictors of postoperative impairment. Variables with an initial value of P < .10 on the individual regression analysis were entered into the multiple variable model. The final model consisted of all variables with a P value <.05. The Omnibus test was used to evaluate model fit.

Results Primary TSA group The preoperative and postoperative strength, ROM, and impairment results in the primary TSA group (n ¼ 87) are summarized in Table II. There was improvement preoperatively to postoperatively across all measures, with significant improvement in strength demonstrated in IR and ER, as well as in all measures (FF, abduction, ER, IR) of ROM

(P < .001). There was also a significant reduction in impairment for the group from 21% to 10% (P < .001). Complications occurred in 3 of the 87 patients (3%) in the TSA group, including postoperative cellulitis in 1, brachial plexus palsy that resolved in 1, and unresolved reflex sympathetic dystrophy in 1.

Primary RSA group The results of the 55 patients undergoing a primary RSA are summarized in Table II. These patients also improved across all measures but demonstrated significant improvement in strength measures for FF at 0 and ER, as well as in 3 of 4 measures of ROM (FF, abduction, IR). The impairment rating also significantly decreased from 25% preoperatively to 15% postoperatively (P < .001) in this group. There were 6 complications in the RSA group. Postoperative infection developed in 3 patients (1 superficial and 2 deep). One patient with infection had a superficial hematoma that required irrigation and debridement 2 weeks after surgery. The other patient with infection received irrigation and debridement with conversion to a long-stem prosthesis and glenosphere exchange at 9 months. This procedure was complicated by an intraoperatively displaced greater tuberosity fracture that was fixed with cerclage wires. The third patient with infection was treated twice with irrigation and debridement at 3 months, one of which involved exchange of the humeral stem polyethylene component with retention of the glenoid component and humeral stem. The nonoperative primary RSA complications included a postoperative acromial fracture and a periprosthetic greater tuberosity fracture (both asymptomatic at final follow-up after conservative treatment), and one readmission for a chronic obstructive pulmonary disease (COPD) exacerbation. None of these patients with complications were excluded from the study, and the preoperative data used were from before their initial surgery.

Revision arthroplasty group The results of the 32 patients undergoing revision arthroplasty are presented in Table II. This group demonstrated the highest average preoperative impairment (28%), but there was still a noticeable, significant (P < .001)

.344 .352 .001y .522 .524 (50, 42 to 140) (50, 27 to 150) (30, 30 to 100) (30, 30 to 90) (12, 29 to 10) (55, 55 to 145) 49 (47, 43 to 135) 45 (10, 55 to 120) 28 (15, 30 to 105) 23 (10, 22 to 13) 11 (140, 45-180) 128 (135, 40-180) .024 54 (130, 40-180) 121 (120, 40-190) .266 52 (50, 10 to 90) 30 (30, 55 to 90) <.001y 10 (30, 0-105) 39 (30, 0-105) .755 21 (8, 1-32) 15 (14, 4-34) <.001y 10 139 127 51 37 10 .004 .007 .684 .049 .001y 20-160) 20-157) 35 to 70) 0-60) 13-37) (68, (65, (20, (15, (23, 74 68 20 18 25 (90, (80, (25, (15, (20, 90 81 23 14 21

15-160) 30-165) 25 to 63) 0-75) 8-35)

.955 .666 .609 .770 .010 .370 (23, 4-95) (14, 0-70) (6, 0-58) (2, 0-56) (22, 0-75) (13, 0-49) 29 18 10 8 25 14

17 8 3 2 20 9

(15, 0-45) (6, 0-30) (0, 0-22) (0, 0-16) (18, 1-54) (8, 0-30)

<.001y,) <.001y,) <.001y,) <.001y,) .046 .001y

35 23 15 11 35 25

(33, 2-92) (20, 0-64) (12, 0-65) (9, 0-50) (32, 4-99) (22, 2-68)

22 13 6 4 22 17

(20, 0-55) (9, 0-87) (3, 0-24) (1, 0-27) (22, 0-47) (17, 0-40)

<.001y <.001y <.001y <.001y <.001y <.001y

6 5 3 3 3 8

(7, (3, (1, (0, (3, (8,

29 26 22 12 29 14

to to to to to to

35) 87) 24) 27) 35) 25)

6 5 5 3 9 10

(8, (5, (2, (2, (8, (9,

65 to 49) 7 to 53) 25 to 65) 32 to 50) 49 to 57) 28 to 40)

P First TSA First RSA P

First TSA

First RSA

P

Pre-op to post-op D

First RSA First TSA

D, change; ER, external rotation; FF, forward flexion; IR, internal rotation; N-m, Newton-meters; ROM, range of motion. ) All values displayed as mean (median, range). y Statistically significant (P < .001).

The results of a direct comparison of impairment in the revision group and the primary TSA and primary RSA

Strength (N-m) FF 0 30 60 90 IR neutral ER neutral ROM, FF Abduction ER IR Impairment rating

Comparison of disease severity between primary and revision procedures

Post-op

To analyze the differences in disease severity between the TSA and RSA patients, we performed a direct comparison of those patients undergoing a primary arthroplasty procedure (Table IV). We observed significant differences between the TSA and RSA populations at both preoperative and postoperative settings in isometric strength, ROM, and degree of impairment. In general, compared with the primary TSA patients, primary RSA patients were weaker, showed worse ROM, and had a higher level of impairment preoperatively. These differences were especially apparent in all 4 of the forward flexion measures of strength, which were significantly lower than the primary TSA group (Table IV). These strength differences became even more pronounced postoperatively, with the primary TSA group showing significantly better strength across all measures as well as significantly better ROM in ER. The differences in impairment also became significant postoperatively, with primary RSA patients continuing to have a greater amount of impairment compared with primary TSA (primary RSA, 15; primary TSA, 0; P < .001). However, despite the differences in disease severity, a comparison of the preoperative with the postoperative changes between the 2 groups (Table IV) demonstrates similar amounts of improvement with each procedure. There was no statistically significant difference in the improvement in the TSA group compared with the RSA group across all measures, with the exception of ROM in ER. In this case, primary TSA patients showed greater than twice the change shown in the primary RSA group.

Pre-op

Comparison of disease severity between primary TSA and primary RSA

Variable

improvement in the postoperative impairment rating (20%). This group also improved on nearly all measures, with significant improvement in strength in FF at 0 and for ROM in FF and abduction measures. However, there were no obvious changes in IR and ER measures of ROM from the preoperative to the postoperative state. Two complications occurred in the revision arthroplasty group. One revision RSA patient developed an acromial fracture that was treated nonoperatively and was asymptomatic at the final postoperative evaluation. The second complication was also in a revision RSA patient who had a failed hemiarthroplasty treated with an allograft-prosthetic composite. This patient had postoperative inflammation and resorption of the allograft suspicious for infection and was later found to have a nickel allergy. One year before the study visit, he underwent an exchange of the glenosphere and polyethylene socket and removal of the nickel cables.

873 Table IV Preoperative, postoperative, and preoperative to postoperative change comparison of strength and range of motion for primary total shoulder arthroplasty and reverse shoulder arthroplasty groups)

Shoulder arthroplasty strength, motion, and impairment

874 groups are summarized in Table III. The primary TSA group was significantly less impaired (P < .001) than the revision group at the preoperative and postoperative assessments. The primary RSA group demonstrated a very strong trend toward less impairment than the revision group at both time points, but this failed to reach statistical significance due to the modification of the level of statistical significance for multiple comparisons.

Effect of patient factors on postoperative impairment To understand what factors were predictive of postoperative impairment in these patients, we performed regression analyses. Age at surgery (P ¼ .941) and sex (P ¼ .472) were nonsignificant in the initial regression models and were excluded from subsequent analyses. The remaining variables with a P value <.10 were entered into the multiple variable model. Side of surgery (dominant vs nondominant hand) became nonsignificant (P ¼ .143) in this model and was also dropped. The final model indicated that revision surgery (b ¼ 4.32, P ¼ .001), RSA (b ¼ 3.69, P < .001), and preoperative impairment (b ¼ .40, P < .001) were all significantly associated with increased postoperative impairment. The Omnibus test for model fit indicated the model was a significant improvement over the interceptonly model (likelihood ratio c2 ¼ 75.9, P < .001).

Discussion In this age of rising health care costs, there is an increased emphasis on cost-effective intervention. Increased understanding of the ability of given surgical interventions to improve on the preoperative disease state is paramount in determining which procedures are worthwhile and potentially cost-effective. Persistent shoulder dysfunction is a common condition associated with high societal cost and patient burden.10 Thus, an improved understanding of preoperative shoulder disease severity as well as the degree to which a given procedure is able to effect a positive change in function should allow for a rational allocation of resources. The purpose of this study was to provide objective evidence regarding severity of end-stage degenerative shoulder disease, with or without an intact rotator cuff, and its response to shoulder arthroplasty. This is the first study comparing RSA and TSA outcomes to use an independent third party, blinded to study design and purpose, to collect data. This design was used to minimize measurement bias, thereby delivering more objective outcomes. The current study examined whether patients undergoing primary RSA had a higher preoperative and postoperative impairment rating compared with patients who were primary TSA candidates. We hypothesized that the RSA patients would exhibit an increased disease severity

B. Puskas et al. preoperatively and would fail to achieve similar reductions in disease severity after surgery compared with the TSA group. Our data demonstrate that primary TSA patients were less impaired postoperatively than primary RSA patients and that patients undergoing revision arthroplasties were more impaired than primary TSA patients. Primary RSA patients did not demonstrate a significant difference in impairment postoperatively compared with the revision group. We also examined whether there would be identifiable patient factors that have a clear effect on postoperative disease severity as measured by the impairment rating. Our regression analysis showed that a higher preoperative impairment rating and having undergone a previous shoulder arthroplasty surgery affected the postoperative impairment rating. This investigation provides evidence that, despite the current grouping within the same Medicare DRG, there are appreciable differences between these 3 patient populations in preoperative and postoperative impairment. The TSA patient population demonstrated less impairment at both time points than the RSA patient population, even when excluding revision RSAs. This is likely reflective of the underlying severity of pathology in the shoulder between these 2 patient populations. In addition, although patients in this study group were not injured workers requiring an impairment rating be provided to settle the legal dispute, this study provides information that may be helpful to the practitioner who is asked to provide information regarding anticipated final impairment ratings. The specific amount of resources required to successfully manage these distinct patient populations is currently unknown and is the focus of ongoing research. Various studies of health economics have tried to compare varying medical conditions using a quality of life model methodology that uses patient-reported measures. However, this introduces a subjective component that may prevent an accurate determination of the severity of a condition. Yet, our objective assessment demonstrates that there are significant differences in what these patients can functionally achieve. The way we use subjective and objective measurements of shoulder function is controversial. This study purposely did not include a subjective score, such as pain, because we wanted to evaluate how ROM, strength, and impairment measured the severity of a patient’s disease process. Most scores incorporate a subjective pain component in patientreported measures. As a result, it appears that these instruments are at risk to underestimate or overestimate a patient’s residual shoulder dysfunction compared with a stringent objective assessment of what a patient can actually achieve in strength and ROM. Recent studies have described the variance that exists between different patientreported outcome measures, as well as how objective physical examination findings may be both underutilized and undervalued.4,8,14 Despite the observed differences between the groups, all 3 groups were noted to experience similar reductions in

Shoulder arthroplasty strength, motion, and impairment impairment from their preoperative to postoperative state. Primary TSAs, primary RSAs, and revisions arthroplasties resulted in improvements in ROM, especially FF and abduction. Improvement in isometric strength was not consistent throughout the 3 groups. This demonstrates that these procedures are similarly effective for improving ROM for their respective end-stage forms of disease, despite more advanced muscular insufficiency in the RSA and revision arthroplasty patients. Although the groups experienced functional gains, the regression analysis demonstrated that type of surgery (primary TSA vs primary RSA) was an independent predictor of postoperative outcome, even after controlling for preoperative impairment. Patients undergoing a primary TSA can expect higher postoperative strength and ROM compared with those treated with a primary RSA. In addition, revision arthroplasty was indicative of a greater amount of postoperative impairment, as evidenced by decreased strength and ROM. Strengths of the current study include its prospective design, a long duration of follow-up, and the method of data collection. The use of blinded therapists and the Biodex dynamometer provides for an accurate depiction of strength and unique information that has not previously been published after shoulder arthroplasty. In addition, videorecorded ROM is a more reliable assessment of motion than surgeon-measured or estimated motion. For these reasons, we believe the information in the current study accurately represents the true functional capabilities of patients both before and after shoulder arthroplasty. Lastly, the use of a calculated impairment rating is unique because it allows for the representation of a large volume of objective outcome data in one easily interpreted value of shoulder function that can be rapidly compared between groups. However, the impairment method may actually minimize the amount of improvement observed after surgery. For example, several patients were noted as having improvement in strength beyond the published normal values at postoperative follow-up. This improvement in strength was therefore not captured because the best performance in the strength component is a zero deficit compared with normal. In addition, patients in a few instances were noted preoperatively as having strength greater than the published norms in one or more positions. In these cases, improvement could not be captured because they began with a strength deficit of zero. Other potential limitations of our methodology are not unique to our study but are common to all studies performing multiple comparisons and using objective strength and motion data. The adjusted P value of <.001 for statistical significance helps reduce the possibility of type I errors; however, several comparisons in Table II and III demonstrated large differences but failed to reach statistical significance. It is important to remember that the failure to reach statistical significance does not mean that

875 there are not meaningful differences on those measures. It is quite possible that these comparisons would also become statistically significant with larger sample sizes, and future studies should certainly continue to measure and report those results for this reason. Second, performance-based measures are generally effort-dependent. As a result, several precautions were established to diminish the effects of bias in patient effort. For example, no one in the treatment team was present during the testing, and the independent therapist minimized verbal cues during strength testing. ROM assessments are similarly subject to bias, as well as problems with interrater variability. As a result, ROM measurements were performed by an independent observer using videorecorded ROM. We lacked normal age- and sex-matched controls. However, we attempted to control for expected age- and sex-related strength changes by expressing strength deficit as a percentage of previously published normative data. As a result, we believe the persisting deficits observed after primary TSA and primary RSA are indicative of the residual effects of shoulder arthroplasty or chronic disease states on ultimate shoulder performance, as opposed to agerelated functional decline. Finally, because this investigation involved a single surgeon’s experience at one institution, the results may not be generalizable to all patients undergoing shoulder arthroplasty. Despite the acknowledged limitations, the current investigation adds evidence to the existing literature demonstrating the effectiveness of shoulder arthroplasty.

Conclusion The current study demonstrates that patients who are candidates for a revision shoulder arthroplasty or a primary RSA have a higher preoperative and postoperative impairment rating compared with patients who suffer from glenohumeral arthritis alone. However, primary TSA and primary RSA demonstrate comparable effectiveness in reducing a patient’s impairment while improving ROM and strength in their respective patient populations at an average of 4 years after shoulder arthroplasty.

Disclaimer This study was funded by an institutional research grant to the Foundation for Orthopaedic Research and Education from the Orthopaedic Research and Education Foundation. Dr. Mark Frankle is a paid consultant for and receives royalties from DJO. The other authors, their immediate families, and any research foundations with which they

876 are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.

References 1. Bryant D, Litchfield R, Sandow M, Gartsman GM, Guyatt G, Kirkley A. A comparison of pain, strength, range of motion, and functional outcomes after hemiarthroplasty and total shoulder arthroplasty in patients with osteoarthritis of the shoulder. A systematic review and meta-analysis. J Bone Joint Surg Am 2005;87:1947-56. http://dx.doi.org/10.2106/JBJS.D.02854 2. Cuff D, Pupello D, Virani N, Levy J, Frankle M. Reverse shoulder arthroplasty for the treatment of rotator cuff deficiency. J Bone Joint Surg Am 2008;90:1244-51. http://dx.doi.org/10.2106/JBJS.G.00775 3. 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;14:471-9. http://dx.doi.org/10. 1016/j.jse.2005.02.009 4. Doornberg JN, Ring D, Fabian LM, Malhotra L, Zurakowski D, Jupiter JB. Pain dominates measurements of elbow function and health status. J Bone Joint Surg Am 2005;87:1725-31. http://dx.doi.org/10. 2106/JBJS.D.02745 5. Florida Uniform Impairment Rating Schedule. Division of Workers’ Compensation as mandated by Rule 69L-7.604, Florida Administrative Code. Florida Department of Financial Services; 1996. p. 3-61. 6. Frankle M, Siegal S, Pupello D, Saleem A, Mighell M, Vasey M. The reverse shoulder prosthesis for glenohumeral arthritis associated with severe rotator cuff deficiency. A minimum two-year follow-up study of sixty patients. J Bone Joint Surg Am 2005;87:1697-705. http://dx.doi. org/10.2106/JBJS.D.02813 7. Hughes RE, Johnson ME, O’Driscoll SW, An KN. Age-related changes in normal isometric shoulder strength. Am J Sports Med 1999;27:651-7.

B. Puskas et al. 8. Lindenhovius AL, Buijze GA, Kloen P, Ring DC. Correspondence between perceived disability and objective physical impairment after elbow trauma. J Bone Joint Surg Am 2008;90:2090-7. http://dx.doi. org/10.2106/JBJS.G.00793 9. Matsen FA 3rd, Boileau P, Walch G, Gerber C, Bicknell RT. The reverse total shoulder arthroplasty. Instr Course Lect 2008;57:167-74. 10. Meislin RJ, Sperling JW, Stitik TP. Persistent shoulder pain: epidemiology, pathophysiology, and diagnosis. Am J Orthop (Belle Mead NJ) 2005;34(12 Suppl):5-9. 11. Mulieri P, Dunning P, Klein S, Pupello D, Frankle M. Reverse shoulder arthroplasty for the treatment of irreparable rotator cuff tear without glenohumeral arthritis. J Bone Joint Surg Am 2010;92:254456. http://dx.doi.org/10.2106/JBJS.I.00912 12. Norris TR, Iannotti JP. Functional outcome after shoulder arthroplasty for primary osteoarthritis: a multicenter study. J Shoulder Elbow Surg 2002;11:130-5. http://dx.doi.org/10.1067/mse.2002.121146 13. Smith AM, Barnes SA, Sperling JW, Farrell CM, Cummings JD, Cofield RH. Patient and physician-assessed shoulder function after arthroplasty. J Bone Joint Surg Am 2006;88:508-13. http://dx.doi.org/ 10.2106/JBJS.E.00132 14. Slobogean GP, Noonan VK, Famuyide A, O’Brien PJ. Does objective shoulder impairment explain patient-reported functional outcome? A study of proximal humerus fractures. J Shoulder Elbow Surg 2011; 20:267-72. http://dx.doi.org/10.1016/j.jse.2010.06.005 15. Torchia ME, Cofield RH, Settergren CR. Total shoulder arthroplasty with the Neer prosthesis: long-term results. J Shoulder Elbow Surg 1997;6:495-505. 16. Wall B, Nove-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;89:1476-85. http://dx.doi.org/10. 2106/JBJS.F.00666 17. Werner CM, Steinmann PA, Gilbart M, Gerber C. Treatment of painful pseudoparesis due to irreparable rotator cuff dysfunction with the Delta III reverse-ball-and-socket total shoulder prosthesis. J Bone Joint Surg Am 2005;87:1476-86. http://dx.doi.org/10.2106/JBJS.D. 02342 18. Wiater JM, Fabing MH. Shoulder arthroplasty: prosthetic options and indications. J Am Acad Orthop Surg 2009;17:415-25.