Background: The number of total shoulder arthroplasties performed in the United States increased slightly between 1990 and 2000. However, the incidence of shoulder arthroplasty in recent years has not been well described. The purpose of the present study was to examine recent trends in shoulder hemiarthroplasty and total shoulder arthroplasty along with the common reasons for these surgical procedures in the United States.
Methods: We modeled the incidence of shoulder arthroplasty from 1993 to 2008 with use of the Nationwide Inpatient Sample. On the basis of hemiarthroplasty and total shoulder arthroplasty cases that were identified with use of surgical procedure codes, we conducted a design-based analysis to calculate national estimates.
Results: While the annual number of hemiarthroplasties grew steadily, the number of total shoulder arthroplasties showed a discontinuous jump (p < 0.01) in 2004 and increased with a steeper linear slope (p < 0.01) since then. As a result, more total shoulder arthroplasties than hemiarthroplasties have been performed annually since 2006. Approximately 27,000 total shoulder arthroplasties and 20,000 hemiarthroplasties were performed in 2008. More than two-thirds of total shoulder arthroplasties were performed in adults with an age of sixty-five years or more. Osteoarthritis was the primary diagnosis for 43% of hemiarthroplasties and 77% of total shoulder arthroplasties in 2008, with fracture of the humerus as the next most common primary diagnosis leading to hemiarthroplasty.
Conclusions: The number of shoulder arthroplasties, particularly total shoulder arthroplasties, is growing faster than ever. The use of reverse total arthroplasty, which was approved by the United States Food and Drug Administration in November 2003, may be part of the reason for the greater increase in the number of total shoulder arthroplasties. A long-term follow-up study is warranted to evaluate total shoulder arthroplasty in terms of patient outcomes, safety, and implant longevity.
1Department of Orthopaedic Surgery, University of California at Davis, 4860 Y Street, Suite 3800, Sacramento, CA 95817. E-mail address: firstname.lastname@example.org
2Musculoskeletal Diseases of Aging Research Group, 4625 Second Avenue, Suite 1002, Sacramento, CA 95817.
3Clinical Epidemiology Research and Training Unit, Boston University, 650 Albany Street, x-215, Boston, MA 02118.
4Department of Orthopaedic Surgery, University of California at Davis, 4860 Y Street, Suite 3800, Sacramento, CA 95817. E-mail address: email@example.com
Shoulder arthritis is common among older adults and has a substantial impact on function and well-being1,2. When nonoperative treatment of arthritis of the glenohumeral joint fails to relieve pain or improve function, either hemiarthroplasty or total shoulder arthroplasty can be performed, depending on surgeon discretion. There is a paucity of data on the incidence of shoulder arthroplasty in the United States (U.S.). A previous study demonstrated a minor increase in the rate of total shoulder arthroplasty from 1990 to 20003. Another study demonstrated that the annual number of procedures increased by 10.6% for total shoulder arthroplasty and 6.7% for hemiarthroplasty between 1993 and 2007 and projected future demand on the basis of exponential growth assumption4. However, the actual pattern of shoulder arthroplasty growth has not been reported.
Since the first documented shoulder arthroplasty in 18935, there have been many advances in shoulder implants, particularly within the last two decades. Modern shoulder implants include both anatomic and reverse shoulder replacements6. We suspected that advances in implants and surgical techniques might have resulted in increased use of shoulder arthroplasty. The objective of the present study was to evaluate the recent trends in the utilization of hemiarthroplasty and total shoulder arthroplasty along with the common reasons for these surgical procedures in the U.S. The present study represented the first step in documenting the need for a long-term follow-up study to evaluate patient outcomes and implant longevity following total shoulder arthroplasty.
Materials and Methods
We studied temporal trends in the utilization of hemiarthroplasty and total shoulder arthroplasty from 1993 to 2008 using the Nationwide Inpatient Sample (NIS). Our institutional review board (IRB) determined that this study did not involve human subject research and that no IRB review was needed.
The NIS, part of the Healthcare Cost and Utilization Project (HCUP), is a survey of hospital inpatient stays conducted annually under the auspices of the Agency for Healthcare Research and Quality (AHRQ). The NIS is the principal source for national data concerning the characteristics of patients discharged from nonfederal, short-stay hospitals7. In 2008, probability samples of 1000 hospitals (∼20% of U.S. community hospitals) were selected to produce a nationally representative sample. All discharge records in the selected hospitals were included in the sample. Approximately five million to eight million discharge records were selected from each year. Details concerning the sampling procedures have been published elsewhere8.
Procedure and Diagnosis Codes
We identified shoulder arthroplasty cases with use of the procedure codes for hemiarthroplasty (81.81) and total shoulder arthroplasty (81.80) from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Both conventional total shoulder arthroplasties and reverse total shoulder arthroplasties had the same ICD-9-CM code during the period of interest (1993 to 2008). The NIS database does not include Current Procedural Terminology (CPT) codes. We are not aware of any CPT or ICD-9-CM codes specific to the resurfacing of the shoulder. Therefore, hemiarthroplasty includes resurfacing of the humeral head, and total shoulder arthroplasty includes resurfacing of the humeral head as well as the glenoid. To ascertain the indications for hemiarthroplasty and total shoulder arthroplasty, we analyzed the primary ICD-9-CM diagnosis codes for the cases identified. We further evaluated the most common comorbid conditions on the basis of secondary diagnosis fields.
We used HCUPnet, a web-based interactive analytical tool, to identify the number of shoulder arthroplasties from 1993 to 20089. Because the NIS is a probabilistic sample survey, the sampling weight and sampling design were considered during estimations of the total numbers of cases and their standard errors. Sampling weights were used to account for the unequal sampling probabilities and to produce estimates for all hospital admissions in the U.S. The standard error was estimated with Taylor linearization. We modeled different linear trends in different time periods with use of piecewise linear regression. Piecewise regression models, also known as broken-stick models, can be used where two or more trend lines are joined at the breakpoints10. To determine the rates of change among different age groups, we calculated the incidence of shoulder arthroplasty for every 100,000 people in each respective age group according to the U.S. Census11.
Source of Funding
There was no external funding for this study.
Shoulder Arthroplasties in the U.S.
Over the past decade, the number of shoulder arthroplasties performed in the U.S. has increased rapidly. Nearly 47,000 shoulder arthroplasties were performed as inpatient procedures in 2008 (Table I). In comparison, only 19,000 shoulder arthroplasties were performed in 1998, representing a 2.5-fold increase over the past decade. The annual number of hemiarthroplasties also grew steadily, but the increase in total shoulder arthroplasties was particularly steep. As a result, the number of total shoulder arthroplasties now exceeds that of hemiarthroplasties in the U.S. As indicated by nonoverlapping confidence intervals, there were significantly more total shoulder arthroplasties than hemiarthroplasties in 2008 (Table I).
Because the number of total shoulder arthroplasties performed jumped discontinuously between 2003 and 2004, we fitted the trend with use of the piecewise linear regression function (R2 = 99%). The number of arthroplasties increased by 373 per year until 2003 (Fig. 1, dashed line). Since 2004, however, the number of shoulder arthroplasties increased by 2922 each year. This difference in rates (or slopes) was significant (p < 0.01). Furthermore, compared with the number of total shoulder arthroplasties performed in 2003, the discontinuous jump (∼4500 cases) in 2004 was significant (p < 0.01).
Annually, more than two-thirds of shoulder arthroplasties were performed for elderly patients (patients who were sixty-five years of age or older). More total shoulder arthroplasties than hemiarthroplasties were performed for elderly patients in more recent years. Taking into account the population size of the respective age groups11, age-specific shoulder arthroplasty rates increased for all age groups for both hemiarthroplasty and total shoulder arthroplasty (Table II). Among the elderly, the increase in the number of total shoulder arthroplasties was particularly evident: between 2000 and 2008, the rate of total shoulder arthroplasty increased 3.5-fold (from fourteen to forty-nine per 100,000 people), whereas the rate of hemiarthroplasty increased 1.5-fold (from twenty-three to thirty-three per 100,000 people). Women underwent hemiarthroplasty and total shoulder arthroplasty more frequently than men even after adjusting for population size (Table II).
Most Frequent Diagnoses
Osteoarthrosis was the most common reason for hemiarthroplasty, constituting 43% of cases (Table III). The next most common reason for hemiarthroplasty was fracture of the proximal end of the humerus, constituting 33% of cases. Osteoarthrosis was also the primary condition leading to total shoulder arthroplasty, accounting for 77% of cases, followed by unspecified arthropathy (most likely for rotator cuff arthropathy), disorders of the bursae and tendons in the shoulder, and rheumatoid arthritis.
The annual number of shoulder arthroplasties in the U.S. increased 2.5-fold between 2000 and 2008. An expanding elderly population (sixty-five years of age or older) may have contributed to the increase. During the same time period, the elderly population increased approximately 11% (from 35.1 million to 38.8 million)11. The number of companies selling shoulder arthroplasty implants also increased, with no fewer than eight manufacturers currently in the U.S. In addition to medical device advances that have made complicated surgical procedures possible, the number of shoulder and elbow surgeons also increased. According to the American Academy of Orthopaedic Surgeons (AAOS), 5392 more orthopaedic surgeons were in practice in 2008 as compared with 2000 (21,658 compared with 16,266). In terms of surgeon density, this change represents a 24% increase (5.8 compared with 7.2 orthopaedic surgeons per 100,000 people)12,13. As demonstrated in the Dartmouth Atlas of Musculoskeletal Health Care14, the density of surgeons bears a very close relationship with the number of surgical procedures in a given area, implying that physician density may drive the number of procedures performed. However, the increase in case volume was much steeper than the growth of the population or the density of orthopaedic surgeons in the U.S. during the same time period, suggesting that multiple factors were responsible for the steep rise in shoulder surgery. In addition, changing attitudes among surgeons and patients also may contribute to the increasing number of shoulder arthroplasty procedures. In anticipation of faster pain relief, patients may preferentially select the surgical option before attempting all available nonsurgical treatments.
Our study revealed a steady and continuous increase in hemiarthroplasty but an abrupt increase in total shoulder arthroplasty. Our findings were consistent with those of a previous study that described a minor increase in the rate of total shoulder arthroplasty between 1990 and 20003. Interestingly, we found a sudden, steep increase in total shoulder arthroplasty between 2003 and 2004, and the number of total shoulder arthroplasties has increased steeply since then. The choice of hemiarthroplasty or total shoulder arthroplasty is determined in part by surgeon preference in addition to diagnostic indications. Given that the number of hemiarthroplasties was growing steadily without an abrupt rate increase during the same time period, the sudden acceleration in total shoulder arthroplasty likely was due to more than a growing acceptance of shoulder arthroplasty. We speculated that the dramatic increase and acceleration in the number of total shoulder arthroplasties since 2004 was possibly due to the Food and Drug Administration (FDA) approval of a reverse total shoulder arthroplasty device in the U.S. in November 2003.
Reverse total shoulder arthroplasty is primarily indicated for patients with rotator cuff arthropathy for which conventional total shoulder arthroplasty is not suitable. Industry reports have demonstrated a dramatically increasing trend in reverse shoulder device sales15,16. An implant market analysis revealed that approximately 2000 reverse total shoulder arthroplasties were performed in the U.S. in 2004, compared with nearly 10,000 in 200716. Moreover, the same market analysis predicted that 15,000 and 30,000 reverse total shoulder arthroplasties would be performed in 2008 and 2012, respectively16. Unfortunately, we were not able to identify from the NIS database the proportion of total shoulder arthroplasties that involved the use of this new device because total shoulder arthroplasty and reverse total shoulder arthroplasty were coded under a single ICD-9-CM number during the period of interest. The Centers for Disease Control and Prevention and the Centers for Medicare & Medicaid Services issued a new code, 81.88, for reverse total shoulder arthroplasty that went into use beginning in October 201017. Future study is warranted to better understand the utilization of reverse total shoulder arthroplasty in addition to its performance and safety.
Compared with hip and knee arthroplasty, shoulder arthroplasty was less frequently performed for nonelderly adults. Approximately 30% of patients who underwent shoulder arthroplasty were nonelderly, compared with nearly 40% of patients who underwent hip or knee arthroplasty18. Additionally, hip and knee arthroplasty are increasing faster among the middle-aged population (forty-five to sixty-four years old) than among the elderly18, a phenomenon that was not observed for total shoulder arthroplasty. Instead, the primary increase in total shoulder arthroplasty was among the elderly. The introduction of reverse total shoulder arthroplasty is one possible explanation for the predominance of elderly patients among those who were managed with total shoulder arthroplasty. Currently, reverse total shoulder arthroplasty is recommended predominantly for patients over the age of seventy years who have disabling rotator cuff arthropathy19. Thus, one can postulate that the acceleration of total shoulder arthroplasty that we observed, primarily in the elderly, may represent a natural correction in the distribution of procedures related to new opportunities to address a previously unmet need. Reverse total shoulder arthroplasty initially was introduced to treat rotator cuff arthropathy, but it now is used to treat a wider scope of problems, such as revision of failed total shoulder arthroplasty or hemiarthroplasty20. Therefore, the continued increase might be, in part, due to the broadening indications for reverse total shoulder arthroplasty. Recent advances in reverse total shoulder arthroplasty have minimized early mechanical failure for patients with rotator cuff deficiency21. The authors of previous survival studies cautioned about the use of reverse total shoulder arthroplasty in younger patients19, but technical advances warrant further studies.
The strength of the present study was the large sample size of the NIS data, which constitutes nearly 20% of all U.S. hospital admissions each year. Despite the use of a probabilistic sampling technique, the large sample size yielded a reasonably small standard error. The primary weakness of this study was the ambiguity of the coding system. The fact that the ICD-9-CM code was the same for both conventional total shoulder arthroplasty and reverse total shoulder arthroplasty did not allow us to characterize the trends and conditions associated with the individual procedures. As noted, the Centers for Disease Control and Prevention and the Centers for Medicare & Medicaid Services issued a new code, 81.88, for reverse total shoulder arthroplasty that went into use in October 201017. Because of the large scale of data collection, NIS data are only available to researchers approximately two years after the referenced year. Therefore, 2011 data are expected to be available to researchers in 2013. To date, there are no separate codes specific to resurfacing of the shoulder, limiting our ability to explore interesting questions related to this device. Although the NIS utilized a multistage probabilistic sampling technique to produce a nationally representative sample8, all sample studies are subject to sampling errors; therefore, the findings of our study should be interpreted with caution. Additionally, we were only able to report the incidence (new cases in a given time period) of shoulder arthroplasty. Because an arthroplasty registry does not currently exist at the national level in the U.S., the prevalence (all existing cases) of shoulder arthroplasty remains unknown.
In conclusion, the number of shoulder arthroplasties performed in the U.S. accelerated rapidly between 2000 and 2008. This finding may be partly explained by the approval of reverse total shoulder arthroplasty in November 2003, allowing for surgical procedures in the elderly who were not previously eligible for shoulder arthroplasty. The rate of hemiarthroplasty also increased during this period, but at a slower rate, consistent with prior growth trends. Given the disability associated with shoulder pathology, particularly in the elderly, further study focusing on patterns of disease, indications for surgery, safety and implant longevity, and outcomes of shoulder arthroplasty is warranted.
NOTE: Authors thank Dr. Betty Guo at University of California at Davis Clinical and Translational Science Center for her editorial advice.
Investigation performed at the University of California at Davis, Sacramento, California
1. Hill CL Gill TK Shanahan EM Taylor AW. Prevalence and correlates of shoulder pain and stiffness in a population-based study: the North West Adelaide Health Study. Int J Rheum Dis. 2010;13:215–22.
2. Vogt MT Simonsick EM Harris TB Nevitt MC Kang JD Rubin SM Kritchevsky SB Newman AB; Health, Aging and Body Composition Study. Neck and shoulder pain in 70- to 79-year-old men and women: findings from the Health, Aging and Body Composition Study. Spine J. 2003;3:435–41.
3. 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;55:591–7.
4. Day JS Lau E Ong KL Williams GR Ramsey ML Kurtz SM. Prevalence and projections of total shoulder and elbow arthroplasty in the United States to 2015. J Shoulder Elbow Surg. 2010;19:1115–20.
5. Bankes MJ Emery RJ. Pioneers of shoulder replacement: Themistocles Gluck and Jules Emile Péan. J Shoulder Elbow Surg. 1995;4:259–62.
6. Flatow EL Harrison AK. A history of reverse total shoulder arthroplasty. Clin Orthop Relat Res. 2011;469:2432–9.
9. Agency for Health Research and Quality. HCUPnet, a tool for identifying, tracking, and analyzing national hospital statistics. http://hcupnet.ahrq.gov/
. Accessed 2010 Oct 27.
10. Montgomery D Peck E Vining G. Introduction to linear regression analysis. Hoboken, New Jersey: Wiley Interscience; 2006.
14. The Center of the Evaluative Clinical Sciences. Dartmouth Medical School. The Dartmouth Atlas of Musculoskeletal Health Care. Chicago: American Hospital Association; 2000.
18. Kim S. Changes in surgical loads and economic burden of hip and knee replacements in the US: 1997-2004. Arthritis Rheum. 2008;59:481–8.
19. Guery J Favard L Sirveaux F Oudet D Mole D Walch G. Reverse total shoulder arthroplasty. Survivorship analysis of eighty replacements followed for five to ten years. J Bone Joint Surg Am. 2006;88:1742–7.
20. Drake GN O'Connor DP Edwards TB. Indications for reverse total shoulder arthroplasty in rotator cuff disease. Clin Orthop Relat Res. 2010;468:1526–33.
21. 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.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.