Epidemiology of musculoskeletal upper extremity ambulatory surgery in the United States

Musculoskeletal disorders of the upper extremity are common reasons for patients to seek care and undergo ambulatory surgery [1‐3]. Ambulatory surgery was introduced because it generally costs less than inpatient surgery and because technological advancements have enabled a safe transition from the in-patient setting [4, 5]. The number of procedures performed in ambulatory surgery centers increased from 380,000 in 1983 to 31.5 million in 1996 and to 57.1 million in 2006 [4, 6, 7]. It was recently reported that 272,148 rotator cuff repair and 577,000 carpal tunnel release procedures were performed in the United States in 2006 on an ambulatory basis [1, 2]. The annual national utilization of shoulder arthroscopy performed for indications other than rotator cuff tear and wrist arthroscopy performed for indications other than carpal tunnel syndrome have not been previously reported.

The National Survey of Ambulatory Surgery (NSAS) is a Center for Disease Control and Prevention (CDC) based population sample survey of ambulatory surgery procedures performed in the United States [8]. The NSAS databases were assembled in 1994–96 and not again until 2006. Thus the 2006 NSAS databases permit contemporary estimates of upper extremity orthopedic surgery. We sought to compare utilization, demographics, indications, and operating room time for most commonly performed ambulatory procedures of the upper extremity. This data is helpful in comparing commonly performed upper extremity orthopedic procedures and may also assist policy makers and hospitals in projecting future needs to accommodate upper extremity procedural volume.

We assessed the utilization of select upper extremity ambulatory procedures in the United States using a national database, NSAS, for the year 2006. We also studied age-stratified utilization rates, indications, and characteristics of patients undergoing these procedures. We found that over one million upper extremity ambulatory procedures of interest were performed. In general middle-aged adults in the age-group 45–64 years had the highest rate of these procedures, and with the exception of carpal tunnel release which had the highest rate in those 75 and older. Private or commercial insurance and Worker’s Compensation were the primary payers in a majority of cases. The most common indications for shoulder arthroscopy were impingement syndrome, periarthritis, and instability/SLAP tears whereas those for wrist arthroscopy were ligament sprains and cartilage disorders.

By way of comparison, in 2006, meniscal procedures of the knee (excision of semilunar cartilage) were performed on an estimated 690,000 cases [6] on an ambulatory basis. In the same year, 984,607 knee arthroscopies were performed [20]. Hence, the resources utilized by upper extremity procedures in the ambulatory setting are substantial and comparable to the most frequently performed orthopedic procedures in the ambulatory setting.

Our data only capture upper extremity procedures performed on an ambulatory basis. Data from the National Hospital Discharge Survey presented by the American Academy of Orthopaedic Surgeons estimated that in 2006 approximately 20,000 rotator cuff repairs were performed on an in-patient basis [21]. Thus, our estimate of 272,148 rotator cuff repairs per year may understate the true frequency by 7%. Similarly, an estimated 9,231 carpal tunnel release procedures were performed in 2006 as per data from the Nationwide Inpatient Samples [2]. Thus, our study likely underestimates the true frequency of carpal tunnel release by 1.6%. We are not aware of data on shoulder arthroscopy performed on ambulatory or non-ambulatory basis. However, a recent study using data from the American Board of Orthopaedic Surgery (ABOS) reported that 87.7% of Bankart repairs in 2006–2008 were performed arthroscopically [22], most likely in an ambulatory setting. Thus, it is likely that the majority of shoulder arthroscopies were captured by the ambulatory database used in our study.

As expected, the use of these upper extremity procedures varied across demographic groups. Shoulder arthroscopy was performed in 15–44 year olds at a higher rate as compared with rotator cuff repairs and wrist arthroscopy. This is likely because instability and SLAP lesions are most frequently operated on in young patients and in those who participate in organized sports due to a high rate of recurrent dislocation [23, 24]. In our study, carpal tunnel release was most commonly performed in post-menopausal women consistent with several studies documenting the incidence of carpal tunnel syndrome to be highest in post-menopausal women [13, 14]. In an Italian population, Mondelli et al. reported an incidence rate of carpal tunnel syndrome of 506 per 100,000 person-years for women, with peak incidence in women between 50–59 years of age [25]. Several reasons such as hormone replacement therapy [26, 27], endogenous hormonal factors [27], presence of menstrual disorders [27], and increased expression of estrogen receptors in the tenosynovium [28] have been postulated to explain these findings. Workers Compensation was a primary payer in 13%-37% of upper extremity ambulatory cases. These work-related injuries were likely due to a traumatic event or from repetitive motion [29].

Shoulder arthroscopy was most commonly performed for a diagnosis of instability, SLAP lesions, and other disorders with an ICD-9 diagnosis code of 726.2. The increasing utilization of shoulder arthroscopy for management of instability [22] has likely enabled treatment of such patients on an ambulatory basis. As per communication with certified coders specializing in shoulder surgery from our institution [30], other disorders that are usually coded using the ICD-9 code 726.2 include impingement syndrome, periarthritis of the shoulder region, and scapulothoracic bursitis (likely subacromial/subdeltoid bursitis leading to subacromial decompression). Patients undergoing shoulder arthroscopy for reasons other than cuff repairs were younger, on average, than patients undergoing arthroscopic cuff repairs. This is expected since arthroscopic treatment for instability and SLAP lesions is usually performed in a younger population. Weber et al. reported that the mean age of male patients undergoing SLAP repairs was 37 years and that of female patients was 40.9 years based on data self-reported by candidates admitted to Part II of the American Board of Orthopedic Surgery Examination [31]. Despite the lack of evidence on treatment of SLAP lesions, Zhang et al. reported a steady increase in SLAP repairs from 2004–2009 in the United States based on billing records from several insurance companies [32]. Wrist arthroscopy was performed in a relatively smaller number of patients and most common indications included sprains and strains, wrist pain, and articular cartilage disorders. These diagnoses likely include sprain/tear of triangular fibrocartilage complex (TFCC), intercarpal ligament injuries, and synovitis that are common indications for wrist arthroscopy [33, 34]. We cannot assess the prevalence of these specific indications since a specific ICD-9-CM diagnosis code is not assigned to these disorders. The high utilization of diagnostic wrist arthroscopies for pain and articular cartilage disorders is likely because MRI/MRA imaging correlation for wrist disorders is still not optimal [35]. TFCC injuries are one of the most common causes of wrist pain. In a specialty practice, Park et al. reported that 43% of patients with ulnar-sided wrist pain required arthroscopic intervention and all of these patients had a TFCC injury on intra-operative assessment [34]. Thus, a large proportion of patients with TFCC injuries require arthroscopic intervention and this likely comprises a substantial proportion of patients undergoing wrist arthroscopy in our study. Our data shows that as total operating room time increases, the ratio of actual surgical time (time from first incision until wound closure) to operating room time also increases. This could be interpreted as an increase in efficiency in the use of total operating time, with rotator cuff repair being the most efficient procedure by this criterion.

It is estimated that 40% of persons over the age of 50 years have cuff tears on imaging as compared with 54% in those >60 years, and 65% in persons over 70 years [36, 37]. Thus, the prevalence of rotator cuff tears on imaging and cadaveric studies rises with increasing age. The prevalence of symptomatic rotator cuff tears across age is unknown. Our results show that the rate of rotator cuff repairs was highest in 45–74 year olds and declined in patients 75 years and older. It is possible that concerns about post-surgical cuff healing and rehabilitation potential of older patients discourages surgeons from recommending repair in these older persons [38, 39]. It is also possible that tears are more symptomatic in younger persons who are traditionally more active. Moreover, patients in the older age-groups are likely not working, may perform less strenuous activities and may be less likely to opt for surgery. Further analysis of indications for surgery across age strata requires investigation.

The limitations of our study include the lack of data from Federal, military, and Department of Veteran’s Affairs hospitals. Therefore, our estimates are likely conservative. It is possible that some procedures were revision surgeries. This information cannot be ascertained from the database. Strengths of our data include a relatively large sample size and the ability to calculate national estimates but our data source lacks more detailed clinical information such as disease severity and patient outcomes.

Rotator cuff repair, shoulder arthroscopy, carpal tunnel release, and wrist arthroscopy were performed an estimated 272,148 times, 257,541 times, 576,924 times, and 3,686 times, respectively, in the United States in the year 2006. The outcomes of rotator cuff repair and of non-operative management of these lesions in older patients needs further evaluation so that informed decisions about surgical intervention are made. The clinical and demographic predictors of the outcomes of many of these upper extremity ambulatory surgeries also remain ill-defined despite the substantial healthcare resource utilization associated with these procedures.