Incidence of Surgically Treated Brachial Plexus Injury in Privately Insured Adults Under 65 Years of Age in the USA

Abstract

Background

Brachial plexus injury (BPI) is a life-altering event, and surgical treatment of BPI is costly ($66 million nationwide in 2006 alone). Despite the sizable costs associated with surgically treated BPI, the incidence of such cases in adults in the USA remains unknown.

Questions/Purposes

We sought to provide an estimate of the national incidence of surgically treated BPI in the USA and to determine whether changes have occurred over time.

Methods

We sought to identify privately insured adults ages 18 through 64 years who underwent BPI surgery between 2008 and 2014 in the IBM® MarketScan® Commercial Database. We then extrapolated those data to the same age group in the general population (all insurance types were considered), using data from the National Health Interview Survey and a study on the proportion of BPI surgery cases covered by private insurance to estimate the national incidence. We also used the Cochran–Armitage trend test to evaluate for statistically significant changes in BPI incidence over time in MarketScan private insurance data.

Results

We identified 966 patients surgically treated for BPI in the MarketScan database in the study period. The average annual incidence of surgically treated BPI in this population of privately insured patients was 0.64 per 100,000 (range, 0.38 to 1.03). When extrapolated to all payers, the average annual incidence was 0.89 per 100,000 people (range, 0.53 to 1.47). In the MarketScan data, there was a statistically significant overall increase in the incidence of surgically treated BPI from 2008 through 2014 (0.47 per 100,000 to 1.03 per 100,000).

Conclusion

The estimated annual incidence of surgically treated BPI is 0.89 per 100,000 persons in the USA, and the incidence has increased over time. These findings can serve as a baseline for further studies of disease epidemiology and societal and financial impact.

Electronic supplementary material

The online version of this article (10.1007/s11420-019-09741-8) contains supplementary material, which is available to authorized users.

Introduction

Traumatic brachial plexus injury (BPI) is a major, life-altering event. Tremendous resources are used to restore function and alleviate pain in patients whose injuries warrant surgical reconstruction. An analysis of national administrative data from 1993 through 2006 revealed that the hospital charges associated with treating BPI were nearly $66 million in 2006 alone [14]. Despite such sizable costs, the incidence of adult traumatic BPI in the USA remains unknown. The publication of large case series from single US centers suggests that BPI has a broad impact [1, 3, 4, 6, 13, 15, 17–20], but a national estimate of incidence can more accurately reflect the burden of disease and guide the allocation of resources for clinical care and research. The goal of our study was to use administrative claims data to estimate the incidence of surgically treated BPI in adults under 65 years of age in the USA and to determine whether changes have occurred over time.

Methods

We used an administrative claims database of commercially insured patients to identify adults ages 18 to 64 years who underwent BPI surgery between 2008 and 2014. Data were abstracted from the IBM® MarketScan® Commercial Database. The MarketScan database comprises medical and outpatient pharmacy claims and enrollment information contributed by employers and health plans for more than 130 million patients. Individuals in the database include employees, their spouses, and dependents covered by employer-sponsored private health insurance. The MarketScan database is a limited dataset, and this study was therefore considered exempt from review by our institution’s human research protection office.

Patients with BPI were identified from the inpatient and outpatient medical claims using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes 953.4 (injury to the brachial plexus) and 353.0 (brachial plexus lesions). BPI surgery was identified using Current Procedural Terminology, 4th Edition (CPT-4) codes 64861 and 64713 (specific to BPI surgery) or any of CPT-4 codes 64708, 64856, 64857, 64859, 64872, 64874, 64876, 64892, 64893, 64897, 64898, 64901, 64902, 64905, or 64907 plus one of the BPI diagnosis codes listed on the same service date as the surgery. If a patient had multiple BPI surgery dates, the first was used as the index BPI procedure.

Statistical Analysis

Incidence estimates of surgically treated BPI were calculated using as the denominator the total number of adult beneficiaries ages 18 to 64 years who were continuously enrolled in health plans for at least 2 years.

The rate, expressed as cases of surgically treated BPI per 100,000 employer-insured people, was then extrapolated to the general population (all insurance types) to estimate the incidence in all patients in the USA between ages 18 and 64 years. To calculate the rate, we used the total proportion of people included in the MarketScan dataset, the percentage of the country’s population between ages 18 to 64 years who were covered by private insurance according to the National Health Interview Survey from the Centers for Disease Control and Prevention [2], and a study published by some of us (C.J.D., M.A.O.) on the proportion of BPI surgery cases in several states that was covered by private insurance (47%) [5]. (This estimate captures only individuals covered by private insurance and does not include those covered by Medicaid, Medicare, or workers’ compensation or those who do not have health insurance.) The Cochran–Armitage trend test was used to assess for statistically significant changes in BPI incidence over time in the primary data (MarketScan private insurance data).

Results

Out of 46,006,703 patients ages 18 to 64 years, we identified 966 patients surgically treated for BPI over 7 years (2008 through 2014) in the MarketScan database who were continuously enrolled in health plans for at least 2 years. The average incidence of surgically treated BPI in the database during that time was 0.64 per 100,000 (range, 0.38 to 1.03 per 100,000) privately insured patients ages 18 to 64; when extrapolated to all payers, the average incidence from 2008 through 2014 was 0.89 per 100,000 population ages 18 to 64 (range, 0.53 to 1.47 per 100,000).

Among the privately insured patients in MarketScan, there was an overall increase in the incidence of surgically treated BPI from 2008 (0.47 per 100,000) through 2014 (1.03 per 100,000) (Fig. 1), which was statistically significant when evaluated using the Cochran–Armitage trend test (p < 0.0001).

Fig. 1.

Annual incidence of surgically treated brachial plexus injury per 100,000 privately insured adults ages 18 to 64 in the USA. Data from the MarketScan database and extrapolated to the general population.

Discussion

After extrapolating the results of our analysis of claims in the MarketScan database to the country as a whole, we estimate the annual incidence of surgically treated BPI in the USA to be 0.89 per 100,000 population ages 18 to 64 years. Two papers in the literature have estimated rates in other countries (the UK and Brazil, respectively); two others have come closest to estimating an incidence of BPI in the USA (Table 1). Lad et al. presented an average number of hospital discharges with BPI listed as a diagnosis over a 14-year period in the USA but did not estimate the disease incidence [14]. Furthermore, the authors identified cases from the Nationwide Inpatient Samples by diagnosis code, rather than a BPI-specific procedure code or diagnosis code with a procedure code, which likely decreases the accuracy of their BPI case definition [7, 8]. Midha found that BPI was present in 1.2% of all multitrauma patients presenting at a single Canadian trauma center over a 9-year period [16]. He also examined the associations between BPI and other traumatic injuries, including closed head injury and spinal cord injury, and extrapolated those proportions to earlier population-based incidences, allowing him to estimate a range of the incidence of BPI in the USA: 0.64 to 3.9 cases per 100,000 population [12]. In our cohort assembly, we attempted to address the shortcomings of these earlier studies by requiring a BPI-surgery code or a combination of a procedure code and a BPI-diagnosis code, which is most likely to accurately reflect patients undergoing BPI reconstruction. Because we required a procedure code in order to increase the validity of our cohort, we were unable to estimate the incidence of patients with BPI treated non-operatively. However, the size and scope of the MarketScan Commercial Claims and Encounters dataset were substantial enough to provide an estimate of the national incidence of BPI treated surgically.

Table 1.

Summary of published papers examining the incidence of BPI

Author(s); title; year of publication	Study design	Estimate	Notes
Goldie BS, Coates CJ. Brachial plexus injury: a survey of incidence and referral pattern. 1992. [11]	Survey of UK orthopedic consultants and chart review at two referral centers	Estimate of 336 traumatic BPI cases in a single year (1986) in the UK	Validity and generalizability limited by study design.
Midha M. Epidemiology of brachial plexus injuries in a multitrauma population. 1997. [16]	Chart review of consecutive multitrauma cases over 9 years at a Canadian trauma center	1.2% of multitrauma patients (1986 to 1994)	Only captures BPI that presented through trauma or emergency departments.
Flores LP. Epidemiological study of the traumatic brachial plexus injuries in adults. 2006. [9]	Prospective assessment of 35 consecutively treated BPI patients in Brazil	1.75 per 100,000 local population per year (2004 to 2005)	Does not account for those who received care in private hospitals.
Lad SP, et al. Trends in median, ulnar, radial, and brachioplexus nerve injuries in the United States. 2010. [14]	Analysis of US administrative database (Nationwide Inpatient Samples)	Average of 2162 BPI discharges per year (1993 to 2006)	Diagnosis codes only (treatment codes not included for BPI). Only captures hospital-based (inpatient) care. Number of BPI-related discharges per year not published.
Current study	Analysis of national commercial insurance database (MarketScan), extrapolated to national population	Average annual incidence of surgically treated BPI is 0.89 per 100,000 people ages 18 to 64 years (2008 through 2014)	National all-payer incidence estimate was extrapolated using previously published payer mix for BPI patients.

BPI brachial plexus injury, UK United Kingdom

Our study is limited by the use of administrative claims data. As mentioned above, we used strict criteria combining diagnostic and procedure codes to assemble a cohort of patients who had undergone surgery for BPI, but we do not have access to clinical records, operative notes, or imaging to confirm the diagnoses. An additional limitation is that our dataset is composed entirely of privately insured beneficiaries and does not include those covered by Medicaid, Medicare, or workers’ compensation or those who do not have health insurance. We derived a national estimate by applying the payer mix of BPI patients reported in our prior work (statewide administrative databases for Florida, New York, and California from the Healthcare Cost and Utilization Project), in which the proportion of BPI surgeries paid by private insurers was 47%; this is similar to that reported by Lad et al. (43%) using the Nationwide Inpatient Samples [5, 14]. The proportions paid by Medicaid (Lad, 16%; Dy, 12%) and Medicare (Lad, 16%; Dy, 18%) were also similar between those studies, supporting the assertion that this payer mix is representative of the country as a whole. We accepted this limitation because without a national all-payer database [10], there is no other feasible way to derive a national estimate of the incidence of traumatic BPI.

There was a statistically significant increase in surgically treated BPI in the USA from 2008 to 2014. This may reflect changes in injury epidemiology but could also reflect shifts in the practice patterns and supply of BPI surgeons. Because of the limitations of administrative claims data, we are unable to perform any additional analysis to investigate reasons for this temporal increase. Our previous administrative data study demonstrated that more than half of BPI surgery cases do not have an identifiable injury record (from a prior emergency department visit or inpatient hospitalization) before surgery [5], indicating that study of BPI injury mechanisms via claims databases may not be sufficiently comprehensive. Further studies with more detailed sources of data may be able to discern whether common causes of BPI in the USA (such as traffic accidents, all-terrain vehicle use, and violent trauma) are becoming more prevalent. The national estimates of adult traumatic BPI incidence we provide here will inform assessments of disease burden, enabling informed allocation of resources to support clinical care and research.

Funding Information

This publication was made possible by Washington University Institute of Clinical and Translational Sciences grant UL1 TR002345, sub-award KL2 TR000450 (C.J.D.), from the National Center for Advancing Translational Sciences (NCATS) components of the National Institutes of Health (NIH) and NIH Roadmap for Medical Research. The Center for Administrative Data Research (M.A.O., K.P.) is supported in part by Washington University Institute of Clinical and Translational Sciences grant UL1 TR002345 from the NCATS and grant R24 HS19455 through the Agency for Healthcare Research and Quality (AHRQ).

Compliance with Ethical Standards

Conflict of Interest

Christopher J. Dy, MD, MPH, FACS; Kate Peacock, BS; and Margaret A. Olsen, PhD, MPH, declare that they have no conflict of interest. Wilson Z. Ray, MD, reports receiving grants from the Department of Defense and from the National Institutes of Health/National Institute of Neurological Disorders and Stroke, and personal fees from Globus Medical and from Depuy Synthes, outside the submitted work. David M. Brogan, MD, MSc, reports receiving grants from American Foundation for Surgery of the Hand, nonfinancial support from Axogen, personal fees from Arthrex and from TriMed, and grants from the University of Missouri Orthopedic Association and from KL2 Research Award, outside the submitted work.

Human/Animal Rights

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2013.

Informed Consent

Informed consent was waived from all patients for being included in this study.

Required Author Forms

Disclosure forms provided by the authors are available with the online version of this article.

Disclaimer

This publication’s contents are solely the responsibility of the authors and do not necessarily represent the official views of the NCATS, ARHQ, or NIH.