Treatment of Lateral Epicondylitis: Is Surgery Still an Option?

© 2024 HMP Global. All Rights Reserved.

Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of ePlasty or HMP Global, their employees, and affiliates.

Lateral epicondylitis (LE) is a common enthesopathy and often a painful condition involving the origin of the extensor carpi radialis brevis tendon (ECRB). The etiology has not been definitively identified, but it is believed to be related to repetitive stain or degeneration changes to the tendon origin.¹ Previous population-based studies report an incidence of 2.4%² and a surgical rate between 2% and 16%^2,3,4,5. Most regard LE as a self-limiting condition with the mainstay of treatment being nonsurgical.^6,7 

In 2018, in a survey-based study of LE by members of the American Society for Surgery of the Hand and American Shoulder and Elbow Surgeons, Niedermeier et al⁸ reported that 98% of responders recommended some form of nonoperative treatment for at least 6 months for the management of LE and that 42% recommended surgical intervention if symptoms persisted more than 6 months. Although many health care specialists consider operative treatment of long-standing LE controversial,^4,9 several surgical techniques have been described for its treatment: open or arthroscopic debridement^10,11 with or without ECRB tendon repair,^12,13 ECRB tendon lengthening,^14,15 open or percutaneous ECRB tenotomy,¹⁶ and denervation procedures.¹⁷

Given the lack of prospective comparative outcome studies, controversy still exists regarding the indications for surgical management of LE. The purpose of this study was to analyze the trends and incidence in LE surgery over a 9-year period prior to the coronavirus pandemic in 2019.

Study Design

This study was exempt from Institutional Review Board approval. With over 25 million residents, Texas is representative of a diverse economic and ethnic population. A 9-year study period prior to the COVID-19 pandemic was selected (2010-2018). 

A retrospective cross-sectional analysis was performed using de-identified outpatient surgical records obtained from the publicly available Texas Health Care Information Collection (THCIC) and outpatient database through the Texas Department of Health Services (TDHS).¹⁸ We included all patients with the diagnosis of LE that underwent surgical intervention at outpatient centers. Records were initially identified through database query using Current Procedure Terminology (CPT) codes, followed by exclusion of records not including an active diagnosis of LE according to International Classification of Diseases (ICD), Ninth Revision, Clinical Modification (ICD-9-CM) codes (2010-2015Q3) or ICD-10-CM codes (2015Q4-2018). 

The resulting records included demographic characteristics, billing codes, all sources of payment, payor status, and zip codes for both patient residences and the health care centers at which the procedures were performed. Population-wide trends in outpatient surgical management of LE were then analyzed.

A secondary analysis was performed to include the incidence of surgery with respect to patient age, primary insurance designation, and patient-reported place of residence (metropolitan, micropolitan, or rural). For the purposes of this study, the platelet-rich plasma (PRP) and cortisone injection groups were reported but did not undergo further analysis.  

Study Population

The CPT, ICD-9-CM, and ICD-10-CM codes along with the query sequence used in this study are summarized in Table 1. Outpatient surgical procedures occurring between 2010 and 2018 were queried using the CPT codes 252.80, 243.05, 243.57, 243.58, 299.99, 29837, 298.38, and 0232T. The resulting data set was refined to include only those cases connected to the ICD-9-CM code of 726.32 (records from 2010-2015Q3) or any of the ICD-10-CM codes M77.10, M77.11, or M77.12. 

Population Estimates

Annual population estimates by zip code for the state of Texas were obtained from the US Census Bureau. Public health regions (PHRs), as defined by the TDHS, were used to estimate regional disease burden. Between 2010 and 2018, Texas PHRs with an average population of more than 4 million were considered metropolitan regions, PHRs with an average population of 2 to 3.9 million were categorized as micropolitan regions, and PHRs with an average population of less than 2 million were categorized as rural areas.¹⁸

Statistical Analysis

All statistical analyses were performed using SPSS Version 26.0 (IBM Corporation). Demographics, including procedure laterality, patient financial income group, primary insurance designation, and patient reported place of residence, were analyzed as subgroups. The per-capita incidence of operative cases per 100,000 persons was calculated for each procedure type. Fold change in the number of procedures relative to 2010 were calculated for 2011-2018. Overall procedural volumes were compared between subgroups using one-way ANOVA with post-hoc Tukey tests. Annual procedural volumes for subgroup were compared year-by-year using one-way repeated-measures ANOVA. Nonparametric follow-up analyses were performed as appropriate. Linear regressions were performed to identify relative rates of change in annual procedure volumes for each subgroup. Logistic regression was performed to assess the predictive value of dependent variables on categorical outcomes (year of surgery), with results reported as odds ratios with 95% confidence intervals. Bonferroni correction was performed for all multivariate analyses. P values less than .05 were considered statistically significant.

A total of 12,802 operative procedures for LE in the ambulatory outpatient surgical setting between 2010 and 2018 met our criteria. These procedures included 11,452 tenotomies/debridements, 672 arthroscopic procedures, 246 tendon lengthening, 378 platelet rich plasma (PRP) injections, and 107 cortisone injections (Table 1). PRP and cortisone injections were included in the data set if performed in conjunction with a LE surgical procedure. Injections performed prior to surgical intervention were not available for review. 

Across all procedures, surgery was more frequently performed on patients ages 45 to 64 years old (P < .001). Procedures were performed on the right limb more often than the left (P < .001) and on women more than men (P < .001). Laterality was determine based on the ICD codes. Procedures were performed more commonly on Caucasian patients and consistent throughout the study period (Table 2).

Over the duration of the study, 93% of the operative procedures encompassed lateral epicondylar debridements with tenotomy (with or without repair), 5% were arthroscopic, and 2% were coded as tendon lengthening. The incidence of arthroscopic and tendon lengthening surgical procedures for LE was consistently low, while tenotomies and tendon origin debridement (with or without repair) occurred most frequently (P < .001).

The overall incidence ranged from 2.4 cases per 10,000 population in 2010 to 1.8 in 2018, and there was no significant statistical change over the studied period of time (P > .05) (Figure 1). Surgical case volume was significantly higher in metropolitan areas (P < .001) and increased at a faster rate when compared with the less populous micropolitan or rural regions. For every increase of 1 case per 100,000 in population, metropolitan areas increased by 1.67 cases, micropolitan areas increased by 1.36 cases, and rural areas increased by 1.19 cases (P < .001) (Figure 2).

Commercial insurance was the most frequent and statistically significant primary payor source encountered in this study (86% of coded procedures, P > .001). Other forms of payments included worker’s compensation (11.5%), Medicare (1.3%), and Medicaid (<1%) (Figure 3). The distribution of surgical incidence compared with insurance category remained constant over the study period. The surgical incidence for worker’s compensation insurance had a slight downward trend from 2015 to 2018, though this was not statistically significant. 

Figure 1. Number of cases per 10,000 population by year and region.

Figure 2. Frequency of operative cases per 100,000 population according to public health region categories. Metro, metropolitan; micro, micropolitan.

Figure 3. Percentage of operative cases according to primary insurance provider.

Lateral epicondylitis is a self-limiting condition that was first described by Runge in 1873¹⁹ and one that is well-managed with conservative management.²⁰ In a 2013 review, Ahmad et al proposed that the principles of treatment for LE include: (1) control of pain; (2) preservation of movement; (3) improvement in grip strength and endurance; (4) return to normal function; and (5) control of further histological and clinical deterioration.²¹ Multiple conservative treatments have been developed for the management in these conditions, including physiotherapy, local injections, non-steroidal anti-inflammatory drugs (NSAIDs), bracing, thermal therapy, extracorporeal shock-wave therapy (ECSW), acupuncture, and botulin toxin A injections.²⁰ However, a recent meta-analysis by Sayegh et al found no convincing evidence that nonoperative treatments were better than no treatment²².

Recommendation for surgical intervention is challenging and still a topic of debate. Currently, the decision to perform surgery is on a case-by-case basis and more on personal or expert opinion than on scientific evidence. A study by Kachooei et al showed that if the initial physician visit is with a surgeon rather than a non-surgeon, the likelihood of receiving surgical treatment is 12 times higher.^23,24

Historically, the most widely performed operation for LE is debridement of the degenerative tissue as described by Nirschl and Pettrone in 1979.²⁵ In 2018, Kroslak and Murrell in a prospective, double-blind study, obtained similar outcomes after a sham procedure consisting of a skin incision and ECRB exposure. However, due to the limited study design the authors couldn’t determine the reason for the results but hypothesize that the “sham surgery” may have stimulated healing in the area or caused local denervation because of the surgical incision.²⁶

As mentioned previously, the surgical rate for the treatment of LE ranges between 2% and 16%.^2,3,4,5 In our study, the incidence of LE operative cases per 10,000 individuals remained low and did not significantly change over the 9-year study period. Using a US population database, Buller et al²⁷ reported an increase in the surgical treatment of LE per 100,000 per capita from 1994 to 2006, while Degen et al³ found no change in the surgical treatment of LE per 10,000 per capita from 2007 to 2014. The results presented in the current investigation are similar to the ones reported by Degen et al, which reflect a plateau in the national and regional incidence of LE surgical intervention since 2007. Differing results were previously reported in a single-county Minnesota observational study from 2000 to 2012,² where the authors reported the tripling of surgically treated cases for LE within 2 years of diagnosis and suggested that insurance reimbursement, as well as local treatment expectations, may have contributed to the increase in surgical intervention. 

Regarding payor mix, our study found a significantly greater number of surgeries performed in patients using commercial/private insurance than those with Medicare, Medicaid, or worker’s compensation. This trend remained constant over the 9-year period of the current study. These results are similar to Buller et al,²⁷ who found that commercial and private insurances were the most common payor for LE surgical treatment. In a case-control study of 580 patients, Knutsen et al⁵ demonstrated 8.1-fold increased odds of surgical management of LE for worker’s compensation claims versus other payor types. In our study, inclusive of a larger cohort of worker’s compensation cases, the incidence of surgery in worker’s compensation did not change over the 9-year analysis and had trended a decrease in incidence since 2015. Whether this finding was related to changes in our state worker’s compensation policies and guidelines or a limitation due to our study design is to be determined. Selection bias must be acknowledged as a limitation of any study analyzing payments for service, as this may be the driver for offering more invasive procedures. 

In our study, ECRB tenotomy and tendon origin debridement with or without repair was the most commonly performed procedure for LE across our 9-year analysis, while arthroscopic management and isolated tendon lengthening procedures remained uncommon. This finding is corroborated by Wang et al,²⁸ who found similar results where debridement alone or debridement with tendon repair were the most common procedures performed. Other epidemiologic studies have examined changes in surgical incidence over time without analyzing specific surgical techniques.³

LE is a condition that commonly affects middle-aged patients,²⁹ which would explain why Buller et al found that the majority of patients who underwent operative management in the United States from 1994 to 2006 were between the ages of 40 and 49 years old.²⁷ In their epidemiological review of ambulatory upper extremity surgery, Jain et al found that patients ages 45 to 65 have the highest incidence rate.³⁰ In our study, we performed subgroup analyses, and based on the total procedural volume over the entire study duration, surgery was most frequently performed on patients ages 45 to 64. We also found that more surgical procedures were performed on women than men and that procedures were performed at a greater incidence on the right limb than the left, although the significance of more right limb surgery is not known. LE is more common in the dominant arm than the non-dominant arm,³¹ and the majority of the population is right-hand dominant, which may account for this difference. 

Surgical case volume was higher in the more populous regions of Texas (counties surrounding Dallas and Houston) when compared with lesser populated areas. After adjusting for population changes, the surgical incidence in metropolitan areas was found to have increased at a significantly faster rate when compared with lesser populated areas year over year. For every increase of 1 case per 100,000 in population, metropolitan areas increased by 1.67 cases, micropolitan areas increased by 1.36 cases, and rural areas increased by 1.19 cases. This trend was statistically significant and reflected a greater incidence for surgery in patients with LE in metropolitan areas. One possible explanation for this trend is the increase on reimbursement on ambulatory hand procedures. In 2021, a retrospective review of 1,914,905 orthopedic procedures performed at outpatient centers showed an annual increase of ambulatory hand procedures from 2012 to 2017. In addition, they found a steady annual increase in the reimbursement of hand and sports procedures. They also noted that orthopedic ambulatory centers were located in counties with a higher average household income and in major metropolitan areas, explaining why more procedures were performed in those areas.³² Extrapolating data from a more commonly performed hand procedure, Hudson et al found that carpal tunnel release, especially endoscopic release, is more commonly performed in metropolitan areas compared with micropolitan and rural areas. The authors hypothesized that a reason for these findings is that fellowship-trained hand surgeons and academic training programs are more likely to be located in larger cities and university hospitals.³³ Fellowship-trained hand surgeons are more likely to perform more complex hand operations, like endoscopic carpal tunnel release, than non–fellowship-trained surgeons.³⁴ Anecdotal experience may also play a significant role in the incidence of surgical intervention for the treatment of LE, as many surgeons may rely on personal or share experience given the absence of good evidence. All these arguments could explain our findings; however, other reasons like population shifts or access to health care may also play a role. 

No other LE studies to date have examined the relationship between population density (metropolitan, micropolitan, and rural) and the incidence of surgical intervention for LE. However, limitations must be acknowledged. Population databases and epidemiological analysis are limited in their scope and can only provide generalized conclusions. Additionally, with any retrospective review of a health care database, the results may be subject to errors in documentation, coding, and data submission. Because the database is used for state health care administrative initiatives, detailed clinical patient information was not available. We were not able to determine patterns of treatment according to surgical specialty, board certification, or fellowship training, and it is unclear how these factors might have influenced the data sets.

In conclusion, although the benefit of surgical treatment of LE has yet to be completely elucidated, the results of this study show that surgical intervention is still being offered. The incidence of surgical treatment of LE remained low over the studied period; however, the volume of cases in metropolitan areas increased at a fast rate. Whether this finding is related to surgeon density, level of training, or anecdotal experience combined with no clear evidence regarding optimal treatment is to be determined. Although prospective comparative studies are ultimately necessary for advancing the understanding of the optimal treatment in LE, retrospective analysis of health care databases show the trends in surgical technique and demographic shifts over time. In the future, further study and development of clinical practice guidelines for the management of LE would be helpful to guide clinical and surgical decision-making.