Outcomes of Hip Arthroscopy in the Medicare Patient: A Growing Population

Abstract

Background:

Although the results of hip arthroscopy in the elderly have been inferior to the results in younger patients, there have recently been some encouraging reports in carefully selected series of older patients. The purpose of this study was to identify the utilization of hip arthroscopy in the Medicare population and to determine the rate and timing of revision arthroscopy and/or total hip arthroplasty (THA) with the goal of identifying risk factors for secondary procedures based on patient demographics, comorbidities and the diagnosis at the time of arthroscopy.

Methods:

The Medicare Standard Analytic Files were reviewed from 2005-2014 for all patients undergoing hip arthroscopy allowing for minimum 2 year follow-up (100% sample). Patients were tracked through the dataset for the occurrence of an ipsilateral THA or revision hip arthroscopy. Rates and timing of the subsequent procedures were then determined within 6 month intervals. Patients less than 65 years old were excluded. Multivariate logistic regression analysis was performed to determine the impact of patient age, sex, obesity or a diagnosis of hip osteoarthritis on need for revision procedures.

Results:

3,320 Medicare patients had a hip arthroscopy during 2005-2014 (0.3% compared to THA). 73 patients (2.2%) underwent reoperation during the follow-up period. Two-thirds (n = 46) of all revision procedures occurred within one year of primary hip arthroscopy. A pre-operative diagnosis of hip osteoarthritis significantly increased the odds of reoperation (OR = 5.3). (Conclusion: Relatively few numbers of Medicare patients underwent hip arthroscopy during the time interval evaluated (0.3% when compared to THA utilization). 2.2% underwent a subsequent revision arthroscopy or THA with many occurring soon after the procedure and for the diagnosis of hip OA demonstrating the need to better define indications in this population. This study should provide baseline utilization and outcome trends for future studies.

Level of Evidence: IV

Introduction

The use of hip arthroscopy has increased exponentially in the last decade.^1-4 Recent research on hip arthroscopy trends using large national databases demonstrate an overall increase from 3.6 procedures per 100,000 in 2005 to 16.7 per 100,000 in 2013.³ Although the largest increase has been observed in ages 35-44, enrollees in the 60 years and greater age group demonstrate a 200% increase in hip arthroscopies from 2007 to 2011.¹ With this rapid increase in hip arthroscopy, concerns regarding indications for the procedure, particularly in older patients, has risen. Initial results of hip arthroscopy in the elderly have been inferior to results in younger patients, with increasing age associated with greater subsequent risk of THA.⁵ Multiple previous studies found that osteoarthritis was the biggest predictor of failed survivorship after hip arthroscopy in this population.^5,6 Recent evaluation of private insurer databases found a 25.2% rate of conversion to THA in patients over 60, and a mean of 25.0 months to THA.³ Success, however, has been demonstrated in carefully selected series of older patients.^7,8

Most studies evaluating the procedure in older patients have been limited by small cohorts, older data, or both.^1,5,9 Additionally, the Medicare population which represents the largest insurer for ages >65 has not been singularly evaluated. The purpose of this study was to evaluate the utilization of hip arthroscopy in the Medicare population from 2005 to 2014 and to (1) determine the rate of revision arthroscopy and total hip arthroplasty (THA) after initial hip arthroscopy, (2) determine how soon after hip arthroscopy a revision procedure was performed, and (3) identify factors that may predict the need for THA after arthroscopic surgery of the hip. Specifically, we targeted risk factors suggested by prior studies including age, gender, obesity status, and presence of osteoarthritis at the time of index arthroscopy.

Materials and Methods

The PearlDiver Research Program (www.pearldiverinc.com; PearlDiver Inc, Fort Wayne, IN) was used to query the Medicare Standard Analytic Files for the information presented in this study. The Medicare Standard Analytic Files represents 51 million covered lives and includes 100% of inpatient and outpatient facility records billed to Medicare in this period. All data within this database are health Insurance Portability and Accountability Act compliant and were thus deemed exempt from institutional review board approval by our institutions Human Subjects Office.

Patients undergoing hip arthroscopy were identified using the Current Procedural Terminology (CPT) codes 29861, 29862, 29863, 29914, 29915, 29916. The CPT code for diagnostic hip arthroscopy (CPT-29860) was specifically excluded as it was felt that this code would bias the results of conversion to THA. Patients less than 65 years of age were additionally excluded. Laterality modifiers were used to ensure ipsilateral revision procedures were captured and thus, records without laterality designation were excluded. Patients were then tracked longitudinally through the dataset at six month intervals for 5 years following index procedure, or for two to five years in those patients operated on after December 2011 for the occurrence of an ipsilateral THA or revision hip arthroscopy procedures. This allowed for determination of the incidence and timing of subsequent THA or revision hip arthroscopy procedure.

To evaluate the impact of specific patient factors on incidence of revision procedures, cohorts were filtered into subgroups based on demographic data including age, gender, and obesity status. Age subgroups included; 65-69 years, 70-74 years, 75-79 years, 80-84 years, and 85 or older. Obesity was determined by an International Classification of Diseases 9th Revision (ICD-9) diagnosis code representing body mass index 30 kg/ m2. Additional subgroup analysis was performed to determine if odds of subsequent revision or THA were different with the presence of hip osteoarthritis, using ICD-9 codes for primary (715.15), secondary (715.25), or unspecified (715.35 and 715.95) osteoarthritis of the hip. The chi-square test was used to compare categorical variables, and odds ratios (ORs) with corresponding 95% confidence intervals were used to evaluate differences in rates of subsequent THA. Multivariate logistic regression analysis was performed to determine the impact of each of these groups (age, sex, obesity, or hip osteoarthritis) on need for revision procedures. Statistical significance was set at P < .05 for all tests.

Results

Overall 3,916 Medicare patients underwent a hip arthroscopy between 2005 and 2014. After the application of laterality modifiers to the hip arthroscopy CPT codes, 3320 (84.7%) hip arthroscopy patients were remaining for longitudinal analysis of a subsequent ipsilateral arthroscopy or THA. Except from 2007 to 2008 there was an increasing number of arthroscopies performed each year, with a 454% increase in procedures from 2005 to 2014 (Figure 1). Of the 3320 arthroscopies included, 2077 (62.5%) were female and 1243 (37.5%) were male. There was a decreasing incidence of hip arthroscopy with age (Table 1), with the majority of procedures (53.5%) being done in patients aged 65-69 (1,779), and 891 (26.8%) in patients aged 70-74. The remaining 20% included 434 procedures in ages 75-79 (13.0%), and 167 procedures (5%) in ages 80-84. There were 57 hip arthroscopies done in patients 85 or A older. The most common procedure documented was chondroplasty (CPT-29862), which was documented in 2,425 (73.0%) of procedures. A preoperative diagnosis of hip osteoarthritis of the hip was present for 30.9% of the hip arthroscopy patients.

Figure 1.

Hip arthroscopies performed in medicare patients per year.

Table 1.

Demographics of Hip Arthroscopy Patients

Demographic	Total (n)	Percent (%)
All Patients	3320	100
Male	1243	37.5
Female	2077	62.5
Obese	329	9.9
Non-Obese	2991	90.1
Age breakdown (years)	Total (n)	Percent (%)
65-69	1779	53.5
70-75	891	26.8
76-79	434	13
80-84	167	5
85+	57	1.7
Year Performed	Total (n)
2005	115
2006	164
2007	217
2008	204
2009	304
2010	396
2011	429
2012	497
2013	499
2014	523

Longitudinal tracking of hip arthroscopy patients found that 73 patients (2.2%) went on to require an additional ipsilateral THA or revision hip arthroscopy procedure. 32 patients (1.0%) underwent an ipsilateral THA, and 43 (1.3%) received a revision arthroscopy. Of patients who had a diagnosis of osteoarthritis at the time of their index procedure, 4.5% (n = 46) went on to require revision scope or THA. There was a significantly increased odds of subsequent procedure for the osteoarthritis group with an OR of 1.73 (1.46-1.99, P < .001). Of the 1% (n=32) of patients who underwent an ipsilateral THA, 84% occurred within two years of the index arthroscopy, and 100% occurred within 4 years of the index procedure. Similarly, 84% of revision arthroscopy procedures occurred within two years of the index arthroscopy, and 97.7% (n=42) were within four years (Table 2).

Table 2.

Rate and Time to Revision Arthroscopy or Total Hip Arthroplasty (THA) after Hip Arthroscopy

Months after Hip Arthroscopy	Number of Patients undergoing Revision Arthroscopy	Percent of Total Revisions	Number of Patients Undergoing THA	Percentage of Total THA
6	19	44.2%	7	21.80%
12	28	65.1%	20	62.50%
18	34	79.1%	25	78.10%
24	36	83.7%	27	84.30%
30	37	86.0%	29	90.60%
36	39	90.7%	31	96.80%
42	41	95.3%	31	96.80%
48	42	97.7%	32	100%
54	42	97.7%	32	100%
60	42	97.7%	32	100%

Multivariate regression analysis was performed to evaluate the potential effects of osteoarthritis, obesity, age sub-group, and gender on need for subsequent procedure. Obesity, gender and age were not associated with need for a revision hip arthroscopy or ipsilateral THA on multivariate analysis. A pre-operative diagnosis of hip osteoarthritis significantly increased the odds of a subsequent procedure with OR of 5.3 (95% CI: 3.2-9.1, p<0.001).

Discussion

Of 3,320 Medicare patients analyzed in this study, following hip arthroscopy approximately 2.3% went on to require a subsequent procedure after hip arthroscopy with 1% undergoing THA and 1.3% undergoing a revision arthroscopy during the years 2005 to 2014. A preoperative diagnosis of hip osteoarthritis was present in 30.9% of patients, and patients with a pre-operative diagnosis of osteoarthritis had significantly higher odds of requiring an additional procedure (OR: 1.73, p<.001). Conversion to THA and revision arthroscopy both occurred quickly after index hip arthroscopy, with over half of secondary procedures occurring within 12 months, and 78-79% occurring within 18 months. Additionally, 100% of THAs and 97.7% of revision arthroscopies occurred within four years. The rate of hip arthroscopy in Medicare patients rose exponentially with an overall 450% increase in the number of hip arthroscopies during the 9 years of the study period.

Limitations of this study include those inherent to administrative claims database studies, which are dependent on coding and documentation accuracy. In addition, Health Insurance Portability and Accountability Act-compliant databases are unable to provide exact numbers when a value lies between 0-10, making analysis of smaller subgroups difficult. Also, patients undergoing arthroscopy towards the end of the study (after 2011) did not have a full five years of post-arthroscopy surveillance. Additionally, the arthroscopy CPT codes 29914, 29915, 29916 (femoroplasty of a cam lesion, acetabuloplasty of a pincer lesion, and labral repair, respectively) were introduced halfway through the study period in 2011, making it difficult to analyze the association of specific procedure codes with subsequent outcomes in this population. Finally, indications for repeat surgery and outcomes other than the need for repeat arthroscopy or hip replacement, including patient A reported outcomes and radiographic outcomes can not be obtained from this type of database.

Strengths of the study include the very recent time period of data and large population of older patients able to be assessed. The use of a laterality modifier also allows for identification of true conversions to THA or revision arthroscopy; this addition is critical and sometimes not taken into account in other database studies. The use of the Medicare database gives an excellent sample of older adults in which to study while many studies of arthroscopy are innately limited by the overall small numbers of eligible older participants, using lower thresholds of age 40 or 50 as their cutoff which create an inherently different patient population.

Overall, the results of this study are encouraging for the use of hip arthroscopy in the older population with only 2.2% of patients requiring repeat arthroscopy or total hip replacement. Bedard et al. found a conversion rate to THA after arthroscopy of 9.7% in patients over 50, while Sing et al reports a rate of 17% for patients over 50 for the years 2007 to 2011.^1,10 Neither of these studies evaluated revision arthroscopy as a second indicator of treatment failure. Other recent meta-analyses evaluating hip arthroscopy in adults over 40 additionally emphasize increased re-operation rates compared to younger counterparts. Griffin et al.⁸ reported overall conversion rate to THA of 18.5% at a mean of 17.5 months following hip arthroscopy; Horner et al.⁷ cites an aggregate conversion to THA rate of 18.1% in patients over 50, and 25.2% in patients over 60 with a mean of 25.0 months to THA. Although substantially higher than rates of conversion observed in our study, both meta-analyses found overall a statistically significant improvement in outcome after hip arthroscopy if the indication included femoral osteochondroplasty, labral repair, or was unspecified.^5,8 Labral debridement, alternatively, has less predictable results.

Our finding of osteoarthritis as a predictor of failure is also well supported in literature. Phillipon et al. in an analysis of 203 patients >50 treated with hip arthroscopy for labral tears and femoroacetabular impingement found a joint space of <2mm to be a major predictor of subsequent THA.¹¹ Alternatively, when patients with osteoarthritis are excluded, results from hip arthroscopy in the older patient are very good. Bryan et al evaluated 201 patients without radiographic evidence of arthritis aged 55 or greater who underwent arthroscopy for FAI; although their ADL scores were below their younger counterparts at 2 years, there was significant improvement in mean Harris Hip Score (mHHS) compared with preoperative values.⁷

Although THA remains the clearest endpoint, inclusion of revision arthroscopy in our statistical evaluation of index treatment failure provides important additional data for consideration. A systematic review of >6000 patients of all ages by Harris et al.¹² found the most common reason for reoperation after hip arthroscopy to be conversion to THA while our study found a higher rate of revision arthroscopies (1.3%) than conversions to total hip arthroplasty (1.0%). One explanation for this finding in those >65 may be increased caution when selecting senior patients for hip arthroscopy; many elderly patients presenting with hip pain perhaps are simply directed towards the certain finality of a THA over the possibility of multiple procedures with arthroscopy. Those who are candidates for joint preservation, i.e. FAI with no or minimal arthritic change, may be the very select few who would therefore also consider revision arthroscopy before proceeding with THA. This low rate overall of conversion to THA in the 65+ age group compared to studies evaluating mostly 40-60 year olds need continued study.

Shared decision-making between patient and surgeon is especially valuable for the older patient in whom multiple procedures may be particularly unfavorable. Overall, however, our data suggests that the large majority of Medicare patients indicated for arthroscopy do not go on to require a revision arthroscopy or ipsilateral THA (only 2.2% of 3320 patients). It does suggest that when hip arthroscopy fails, it does so relatively quickly after the index procedure. As supported by prior literature, patients who have osteoarthritis at the time of hip arthroscopy are more likely to go onto conversion to THA and/or revision arthroscopy. These results provide needed age-specific data for hip arthroscopy failure rates in senior adults, and emphasize that risk factors such as osteoarthritis should be carefully taken into consideration when indicating those >65 for hip arthroscopy.