Rates and Risk Factors for Revision Hip Arthroscopy

Abstract

Background:

Revision hip arthroscopy often serves as a measure for a failed primary hip arthroscopy procedure. The purpose of this study was to examine the rate, timing, and risk factors for revision hip arthroscopy using a large national database.

Methods:

The Humana administrative claims dataset was reviewed from 2007 through the second quarter of 2015 to identify patients undergoing hip arthroscopy using Current Procedural Terminology (CPT) codes and laterality modifiers. Patients with subsequent ipsilateral revision hip arthroscopy were identified and the rate and timing of these revisions determined. Subgroup analysis was performed to determine effects of gender, age, body mass index (BMI), osteoarthritis diagnosis, and specific hip arthroscopy procedure on revision rates.

Results:

There were 1807 hip arthroscopy procedures identified with a revision rate of 4% (total of 72 procedures). Of the revision procedures, 43% occurred within 6 months after the index procedure, and 86% occurred within 18 months. Age < 50 years was the only significant predictor of revision hip arthroscopy (OR 2.03, CI 1.173.53) with an even distribution across younger age groups. An osteoarthritis diagnosis, gender, and BMI did not represent significant risk factors for revision (OR 0.87, 0.98, 0.9 and CI 0.5-1.51, CI 0.6-1.59, CI 0.37-2.12, respectively). Chondroplasty was the most common procedure leading to revision (46%) followed by labral repair (37%). The most common revision procedures were chondroplasty (44%) followed by femoroplasty (38%).

Conclusions:

Overall, 4% of hip arthroscopy procedures underwent revision arthroscopy over the 8-year period. Revision was associated with age < 50, and revisions were most frequently performed for femoroacetabular impingement. The majority of revisions occurred within 18 months after the index procedure.

Level of Evidence: IV

Introduction

Arthroscopic visualization of intraarticular structures in the hip was described as early as 1931.¹ The advent of hip arthroscopy as standard practice for treatment of various pathologies in the hip is a relatively recent phenomenon and has gained traction particularly over the last two decades with an exponential increase in utilization and indications.^2-5 The increase in hip arthroscopy over this time period has been shown to be in large part due to increasing recognition and treatment of femoroacetabular impingement (FAI) syndrome.² Recent studies have shown a 365% increase in utilization from 2004-2009, and this rise has continued with a 250% increase from 20072011.^4,5 These studies have also shown that although the longest component of this increase has been demonstrated in age groups under 30, there has been a significant increase in hip arthroscopy procedures in all age groups, including those over 60.⁵

In light of this increase in the use of hip arthroscopy, there has been an increasing focus on outcomes, reoperation rates, and complications associated with these procedures.^6-10 Following hip arthroscopy, subsequent procedures that have been consistently monitored include revision hip arthroscopy or arthroplasty. Conversion to total hip arthroplasty (THA) has been the subject of multiple investigations ranging from small cohorts to large national database studies with rates ranging from 2.9%-16%.^11-16 One of these studies that queried the same database that is the subject of the current work found a conversion rate of 5.3% with most of those occurring within 18 months after the index procedure.¹¹

Causes of failed hip arthroscopy include FAI, iatrogenic chondrolabral injury, adhesions, and postoperative microinstability to name a few. Reoperation rates following hip arthroscopy have been reported with a wide range from 1.3% to greater than 50%.^6,15,17-19 Revision hip arthroscopy is among the more common procedures performed after primary hip arthroscopy. Harris et al., in their large systematic review of 6,134 hip arthroscopy procedures, found that among 6.3% of patients that required reoperation, which included conversion to THA, 30% of those went on to need a revision arthroscopic procedure.¹⁴

With the continued increase in hip arthroscopy procedures and the subsequent increase in revision hip arthroscopies, the purpose of the present study was to use a large national database to evaluate the rate, timing, and potential risk factors for revision hip arthroscopy. We hypothesized that the overall rate of revision hip arthroscopy would be low, and would be predicted by obesity, female sex, advanced patient age.

Methods

The data for the present study were obtained from the Humana Inc administrative claims database using the PearlDiver Research Program (www.pearldiverinc.com; PearlDiver Inc, Fort Wayne, IN). This data set represents over 16 million covered lives between 2007 and 2015 composed of private/commercially insured patients as well as Medicare/Medicare Advantage plans. These data are compliant with the Health Insurance Portability and Accountability Act and therefore were exempt from institutional review board approval.

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 (29860) was excluded. Laterality modifiers were used to determine whether the index procedure was done on the left or the right and records without laterality were excluded. Patients were then tracked longitudinally for the occurrence of an ipsilateral hip arthroscopy using the same CPT codes and laterality modifiers to ensure that the revision procedure was performed on the same hip as the index hip arthroscopy. The database was queried for ipsilateral hip arthroscopy at 6 month intervals for a period of 5 years postoperatively. We evaluated subsequent arthroscopy after index arthroscopy procedures and did not consider other revision procedures, preservation procedures such as periacetabular osteotomy, or conversion to THA. Further subgroups of these cohorts were then assessed for their effect on the rate of revision hip arthroscopy including specific procedure, gender, age, diagnosis of osteoarthritis, and body mass index (BMI). Revision rates were assessed for ages 0-90 years old divided into 5 year age ranges. Diagnosis of osteoarthritis at the time of index procedure was assessed using International Classification of Diseases 9th Revision (ICD-9) codes 715.15, 715.25, 715.35, 715.95 to include primary, secondary, or unspecified osteoarthrosis of the pelvic region and thigh. Effects of BMI were assessed with obesity defined as BMI > 30 with ICD-9 codes 27800, V853, V8530, V8539, 27801, V854, V8541, and V8545. Basic demographic data (age and gender) of the entire cohort were also identified. Odds ratios (ORs) with corresponding 95% CIs were used to evaluate differences in rates of subsequent revision hip arthroscopy. Statistical significance was set at p < 0.05.

Results

During the years included in the time period investigated, 1,807 hip arthroscopy procedures coded for laterality were identified and used for longitudinal analysis. Within this group, the most common procedure performed was chondroplasty at 46.1% (n=833) followed by femoroplasty at 37.2% (n=673) and labral repair at 31.8% (n=575). Overall, 668 patients (37%) were male and 1139 (63%) were female. Hip arthroscopy was most commonly performed between the ages of 30 and 60 (61%) with the largest age groups being 40-44 (12%), 50-54 (12%), and 45-49 (11%) (Figure 1). Of this cohort 63% (n= 1132) were < 50 years old and 37% (n=675) were > 50. Obese patients comprised 9.1% (n=164) of hip arthroscopy patients, and 90.9% (n=1643) had a BMI < 30. A diagnosis of hip osteoarthritis was present at the time of the index procedure in 37.5% (n=678) of patients, and 62.5% (n=1129) did not carry that diagnosis (Table 1).

Figure 1.

Hip arthroscopy procedures for given age groups.

Table 1.

Procedures and Demographics of Total Number of Procedures (n = 1807)

Procedure	Number of Procedures	%
Loose Body Removal (29861)a	136	7.5
Chondroplasty (29862)	833	46.1
Synovectomy (29862)	317	17.5
Femoroplasty (29914)	673	37.2
Acetabuloplasty (29915)	424	23.0
Labral Repair (29916)	575	31.8
Demographics
Male	668	37.0
Female	1139	63.0
Obese	164	9.1
Non-Obese	1643	90.9
Diagnosis of osteoarthritis No diagnosis of osteoarthritis	678 1129	37.5 62.5

^aCurrent Terminology Procedural Codes

Longitudinal tracking of patients identified 72 patients (4.0% of 1,807 patients) that underwent a subsequent ipsilateral hip arthroscopy. Among these revisions, 43.1% were performed between 6 and 18 months postoperatively, and 86.1% were performed within the first 18 months postoperatively. Of the total number of revisions, 55 (76%) were < 50 years old, and 17 (24%) were > 50 years old. Of the criteria that were evaluated for subgroup analysis, age < 50 years was the only significant predictor of revision hip arthroscopy (OR 2.03, CI 1.17-3.53).

Osteoarthritis, gender, and BMI were not associated with the need for revision arthroscopy (OR 0.87, 0.98, 0.9 and CI 0.5-1.51, CI 0.6-1.59, CI 0.37-2.12 respectively) (Table 2). Procedures performed at the time of revision arthroscopy included chondroplasty in 44% and femoroplasty in 38%.

Table 2.

Comparison Between Revision and Non-Revision Groups^a

		Revision	No Revision
Age	> 50	55	1022	OR - 2.03
	< 50	17	641	CI: 1.17 - 3.53
Diagnosis of Osteoarthritis	Yes	20	638	OR - 0.87
	No	38	1053	CI: 0.5 - 1.51
Obesity	BMI > 30	6	152	OR - 0.9
	BMI < 30	66	1511	CI: 0.37 - 2.12
Gender	Male	45	1049	OR: 0.98
	Female	27	614	CI:0.6 - 1.59

^aData are reported as number of index procedures coded. OR, Odds Ration. CI, Confidence Interval

Discussion

The present study presents data collected from a large national database representing a large patient population, which is helpful when investigating relatively uncommon procedures such as revision hip arthroscopy. There was a 4.0% incidence of revision hip arthroscopy. This is comparable to revision rates reported by others. Harris et al. reported 1.8% of patient underwent revision hip arthroscopy in a systematic review of over 6,000 patients that underwent hip arthroscopy.¹⁴ Truntzer et al showed a 5.31% revision hip arthroscopy rate within 6 months of the index procedures.²⁰ By comparison, those that have investigated indications and outcomes after revision hip arthroscopy have found that after revision hip arthroscopy there is a 5-8% re-revision rate.^21,22 As has been shown in other studies, our data supports the conclusion that the most common reason for revision arthroscopy is residual femoroacetabular impingement with chondroplasty and femoroplasty being the most common revision procedures.⁶

Revision hip arthroscopy is associated with younger age which was the only significant variable in the current work. Interestingly, BMI, a preoperative diagnosis of osteoarthritis, and gender were not significant predictors of revision. While others have shown a female predominance for revision hip arthroscopy,²² our data did not reveal gender as a significant risk factor. These findings may be partially explained by a treatment bias applied to those with arthritis and age > 50 as they may be more likely counseled toward arthroplasty. ¹¹

In a previous study, the conversion from hip arthroscopy to THA was evaluated in similar data set.¹¹ In that study, conversion rate to THA was 5.3% after failed hip arthroscopy, and conversion to THA was found to occur mostly in older patients with osteoarthritis. Both the timing and age of patients undergoing conversion to THA correlate to the findings for revision hip arthroscopy reported in the present study.

This work has limitations that are inherent to all database studies that are reliant on insurance administrative claims data and therefore accurate CPT and ICD-9 coding. Indications for individual procedures, intraoperative findings and complete descriptions of the procedures were not available using deindentified registry data and so associating specific treatments with specific diagnoses is not possible using these data. This is further complicated by the fact that CPT codes have expanded in recent years to more specific codes for femoroplasty (29914), acetabuloplasty (29915), and labral repair (29916), although older codes are still routinely used. No outcome data are available using this data set other than reoperation rates for the population. This study only evaluated subsequent arthroscopy after index arthroscopy procedures and did not consider other revision procedures, preservation procedures such as periacetabular osteotomy, or conversion to THA.

In summary, with rates of hip arthroscopy continuing to rise, there is a need to identify factors that increase the risk for revision procedures. Our ability to detect significant variables affecting outcomes with large database cohorts as in the present study should improve with the recent addition to the number of codes for operative treatment and the conversion to ICD-10 diagnostic coding. This information should help surgeons and patients be better informed in their shared decision making when deciding on hip arthroscopic procedures.

Overall, 4% of hip arthroscopies underwent revision arthroscopy A over the 8-year period. Revision was associated with age less than 50, and revisions were most frequently performed for femoroacetabular impingement. The majority of revisions occurred within 18 months after the index procedure.