Higher revision rates in primary total hip arthroplasty among patients with rheumatoid arthritis compared with osteoarthritis

Abstract

Aims

Although therapies for rheumatoid arthritis (RA) have advanced, it remains unclear whether RA continues to carry an elevated risk for revision surgery after total hip arthroplasty (THA). Therefore, the aim of this study was to compare revision rates of patients with RA undergoing primary THA with those of patients with osteoarthritis (OA).

Methods

This observational cohort study was based on data from the German Arthroplasty Registry from November 2012 to March 2024. Primary THA procedures in RA patients were compared with those in OA patients. Cumulative revision rates over nine years were calculated using Kaplan-Meier estimations. Differences by type of revision (major/minor), cause (aseptic/septic), and implant fixation were tested using the log-rank test and multivariate Cox proportional hazard analyses.

Results

A total of 12,750 THAs in RA patients were compared with 528,435 in OA patients. Overall, 17,434 revisions were recorded (RA 604; OA 16,830). At nine years, higher cumulative revision rates were observed in RA patients for major (5.1% vs 3.1%; p < 0.0001) and minor revisions (1.7% vs 1.1%; p < 0.0001), regardless of cause. After adjustment for demographic characteristics and fixation type, RA was associated with increased risk of major (hazard ratio (HR) 1.48, 95% CI 1.34 to 1.63; p < 0.001) and minor revision (HR 1.62, 95% CI 1.40 to 1.88; p < 0.001). In RA patients, hybrid fixation was linked to lower major revision risk compared with uncemented fixation (HR 0.56, 95% CI 0.42 to 0.74; p < 0.001). Cementless fixation predominated in both groups (RA 69%, OA 78%).

Conclusion

An increased risk of major and minor revision after primary THA was found in RA patients compared with OA patients, independent of cause. Although hybrid fixation was linked to lower major revision risk than uncemented fixation, cementless fixation remained the most common method in RA patients.

Cite this article: Bone Jt Open 2025;6(12):1626–1633.

Introduction

The annual number of total hip arthroplasty (THA) procedures is expected to grow substantially over the coming decades.¹ Therefore, to manage the burden of revision surgeries, thorough investigation of the risk factors for such revisions is essential. Rheumatoid arthritis (RA) has been shown to contribute to decreased joint mobility, bone erosion, and pain, ultimately leading to THA, but also to increase the risk of revision surgery following these procedures.^2,3 Several areas have been identified as potential contributors to this increased risk for revision.

Both soft-tissue quality and bone mineral density are compromised in RA.^4,5 Increased joint hypermobility and laxity, as a result of soft-tissue abnormalities, contribute to a higher rate of THA dislocations in RA patients compared to those with OA.^2,6,7 Compromised bone mineral density, often stemming from chronic glucocorticoid use and poor adherence to anti-osteoporotic therapy, elevates the risk of osteoporotic fractures.^8-12 Together, these factors may explain an increased risk of aseptic revision surgery in RA patients compared to those with OA.

Furthermore, RA patients experience a higher frequency of infections compared to non-RA controls.^13,14 This vulnerability is likely due to several immune system alterations, such as neutropenia^15-17 and immunosuppressive therapies, including glucocorticoids.¹⁸ These factors may contribute to an elevated risk of septic revision surgery in RA patients compared to those with OA.

Recently, the development of new antirheumatic agents has led to significant therapeutic advancements. The latest generation of treatments, including biological disease-modifying antirheumatic drugs (bDMARDs) and targeted synthetic disease-modifying antirheumatic drugs (tsDMARDs), has improved disease activity and reduced the incidence of THA in RA patients.^19,20 Simultaneously, guidelines for the perioperative management of RA patients on DMARDs are evolving, aiming to effectively control the disease while minimizing surgical risks.²¹ However, the effect of these improvements on revision surgery rates in THA remains unclear.

Therefore, the aim of this study was to report on the current aseptic and septic revision rates in RA patients undergoing THA compared with OA patients, utilizing a large dataset from a national joint registry. 

Methods

Patient data

Patient data for this prospective observational cohort study was retrieved from the German arthroplasty registry (Endoprothesenregister Deutschland (EPRD)). The EPRD amalgamates data from three independent sources: over 720 German hospitals, the two largest public health insurance companies in Germany, and the product database of implant manufacturers, the German Medical Technology Association.²² Data linkage among these three entities covers approximately 75% of publicly insured patients in Germany and ensures nearly 100% follow-up for patients who underwent primary and revision surgery at hospitals participating in the EPRD.

All primary THA procedures between 1 November 2012 and 31 March 2024 were included. Patients were divided into two groups: those with RA and those with OA, according to the tenth revision of the International Classification of Diseases of the World Health Organization (ICD-10)²³ codes recorded individually for each patient by the discharging physician at the time of index surgery (Supplementary Material).

Patients with RA were more likely to be female (RA 75%; OA 63%) and slightly older (median age: RA 71 years (IQR 62 to 78); OA 70 years (IQR 62 to 77)). Demographic data by group are provided in Table I .

Table I.

Demographic data.

Variable	OA, n = 528,435	RA, n = 12,750
Median age at index surgery, yrs (IQR)	70 (62 to 77)	71 (62 to 78)
Sex, n (%)
Female	332,469 (63)	9,589 (75)
Male	195,966 (37)	3,161 (25)
BMI, kg/m 2 , n (%)
< 18.50	3,031 (0.6)	160 (1.3)
18.50 to 24.99	110,523 (21)	2,998 (24)
25.00 to 29.99	157,458 (30)	3,564 (28)
30.00 to 34.99	89,448 (17)	2,021 (16)
35.00 to 39.99	32,543 (6.2)	684 (5.4)
> 40.00	13,049 (2.5)	278 (2.2)
Missing	122,383 (23)	3,045 (24)
Elixhauser score, n (%)
( ≤ 0)	81,483 (15)	1,968 (15)
(1 to 3)	334,060 (63)	7356 (58)
( ≥ 4)	112,892 (21)	3,426 (27)

OA, osteoarthritis; RA, rheumatoid arthritis.

Revision surgery

The type of revision surgeries was differentiated into major revisions and minor revisions based on the German procedure classification system (Operationen- und Prozedurenschlüssel (OPS)) retrieved from hospital billing data. Major revision was defined as revision surgery involving the removal and/or exchange of a bony implant component, such as femoral component exchange, while minor revision was defined as any other revision surgery without removal and/or exchange of a bony implant component.

Index surgeries were further categorized by implant fixation (fully cemented, cemented femoral component, or cementless) and cup design (e.g. dual-mobility or non-dual mobility), while revision surgeries were classified by cause (aseptic or septic), based on data recorded for both index and revision surgeries in the EPRD. Septic revision was reported when the ICD-10 code T84.5 'Infection and inflammatory reaction due to internal joint prosthesis’ was recorded for the case.

Statistical analysis

Descriptive data were reported as median and IQR for numerical variables and as frequency and percentage for categorical variables. Mann-Whitney U tests and chi-squared tests were applied to test for associations between numerical and categorical variables, respectively. To report cumulative revision rates (CRR), Kaplan-Meier estimations were calculated with 95% log-log CIs, defining revision surgery as the data point. Comparisons of median survival rates between groups were carried out using the log-rank test and of fixation methods using the Fisher’s exact test. A multivariate Cox proportional hazards models were used to compare the risk of revision between patients with OA and RA, adjusting for potential confounding variables. In addition, a separate multivariate Cox proportional hazards models were employed to assess the association of fixation type, and the risk of revision adjusting for potential confounding variables, specifically in RA patients. The significance level was set at 5%, and analyses were performed using R v. 4.2 (R Foundation for Statistical Computing, Austria).

Results

In total, 541,185 THA procedures were included during the study period. Of these, 12,750 (2.4%) were carried out in RA patients, and 528,435 (97.6%) in OA patients. For implant fixation, cementless implant fixation was the most commonly used fixation technique in both groups (RA 69%; OA 78%), as displayed in Table II.

Table II.

Fixation techniques.

Implant fixation	OA (n = 528,435), n (%)	RA (n = 127,500, n (%)	p-value*
Fully cemented	20,550 (3.9)	842 (6.6)	< 0.001
Fully uncemented	414,478 (78)	8,741 (69)
Hybrid
Cemented component	88,714 (17)	2,999 (24)
Cemented cup	4,573 (0.9)	159 (1.2)
Unknown	120 (< 0.1)	9 (< 0.1)

^*Fisher’s exact test.

OA, osteoarthritis; RA, rheumatoid arthritis.

Overall, 17,434 revision procedures (RA 604 (3.5%); OA 16,830 (96.5%)) were observed. Patients with RA showed higher CRRs at nine years for both major (RA 5.1%; OA 3.1%; p < 0.0001) and minor revisions (RA 1.7%; OA 1.1%; p < 0.0001) compared to those with OA (Figure 1). Differences in CRRs by group and cause of revision are presented in Figure 2 and Table III. After adjusting for age, sex, Elixhauser score, BMI, and fixation type, comparable differences in risk of revision between RA and OA were found for both major and minor revisions (hazard ratio (HR) 1.48; 95% CI 1.34 to 1.63; p < 0.001 and HR 1.62; 95% CI 1.40 to 1.8: p < 0.001, respectively) (Table IV).

Fig. 1.

Cumulative revision rates for overall major (top) and minor revision surgery (bottom). OA, osteoarthritis; RA, rheumatoid arthritis.

Fig. 2.

Cumulative revision rates for septic major (top) and aseptic major revision surgery (bottom). OA, osteoarthritis; RA, rheumatoid arthritis.

Table III.

Cumulative revision rates at nine years.

Revision type	OA, % (95% CI)	RA, % (95% CI)	p-value*
Major revision overall	3.1 (3.0 to 3.2)	5.1 (4.2 to 6.1)	< 0.001
Aseptic	2.3 (2.2 to 2.4)	3.6 (2.9 to 4.4)	< 0.001
Septic	0.9 (0.8 to 0.9)	1.5 (1.0 to 2.3)	< 0.001
Minor revision overall	1.1 (1.0 to 1.1)	1.7 (1.4 to 1.9)	< 0.001
Aseptic	0.6 (0.5 to 0.6)	1.0 (0.8 to 1.2)	< 0.001
Septic	0.5 (0.5 to 0.6)	0.7 (0.6 to 0.9)	0.008

^*Log-rank test.

OA, osteoarthritis; RA, rheumatoid arthritis.

Table IV.

Adjusted hazard ratio of the entire nine-year follow-up period.

Revision type	OA	RA, % (95% CI)	p-value*
Major revision overall	Ref	1.48 (1.34 to 1.63)	< 0.001
Aseptic	Ref	1.45 (1.29 to 1.63)	< 0.001
Septic	Ref	1.56 (1.30 to 1.86)	< 0.001
Minor revision overall	Ref	1.62 (1.40 to 1.88)	< 0.001
Aseptic	Ref	1.88 (1.54 to 2.30)	< 0.001
Septic	Ref	1.39 (1.12 to 1.73)	0.003

^*Multivariate Cox proportional hazard model, adjusted for sex, age, Elixahauser score, BMI, and fixation type.

OA, osteoarthritis; RA, rheumatoid arthritis.

A difference in CRRs for major revision between RA and OA was found in cases of cementless fixation (RA 5.4%, OA 3.2%; p < 0.0001) compared to cases with cemented femoral component fixation (RA 4.4%, OA 2.6%; p < 0.0001), as displayed in Figure 3. Furthermore, in RA patients, the multivariate Cox proportional hazards models showed that hybrid fixation was associated with a significantly lower risk of major revision compared with uncemented fixation (HR 0.56, 95% CI 0.42 to 0.74; p < 0.001), while no difference was found for risk of minor revision (HR 0.76, 95% CI 0.52 to 1.11; p = 0.159) (Table V). Fully cemented and cemented cup fixation as well as dual-mobility cases were not analyzed separately regarding their CRR due to the low number of cases and inadequate power.

Fig. 3.

Cumulative revision rates for overall major revision surgery according to group and fixation technique. OA, osteoarthritis; RA, rheumatoid arthritis.

Table V.

Hazard ratios of individual characteristics of the entire nine-year follow-up for rheumatoid arthritis patients.

Variable	Major revision									Minor revision
	All			Aseptic			Septic			All			Aseptic			Septic
	HR	95% CI	p-value*	HR	95% CI	p-value*	HR	95% CI	p-value*	HR	95% CI	p-value*	HR	95% CI	p-value*	HR	95% CI	p-value*
Age	1.01	1.00 to 1.02	0.055	1.01	1.00 to 1.03	0.027	1.0	0.98 to 1.02	0.910	1.02	1.00 to 1.03	0.048	1.03	1.00 to 1.05	0.021	1.0	0.98 to 1.03	0.717
Sex
Female	Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref
Male	0.77	0.60 to 0.97	0.030	0.7	0.52 to 0.94	0.017	0.95	0.62 to 1.44	0.798	0.88	0.62 to 1.25	0.475	0.7	0.42 to 1.15	0.160	1.13	0.70 to 1.85	0.614
Elixhauser score	1.04	1.02 to 1.06	< 0.001	1.02	1.00 to 1.04	0.115	1.09	1.06 to 1.12	<0.001	1.04	1.01 to 1.07	0.003	1.03	0.99 to 1.07	0.191	1.06	1.02 to 1.10	0.003
BMI, kg/m 2
Normal	Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref
Underweight	1.02	0.37 to 2.80	0.963	1.02	0.32 to 3.26	0.973	1.05	0.14 to 7.89	0.962	1.08	0.26 to 4.49	0.916	0.91	0.12 to 6.77	0.929	1.34	0.18 to 10.3	0.775
Pre-obese	1.15	0.86 to 1.55	0.339	1.11	0.79 to 1.55	0.556	1.31	0.73 to 2.36	0.370	1.16	0.76 to 1.77	0.493	0.76	0.42 to 1.38	0.367	1.81	0.96 to 3.42	0.066
Obese 1	1.16	0.83 to 1.64	0.379	1.01	0.67 to 1.50	0.980	1.68	0.88 to 3.20	0.117	1.18	0.72 to 1.94	0.503	0.94	0.48 to 1.83	0.850	1.6	0.76 to 3.37	0.215
Obese 2	1.37	0.86 to 2.19	0.186	0.9	0.48 to 1.68	0.749	2.97	1.39 to 6.31	0.005	1.04	0.48 to 2.25	0.912	0.62	0.19 to 2.08	0.441	1.75	0.63 to 4.89	0.284
Obese 3	2.65	1.58 to 4.44	< 0.001	1.47	0.70 to 3.09	0.306	6.8	3.11 to 14.8	< 0.001	2.95	1.41 to 6.14	0.004	1.53	0.46 to 5.14	0.490	5.45	2.08 to 14.3	0.001
Missing	1.16	0.87 to 1.56	0.321	1.07	0.76 to 1.50	0.703	1.45	0.81 to 2.61	0.209	1.22	0.80 to 1.88	0.355	1.4	0.83 to 2.37	0.211	0.89	0.42 to 1.91	0.773
Fixation
Cementless	Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref		Ref	Ref
Hybrid	0.56	0.42 to 0.74	< 0.001	0.55	0.40 to 0.77	< 0.001	0.57	0.34 to 0.97	0.039	0.76	0.52 to 1.11	0.159	0.58	0.34 to 0.98	0.042	1.06	0.62 to 1.83	0.832
Reverse-hybrid	2.09	1.19 to 3.65	0.010	1.85	0.91 to 3.77	0.091	2.69	1.08 to 6.71	0.034	2.16	0.94 to 4.93	0.069	1.78	0.55 to 5.69	0.333	2.77	0.85 to 8.98	0.090
Cemented	0.67	0.44 to 1.02	0.060	0.55	0.32 to 0.95	0.031	0.98	0.49 to 1.95	0.952	0.42	0.19 to 0.92	0.030	0.27	0.08 to 0.87	0.028	0.7	0.25 to 2.01	0.512

^*For Cox proportional hazards regression, Wald test was used.

HR, hazard ratio.

Discussion

The most important finding of this study was that patients with RA, compared to those with OA, continue to face a higher risk of both major and minor revision surgery after primary THA, encompassing both aseptic and septic revisions.

Historically, the overall revision rate of THA in RA has seen a continuous decrease from 12.9% between 1980 and 1989 to 6.2% between 2010 and 2019.²⁴ With the CRR among RA patients in the present study at 5.1% (overall major revision), this trend appears to be ongoing. However, the absence of this trend in the revision rate for THA in OA suggests that this is an RA-specific development.^25,26 The management of RA has seen several fundamental innovations over the last few decades potentially responsible for this development. The optimization of methotrexate therapy, earlier diagnosis through reliable clinical assessment tools, and prompt initiation of therapy have led to better outcomes and higher remission rates.^27-30 However, a causal relationship between these developments and the decrease in THA revision rates among RA patients has still to be established.

In RA, septic complications have historically accounted for the majority of revisions.²⁴ Yet, the CRR for septic revisions in RA found in the present study was lower than that for aseptic revision in both major (septic 1.5%; aseptic 3.6%) and minor revisions (septic 0.7%; aseptic 1%). Furthermore, RA and OA patients showed a similar distribution of causes for revision (RA: aseptic 65.6%, septic 34.4%; OA: aseptic 64.0%, septic 36.0%). In light of an improved management of disease activity this may suggest a shift in the predominant cause of higher revision rates in RA, from septic towards aseptic complications. While the direct impact of new therapies on revision rates has not been definitively proven, some early evidence suggests a degree of causality. For septic complications, bDMARD therapy has been linked to a reduced risk of surgical site infection compared with glucocorticoid therapy.³¹ Moreover, some reports suggest that the choice of therapeutic agent may also influence aseptic revision causes, with bDMARDs being linked to less radiological loosening in arthroplasty compared with conventional DMARDs.³²

Apart from pharmaceutical therapies and reduced bone mineral density in RA patients, implant fixation methods may also play a role in revision rates.^8-10,32 The optimal choice of fixation methods for RA patients remains a topic of debate. While older studies have traditionally supported cemented femoral and acetabular components as providing better outcomes,³³ recent research has shown promising results with cementless³⁴ or hybrid³⁵ implants for this specific patient group. In the present study, cementless fixation was reported as the most common implant fixation in both groups. However, we also found a higher rate of major revisions involving removal and/or exchange of bony implant components in patients with cementless fixation compared to those with cemented femoral component fixation (cementless 5.4%; cemented component 4.4%; p < 0.001) and a HR of 0.55 for major aseptic revision in patients with hybrid fixation compared with cementless fixation. Furthermore, while the risk of septic major revision remained relatively stable over the study period in both RA and OA patients, the risk of aseptic major revision in RA continued to increase steadily (Figure 2). This trend, while not directly analyzed in the present study, may reflect the persistent impact of mechanical risk factors described in the literature, such as reduced bone mineral density and periprosthetic fractures in this patient population.^8-10,36 Nonetheless, due to a lack of randomized controlled trials, along with conflicting results from cohort studies with high heterogeneity in implants and cementing techniques, the impact of fixation technique on revision rates in RA is yet to be determined.^33-35,37-40

A strength of the present study lies in its large dataset, derived from three independent sources. Registry studies excel at identifying trends and complications with low incidence rates. However, due to the lack of detailed information on individual patients, they offer limited insight into causality regarding specific aspects of the diseases investigated. As a result, no data on rheumatic disease activity, antirheumatic treatment, or intra-articular symptoms were available and their potential impact on the CRR could not be investigated in the present study.

In conclusion, patients with RA should be informed that they continue to carry an increased risk of revision surgery following primary THA. While the risk of septic complications is at a historical low, challenges such as compromised bone mineral density and implant integration persist, suggesting a preference for cemented implant fixation in this patient population. Future studies should aim to clarify the causal relationship between new antirheumatic therapies and revision rates in arthroplasty.

Take home message

- Patients with rheumatoid arthitis carry an increased risk of revision surgery following primary total hip arthroplasty.

- While the risk of septic complications is at a historical low, challenges such as compromised bone mineral density and implant integration persist.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.