Background Treatment of patients with rheumatoid arthritis (RA) has improved markedly over the past 25 years.
Objective To investigate whether rates of joint surgery, a long-term consequence of poorly controlled RA, have changed over this period.
Methods In this population-based, serial cross-sectional study of patients with RA aged ≥40 years in California, trends in annual rates of total knee arthroplasty, total hip arthroplasty, total ankle arthroplasty or arthrodesis and total wrist arthroplasty or arthrodesis from 1983 to 2007 were examined.
Results Rates of joint surgery peaked in the 1990s and since have decreased. Among patients aged 40–59 years, rates of knee surgery in 2003–2007 were 19% lower than in 1983–1987 (adjusted rate ratio 0.81; 95% CI 0.74 to 0.87, p<0.0001), while rates of hip surgery in 2003–2007 were 40% lower (p<0.0001). Rates of knee and hip surgery did not decrease in patients aged ≥60 years but increased as observed in the general population. Compared with rates of ankle and wrist surgery in the mid-1980s, rates in the mid-2000s decreased signifi cantly in both age groups.
Conclusions Rates of joint surgery in RA peaked in the 1990s and have declined thereafter, suggesting that longterm outcomes of RA are improving.
Statistics from Altmetric.com
A long-term sequela of inadequately controlled chronic synovitis in patients with rheumatoid arthritis (RA) is joint damage severe enough to require surgery.1 Several studies have suggested that rates of arthroplasty among patients with RA have decreased in recent decades.2,–,7 Although the reasons for this are uncertain, these decreases mirror changes in other measures of RA severity and followed the shift to more aggressive antirheumatic treatment.8 We previously found that rates of total knee arthroplasty in patients with RA in California declined in 1998–2001 after peaking in the mid-1990s.3 It is unknown if these rates have continued to decrease. In this population-based, serial cross-sectional study, we examined rates of total knee arthroplasty through 2007 and investigated whether rates of total hip arthroplasty and ankle or wrist surgery among patients with RA in California decreased between 1983 and 2007.
Source of data
We used data on inpatient hospitalisations from the California Offi ce of Statewide Health Planning and Development9 from 1983 (the fi rst year of the database) to 2007. This offi ce receives discharge abstracts, which include a record of diagnoses and procedures, for all patients admitted to hospital and performs extensive reliability checks to ensure data accuracy. The study protocol was exempted from human subjects review by the institutional ethics committee.
Patients and procedures
We tabulated the number of hospitalisations in which total knee arthroplasty, total hip arthroplasty, ankle arthrodesis or arthroplasty (hereafter ‘ankle surgery’), or wrist arthrodesis or arthroplasty (hereafter ‘wrist surgery’) were recorded. We limited the analyses to patients with RA aged ≥40 years because the inclusion of younger patients at low risk for surgery would have decreased our ability to detect changes over time. Because unique patient identifi ers were not available before 1991 or after 2000, analyses were based on the number of procedures and not the number of patients.
We computed the annual crude rate of each procedure by dividing the number of procedures in each year by the estimated number of patients with RA in California in each year. We estimated the number of patients with RA by applying age, sex and racespecifi c prevalences of RA (based on data from the Third National Health and Nutrition Examination Survey10) to yearly estimates of the population of California.2 11 We standardised crude rates to the age, sex and race distribution of the RA population in 2000. To provide more stable estimates, we pooled data in fi ve equal 5-year periods and used Poisson regression models to compare adjusted rates across time periods. Adjusted rate ratios were computed using both the fi rst 5-year study period (1983–1987) or the peak time period as the referent group. We used a similar approach to calculate adjusted rates of procedures from 1998 to 2007 among patients aged ≥40 years in the general population. Data on hospitalisations not related to RA for years before 1998 were not available. Analyses were performed using SAS version 9.1.3 (SAS Institute, Cary, North Carolina, USA).
Joint surgery in patients with RA
The adjusted rate of total knee arthroplasty was 459 per 100 000 people with RA in 1983, attained a maximum of 556 per 100 000 people in 1997 and then decreased to 467 per 100 000 people in 2007 (figure 1A). The relative risk of total knee arthroplasty was 9% higher in 1993–1997 than in 1983–1987, but was signifi cantly lower by 8% in 1998–2002 (table 1).
The rate increased slightly in 2003–2007. The relative risk of total knee arthroplasty was 12% lower in 2003–2007 than in the peak period of 1993–1997. The recent increase in rates was concentrated among patients aged ≥60 years. Among those aged 40–59 years, there was a progressive decrease in rates overtime, with a relative risk of total knee arthroplasty that was 19% lower in 2003–2007 than in 1983–1987.
Rates of total hip arthroplasty steadily increased from 1983 to 1996, and then declined (figure 1B). The risk of total hip arthroplasty was 32% higher in 1993–1997 than in 1983–1987, and nearly the same in 2003–2007 as in 1983–1987 (table 1). Compared with the peak period of 1993–1997, the risk of total hip arthroplasty was 23% lower in 2003–2007. Similar to the fi ndings for total knee arthroplasty, rates of total hip arthroplasty decreased progressively overtime among those aged 40–59 years, with a relative risk that was 40% lower in 2003–2007 than in 1983–1987.
Rates of ankle surgery peaked in the mid-1990s (figure 1C). The relative risk of ankle surgery was 21% higher in 1993–1997 than in 1983–1987 (table 1). Thereafter, the risk decreased, such that the relative risk was 39% lower in 2003–2007 than in 1983–1987.
Adjusted rates of wrist surgery in patients with RA increased from 1983 to 1990 before decreasing in the most recent periods (figure 1D). The relative risk of wrist surgery was nearly twofold higher in 1988–1992 than in 1983–1987, but 59% lower in 2003–2007 than in 1983–1987 (table 1). Trends in rates of ankle and wrist surgery were similar in younger and older patients. For all four procedures, trends were similar in men and women.
Joint surgery in the Californian general population
Between 1998 and 2007, annual adjusted rates of total knee arthroplasty among the general population increased steadily, from 260 per 100 000 people in 1998 to 360 per 100 000 people in 2007. Adjusted rates of total hip arthroplasty decreased slightly from 332 per 100 000 people in 1998 to 290 per 100 000 people in 2007, while adjusted rates of ankle surgery were stable (5.8 per 100 000 people in 2000 to 5.1 per 100 000 people in 2007). Adjusted rates of wrist surgery decreased slightly from 0.4 per 100 000 people in 2000 to 0.2 per 100 000 people in 2007.
Rates of joint surgery in patients with RA in California peaked in the 1990s and have since decreased. Rates of ankle and wrist surgery were markedly lower in the mid-2000s than in the mid- 1980s. Rates of total knee arthroplasty and total hip arthroplasty also decreased substantially (19% and 40%, respectively) among patients aged 40–59 years. Rates of total knee and hip arthroplasty did not decrease in older patients, and paralleled increases seen in the general population.
Our results are consistent with other recent reports of temporal trends in rates of orthopaedic surgery among patients with RA.4,–,7 In a population-based retrospective study of patients with RA in Central Finland, rates of total hip or knee replacement from 1986 to 2003 were essentially stable.4 Rates of joint arthroplasty among patients with chronic infl ammatory arthritides (86% with RA) in Norway from 1994 to 2004 either decreased (hip) or were stable (knee), while rates of ankle and wrist arthrodesis signifi cantly decreased.5 A study of the Swedish National Hospital Discharge Register found declining rates for lower extremity surgeries from 1987 to 20016 and upper extremity surgeries from 1998 to 2004 among hospitalised patients with RA.7
Reasons frequently proposed to account for the recent declines in radiographic severity, functional impairment and rates of joint surgery in RA are either that these represent responses to improvements in the way antirheumatic drugs are used, or that RA is evolving to become milder.12,–,14 Both changes in treatment and changes in the natural history may have an infl uence on the severity of RA, but the pattern of a gradual increase in rates of joint surgery through the 1990s, followed by a more recent decline, is more consistent with a treatment effect. A gradually progressive decline in rates might have been expected if a change in the natural history of RA was primarily responsible. The timing of the decline is also consistent with the time necessary for the new treatment strategy to be widely adopted, and for the fi rst generation of patients treated under the new strategy to replace previous generations and reach the stage of RA when joint operations become more common.15 This may also account for the observation that decreases in rates of joint surgery were more pronounced among younger patients. This time lag also suggests that the effect of biological agents on rates of joint surgery at the population level may not be evident for some time.
The strengths of our study include the investigation of rates of four types of joint surgery over a 25-year study period in a large population-based sample. Potential limitations of our study include possible coding errors in the database which might have misclassifi ed some patients. However, validation studies showed that major diagnoses were accurately coded.3 We also could not determine whether the same patient underwent multiple operations (eg, bilateral joints or revisions). Therefore, our rates were probably higher than rates based on the number of patients with joint surgery.
Our results demonstrate that rates of joint surgery in patients with RA peaked in the 1990s and have since decreased. Wrist and ankle surgery declined markedly, while progressive decreases in rates of total knee arthroplasty and total hip arthroplasty occurred in younger patients. Collectively, these fi ndings suggest that long-term health outcomes of RA are improving.
Funding This work was supported by the Intramural Research Program, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health.
Competing interests None.
Ethics approval The study was exempted from human subjects review by the National Institutes of Health Offi ce of Human Subjects Research.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.