Article Text
Abstract
Objectives To determine whether men and women with rheumatoid arthritis are prescribed anti-tumour necrosis factor (anti-TNF) treatment at different levels of disease activity.
Methods Data from the Swedish national biologics registry ARTIS were used to analyse characteristics of patients' disease at the start of the first anti-TNF treatment. Means for men and women were compared using t-tests, and non-normally distributed covariates were compared using the Wilcoxon rank-sum test. Linear regression models, adjusted for age and calendar year, were used to investigate the association between sex and each disease activity measurement.
Results Women were younger and had longer disease duration at treatment start than men. Tender joint count, erythrocyte sedimentation rate, patient's global assessment, patient-reported pain and health assessment questionnaire scores were significantly higher in women, whereas men had a higher level of C-reactive protein (p<0.05 for all comparisons). Swollen joint count and physician's global assessment did not differ by sex.
Conclusions For women with rheumatoid arthritis, treatment with anti-TNF therapy was initiated at a higher level of subjective disease activity than for men, but at the same level of physician-reported disease activity. These data imply that patients' subjectively experienced disease activity may be discounted in the treatment decision.
Statistics from Altmetric.com
Introduction
The prevalence of rheumatoid arthritis (RA) is estimated to be two to four times higher in women than in men, depending on age, and women are younger at RA onset.1 In some studies, women's long-term disease outcomes have been shown to be worse compared with men's; women have lower remission rates and a worse prognosis in terms of functional capacity.2,–,6 These sex differences might be related, at least partly, to differences in normal physiological processes in men and women and to sex hormones involved in these processes.3 4 7,–,10 However, the differences in disease outcomes between the sexes depend on how the outcomes are defined. Some studies have suggested that the measures used to ascertain disease activity, particularly the 28-joint-count disease activity score (DAS28), are not sex neutral and might be influenced by factors other than true disease activity.11,–,13
Differences in disease outcomes may be physiological, or due to how the outcome is measured or how patients are treated. Since their introduction in 1999, anti-TNF drugs have greatly improved the therapeutic possibilities for patients with RA. TNF plays a key role in the pathophysiological processes in RA and is a major therapeutic target. Therefore, sex differences in the use of biologics could lead to differences in the long-term outcomes of RA between men and women. Assessing whether men and women receive biologics at the same time-point in the disease course becomes more important as the treatment is shown to be the single strongest determinant of outcome for patients with RA.5
In this study, we examined measurements of disease activity in patients at anti-TNF treatment initiation using the Swedish biologics registry, Anti-rheumatic Therapy in Sweden (ARTIS),14 in order to determine whether female patients receive their first anti-TNF prescription at the same level of disease activity as male patients, or if a sex bias in prescription practices is present.
Methods
Participants
Anti-TNF treatment in Sweden has been systematically registered in the Swedish biologics registry, ARTIS, since such therapies first became available in 1999, providing for comprehensive data on disease characteristics in the patients for whom these therapies were chosen. ARTIS has been estimated to cover 87% of all patients with RA treated with biologics,14 and in the Swedish RA population in 2007 the penetration of biologics was estimated as 15–18% overall and 20–25% in working-age patients.1 Anti-TNF biologics in Sweden are provided within the national healthcare system and financed through a universal health insurance. Adult patients with RA in ARTIS who initiated treatment with the anti-TNF medications infliximab, etanercept or adalimumab (not available until 2003) from January 1999 through 2008 were identified, as described in greater detail elsewhere.15
Disease activity measurements
In ARTIS, clinical data at each patient's visit were recorded by the treating rheumatologist. Patient-reported outcomes included the patient's assessment of global disease activity, pain by visual analogue scale, tender joint count (TJC) and results from the health assessment questionnaire (HAQ) score. Swollen joint count (SJC) and physician's global assessment were reported by the physician. Objective measurements collected were erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels. DAS2811 was calculated using four components: the TJC, SJC, ESR and patient's global assessment. If all four components were not available, DAS28 was calculated using three core variables.16
Statistical methods
We calculated summary statistics of disease characteristics at initiation of anti-TNF treatment by treatment start year in men and women separately. Means for men and women were compared using t-tests, and non-normally distributed covariates were compared using the Wilcoxon rank-sum test; p<0.05 was considered statistically significant. Linear regression models adjusted for age and calendar year of treatment start were used to investigate whether sex was associated with disease activity measurements, transforming variables (ESR and CRP) if necessary. Unadjusted means and medians, as appropriate, were plotted over time for men and women.
Results
During the study period, 9302 patients began treatment on an anti-TNF, of which 7098 patients (76%) were women. Women were significantly younger (55.1 vs 56.3 years) and had longer disease duration at treatment start (11.5 vs 10.5 years; p<0.05).
Of the male patients, 47% began treatment with infliximab, 36% with etanercept and 17% with adalimumab. Infliximab was also the most common treatment prescribed among women (44%) followed by etanercept (39%) and adalimumab (17%). Although over the study period infliximab was the most commonly prescribed first anti-TNF, in the later years etanercept dominated.15 Non-steroidal anti-inflammatory drugs, glucocorticoids or methotrexate were being taken by 87% of all patients at start of anti-TNF treatment. An equal proportion of women and men were taking non-steroidal anti-inflammatory drugs (53%). Glucocorticoids were used by 50% of female patients and 55% of male patients (p<0.05). Concomitant methotrexate was received by 65% of female patients and 67% of male patients (p=0.1). Data on concomitant disease modifying antirheumatic drugs were missing for 265 patients (3%).
All disease characteristics at biologic initiation differed statistically significantly by sex (p<0.05), except for SJC and physician's global assessment of disease activity (table 1). The mean DAS28 at anti-TNF initiation was 5.4 for women and 5.2 for men. TJC was also somewhat higher for women than for men (8.8 vs 7.8). Median CRP was significantly higher in men (21.0 vs 18.0 in women). The patients' assessment of global health was significantly higher for women than for men (60.2 vs 57.1), as was the assessment of pain (59.1 vs 56.4). HAQ values also differed significantly, 1.4 versus 1.1 for women and men, respectively. Adjusting for calendar year and age did not change the interpretation of the association between sex and each disease measurement, with the exception of ESR, which was significantly associated with sex after adjustment for age and calendar year of anti-TNF initiation. Disease activity at treatment start was associated with calendar year with a progressively lower disease activity at anti-TNF initiation in later years. The trends of disease characteristics over time (not adjusted for age) are shown in figures 1 and 2.
Discussion
In this study, we examined RA patients' characteristics when the decision to initiate anti-TNF therapy was made in order to determine if such therapy decisions are made in a sex-neutral fashion. We found that several disease activity measures, including the DAS28, TJC, patient's global health assessment, patient's assessment of pain, ESR and HAQ were significantly higher (ie, worse) for women than for men at this time-point. By contrast, physician-assessed SJC and physician's global assessment did not significantly differ by sex, and CRP was higher in men. The difference in DAS28 was explained mostly by the components with a more subjective nature, namely, the TJC and the patient's global assessment of disease activity. Our study's focus was on the measures of disease activity at biologic start in a clinical setting. The strength of this simple analysis lies in the large number of patients over 10 years from a population-based registry in a country where biologics use is largely unrestricted.
The South Swedish Arthritis Treatment Group's paper on anti-TNF response presented similar differences between the sexes in disease activity measurements, although the focus was not on sex differences.17 In comparison with the southern Sweden data, which was a subset of our data, our study had a longer follow-up and included patients from all regions of Sweden, therefore giving us the power to detect small differences and the ability to adjust for age and calendar year.
Our results indicate that for female patients with RA, treatment with anti-TNFs is started at a higher level of disease activity when measured by subjective but not by objective measures. This could mean that patients' experience and reporting of pain/discomfort is different in men and women. Sex differences in the processing of peripheral pain are well established18 and may be explained by biological or social factors.18 19 In contrast to the patient-reported disease activity measurements (HAQ, pain, patient global), using the physician's global assessment as a representation of the physician's point of view, treatment is initiated for male and female patients at the same level of the physician's perceived disease activity. Thus, if the physician's perception of disease does not differ by sex while the perception of patients who experienced pain and tenderness differs, perhaps physicians are not including the patient's experiences in their assessment to a sufficient degree. Another interpretation may be that physicians consider anti-TNFs specifically in the context of controlling objective disease manifestations, and in particular the long-term destructive process, and not (or not as much) in the context of trying to alleviate subjective symptoms. As the ultimate goal of treatment in chronic painful diseases such as RA is the alleviation of subjectively experienced discomfort, it could be argued that the patient's subjectively experienced symptoms should play a more important part in the treatment decision.
In summary, our data imply that the subjective aspects of disease activity are to some extent discounted in the treatment decision, so that women with RA are receiving anti-TNFs at a higher level of subjective disease activity than men.
Acknowledgments
The ARTIS Study Group conducts scientific analyses using data from the Swedish Biologics Register ARTIS run by the Swedish Society for Rheumatology. For the maintenance of this register, the Swedish Society for Rheumatology has received funding, independent of the conduct of these scientific analyses, from Schering-Plough, Bristol-Myers Squibb, Wyeth, Abbott Laboratories and Roche. The authors were in charge of and solely responsible for all data collection, analysis, interpretation and writing of the manuscript, without any constraints exerted from the agencies or companies that provided financial support for the study, or for ARTIS, respectively. The authors thank all the clinicians who have registered patients, and the reviewers for very helpful comments. The ARTIS study group : E Baecklund, L Coster, N Feltelius, C Forsblad-Elia, P Geborek, L Jacobsson, L Klareskog, S Lindblad, S Rantapaa-Dahlqvist, T Saxne and R van Vollenhoven.
References
Footnotes
-
Ethics approval The study was approved by the ethics committee of Karolinska Institutet.
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.