Rheumatoid arthritis (RA) is a chronic inflammatory disease with a prevalence of approximately 0.8 (0.3–1.0) per 100 adults. The peak in the onset of the disorder ranges between 55 and 64 years of age.1–3 As a consequence of this condition, RA patients consume a great deal of health care resources, and the costs are high. Given the cost of treatment and the loss in productivity, this disease has a great impact on society. Studies on the costs of the disease conducted in Europe and the United States estimate a figure of around €4000–€6000 per patient per year.4–6

In Spain, the management of inflammatory rheumatic diseases has undergone marked advances due to several factors such as the availability of new treatments, a deeper understanding of the processes and progress in the care of the rheumatic patient.

The increase in the use of biological agents in routine rheumatology practice has raised the direct costs 3 to 6-fold in several European countries.7,8 However, changes in the patterns of treatment have achieved a decrease in the mean disease activity (Disease Activity Score 28 joints [DAS28]), hospital stays, lost work days and work disability,9 thus resulting in a reduction in the indirect costs related to the disease.

In a number of studies performed in the 1980s and 1990s,10–13 the major medical expense was hospital admission, which, in those articles, represented a third of the cost. In later studies, the greatest cost corresponded to the drugs prescribed.14 This change was associated with a growing percentage of patients treated with biological agents.

The purpose of the study entitled “Variability in the management of rheumatoid arthritis and spondyloarthritis in Spain” (EMAR-II) was to describe variability in terms of health care resource consumption and the use of techniques and treatments and, thus, made it possible to define these changes in this country. Hospitals of different autonomous communities participated in the study, which had a duration of 2 years (Appendix A). This enabled us to recruit a representative cohort of the disease in Spain.

There are few studies on the costs of RA in the biological era, and those that exist have been conducted in settings in which these agents were still little employed. Thus, the objective of our report was to estimate the direct medical costs corresponding to the EMAR-II cohort, analyze factors that may be predictive of those costs and evaluate the predictive factors in 2 distinct subpopulations: patients who were treated with biological agents during the study period and those who did not receive them.

The final study sample consisted of a total of 1095 patients for whom all of the data on medical costs were available.

In all, 26% of the patients included in the study were men. The mean age was 62±14 years. With respect to the clinical characteristics of the sample, the duration of the disease was 10.2±8.9 years. Overall, 49.8% of the patients had some type of associated comorbidity, distributed as follows: nodules (12.16%), Sjögren's syndrome (10.53%), carpal tunnel syndrome (6.82%), Raynaud's phenomenon (2.5%), C1–C2 subluxation (2.5%), pulmonary fibrosis (2.3%), pleuritis (1.4%), vasculitis (0.9%), scleritis (0.9%), Felty's syndrome (0.4%) and amyloidosis (0.15%). The highest DAS28 score during the study period was recorded in patients with low activity (DAS≤3.2) in 27.76%, moderate activity (DAS 3.2–5.1) in 45% and high activity (DAS>5.1) in 27.24%. The mean minimum ESR was 15.26±14.10 and the maximum values were 38.30±27.13. The SPA had minimum values of 15.26±14.10 and the maximum values were 38.30±27.13. Lastly, for the disease activity variable generated for this study (a combination of DAS28 in active disease [≥3.2] in 2 visits or an elevated ESR [≥50mm/h] in 2 measurements or an elevated SPA [≥moderate level] in 2 visits), we found that 26.14% of the patients had high activity and 73.86% of the patients had low activity.

In all, 67.3% of the participants took NSAID during the study period and 67.6% received corticosteroids. Overall, 20.7% of the patients had received 2 or more DMARD simultaneously over the preceding 2 years. The DMARD most widely administered during the study period was methotrexate (59.6%), followed by leflunomide (22.1%), antimalarial drugs (12.2%) and sulfasalazine (3.1%). Altogether, 36.95% of the sample received some type of biological therapy during the study period. The agents most widely used were adalimumab (27.3%), infliximab (20.5%); etanercept 50mg (19.9%), etanercept 25mg (12.2%) and rituximab (7.6%).

The costs were calculated using the monetary value assigned to the health services according to the different data sources consulted. The total direct medical cost per patient was €24,291±45,382 throughout the 2 years of the study. The total expenditure for drugs (DMARD, biological agents, NSAID, corticosteroids and analgesics) was €19,567±43,560. If we exclude the cost of biological drugs, the mean cost per patient was €3742±3711. The distribution of the costs is shown in Figure 1. There was a statistically significant gender-related difference, with an average cost of €17,330±1993 for men and of €26,814±1720 for women (P=.002).

Fig. 1.

Distribution of the direct medical costs over the 2-year EMAR-II study.

(0.12MB).

With respect to differences between the groups, the patients who received biological agents had a significantly higher cost in relation to the following variables: visits to the rheumatologist; visits to the nursing service; orthopedic surgery without prosthesis; orthopedic surgery of the spine; social service counseling; psychology unit; rehabilitation services; complete blood counts; laboratory testing; sedimentation; protein electrophoresis; serology; cultures; rheumatoid factor; ESR; CRP; anti-cyclic citrullinated peptide antibodies; chest radiography; axial bone radiography; magnetic resonance imaging; Mantoux test; pulmonary function tests; arthrocentesis; and hospital admission.

There were no differences in the costs for the other variables evaluated. Although it was not statistically significant, the patients who received biological agents were less costly in terms of visits to other specialists, tests for human leukocyte antigen B27, abdominal radiographs, certain other types of radiographs and abdominal ultrasound.

Bivariate linear regression analysis showed that the variables associated with direct medical costs were patient sex—with higher costs for women (P=.002), patient age (P=.00), disease duration (P=.00), disease activity (P=.00) and treatment with biological drugs (P=.00).

Multivariate regression analysis was subsequently performed with all of the variables that had been significant in the bivariate analysis, which were adjusted for the sociodemographic variables (Table 1). In the analysis of the predictors of medical costs for the sample as a whole, the predictive factors were biological agents (P=.00), comorbidity (P=.00) and patient age (P=.01).

A bivariate model was utilized in the analysis by subgroup. For the group of patients who did not receive biological drugs, the variables associated with costs were comorbidity (P=.00) and patient age (P=.01). For the group of patients who took biological drugs, the variables that were associated with costs were disease duration (P=.02), patient sex—with higher direct medical costs for women (P=.01)—and disease activity (P=.00).

Finally, 2 models of multivariate regression were built, one for each group of patients, including the variables that were statistically significant in the bivariate analysis. In the analysis of the direct medical cost predictors in the group of patients that did not receive biological agents during the study period, the independent predictive factors identified were comorbidity (P=.00) and patient age (P=.02) (Table 2). In the analysis of the predictors of direct medical costs in the group of patients who took biological drugs during the study period, the independent factors detected were patient sex (P=.03) and disease activity (P=.02) (Table 3).

This retrospective study calculates the direct costs involving the recourses utilized by the patients. The results of this report provide an extensive characterization of RA patients in Spain. Their distribution in terms of age and sex was similar to that of other published series,24 the mean age of the patients was somewhat over 60 years of age and 26% were men.

The costs (excluding biological drugs), with a mean of €3742 during the study period, were lower than those reported for patients in other studies, in which the annual cost was around €3980.25

It appears to be clear that biological agents are the bulk of these direct medical costs, although other factors have a role, like disease activity, in the case of patients taking biologicals. In individuals who do not receive these drugs, patient age and comorbidities are major predictors of direct medical costs.

There are studies that show that the increase in the cost of therapy produced by the introduction of biological agents has decelerated since 200926 since the rate of prescriptions is growing, but more sequentially. This has also resulted in the optimization of the drugs with less rigid protocols that can be better adapted to the characteristics of the patient. Given the trend toward early treatment, it is expected that the costs will stabilize over time.

It seems evident that the direct medical costs of RA patients have risen due to the introduction of new treatments, as well as to the aging of the patients, which is increasing in parallel to the age of the general population. However, although this study does not focus on indirect costs, other reports indicate that they are being compensated by reductions in other costs. In fact, a recent study shows a continuous increase in annual medical costs for patients with RA who were being treated by German rheumatologists between 2002 and 2011, caused by the growing use of biological agents. However, the rise was partially offset by the decrease in the costs of hospital admissions, sick leave and work disability, with a reduction in lost work days and fewer cases of permanent disability.27

On interpreting the results of this study, certain limitations must be taken into account. Data were collected for only 2 years and, thus, the interpretation of the findings should be done with caution. Moreover, it should be kept in mind that certain data may not be extractable from the medical records or not be documented, both of which influence the validity of the results.28 Likewise, the cost data are not exact, but rather are assertions taken at the national level. The costs per patient can be incorrectly estimated, especially because differences related to the community and center were not taken into account. It is important to point out that, because of the type of study, it was impossible to establish the indirect costs associated with work productivity and other causes. Therefore, we were not able to estimate the impact on the indirect costs that could have resulted from a better therapy management or the introduction of biological drugs, as has been seen in other studies.

In conclusion, this study presents a detailed map of the direct medical costs associated with RA patients in Spain. These data may contribute to the understanding and estimation of the economic impact of the disease, to their comparison with data from other geographic areas, to the development of strategies for the reduction of these costs and to an increase in quality of care. Along this line, recommendations on the optimization of drugs and treatment profiles adapted to the patient are of major importance, as well as the possibility of establishing nurse counseling to instruct patients and broaden their knowledge of the disease, thus reducing medical visits and unnecessary tests in many cases.

The EMAR-II study was promoted by the Spanish Society of Rheumatology and financed by the pharmaceutical company AbbVie.

The EMAR-II study was financed by the pharmaceutical company AbbVie.