Variability in the prescription of non-biologic disease-modifying antirheumatic drugs for the treatment of spondyloarthritis in Spain

Abstract

Objective

To describe the variability in the prescription of non-biologic disease-modifying antirheumatic drugs (nbDMARDs) for the treatment of spondyloarthritis (SpA) in Spain and to explore which factors relating to the disease, patient, physician, and/or center contribute to these variations.

Methods

A retrospective medical record review was performed using a probabilistic sample of 1168 patients with SpA from 45 centers distributed in 15/19 regions in Spain. The sociodemographic and clinical features and the use of drugs were recorded following a standardized protocol. Logistic regression, with nbDMARDs prescriptions as the dependent variable, was used for bivariable analysis. A multilevel logistic regression model was used to study variability.

Results

The probability of receiving an nbDMARD was higher in female patients [OR = 1.548; 95% confidence interval (CI): 1.208–1.984], in those with elevated C-reactive protein (OR = 1.039; 95% CI: 1.012–1.066) and erythrocyte sedimentation rate (OR = 1.012; 95% CI: 1.003–1.021), in those with a higher number of affected peripheral joints (OR = 12.921; 95% CI: 2.911–57.347), and in patients with extra-articular manifestations like dactylitis (OR = 2.997; 95% CI: 1.868–4.809), psoriasis (OR = 2.601; 95% CI: 1.870–3.617), and enthesitis (OR = 1.717; 95% CI: 1.224–2.410). There was a marked variability in the prescription of nbDMARDs for SpA patients, depending on the center (14.3%; variance 0.549; standard error 0.161; median odds ratio 2.366; p < 0.001). After adjusting for patient and center variables, this variability fell to 3.8%.

Conclusion

A number of factors affecting variability in clinical practice, and which are independent of disease characteristics, are associated with the probability of SpA patients receiving nbDMARDs in Spain.

Introduction

Spondyloarthritis (SpA) encompasses a group of disorders with a common genetic background and similar clinical and radiological features. This group is divided into five subtypes, which includes the following: ankylosing spondylitis (AS), which is the major subtype; psoriatic arthritis (PsA); inflammatory bowel disease (IBD)-associated arthritis; reactive arthritis (ReA); and undifferentiated SpA (USpA) [1]. Their prevalence varies between 0.3% and 1.1%, depending on the study in question [2], [3], [4]. The most important clinical features of SpA are inflammatory back pain, asymmetrical oligoarthritis, which predominates in the lower limbs, enthesitis and dactylitis, and specific extraskeletal manifestations such as psoriasis, uveitis, and chronic IBD. Pain, morning stiffness, progressive functional limitation, fatigue, and diminished quality of life characterize the clinical patterns of these patients [5]. Goals for SpA treatment include controlling pain and stiffness, maintaining function, treating and preventing extraskeletal manifestations, slowing or halting radiographic progression, controlling inflammation and, finally, avoiding work loss and improving quality of life [6], [7], [8]. Until the development of tumor necrosis factor (TNF)-blocking agents, the therapeutic armamentarium in SpA was limited to non-steroidal anti-inflammatory drugs (NSAIDs) [9]. For patients refractory or intolerant to NSAIDs, non-biologic disease-modifying antirheumatic drugs (nbDMARDs), by definition in fact, used mainly in rheumatoid arthritis (RA), can serve as a second-line approach [10]. Although no single nbDMARD has proven to be completely effective, either in improving symptoms or in altering the natural history of the disease in SpA patients [11], they are still the most frequently prescribed drugs in certain parts of the world [12], [13]. The Assessment in Ankylosing Spondylitis International Society/European League Against Rheumatism (ASAS/EULAR) recommendations for the management of AS state that there is no evidence for the efficacy of nbDMARDs for the treatment of axial disease and that sulfasalazine (SSZ) may be considered in patients with peripheral arthritis [14].

Variability in clinical practice is defined as the differences in the care process and/or outcome of care of a particular clinical problem among different providers after controlling for demographic and sociocultural circumstances and health status [15]. This variability is influenced by several factors, including those related to the demand of care, characteristics of the health professionals, and the health system itself [16], [17], [18]. The lack of agreement about specific interventions, such as the use of nbDMARDs for SpA, may also result in different approaches to the same problem. Although the use of healthcare resources and indications of procedures should mainly be driven by patient- and disease-related characteristics, care is often strongly influenced by subjective factors related to the attitudes of individual physicians, institutional structures, and/or healthcare service features and policies [19]. This situation leads to the subsequent mismanagement of resources, patient care, and funding for the health system [20].

The prescription of nbDMARDs for SpA in Spain has not been extensively documented, and only a few studies in rheumatology settings have been carried out [21]. In order to design strategies to help decrease the variability and improve the quality of clinical practice, we first need to understand the dimensions of the problem and examine its determinants and effects. The objective of this study, therefore, is to describe the variability in the prescription of nbDMARDs for the treatment of SpA in Spain and to explore which factors related to the disease, patient, physician, and/or environment contribute to these variations.