Algorithm Uncovers New Insights into Rheumatoid Arthritis Diagnoses

Algorithm Uncovers New Insights into Rheumatoid Arthritis Diagnoses

shutterstock_210108730A new study entitled “Validation of the rheumatoid arthritis diagnosis in the Swedish National patient register: a cohort study from Stockholm County” reports than the combination of patients’ data registered in National Patient Register databases, together with algorithms is a useful tool to perform epidemiological studies for rheumatoid arthritis. The report was published in the journal BMC Musculoskeletal Disorders.

In this study, the authors used the Swedish health information network, specifically the National Patient Register, to identify both prevalent and incident cases of rheumatoid arthritis (RA). The National Patient Register includes all hospital discharges from patients whose condition required admission to a hospital (inpatient care), as well as patients’ visits to a hospital for diagnosis or treatment but who are not hospitalized (outpatient care). The register also includes diagnosis information, but while validation studies were already performed for the inpatients, for the outpatients the validity of the diagnosis performed with the algorithms of the register is still lacking. Thus, in this study the authors assessed the validity of the RA diagnosis by the Swedish National Patient Register and also to determine if the algorithms can identify incident RA cases.

The authors analyzed 211 individuals – 100 defined as RA prevalent patients (with two or more visits to a rheumatologist between 2005 and 2008) and 111 individuals defined as RA incident patients (who had their first visit to a rheumatologist in 2008). The authors analyzed diagnosis-related parameters retrieved from patients medical records and determine patients fulfilling RA criteria according to the 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) classification criteria and the 1987 ACR classification criteria. The incident group was defined using a strict algorithm and required the fulfillment of two additional criteria, besides the above mentioned – within one year of the first visit, the patient had a second visit to the hospital; and no disease-modifying anti-rheumatic drugs (DMARDs) therapeutics within 6 months before the first visit.

Researchers identified that the register-based RA diagnosis was valid for both prevalent patients, as well as those with new-onset RA, with approximately 90% validity in both cases. Thus, using specific algorithms together with data from the National Patient Register is a useful new tool to identify RA individuals, which will allow for reliable epidemiological studies and contribute to our deeper understanding of RA epidemiology and therapeutical outcome.

Leave a Comment