medwireNews: An estimate of the number of individuals affected by arthritis in the USA, based on an expansive surveillance definition, has found a prevalence that is substantially higher than previously reported.
S Reza Jafarzadeh (Boston University School of Medicine, Massachusetts, USA) and David Felson (University of Manchester and Central Manchester NHS Foundation Trust, UK) used data from the 2015 US National Health Interview Survey to develop a Bayesian model to estimate the error-adjusted prevalence of arthritis based on three criteria: doctor-diagnosed arthritis; chronic joint symptoms; and the presence of symptoms for longer than 3 months.
The model, which is described in full in Arthritis & Rheumatology, estimated that 91.2 million adults in the USA, comprising 36.8% of the population, were affected by arthritis in 2015. This compares with previous estimates of 52.9–54.4 million (21.4–22.7%) adults from surveys that relied on a single question about doctor-diagnosed arthritis.
The researchers note that the higher prevalence they report “is due in large part to the previous underestimate of arthritis in adults between 18–64 years of age,” finding that 61.1 million (30.6%) adults in this age group were affected by arthritis in 2015.
An assessment of the performance of the surveillance criteria indicated that, in individuals aged 18–64 years, the criterion of doctor-diagnosed arthritis had very low sensitivity, while that of symptoms duration had the highest sensitivity (albeit the lowest specificity). Thus, the researchers comment, “a substantial (i.e. 65–80%) fraction of the population with arthritis, who are between 18–64 years of age, but are misclassified as healthy by the doctor-diagnosed arthritis criterion due to low sensitivity, are captured by the two remaining questions on joint pain, aching or stiffness.”
The study authors acknowledge some limitations of their model-based approach, in particular the inability to directly apply weighting, which “may have introduced inaccuracy with regard to the precision of our estimates.” But they say “the gain in accuracy with regard to misclassification bias is so substantial that we believe it justifies our choice.”
In an accompanying editorial, Jeffrey Katz (Brigham and Women’s Hospital, Boston, Massachusetts, USA) commented: “[This work] moves our field a large step forward, both by introducing a set of techniques that permits more nuanced use of existing survey data and by highlighting the limits of the specific questions deployed at present in national prevalence surveys.
“While additional research will continue to refine the estimates of Jafarzadeh and Felson, the adjusted prevalence estimates these authors present are sufficiently robust and concerning to justify a policy agenda encouraging greater investment of scarce resources into the pathogenesis, prevention and treatment of arthritis conditions.”
By Catherine Booth
medwireNews is an independent medical news service provided by Springer Healthcare. © 2017 Springer Healthcare part of the Springer Nature group