Dr. Jafarzadeh notes the true accuracy of the doctor-diagnosed arthritis survey question is even less in the real world because the NHIS participants’ access to healthcare may have been more limited than people who participated in the validation study. “This means the rate of misclassification by using a single survey question on doctor-diagnosed arthritis is potentially even higher than what was reported in the validation study,” he says.
One strategy to increase accuracy is to combine results of individual diagnostic criteria such as self-reports of arthritis diagnosis and arthritis-like symptoms, the article states. This would address concerns about participants 64 years old or younger who may have arthritis yet were misclassified as healthy under the single criterion of doctor-diagnosed arthritis.
The Nuts & Bolts
In their study, Dr. Jafarzadeh and David T. Felson, MD, MPH, of the Boston University School of Medicine, analyzed data from the 2015 survey using expanded surveillance criteria to reexamine the prevalence of arthritis in the United States. To account for arthritis cases that might have been missed in the current national figures, they developed a Bayesian multinomial latent class model that allowed them to estimate prevalence without knowing the true disease status of each survey participant.
“Our analytic approach explicitly adjusts for the misclassification driven by the imperfect accuracy of the arthritis-related questions in the NHIS survey,” Dr. Jafarzadeh writes in an email. “Further, our study uses two additional arthritis-related questions from the NHIS survey on joint symptoms that are consistent with arthritis (i.e., aching or stiffness). These additional questions were also studied in the [validation] study of 2005. Our analytic approach allows estimation of the true prevalence of arthritis from aggregate-level data that are corrected for misclassification, despite using three imperfect criteria.”
In addition to the CDC-defined cases of doctor-diagnosed arthritis, Dr. Jafarzadeh charted positive survey responses to chronic joint symptoms within the past 30 days, and whether onset had exceeded three months. Those symptoms included pain, aching or stiffness felt in or around a joint but not in the back or neck.
Out of 33,672 people in the survey who were 18–64 years of age, 19.3% of men and 16.7% of women had experienced joint symptoms but responded “no” when asked if they had doctor-diagnosed arthritis, according to the article. For those 65 or older, 15.7% of men and 13.5% of women said they had chronic joint symptoms without a report of doctor-diagnosed arthritis.
“Implicit in the question on doctor-diagnosed arthritis when the response is positive, is that the surveyed individual sought or had access to medical care from a health professional,” states the article. “However, a negative response to the doctor-diagnosed question could be the result of either lack of medical attention to joint symptoms, or a truly negative diagnosis. Moreover, an individual … diagnosed by a health professional to have arthritis may never be explicitly informed of the diagnosis.”
