I learned something new while reading this article on PMR:
It says "The European League Against Rheumatism (EULAR) and American College of Rheumatology (ACR) criteria can distinguish polymyalgia rheumatica from mimicking conditions."
However, I then clicked on the tab marked "Responses" and found the following comment by Dr Megan R W Barber MD PhD:
"I read this article with interest. Thank you for drawing attention to this often delayed diagnosis.
"Regarding point 3, the authors do not appropriately distinguish between classification and diagnostic criteria when discussing the 2012 EULAR/ACR provisional classification criteria for PMR.
"Classification criteria are designed solely to select a homogenous sample of patients for clinical research studies and are not intended to diagnose patients. They are often not sensitive enough for this purpose and fail to capture the heterogeneity intrinsic to rheumatic diseases (Aggarwal et al. 2015).
"In fact, the 2012 EULAR/ACR provisional classification criteria to which the authors refer had a sensitivity of just 68% for a score ≥4 in their original cohort (Dasgupta et al. 2012) and the manuscript clearly notes that “these criteria are not meant for diagnostic purposes.”
"When diagnosing PMR, rheumatologists take into consideration many factors that are not reflected in the provisional classification criteria, such as patient response to a trial of steroids.
"Confusion between classification and diagnostic criteria is common, but it is a crucial distinction. Relying on classification criteria at the bedside results in missed diagnoses."
I then found the quoted article by "Aggarwal et al. 2015":
acrjournals.onlinelibrary.w...
This explains in detail why most epidemiological studies should apply strict Classification Criteria when choosing patients to study against a control group. This is to minimise the chance that misdiagnosed cases invalidate the results.
Doctors should use much looser Diagnostic Criteria to avoid missing cases that do not meet all the Classification Criteria. EULAR/ACR responded that they do not produce such DC, only CC.
This confusion may go some way to explain the huge variations in diagnostic criteria claimed for PMR and other diseases. At the foot of the Responses tab, Dr Patricia A Forbes MD Chb describes her own experience of PMR despite not testing positive for CRP or ESR:
"I had a classical case which responded to prednisone. I had a shorter recurrence several months later. I was one of the percentage of patients who had normal ESR and CRP. No further recurrence in 2 years. I do have finger arthritis."
According to the original study, she should not have been diagnosed or treated for PMR because she did not meet the EULAR/ACR criteria. It would seem those denying the existence of atypical PMR may be relying on the wrong set of published criteria. I wonder whether that includes this very strict NHS guidance, which demands "all" criteria be met? It reads very much like a set of CC: