Accuracy of diagnostic criteria in knee OA



The reliability of osteoarthritis (OA) diagnostic criteria has been questioned as they have been mostly developed based on experience with patients in secondary care. In the current study, ACR, EULAR and NICE criteria were applied to evaluate longitudinal clinical and radiographic data from a cohort of patients, initially recruited for the CHECK (Cohort Hip and Cohort Knee) Study. The present study focused on patients with symptomatic knees at baseline, for whom follow up data for 5 and 10 years were available, and invited general practitioners and secondary care clinicians to evaluate clinical and radiographical data in order to address the sensitivity, specificity, and predictive values of the ACR, EULAR and NICE criteria.

Synthesis – key messages

  • A ~70% overlap between the three diagnostic criteria was found, which the authors classified as moderate.
  • In clinical practice, 25% to 33% cases of OA would be inconsistently diagnosed when ACR, EULAR or NICE criteria are applied.
  • EULAR criteria have been shown to exhibit the highest specificity, making them more appropriate when seeking to identify clinically relevant OA knees for research.
  • NICE criteria had the highest sensitivity, thus they could be more suitable for initial diagnosis of the condition.

Details of the study

“Evaluation of the diagnostic performance of American College of Rheumatology, EULAR, and National Institute for Health and Clinical Excellence criteria against clinically relevant knee osteoarthritis: data from the CHECK cohort – Wang et al.”(1)


A total of 833 knees from 539 participants were included in the analysis. Only patients with available data for their 5-year, 8-year and 10-year follow-up were included in the study. General practitioners and secondary care physicians were asked to determine whether a clinically relevant knee OA diagnosis could be established based on clinical and radiographic data. This was compared to the ACR/ EULAR/ NICE criteria-based diagnosis for which the following classification was made: F1 meant if fulfilled the criteria in one or more time points, F2 for two or more time points and  F3 for three or more time points.

Main results

Within the observational period, the prevalence of radiographic knee OA diagnosis increased from 38% to 60% while the prevalence of ACR, EULAR and NICE criteria diagnosis decreased. Authors hypothesized that this decrease when looking at diagnosis established through the ACR/EULAR/NICE criteria could be related to the fact that OA symptoms are intermittent and could have fluctuated during the 5-year follow-up period.

The overlap for a positive or negative diagnosis among the three criteria was moderate reaching an approximate 70% (67% for F1, 67% for F2 and 74% for F3). EULAR were the criteria showing higher specificity while NICE those with higher sensitivity.

about one of four to one of three knees would receive inconsistent diagnosis by the ACR, EULAR, and NICE criteria. With the clinical experts’ diagnosis of clinically relevant knee OA as the reference, no remarkable differences were observed for the diagnostic performances of the three criteria.

Between 25% and 33% of OA knees are subject to inconsistent diagnosis applying the ACR, EULAR and NICE criteria, manifesting that there are limitations to overcome in the field of OA diagnosis as different criteria may be suitable for different clinical diagnosis situations. EULAR were the criteria showing higher specificity while NICE those with higher sensitivity

1.           Wang Q, Runhaar J, Kloppenburg M, Boers M, Bijlsma JWJ, Bierma-Zeinstra SMA, et al. Evaluation of the Diagnostic Performance of American College of Rheumatology, EULAR, and National Institute for Health and Clinical Excellence Criteria Against Clinically Relevant Knee Osteoarthritis: Data From the CHECK Cohort. Arthritis Care Res. 7 de noviembre de 2023;

Link to the full study

This article is a summary based on the following study. For more information and details, please consult the full study. Please do not hesitate to contact us if you have any comments.

Related Content

A promising eHealth tool to manage musculoskeletal pain in rural areas


Knee OA care: Focusing on lifestyle


Projections for the rising global burden of osteoarthritis

Scroll to Top