Background Knee Osteoarthritis (OA) causes large economic burden worldwide. Accurate prevalence estimates helps evaluation of its impact and guide resource allocation. Screening questionnaires with reasonable accuracy offer less costly alternative to confirming diagnosis by physical examination and radiography.
Objectives We aim to evaluate the performance and validity of three sets of screening questionnaires for symptomatic knee osteoarthritis (OA) among elderly in Singapore.
Methods Three sets of questionnaires for screening symptomatic knee OA that have been validated in other populations were evaluated. These questionnaires asked signs, symptoms and prior diagnosis of OA knee. Subjects in a community elderly centre were recruited to answer three sets of screening questionnaires presented in a random order. Half of subjects screened positive and negative were invited to physical examination and radiography of knee in a rheumatology clinic. Symptomatic OA knee was classified according to the American College of Rheumatology (ACR) clinical and radiographic criteria. Sensitivity and specificity of each set of questionnaires were evaluated. Predictive algorithms were created to find the best performing screening test.
Results Of 120 subjects enrolled, 103 were evaluated in clinic. ACR clinical and radiographical symptomatic knee OA was confirmed in 37 subjects. Only 10% of confirmed knee OA subjects knew their diagnosis and thus reducing the sensitivity of prior diagnosis of knee OA as a screening item. None of the screening instruments had adequate performance to serve as a single step evaluation of presence or absence of symptomatic knee OA. The most sensitive (91.9%, CI 77- 97.9%) screening tool was relatively non-specific (47%, CI 34.7- 59.6%) in identifying symptomatic OA knee. A screening algorithm developed from one of the screening instruments, which consisted of three knee OA symptoms or one symptom together with physician diagnosis of knee OA gave higher sensitivity (48.6%, CI 32.2- 65.3%) without losing the specificity (95.5%, CI 86.4%>98.8%).
Conclusions One of the three sets of screening instruments offered high sensitivity and moderate specificity in identifying symptomatic OA knee. This would be an useful instrument to recruit potential subjects with symptomatic knee OA into clinical trials, in which a second step confirmation with physical examination and radiography is necessary. An algorithm developed from this instrument offered high specificity and moderate sensitivity. When this algorithm is applied to the general population, it may offer a potential of estimating the prevalence of symptomatic knee OA in Singapore.
Acknowledgements The authors appreciate the support of Mr Firman Loo and Mr Peter Chou of the Nice Organization in performing the screening questionnaires, and Ms Tan Hwee Kheng of the NTUC Elder care for patient recruitment.
Disclosure of Interest None Declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.