The reliability and validity of instruments measuring pattern identification in Korean medicine: A systematic review

Introduction: Pattern identification (PI) is a unique concept in traditional East Asian medicine that refers to diagnostic process. This study aims to summarize and critically evaluate the reliability and validity of all Korean designed PI instruments. Methods: A systematic literature search was conducted in MEDLINE, EMBASE, and eight Korean medical databases from the study's inception to August 2017 to identify all studies that described Korean PI instruments. We included PI instruments without restriction on the types of diseases or, conditions of participants. General characteristics of the included instruments, reliability verification, internal consistency, and the types of validity, including construct validity, content validity, and criterion validity, were reviewed. Results: Thirty-six PI instruments were identified and analyzed. Ten of them were PI diagnostic tools for specific diseases. Fifteen were related diagnoses of PI for non-specific diseases. Four instruments involved the diagnosis of PI for specific decoction treatment and seven were related to the diagnosis of constitution, including Sasang constitution. The most commonly used statistical test was Cronbach's alpha, an estimator of internal consistency. A total of ten studies conducted test-retest reliability with varying time intervals, and the test-retest coefficient was moderate to good. Conclusions: This article examined the reliability and validity of the PI instruments used in Korea and the need to improve the standardisation of PI diagnosis. Almost studies reported the value of internal consistency only. Clinical studies on the application of future questionnaires will be needed in the future.