Methodological challenges in mapping chinese medicine syndrome with conventional diagnosis: Implications for multi-centre trials in integrative medicine

Abstract

IntroductionChinese medicine (CM) diagnostic instruments are increasingly developed to classify patients into different syndrome subtypes, potentially allowing for subgroup analyses in randomized trials. However, its external validity has been questioned. We illustrated this problem by evaluating a previously validated Functional Gastrointestinal Disorder CM Syndrome Differentiation Instrument (FDCMI) on a heterogeneous sample of functional dyspepsia (FD) patients. MethodsA cross-sectional study was performed using two questionnaires to diagnose FD patients from both conventional and CM perspectives. CM diagnostic subtypes were classified into Spleen-Stomach Qi Deficiency (QD), Liver Qi Stagnation (LQS), Dampness Syndrome (DS), and Qi Stagnation (QS) by the FDCMI. Descriptive analyses were used to illustrate relationships between conventional and CM diagnostic criteria. Explanatory factor analysis (EFA) was performed to explore the underlying factor structure of FDCMI in the current sample. ResultsData were successfully collected from 224 FD patients, with a response rate of 83.7%. Respectively 49 (21.9%) and 162 (72.3%) patients qualified for conventional FD diagnostic criteria under the Rome III (R-FD) criteria and modified Rome III (M-FD) criteria. Respectively 118 (52.7%), 116 (51.8%), 116 (51.8%) and 107 (47.8%) patients qualified for a diagnosis of QD, LQS, DS and QS under the FDCMI criteria using a 30% cutoff threshold of the total subscale score for each CM diagnostic subtype. Forty patients (17.9%) did not qualify for any single CM subtype under this threshold. Most patients belonged to more than one CM subtype. EFA results revealed an 8-factor structure that accounts for only 59.4% of the total variance. ConclusionsDue to population heterogeneity, differences in disease phenotype, diagnostic performance of conventional tests, co-morbidity and cut-off threshold, CM diagnostic instruments may have limited external validity, potentially reducing their applicability in multicenter trials.