Use of traditional medicine and complementary and alternative medicine in Taiwan: a multilevel analysis.

As the selection of a medical modality is not completely independent, environmental and sociocultural contexts of ecological validity are desired. This study aimed to apply a multilevel analysis using the Hierarchical Linear Modeling software to examine predictors of traditional medicine (TM)/complementary and alternative medicine (CAM) use in Taiwan on both individual and division levels. Individual-level data were obtained from the government database involving TM/CAM use and its impact on the population, whereas division-level data were obtained from a government annual report. A total of 2310 individuals from 22 administrative divisions of Taiwan were evaluated in the data analysis, of which 86.9% had used at least 1 TM/CAM modality in the past year. The average division of TM/CAM use was 2.86 modalities in the null model and 4.15 in the full model. Significant relationships were found between TM/CAM use and individual-level variables of gender, educational level, monthly income, perceived health status, experience with Western medical treatment, and the cost, effect, and degree of satisfaction with TM/CAM. At the division level, TM/CAM use was significantly related to aging population, employment status, and the number of medical institutions. With a simultaneous evaluation of the individual-level and division-level influences, it was found that the average division of TM/CAM use increased significantly. The place of residence is an important predictor of TM/CAM use. The age factor in predicting TM/CAM use in this study may be overestimated in the population of 26 to 60 years of age, whereas an aging population is important in the average division of TM/CAM use. Efforts to reform health insurance to completely cover the costs of TM/CAM and to better facilitate equality of access of health care in rural and remote areas are deemed necessary.