The trends of utilization in traditional Chinese medicine in Taiwan from 2000 to 2010: A population-based study.

INTRODUCTION: There is no study exploring the trend of utilization in traditional Chinese medicine (TCM) from 2000 to 2010. The objective of this study was to investigate the trends of TCM utilization among 3 cross-sectional cohorts of 2000, 2005, and 2010.
METHOD: This study was a cross-sectional analysis of TCM utilization over time. We compared the mean TCM visits among 3 cohorts of 2000, 2005, and 2010. We derived 3 randomly sampled cohorts of nearly 1 million representative beneficiaries in each of 2000, 2005, and 2010 from National Health Insurance Research Database for this research. Multivariate logistic regression was performed to evaluate the relative relationship in categorical variables correlating to TCM users. The percentage change (% change) in mean TCM visits between 2000 and 2005 (2010) was used to evaluate the trends of TCM utilization during the period.
RESULTS: The ratio of TCM users increased throughout cohorts. The ratio of TCM users among women was more than that among men in all cohorts of 2000, 2005, and 2010 (adjusted odds ratio = 1.47; 1.52; 1.62). The mean TCM visits increased from 2000 to 2010. The percentage change in mean TCM visits among women was more than that among men. The group aged less than 20 years had the least percentage change in mean TCM visits (18.8%); nevertheless, the group aged 20 to 34 years had the largest change (30.2%). The high socioeconomic status group had the largest percentage change in mean visits to TCM, whereas the central region had the least percentage change. Neoplasms had the greatest increase in percentage change in mean TCM visits among all disease categories; in contrast, diseases of the respiratory system had the greatest decrease.
CONCLUSION: Both the ratio of TCM users and mean TCM visits increased gradually from 2000 to 2005 and further to 2010. Women used TCM more than men, and this is expected to continue in the future. The high socioeconomic status group used TCM more and more over time. The picture of TCM need among different types of cancer patients should be explored in further research because of the substantial increase in TCM utilization for the disease category of neoplasms.

Introduction

Recently there has been an increasing trend in utilization of complementary and alternative medicine (CAM).[ Herbs are being used by 75% of the people in the world for their basic healthcare needs. [ Medical doctors should know about patients’ use of CAM concurrently when taking their medical history, because the frequency of CAM use has greatly increased.[ This increase is due to both the interest of patients in CAM for seeking help as well as exploration of the effectiveness of various therapies and interventions by researchers.[

Traditional Chinese medicine (TCM), an important category of CAM, is famous for Chinese herbal medicine (CHM) and acupuncture, and is increasing in popularity in many countries.[ Some pharmaceutical drugs composed of a form derived from CHM like artemisinin and 3-n-butylphthalide are also popular globally and successfully marketed.[ People go to TCM clinics to seek treatment of diseases or to augment or replace other treatments or Western medicine (WM).[ People use TCM not only because of searching for treatment of disease or regarding as adjunct to WM, but also because of the few side effects of TCM, need of tonic care or health promotion, and expectation of removing the root of diseases.[

The utilization of TCM is common among Asian immigrants in Western countries and in Asian, including China, Hong Kong, Singapore, Korean, Japan, and Taiwan.[ Usage of TCM is extensive in Taiwan not only because TCM is a part of Chinese culture left from ancient Chinese, but also because TCM is an important part of the medical system in Taiwan.[

The National Health Insurance (NHI) program, a milestone of the medical system, was implemented in 1995 in Taiwan.[ All residents with a registered dwelling in Taiwan are mandated to join the universal health insurance program.[ More than 99% of residents in Taiwan were enrolled in the NHI program by the end of 2010.[ The NHI program covers both WM and TCM, but there is different insurance coverage between them under the program.[ NHI covers both inpatient and ambulatory care of WM, and ambulatory care of TCM, but excludes inpatient TCM care.[ The enrollees under the NHI program can seek medical care from either WM or TCM or both, and from public or private medical facilities or both.[

The data of the NHI program are an administrative dataset.[ All the claims data and file of registry in the NHI program are collected in the National Health Insurance Research Database (NHIRD), which is maintained by the National Health Research Institutes (NHRI), and which provides an optimal platform for research.[ The NHIRD has been used by researchers to explore some issues and publish articles in Taiwan.[

TCM utilization has been discussed in several articles, but the trends of TCM utilization have seldom been explored. TCM usage by children between 2 cross-sectional cohorts has been published recently.[ However, no study has investigated the trend of TCM utilization by the whole population among 3 cohorts. This issue is important, because TCM utilization of the whole population can provide a much more extensive picture of TCM utilization than can data only on children.[

The objective of this study was to investigate the trends of TCM utilization from 2000 to 2010. We compared the mean TCM visits among 3 cohorts of 2000, 2005, and 2010, and derived 3 randomly sampled cohorts of nearly 1 million representative beneficiaries in 2000, 2005, and 2010 from NHIRD for this research. The results of this study may serve as a reference for medical providers to improve preparedness and inform the health policies of government.

Method

Data source

This study was a cross-sectional analysis of TCM utilization over time. All data of the NHI program used in this study were derived from the NHIRD, which is maintained by the NHRI of Taiwan.[ The NHIRD contains patient sex, date of birth, all records of clinical visits and hospitalizations, drugs prescribed, and their dosages and diagnosis codes, which are encoded with the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).[ For the sake of confidentiality and ethical considerations, the identification numbers of all individuals and facilities of the dataset, which we obtained from NHIRD, were encrypted and transformed with a random alphanumeric string.[

Study samples

For this research, we derived the dataset of randomly sampled groups totaling almost 1 million insured beneficiaries from NHIRD in each of 3 cohorts: 2000, 2005, and 2010. All ambulatory TCM visits used by these representative beneficiaries in the 3 cohorts were analyzed for trends of TCM utilization. Diagnosis of medical records in the NHIRD was coded according the ICD-9-CM. The distribution of ambulatory visits in different disease categories was grouped by the primary diagnosis. In order to compare the mean TCM visits to WM visits among different disease categories, we also derived the mean WM visits for the 3 cohorts from NHIRD. In each cohort, the representative beneficiaries who had at least 1 ambulatory TCM visit were defined as TCM users, whereas the representative beneficiaries who had no ambulatory TCM visits were defined as non-TCM users.

Study variable

Gender, age, region, and income were chosen as independent variables to explore their effects on TCM utilization. Age was grouped into 5 groups: <20, 20 to 34, 35 to 49, 50 to 64, and ≧65 years. Region was divided into 6 geographic regions: Taipei, Northern, Central, Southern, Kao-ping, and Eastern regions. Income was used as an indicator of socioeconomic status (SES) and classified into 4 levels. We classified sampled beneficiaries with well-defined monthly income into 3 categories: low, middle, and high levels of SES. The ranges of income between low to middle levels of SES and between middle to high levels of SES were the same. Those without defined income were regarded as “other” level of SES.

Statistical analysis

SAS software, version 9.3 (SAS Institute Inc, Cary, North Carolina) was used to integrate, manage, and analyze the data. The data analysis comprised descriptive statistics, including the demographic characteristics of TCM users and non-TCM users. A chi-square test (χ2) was used to examine the relationships between the categorical variables and the differences between TCM users and non-TCM users. The relative relationship of ratio of TCM users in different demographic factors was calculated by multivariate logistic regression and expressed by adjusted odds ratio (AOR). Mean visits of all sampled enrollees (or TCM users) was the value derived from the number of all visits divided by all sampled enrollees (or TCM users). The percentage change (% change) in mean TCM visits between 2000 and 2005 (2010) was calculated by dividing the difference in mean TCM visits between 2000 and 2005 (2010) by mean TCM visits in 2000.

Ethic consideration

All the names and identification numbers of enrollees and names of medical facilities in the dataset from NHIRD in our study were encrypted as a random alphanumeric series to protect the privacy of study subjects and fulfill ethical considerations. So no one can identify any enrollee or facility from the dataset; therefore, the approval of Institutional Review Board is exempted.

Results

Ratios of TCM users among different Cohorts of 2000, 2005, and 2010

The final samples contained 922,176 beneficiaries in 2000, 999,398 in 2005, and 998,432 in 2010 after removing incomplete data. Table 1 shows that the ratio of TCM users increased from 26.59% in 2000 to 28.29% in 2005 and further to 28.66% in 2010. The ratio of TCM users among women was more than that among men in all 3 cohorts. This difference between genders in the number of TCM users increased gradually from 2000 to 2010 (AOR = 1.47 in 2000; 1.52 in 2005; 1.62 in 2010).

Table 1

Ratios of TCM users of selected enrollees under NHI in 3 cohorts of 2000, 2005, and 2010.

The ratio of TCM users in all age groups increased consistently from 2000 to 2005 and further to 2010, except the 2 groups aged 35 to 49 years and 50 to 64 years, which both increased from 2000 to 2005 and decreased slightly in 2010. Compared with the group aged <20 years, the group aged 35 to 49 years had the highest ratio of TCM users in 2000 and 2005. (AOR = 1.68; 1.61); the group aged 20 to 34 years had the highest ratio of TCM users in 2010 (AOR = 1.60).

The ratio of TCM users in all SES groups except the other SES group increased constantly from 2000 to 2010. Compared with the other SES group, the middle SES group had the highest ratio of TCM users in 2000 and 2005 (AOR = 1.37; 1.32); the middle and high SES groups had the highest ratio of TCM users in 2010 (AOR = 1.40).

The ratio of TCM users in all regions increased from 2000 to 2005 and further to 2010 except for those in the central and southern regions, which increased from 2000 to 2005, but decreased by 2010. Compared with the northern region, the central region had highest ratio of TCM users in all 3 cohorts (AOR = 1.75; 1.71; 1.63). This indicates that the ratio of TCM users in the central region decreased gradually from 2000 to 2010.

The trend of mean TCM visits from 2000 to 2010

Table 2 displays the percentage change in mean TCM visits from 2000 to 2005 (2010). The mean TCM visits per enrollee was 1.22 in 2000 then increased to 1.46 in 2005, and 1.56 in 2010. The percentage change in mean TCM visits was 19.7% and 27.9%, respectively, from 2000 to 2005 (2010). The percentage change in mean TCM visits from 2000 to 2005 was larger than that from 2005 to 2010. It could be seen that the increasing trend of mean TCM visits was less steep from 2005 to 2010.

Table 2

Trend of mean TCM visits among 3 cohorts of 2000, 2005, and 2010.

Mean TCM visits among both women and men increased over time. The mean TCM visits among women increased more than that of men from 2000 to 2005 (2010). The mean TCM visits in all age groups increased through cohorts. The group aged less than 20 years had the least percentage change in visits to TCM from 2000 to 2005 (2010), while the group aged 20 to 34 years had the most. The percentage change in mean TCM visits of all SES groups increased from 2000 to 2005 (2010). The middle SES group had the greatest percentage change in mean TCM visits from 2000 to 2005; however, the high SES group had the largest percentage change from 2000 to 2010. The percentage change in mean TCM visits in all regions increased over time. The central region had the least percentage change in mean TCM visits from 2000 to 2005 (2010), while the Taipei region had the most.

The comparison of mean TCM visits among different disease categories

Table 3 exhibits the mean TCM visits per thousand enrollees in different disease categories. The top 5 disease categories of mean visits for TCM were diseases of the respiratory system (302.3, 287.6, and 274.1); diseases of the musculoskeletal system and connective tissue (198.8, 220.8, and 204.6); injury and poisoning (187.5, 232.8, and 215.9); diseases of the digestive system (134.2, 165.6, and 190.3); and symptoms, signs, and ill-defined conditions (171.2, 284.4, and 332.2) in 2000, 2005, and 2010. The top disease category in terms of most mean TCM visits was diseases of the respiratory system in 2000 and 2005 (302.3 and 287.6) and symptoms, signs, and ill-defined conditions (332.2) in 2010.

Table 3

Mean TCM and WM visits per thousand enrollees and % change in different disease categories from 2000 to 2010.

The 5 disease categories behind the lowest numbers of TCM visits in 2000, 2005, and 2010 were infectious and parasitic diseases (5.6, 6.3, and 5.2); diseases of the blood and blood-forming organs (2.4, 3.4, and 4.1); complications of pregnancy, childbirth, and the puerperium (1.0, 2.7, and 2.1); congenital anomalies (0.7, 1.7, and 2.1); and neoplasms (2.6, 6.5, and 11.5). The 3 disease categories with most upward percentage change of mean visits for TCM from 2000 to 2005 and from 2000 to 2010 were neoplasms (152.4 and 243.7); congenital anomalies (157.2 and 216.6); and complications of pregnancy, childbirth, and the puerperium (175.3 and 109.1). Above all, the disease category of neoplasms had greatest upward percentage change in mean TCM visits from 2000 to 2010 (343.7); nevertheless, the disease category of respiratory system had the greatest downward percentage change in mean TCM visits from 2000 to 2005 and 2000 to 2010 (−4.9, −9.3).

Discussion

Past studies have investigated changes in TCM utilization between several years in 1 cohort. Our study explored TCM utilization among 3 cohorts of 2000, 2005, and 2010. This research method makes the sample more representative of the population. To the best of our knowledge, this is the first study to explore the trend of TCM utilization among 3 different cohorts with a large-scale sample.

The increase over time in ratio of TCM users has been reported in previous studies.[ The trend of increasing TCM utilization was similar to the trend of TCM or CAM utilization reported in previous studies in Taiwan, the United States, and Europe.[ Increase of ratio of TCM users and mean TCM visits may be related to the following reasons. First, the NHI coverage for TCM makes medical service of TCM cheaper and easily accessed for enrollees under the NHI program.[ Second, development of TCM began more than 2000 years ago, and it is part of the culture and daily life.9 Third, people regard TCM as a substitute for or adjunct to WM treatment, and believe that TCM can treat the root of diseases.[

TCM users among women were more than those among men, and this difference between genders increased as time went on. This result of this study that the increase of percentage change in mean TCM visits was higher for women than that for men was consistent with studies in the United States and in Taiwan.[ Women seemed to prefer TCM more than men did.[ There is no doubt that the ratio of TCM users among women will be more than that among men in the future.

TCM users were mainly the enrollees in the middle groups by age, 35 to 49 and 50 to 64 in this study, similar to findings in past studies.[ By 2010, TCM users increased faster among younger adults (age 20–34) than other age groups, which might be related to a positive attitude toward TCM among the younger generation.[ The percentage change of mean TCM visits increased most in the group aged 20 to 34 years from 2000 to 2010, and this is comparable to the result shown in Table 1 that the group aged 20 to 34 years had the highest within-group ratio of TCM users in 2010.

Consistent with the result of previous studies, the within-group ratio of TCM users was lowest in the group aged less than 20 years in all 3 cohorts.[ Because the mean TCM visits and the percentage change in mean TCM visits were both least in the group aged less than 20 years, it is expected that the group aged less than 20 years will be the age group with the least TCM utilization in the future. This result indicates that young people preferred WM over TCM when seeking medical service. A similar phenomenon has been reported among residents in Hong Kong.[ This effect might be influenced by parents, and indicates that pediatric TCM should be further developed and popularized in the future.

The highest TCM utilization in the group aged ≧ 65 years of TCM users might be caused by greater susceptibility to infectious diseases and aging with chronic conditions.[ Physicians of TCM should make efforts to take care of the related diseases of the aged to fulfill their need for TCM treatment.

In the past, CAM was used mainly by people with middle or high SES, as they were the only ones who could afford it.[ However, as enrollees under NHI, more than 99% of residents in Taiwan can afford TCM treatment now.[ The preference for TCM by people with high SES has been increasing recently. The reasons for TCM usage by people with high SES may be the same in the past as at present—belief in TCM and using TCM as an adjunct to WM.[

The preference for TCM by people with high SES reflects the increasing ratio of TCM users in the group with high SES in 2010. Children with high SES also had the highest TCM utilization, and this may be because the behavior of seeking medical service and the SES of children are both related to their parents.[ Recently, the rise in TCM utilization by people with high SES could result from belief in TCM and using TCM as an adjunct to WM as well as expectation of better quality of life and improvement of constitution through supplementary TCM treatment.[

There are 6 regional divisions under National Health Insurance in charge of regional affairs. They are the Taipei, Northern, Central, Southern, Kao-ping, and Eastern divisions.[ The medical sources, public transportation, lifestyles, and configuration of the population are different in these 6 regions.[ It is reasonable to discuss differences in TCM utilization by region as the outcomes of different regional responses to the TCM needs there, which may be used as a reference for policy making by the National Health Insurance Administration (NHIA).

The ratio of TCM users in the central region was still the highest among the 6 regions in 2010, which could be due to the presence there of the earliest professional TCM physician training school, China Medical University. Establishment of China Medical University, resulting in a greater number of TCM physicians and more TCM clinics and TCM departments in hospitals in the central region. This may have given rise to more TCM users in the region.

The central region had the weakest increasing trend of percentage change in mean TCM visits, which implies that the first priority for selection of workplace for TCM physicians should be to avoid the central region. The Taipei region had the highest increase of percentage change in mean TCM visits, which may have resulted from the convenience of transportation, high accessibility of TCM service, and high population density around Taipei and New Taipei cities.[

The disease categories with the top 5 mean TCM visits were all the same in 2000, 2005, and 2010 in this study, similar to the results of related studies.[ As a previous study stated, respiratory tract infection was the most common cause of ambulatory visits by children.[ On the basis of the reduction in percentage change of mean visits in both TCM and WM ambulatory visits for diseases of the respiratory system from 2000 to 2010, we deduced that the reduction in ambulatory TCM visits for diseases of the respiratory system resulted from a decrease in medical need due to diseases of the respiratory system, not the transfer of medical care from TCM to WM. The reduction of ambulatory visits for diseases of the respiratory system might be due to climate change or advertisements fostering disease prevention.

Symptoms, signs, and ill-defined conditions surpassed diseases of the respiratory system in the mean TCM visits by 2010 and became the most common disease category in TCM usage.[ Utilization of WM for symptoms, signs, and ill-defined conditions also increased by 2010. It will be worth exploring the increase of medical requirements by disease type within the disease category of symptoms, signs, and ill-defined conditions.

The 5 disease categories with the fewest mean TCM visits were similar to those found by a previous study.[ The increase of 343.70% in TCM utilization for neoplasms from 2000 to 2010 may have resulted from the implementation of screening for 4 major cancers by the government to detect potential cancer patients. TCM may have become an alternative option or adjunct to WM for cancer treatment, consequently pushing up TCM utilization for neoplasms.[ However, it would be worthwhile to explore the kinds of cancer patients who seek TCM treatment. Such information could be useful for policymaking and arrangement of medical resources.

The mean TCM visits for diseases of the skin and subcutaneous tissue; diseases of the blood and blood-forming organs; and diseases of the genitourinary system vastly surpassed mean WM visits from 2000 to 2010. Many diseases in these 3 disease categories are related to chronic conditions or degenerative organ function.[ TCM are considered to have the following characteristics: tonic care, health promotion, suitability to different health needs, clearing the root of the disease, and few side effects.[ These characteristics of TCM encourage patients with diseases in these 3 disease categories to use TCM. Therefore, TCM utilization for these 3 disease categories rose more than WM utilization did.

Even though this population-based investigation can minimize the bias due to sufficiently large national representative samples from population. There were still several limitations in this study. First, visits to medical facilities, which were not contracted with the NHIA were not included in the claims data from NHIRD in this study. Second, people who went to visit medical facilities contracted with NHIA and took CHM (yin-pian) rather than scientific granules or powder were not recorded in the claim data we analyzed. Third, the claims data from NHIRD we analyzed did not include the visits to traditional Chinese pharmacies, from which people bought CHM to add to their diet for the purpose of tonic supplement for strong constitution, health care, and taste of food.[

Conclusions

Both the ratio of TCM users and mean TCM visits increased consistently from 2000 to 2005 and continued rising, albeit less steeply, through 2010. It can be predicted that women will have higher TCM utilization than men in the future. Preference for TCM increased among those with high SES during the study time period. The central region had the largest mean TCM visits, but the least increase of percentage change in mean TCM visits; therefore, the Taipei region had the greatest increase of percentage change in mean TCM visits from 2000 to 2010. Diseases of the respiratory system had the highest mean TCM visits in 2000, and this decreased over time. Among all disease categories explored in this study, neoplasms had the highest percentage change in mean TCM visits from 2000 to 2010.