Utilization of Chinese medicine for respiratory discomforts by patients with a medical history of tuberculosis in Taiwan

Background Tuberculosis (TB) is one of the world’s major communicable infectious diseases, and it still imposes a great health burden in developing countries. The development of drug-resistant TB during the treatment increases the treatment complexity, and the long-term pulmonary complications after completing treatment raise the epidemic health burden. This study intended to investigate the utilization of Chinese medicine (CM) for respiratory symptoms by patients with a medical history of TB in Taiwan. Methods We analyzed a cohort of one million individuals who were randomly selected from the National Health Insurance Research Database in Taiwan. The inclusion criteria of patients (n = 7905) with history of TB (ICD-9-CM codes 010–018 and A02) were: (1) TB diagnosed between January 1, 1997 and December 31, 2010 (2) 18 years old or over (3) Clinical records for at least 2 months with complete demographic information (4) Record of treatment with first-line TB medication prescriptions. CM users for conditions other than respiratory discomforts (n = 3980) were excluded. Finally, a total of 3925 TB patients were categorized as: CM users for respiratory discomforts (n = 2051) and non-CM users (n = 1874). Results Among the 3925 subjects, 2051 (52.25%) were CM users, and 1874 (44.753%) were non-CM users. Female patients and those who were younger (18–39 y/o) and who lived in urbanized areas relatively tended to be CM users (p < .0001). Most of the CM users (1944, 94.78%) received Chinese medicines. The most commonly prescribed herbal formulas and single herbs were Xiao-Qing-Long-Tang and Radix Platycodonis (Jie-Geng), respectively. The core pattern of Chinese medicines for TB patients consisted of Ma-Xing-Gan-Shi-Tang, Bulbus Fritillariae Thunbergii (Bei-Mu), Radix Platycodonis (Jie-Geng) and Semen Armeniacae (Xing-Ren). Conclusions The use of CM is popular among patients with a medical history of TB complicated with long-term respiratory discomforts in Taiwan. Further pharmacological investigations and clinical trials are required.


Introduction
In the twenty-first century, tuberculosis (TB) continues to be one of the world's major health challenges. TB is a deadly communicable disease that is primarily transmitted from human to human by droplet infection, and it is a major health burden in both developed and developing countries. Approximately 9.0 million people developed TB, and 1.5 million died from the disease in 2013 [1]. Of the 9.0 million estimated cases, 56% were in Southeast Asian and Western Pacific regions [2]. China accounted for 11% of the 9.0 million estimated TB cases, and 11,528 TB cases were confirmed in Taiwan [3]. Male, elderly and immunocompromised patients, including those with HIV, patients with poorly controlled diabetes mellitus or chronic kidney disease or those receiving immunosuppressant treatments, are generally more susceptible to the disease [4]. Healthcare workers who are exposed to TB also have higher risks of being infected [5].
Due to the side effects of the drugs [6,7], poor medication compliance and the problem of development of drug-resistant TB, global investment in TB research is needed [8]. Among the first-line anti-TB drugs, side effects of hepatotoxicity have been reported to be associated with isoniazid, rifampin and pyrazinamide; cutaneous reactions with isoniazid, pyrazinamide and ethambutol; gastrointestinal intolerance with rifampin; retrobulbar neuritis with ethambutol; and ototoxicity with streptomycin [9][10][11]. Long-term respiratory symptoms and impairment of pulmonary function affect patients' quality of life after completing TB treatment. The respiratory comorbidities were identified in more than half of microbiologically cured TB patients [12]. Despite successful treatment of TB, residual impairments persist, and chronic complications or sequelaes can arise from structural or vascular alterations at disease sites [13,14]. Many patients with chronic illness may seek complementary therapies [15], especially Chinese Medicine (CM) in Asian countries [16,17]. CM, which originated in ancient China, is defined as comprehensive healthcare skills and practice for the maintenance of health and treatment of disease based on the beliefs of holism [18] and experiences of pattern identification/ syndrome differentiation handed down from generation to generation [19]. In Taiwan, CM is important and popular among different categories of complementary therapies, and is regarded as one of the mainstream therapies with coverage of the National Health Insurance (NHI) program [17]. "CM" usually refers to the overall treatment modalities, and services of CM covered by the NHI program, which includes Chinese medicines, acupuncture, moxibustion and Chinese traumatology therapy [20]. "Chinese medicines" usually refers to herbs or herbal products which can be classified into single-herb products and herbal formulas (multi-herb products) [21]. In Taiwan, herbal products (concentrated scientific herbal granules) covered by the NHI program are manufactured by GMP-certified pharmaceutical companies [22].
The ancient CM literature did not use the word "TB", but rather considered this pulmonary condition to be related to a syndrome of lung consumption characterized by cough, hemoptysis, tidal fever, night sweats and emaciation due to consumptive or exhaustive overstrain. It was reported that Chinese medicines combined with conventional medicine has beneficial effects for inhibiting Mycobacterium, strengthening the immune system of the body, enhancing the effect of anti-TB drugs, reducing drug resistance, and being relatively safe [23][24][25].
While several studies have investigated the use of CM in the care of TB patients [23], there is a lack of this kind of survey on the use of CM among patients with a medical history of TB and complicated with long-term respiratory discomforts.
We intended to investigate the utilization of CM among patients with a medical history of TB complicated with long-term respiratory discomforts in this study. The results of this study will be useful for future clinical trials and pharmacological investigations regarding efficacy and safety.

Data sources
The NHI program was established in Taiwan in 1995. The program was highly representative of samples of Taiwan's general population because the reimbursement policy is universal and mandatory. The coverage rate of this compulsory health insurance program reached almost 95% in 1997 [26] and 99.4% at the end of 2010 in Taiwan [27]. CM services that are reimbursed in the NHI program include Chinese medicines, acupuncture, moxibustion, and Chinese traumatology therapy [28]. The National Health Insurance Administration provided the registration files and original claims data to the National Health Research Institutes, which established and managed the National Health Insurance Research Database (NHIRD) for research purposes. This study was based in part on data from the NHIRD. It contains data comprising demographic characteristics, medical care facilities, outpatient and inpatient visits, visit dates, diagnostic codes, management, prescriptions and medical expenditures. We acquired a randomly selected sample consisting of one million individuals (Longitudinal Health Insurance Database 2000; LHID 2000) from the NHIRD managed and released by the National Health Research Institutes, Taiwan. The diagnostic codes were from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) formats.

Ethical consideration
This study followed the ethical standards of the responsible committee and with the Helsinki Declaration of 1964 and later versions. All the datasets were de-identified and encrypted before released by the National Health Research Institutes, Taiwan. All of the individuals or care providers could not be identified in the database. Patient consent was exempted for the total anonymity of all research data in this study. Therefore, the Research Ethics Committee of China Medical University and Hospital approved this study and waived the requirement for informed consent (CMUH104-REC2-115).

Study population
Patients newly diagnosed with TB (ICD-9-CM codes 010-018 and A02) (n = 36,660) were selected from the database. To avoid the inclusion of patients who did not have the disease, we set the inclusion criteria to allow only patients who were newly diagnosed with TB from January 1, 1997, to December 31, 2010, with clinical records from at least 2 months and treatment with first-line TB medication prescriptions (isoniazid, rifampin, pyrazinamide & ethambutol) (n = 8306). Because the prevalence and etiology of TB infections in children were different, and the medical services-seeking behaviors among children or adolescents were mainly dominated by their parents, children with TB infections were not included in our objectives. Therefore, we further excluded patients who were less than 18 years of age or who had missing information related to birth and sex (n = 401). A total of 7905 TB patients were then included. Therefore the sample inclusion criteria is as following: (1) TB diagnosed between January 1, 1997 and December 31, 2010. (2) 18 years old or over. (3) Clinical records for at least 2 months with complete demographic information (4) Record of treatment with first-line TB medication prescriptions. CM users were defined as those who had visited CM clinics and had CM outpatient clinical records of TB or respiratory diseases (ICD-9-CM codes 786.x or 460-519). Non-CM users were defined as those who never visited CM clinics after the initial diagnosis of TB. To investigate TB patients complicated with long-term respiratory discomforts, we further excluded those clinical visits with a non-TB or respiratory diagnosis (n = 3980). Finally, a total of 3925 TB patients were then categorized into CM users (n = 2051) and non-CM users (n = 1874) (Fig. 1). Taiwan is a country with a population of 23 million. Taiwan has a geographical area of 36,000 km 2 . There were 23 cities and 359 townships in Taiwan and many of its residents live within urban cities. The residential areas of 23 cities in Taiwan were classified into 4 levels of urbanization based on the population density (people/km 2 ), the ratio of the population with varying educational levels, and the number of physicians per 100,000 people. Level 1 represents the highest urbanized level, while 4 represents the lowest level. Levels 1 and 2 of this urbanization were defined as urban areas, while levels 3 and 4 were classified as rural areas [29].

Prescription of Chinese medicines
We listed the herbal formulas by their phonetic pin-yin name and single herbs by their phonetic pin-yin name, Chinese material medica name and plant name. Indications for the herbal formulas and single herbs were based on CM theory [30,31]. Full botanical names were in accordance with the International Plant Names List (IPNI; http://www.ipni.org) and The Plant List (http:// www.theplantlist.org/) [32]. An open-sourced freeware Node XL (http://nodexl.codeplex.com/) was utilized to identify the core patterns of prescriptions for TB patients, and the most common co-prescribed Chinese medicines were demonstrated in this network analysis. As previously described [33], the larger of the spots with thicker-line widths indicated significant prescription patterns and counts of close connections between formulas and herbs in the network figure.

Statistical analysis
All statistical analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA). Univariate analysis was utilized to compare the CM users with the non-CM users. Data analysis comprised descriptive statistics, including the frequency of herbal prescriptions, patient demographic characteristics, indications for the TCM prescription, and the most frequently prescribed herbal formulas and herbs for treating TB. The chi-square test was used to examine the relationships between the categorical variables and to examine the differences between CM users and non-CM users. A P-value of < 0.05 was considered statistically significant.

Results
In this study, we identified 52.25% (n = 2051) patients with a medical history of TB who had visited CM clinics and had CM outpatient clinical records of TB or respiratory diseases (Fig. 1). Those who were female, younger (18-39 y/o), and lived in higher urbanized areas (levels 1 and 2) were relatively more likely to use CM. The median duration between newly diagnosed TB and the first CM consultation was 1068 days (Table 1).
Regarding the treatment modes employed among the CM users, approximately 95% of CM users (n = 1944) received only Chinese medicines, 5.2% of patients received combined treatment with both Chinese medicines and acupuncture, and the remaining 0.2% of patients received only acupuncture treatments. Among all CM users (patients with a medical history of TB who had visited CM clinics and had CM outpatient clinical records of TB or respiratory diseases; n = 2051), 90.15% visited CM clinics 1-3 times annually, while 4.58% patients consulted CM doctors more than 6 times/year ( Table 2).
The frequency distribution of disease categories related to respiratory discomforts and comorbidities that patients who had been diagnosed TB was analyzed (Table 3). Compared to the non-CM users, the frequencies of outpatient visits among TB patients who used CM were statistically higher across almost all discomforts and comorbidities, except swelling, mass, and lump in chest (Table 3). More than 70% of TB patients used CM because of acute respiratory infections, cough, chronic obstructive pulmonary disease and allied conditions, and other diseases of upper respiratory tract. These results could indicate the comorbidities and complications of TB and may serve as an explanation for why TB patients complicated with long-term respiratory discomforts turn to CM doctors for help (Table 3). We further investigated the prescription pattern of Chinese medicines and identified the ten most commonly prescribed formulas and single herbs. The ten most commonly prescribed single herbs and herbal formulas were analyzed and are listed in Tables 4 and 5, respectively. The most commonly prescribed herbal formulas and single herb were Xiao-Qing-Long-Tang, and Radix Platycodonis (Jie-Geng), respectively.
We further conducted a network analysis and found that the core patterns of herbal formulas and herbs prescribed for TB patients consisted of Ma-Xing-Gan-Shi-Tang, Bulbus Fritillariae Thunbergii (Bei-Mu), Radix Platycodonis (Jie-Geng), and Semen Armeniacae (Xing-Ren) (Fig. 2). We further summarized these findings in Fig. 3.

Discussion
In this study, we first observed that the median duration between newly diagnosed TB and the first CM consultation was 1068 days. Many TB patients complicated with long-term respiratory discomforts sought CM services. We then determined that Ma-Xing-Gan-Shi-Tang, Bulbus Fritillariae Thunbergii (Bei-Mu), Radix Platycodonis (Jie-Geng), and Semen Armeniacae (Xing-Ren) were the core prescriptions for patients with TB in Taiwan. To the best of our knowledge, this report describes the first nationwide population-based cohort study investigating CM utilization patterns among TB patients complicated with long-term respiratory discomforts. However, whether CM combined with modern conventional medicine has beneficial effects or risks requires further investigation.
The phenomenon that female patients with TB were relatively more likely to seek CM services is comparable to a previous Taiwanese survey demonstrating that women had a greater tendency to seek CM consultations [34]. It has been reported that women are more susceptible to side effects from anti-TB drugs due to higher CYP3A activity [11,35], although men are more susceptible to TB than women [36]. Women may exhibit more help-seeking behaviors than men and are more likely to seek complementary therapies when they suffer from chronic or catastrophic illness [37,38].
We also observed that younger patients (18-39 y/o) and those who lived in more urbanized areas were relatively more likely to choose CM, in accordance with previous studies [38,39]. Patients younger than 40 years old are likely to be more aware of alternative therapeutic choices. A previous report had shown that women, people with high socioeconomic status, higher education levels, self-perceived poor health status, and people who exercise regularly tend to visit CM services more often in Taiwan [40]. Patients who reside in more urbanized areas usually have a higher socioeconomic status and fewer barriers to accessing medical services [41,42]. The time and money spent on transportation may also lead to a significant inequality in contact with the health system [43][44][45]. Therefore, people who live in urbanized areas have a greater tendency to use CM services.
In Taiwan, patients with a diagnosis of TB are required to receive pharmacological therapy to control TB infection  with monitoring by local health officers based on the government's health policy [46]. The included patients who had been diagnosed with TB should have received standard anti-TB treatment. In Taiwan, recommended standard treatment for adult respiratory TB consists of a regimen of isoniazid, rifampicin, pyrazinamide, and ethambutol for 2 months, followed by isoniazid, rifampicin, and ethambutol for 4 months with monitoring by local health officers according to governmental policy [46]. The total treatment course takes approximately 6 months in newly diagnosed patients [46]. The median duration between newly diagnosed TB and the first CM consultation was 1068 days. During the follow-up period, patients might seek CM consultation due to symptoms of respiratory discomfort or complications. In addition to respiratory illness, they may also experience other symptoms related to the anti-TB    treatment, such as skin rash or drug fever, GI intolerance, nausea, vomiting, visual toxicity, hearing disturbances and arthralgia [9,[47][48][49]. The NHI insurance coverage for CM treatments may play a considerable role in patients' tendency to seek CM consultations. Medical services in both Western medicine and CM have been promoted due to the high coverage of the NHI program. Furthermore, the cost of co-payment for concentrated scientific herbal granules is approximately $15 U.S. dollars per month under the NHI Program in Taiwan. The tendency of patients to seek CM may continue to increase since the co-payment for Chinese medicines is relatively low [50][51][52][53]. Among patients with TB, most CM users (94.78%) received only Chinese medicines, which is consistent with previous studies investigating other respiratory diseases for which herbal products are frequently used as treatment in CM visits in Taiwan [52,54]. In patients with adult-onset asthma, 76.7% patients who visited CM physicians received Chinese medicines [54], and 97.1% of the CM-treated rhinosinusitis subjects were prescribed Chinese medicines [52].
Regarding the disease categories of the CM users identified in Table 3, most of TB patients used CM because of acute respiratory infections, cough, chronic obstructive pulmonary disease and allied conditions, and other diseases of upper respiratory tract. It was reported that complications of TB include bacterial pneumonia, cor pulmonale, pneumothorax, and acute respiratory failure. Complications of TB could result in progression of functional pulmonary impairment and structural lung destructions [14,55]. Patients with a prior TB diagnosis were more than twice as likely to have lung airflow obstruction [56]. Furthermore, the complications of respiratory diseases can be independent predictor of shorter survival among patients with TB [57]. TB patients might also turn to CM doctors in the hopes of relieving their symptoms and improving their health [58].
It must be noted that TB did not appear as a disease name in the records of the Chines medical literature. However, ancient people had observed the disease, and the symptoms and signs of TB and its related illness had been recorded and discussed in the Chinese medical literature. The prescriptions of CM were in accordance with the concepts of TB in CM, which were related to chronic cough with symptoms and signs of consumption or exhaustion. The prescriptions were prescribed after pattern identification and syndrome differentiation based on the theory of CM. Most of the prescriptions listed in Tables 4 and 5 are helpful for relieving symptoms caused by yin and qi deficiency of the lung with wind contraction, which corresponds to the respiratory symptoms related to long-term respiratory discomforts [19].
Among the single herbs identified in the core prescriptions, the herb Bulbus Fritillariae Thunbergii has been commonly used as an antitussive agent [59]. Its isosteroidal alkaloids have been reported to have tracheobronchial relaxant activity [60]. Radix Platycodonis has been used for bronchitis, tonsillitis, and laryngitis in ancient times. The pharmacological constituent Platycodin D in Radix Platycodonis has demonstrated anti-inflammatory, antitumor [61], antinociceptive and immunomodulatory activities. Immunostimulatory effects have been revealed via the suppression of IL-6 and TNF-α contents [62] and enhancement of the killing activities of natural killer cells [61]. Radix Platycodonis also inhibits OVA-induced airway inflammation and regulates the production and secretion of airway mucin [63]. Furthermore, Radix Platycodonis has often served as an expectorant in diverse inflammatory pulmonary diseases [64]. The other herb of the core prescriptions, Semen Armeniacae, has anti-inflammatory and analgesic effects by suppressing PGE2 and NO production [65]. The amygdalin found in it has been used as an antitussive, anti-asthmatic, anticancer agent [66,67].
In addition to the above three herbs, other herbs have been reported to relieve respiratory symptoms or reduce inflammation. Herba Houttuyniae (Yu-Xing-Cao) has been reported to relieve fever, cough and asthma with anti-inflammatory [68] and antiviral activities [69]. This herb inhibits the production of pro-inflammatory cytokines via inhibition of the NF-κB signaling pathway in HMC-1 human mast cells [70]. Semen Trichosanthis (Gua-Lou-Ren) has anti-inflammatory, and antibacterial activities [71] and immunomodulatory effects [72]. Ophiopogonin-D in Radix Ophiopogonis (Mai-Men-Dong) may be beneficial for reducing the excitability of parasympathetic ganglionic neurons in the airways and cholinergic control of airway function, and its antitussive effects might be due to the activation of K+ channels and hyperpolarization of paratracheal neurons [73]. Radix Salviae Miltiorrhizae (Dan-Shen) could be helpful for reducing lung fibrosis [74]. This species has been reported to exhibit anti-inflammatory and antioxidative effects [75]. Radix Scutellariae (Huang-Qin) has shown Fig. 2 The 100 most commonly prescribed herbal formulas and single herbs for patients with tuberculosis were assessed, and the core pattern of the prescriptions shows that Ma-Xing-Gan-Shi-Tang, Bei-Mu, Jie-Geng, and Xing-Ren were among the most frequently prescribed combinations pharmacological effects that are anti-inflammatory, antitumor and hepato-protective. Additionally, these extracts have antioxidant, antibacterial and antiviral effects [76]. Research has indicated that a decoction containing Fructus Schisandrae (Wu-Wei-Zi) helps to prevent alveolitis and the development of pulmonary fibrosis [77]. Liver-protective effects of Fructus Schisandrae via the inhibitory effect of CYP-3A4 activity have been reported [78] and may also ameliorate the side effects of hepatitis resulting from the use of anti-TB medication.
Among the herbal formulas identified in this research, Ma-Xing-Gan-Shi-Tang is one of the core prescriptions. It has been used to treat asthma [79] and fever associated with pneumonia. It possesses anti-asthmatic, anti-pyretic, anti-inflammatory, and antitussive effects [80,81]. Xiao-Qing-Long-Tang has anti-inflammatory, antiviral, and anti-allergy activities [82], and it can reduce inflammation in lung tissue via an anti-apoptotic effect and inhibition of cytokine release, as well as prevent pulmonary fibrosis [83][84][85][86]. Ding-Chuan-Tang has shown anti-asthmatic effects in clinical studies [87]. Its protective effects against lung injury in asthma occurs by suppressing the production of pro-inflammatory cytokines with significantly reduced inflammatory cell infiltration, goblet cell proliferation, collagen deposition, and damage in the bronchi and alveoli [88]. Xin-Yi-Qing-Fei-Tang has been commonly used for rhinitis [51,52] and has been shown to reduce eosinophil, serum IgE and IL-4 levels [89]. Clinical studies have reported that Yi-Qiao-San may be helpful for patients with H1N1 influenza to reduce the duration of fever [90]. Lung cancer patients displayed better physical function, role function and cumulative survival after receiving combination treatments of a formula containing Bai-He-Gu Jin-Tang with Western medicine [91].
There are some limitations to our study. First, the herbal products purchased at the patients' own expense in addition to the NHI program products were not included in this study. However, because the co-payment of herbal products through the NHI program was much less than the market cost, the possibility of purchasing herbs outside the NHI program was relatively low. In addition, compliance to prescription regimens was difficult to measure. Finally, this study focuses on the utilization of CM in patients with TB. In the future, a high-quality randomized controlled clinical trial with imaging data and species cultures to determine the efficacy of Chinese medicines for TB is expected.

Conclusions
Our study found that many TB patients complicated with long-term respiratory discomforts sought CM services in Taiwan. Those who were 18-39 y/o, female, and who lived in urbanized areas tended to use CM. The prescription patterns identified in this study could be useful for future clinical studies or pharmacological investigations. Availability of data and materials The datasets analyzed from NHIRD were provided by the National Health Insurance Administration and maintained by the National Health Research Institutes of Taiwan. The use of NHIRD is limited to research purposes only. Applicants must follow the Computer-Processed Personal Data Protection Law (http://www.winklerpartners.com/?p=987) and related regulations of the National Health Insurance Administration and National Health Research Institutes.
Authors' contributions STY conceptualized the study. JHC performed the statistical analysis. STY, YRL, MYW, PSY and HRY contributed to the interpretation of the TCM data and pharmacological mechanisms. TCH contributed to the interpretation of Western medicine data. STY, YRL and HRY drafted the manuscript. STY and HRY finalized the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate This study followed the ethical standards of the responsible committee and with the Helsinki Declaration of 1964 and later versions. All the datasets were de-identified and encrypted before released by the National Health Research Institutes, Taiwan. All of the individuals or care providers could not be identified in the database. Patient consent was exempted for the total anonymity of all research data in this study. Therefore, the Research Ethics Committee of China Medical University and Hospital approved this study and waived the requirement for informed consent (CMUH104-REC2-115).

Consent for publication
Not applicable.