Article Text

The healthcare costs of secondhand smoke exposure in rural China
1. Tingting Yao1,2,
2. Hai-Yen Sung1,
3. Zhengzhong Mao3,
4. Teh-wei Hu4,5,
5. Wendy Max1
1. 1Institute for Health & Aging, University of California, San Francisco, California, USA
2. 2Center for Tobacco Control Research and Education, University of California, San Francisco, California, USA
3. 3Huaxi School of Public Health, Sichuan University, Chengdu, China
4. 4Center for International Tobacco Control, Public Health Institute, Oakland, California, USA
5. 5School of Public Health, University of California, Berkeley, California, USA
1. Correspondence to Dr Tingting Yao, Institute for Health & Aging, University of California, San Francisco, 3333 California Street, Suite 340, San Francisco, CA 94118, USA; tingting.yao{at}ucsf.edu

## Abstract

Objective The goal of this study was to assess the healthcare costs attributable to secondhand smoke (SHS) exposure among non-smoking adults (age ≥19) in rural China.

Methods We analysed data from the 2011 National Rural Household Survey which was conducted among adults in five provinces and one municipality in China (N=12 397). Respondents reported their smoking status, health conditions and healthcare expenditures. Relative risks were obtained from published sources. Healthcare costs included annual outpatient and inpatient hospitalisation expenditures for five SHS-related diseases: asthma, breast cancer (female only), heart disease, lung cancer and tuberculosis. SHS-attributable healthcare costs were estimated using a prevalence-based annual cost approach.

Findings The total healthcare costs of SHS exposure in rural China amounted to $1.2 billion in 2011, including$559 million for outpatient visits and $612.4 million for inpatient hospitalisations. The healthcare costs for women and men were$877.1 million and $294.3 million, respectively. Heart disease was the most costly condition for both women ($701.7 million) and men ($180.6 million). The total healthcare costs of SHS exposure in rural China accounted to 0.3% of China's national healthcare expenditures in 2011. Over one-fifth of the total healthcare costs of SHS exposure in rural China were paid by health insurance. The out-of-pocket expenditures per person accounted for almost half (47%) of their daily income. Conclusions The adverse health effects of SHS exposure result in a large economic burden in China. Tobacco control policies that reduce SHS exposure could have an impact on reducing healthcare costs in China. • Economics • Global health • Low/Middle income country • Public policy • Secondhand smoke ## Statistics from Altmetric.com ## Request Permissions If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways. ## Introduction China is the largest producer and consumer of tobacco in the world. According to the 2010 Global Adult Tobacco Survey, 301 million Chinese adults were current smokers, and the prevalence of smoking among adults in China was 52.9% for men and 2.4% for women in 2010.1 In addition, a report from the Chinese Center for Disease Control and Prevention indicates that over 556 million adults were exposed to secondhand smoke (SHS) in China in 2010,2 a number that was greater than the number of current smokers. China's adult smoking prevalence has decreased recently, dropping from 31.1% in 20022 to 28.1% in 2010.1 However, exposure to SHS in China did not decline in the past decade.2 In rural areas, SHS exposure among those aged 15 years and above increased substantially from 54% in 2002 to 74.2% in 2010.2 ,3 A recent study found that 68% of children and 67.7% of women living in rural areas were exposed to SHS at home in 2008.4 SHS exposure has been linked to several illnesses, including respiratory diseases and other adverse health effects in children, and lung cancer and heart disease in adults.5 ,6 The health effects of SHS exposure could result in excess economic costs. A few studies have estimated the economic costs of SHS exposure in China. Leung et al7 examined the impact and economic costs associated with SHS exposure among infants in Hong Kong with non-smoking mothers. They found that postnatal exposure to SHS at home was linked to higher rates of hospitalisations for any illness compared with unexposed infants (OR=1.1), leading to 662 extra hospitalisations and$0.9 million excess inpatient costs in 1997. Gan et al8 estimated the disease burden of SHS exposure among Chinese adults aged 30 years or above and found that SHS exposure was responsible for the loss of nearly 230 000 years of healthy life from lung cancer, and more than one-quarter of a million years of healthy life from ischaemic heart disease. Li et al9 estimated that the total cost of SHS exposure in China was ¥14.5 billion (US$2.1 billion) in 2000, and ¥29.4 billion (US$ 4.3 billion) in 2005. No study to date has analysed the healthcare costs attributable to SHS exposure in rural China.

Since persons living in rural China often have lower income levels than urban dwellers, they may suffer a heavier economic burden from SHS exposure. Therefore, estimates of the health-related costs attributable to SHS exposure in rural China are needed to understand the economic impact of SHS exposure, and to motivate policymakers to implement smoke-free and other policies to reduce the health and economic toll. Accordingly, the objective of this study is to estimate the healthcare costs attributable to SHS exposure for non-smoking adults (age ≥19) in rural China.

## Methods

We used a prevalence-based, disease-specific approach to estimate SHS-attributable healthcare costs.

### Data sources

All analyses in this study were conducted using data from the 2011 National Rural Household Survey (NRHS). The NRHS is a nationally representative face-to-face survey of Chinese rural households conducted by the China National Health Development Center in 2011 in five provinces (Qinghai, Anhui, Hubei, Yunnan, Jiangsu) and one municipality (Chongqing) in China. These provinces and municipality cover about one-sixth of the geographic area of China with locations around the country, covering a range of rural areas in China. In the five provinces, three rural villages were randomly selected based on different income levels (high, middle and low). In Chongqing, two districts were randomly selected, and three rural villages were then randomly selected in each district.

The head of each household was interviewed, and he/she also reported on behalf of all household members. The questionnaire consisted of three sections with 114 total questions. The following variables from the NRHS were used in our study: (1) gender; (2) smoking status (those who answered ‘yes’ to the question ‘Did you smoke cigarettes in the previous 30 days?’ were classified as a ‘current smoker’); (3) SHS exposure—defined by whether a participant lived with a current smoker or not; (4) chronic conditions in the past 12 months; (5) disease-specific costs for outpatient visits in the past month, and disease-specific costs for inpatient hospitalisations in the past 12 months; and (6) how much of outpatient visits and inpatient visits was covered by health insurance.

#### Study sample

A total of 4249 households (including 13 933 adults aged 19+) were interviewed in 2011. After excluding respondents with missing information on gender, smoking status, chronic conditions and healthcare expenditures, a total of 12 397 adults were included in the final study sample.

### Measures

#### SHS exposure

Our study focused on non-smokers, that is, those who reported not being current smokers. Though smokers may also suffer harmful health effects from SHS exposure, it is difficult to separate the effects from those due to active smoking. We defined non-smokers who are exposed to SHS as those non-smokers who live in a household with at least one current smoker.

#### SHS-related diseases

We included five SHS-related diseases for adults based on the 2006 Surgeon General's report5 and the 2005 California Environmental Protection Agency (EPA) report.6 These diseases are asthma, breast cancer (for females aged 19–50 years only), heart disease, lung cancer and tuberculosis (TB), as shown in table 1. In the NRHS, respondents could report one to three diseases diagnosed by doctors. Interviewers then recorded these disease names in the data set. We considered respondents to have SHS-related diseases if they reported any of the five diseases just discussed.

Table 1

Relative risk of secondhand smoke (SHS)-related diseases

#### Relative risks

The relative risk (RR) of a disease from SHS exposure among non-smokers is defined as the ratio of the disease rate for exposed non-smokers to the disease rate for unexposed non-smokers. The RRs of SHS-related diseases were obtained from published studies conducted in China except for asthma. Since RR estimates for asthma have not been published for the Chinese population, we used the RR estimates from a Finnish study,10 as cited in the California EPA report.6 The RRs by disease and gender are shown in table 1.

#### SHS-attributable fraction

A smoking-attributable fraction is commonly used for the estimation of health or economic burdens attributable to smoking, such as disease incidence, healthcare utilisation and healthcare expenditures. Similarly, the SHS-attributable fraction (SAFshs) measures the proportion of health or economic burden in the specific population (eg, non-smoking adults) that can be attributed to SHS exposure. In this study, the SAFshs for each subgroup stratified by disease (i) and gender (j) was estimated using the standard epidemiological formula (1)16: 1

where Pshs is the prevalence (%) of SHS exposure among non-smokers and RRshs the RR of illness for non-smokers who were exposed to SHS compared to those who were not exposed.

### Healthcare costs

In this study, healthcare costs include expenditures for outpatient visits and inpatient hospital stays. Outpatient visits included ambulatory care visits at outpatient departments of hospitals and doctor visits at clinics. Expenditures for prescribed medicine during the outpatient visits or inpatient stays were included in the respective outpatient or inpatient expenditure categories. In the NRHS, respondents were asked to report (1) the average outpatient expenditures per month for each disease treated in the past 6 months, and (2) the disease name and total inpatient expenditures, including out-of-pocket payment and insurance payment, for the most recent hospitalisation stay during the past 12 months. The SHS-attributable healthcare costs (SAEshs) among the rural population was estimated by disease (i) and gender (j) by multiplying the SAF by the total healthcare expenditures (THE) for the non-smoking rural population according to the following formula: 2where POP denotes the rural adult population aged ≥19; PNS is the percentage of rural adults who were non-smokers; DRATE is the disease prevalence rate among non-smoking rural adults; INPX denotes the average annual inpatient expenditures per person among rural non-smoking adults who had that SHS-related disease; and OUTX denotes the average monthly outpatient expenditures per person among rural non-smoking adults who had that SHS-related disease. POP was obtained from the 2011 Statistical Yearbook,17 while all other variables were estimated from the 2011 NRSH data.

### Health insurance coverage

In the NRHS, respondents were asked to report the annual outpatient expenditures covered by health insurance for each disease treated in the past 6 months and the inpatient expenditures covered by health insurance for the most recent hospitalisation stay during the past 12 months. We first calculated the mean value of costs covered by health insurance for each disease and then applied them to formula (2) to get the costs covered by health insurance among total SHS-attributable costs.

### Sensitivity analysis

Three sensitivity analyses were conducted. The first one was performed based on the lower bound and upper bound of the 95% CIs of RR estimates in table 1. The second one was performed using the RRs of SHS-related diseases estimated from US studies5 ,6 ,18 ,19 except for asthma, because the US RR estimates for asthma were not available. The third one was performed by using published disease rates for the Chinese population.20–22 This was carried out because the disease rates reported in the NRHS are lower than those generally published.

## Results

### Number of non-smoking adults living in rural China

The final study sample of the 2011 NRHS had nearly equal numbers of men and women, as shown in table 2. Women were much more likely to be non-smokers, and among non-smokers the prevalence of SHS exposure was 62.2% for women and 35% for men.

Table 2

Prevalence of non-smoking and number of non-smoking adults living in rural China, 2011

### Number of persons with SHS-related diseases among non-smoking adults in rural China

Table 3 shows the number and disease rates of SHS-related diseases in rural China by gender, type of healthcare service used and disease. Outpatient visits and inpatient hospitalisation were most often reported by people with heart disease (especially women), followed by TB and asthma. There were no inpatient hospitalisations reported for lung cancer or breast cancer.

Table 3

Disease prevalence rate and number of persons with secondhand smoke (SHS)-related diseases among non-smoking adults in rural China, 2011

### SHS-attributable healthcare costs among non-smokers in rural China

Table 4 shows the estimated SAFshs by gender, disease and healthcare service type. The SAFshs values for all diseases were larger for women than men. Asthma had the highest disease-specific SAFshs for both women (0.38) and men (0.25), while lung cancer showed the lowest SAFshs for both women (0.07) and men (0.04).

Table 4

Secondhand smoke (SHS)-attributable fraction (SAFshs) and SHS-attributable healthcare costs among non-smokers in rural China, 2011

Table 4 also shows the healthcare costs of SHS exposure by gender, disease and healthcare service type estimated as the product of the total number of persons with SHS-related disease, the average annualised costs per person with the SHS-related disease, and the SAFshs. The total healthcare costs of SHS exposure in rural China in 2011 amounted to $1.2 billion:$559 million for outpatient visits and $612.4 million for inpatient hospitalisations. The total cost for women ($877.1 million) was almost three times as high as the cost for men ($294.3 million). For women, heart disease was the most costly disease ($701.7 million), followed by TB ($90.6 million), asthma ($44.5 million), breast cancer ($40.2 million) and lung cancer ($0.1 million). For men, the results showed a similar pattern: heart disease was the most costly disease ($180.6 million), followed by TB ($70.3 million), asthma ($41.9 million) and lung cancer ($1.5 million). About one-fifth of healthcare costs ($237.7 million) were covered by health insurance, including$28.5 millon for outpatient visits and $209.2 million for inpatient visits. ### Sensitivity analyses Table 5 shows that the total healthcare costs of SHS exposure in rural China in 2011 ranged from$573.6 million to $1.7 billion using the upper and lower bounds of the RR estimates. Costs almost doubled ($2.1 billion) when using the RRs for the US population and slightly increased to $1.22 billion when using the published disease rates for the Chinese population. Table 5 Sensitive analyses of secondhand smoke (SHS)-attributable healthcare costs among non-smokers in rural China, 2011 ## Discussion Our findings indicate that SHS exposure in rural China imposes a large economic burden ($1.2 billion) on China, accounting for 0.02% of China's gross domestic profit ($6.7 trillion in 201117) and 0.3% of the national healthcare expenditures in 2011 ($385 billion in 201117). Given the total rural population of 650 million in 2011,17 this amounted to $1.80 per person in China. Nearly three-quarters of the total healthcare cost of SHS exposure in rural China was for women because a majority of women are non-smokers who live with smokers. Over one-fifth of the total healthcare cost of SHS exposure in rural China was paid by health insurance. The out-of-pocket expenditures per person ($1.44) accounted for almost half (47%) of their daily income ($3.03) in rural China. Our estimates most likely underestimated the economic burden of SHS exposure for several reasons. First, our calculation only included five SHS-related diseases for adults. Children are also vulnerable to SHS exposure and suffer from diseases that are related to SHS exposure including low birth weight,6 middle ear disease6 and attention deficit hyperactivity disorder.23 According to the SHS exposure rate for children (68%) in rural China,4 we estimated that about 100.4 million children were exposed to SHS in rural China in 2011. Therefore, the total costs would be much higher if the costs of children's SHS exposure could be also included. Second, our study was limited to direct healthcare costs only. Future studies that include indirect morbidity and mortality costs attributable to SHS exposure are needed. Third, besides the home environment, people are also exposed to SHS in workplaces and public places. A recent report in 20102 found that 73% and 63% of Chinese were exposed to SHS in public places and workplaces, respectively. However, we were not able to separate the impact of SHS exposure at home from exposure in other settings. Fourth, our study is based on self-reported data and only the head of the household was interviewed, which may create recall bias. Fifth, the RRs we used were lower than those from western countries. For example, the RR of heart disease among men was 1.22 for the Chinese population compared to 1.50 for the US population.5 The RR of TB was 1.55 for the Chinese population compared to 2.33 for the US population.19 According to the sensitivity analysis which assumed the RRs for the US population, we found that the estimated total healthcare costs attributable to SHS in rural China would almost double from$1.2 billion to $2.1 billion if we used the US RRs, and the costs would then account for 0.6% of China's national healthcare expenditures in 2011. The higher RRs in the US population reflect a more mature epidemic of smoking, and suggest that risks of smoking-related disease in China may increase in the future. Sixth, very few cases of SHS-related diseases were reported in the NRHS, with the exception of heart disease. It is possible that these diseases were under-reported due to the lack of a formal diagnosis. Therefore, we conducted another sensitivity analysis using disease rates from published Chinese studies and found that the estimated SHS-attributable total healthcare costs in rural China increased to$1.22 billion. In addition, the rate of outpatient visits for heart disease in our study was much higher than that from a published Chinese study (2.89% vs 0.77%). This might be because the latter study included all of China and our study focuses only on rural areas. More studies are needed to address this. Finally, the RRs used in this study were based on findings from previous studies, which did not control for exposure to air pollution. This may cause an upward bias of our estimated SHS-attributable fractions and healthcare costs.

The study by Yang et al24 provided the most recent estimate of the total healthcare cost of active smoking in China at $6.2 billion in 2008, including$2.8 billion for urban areas and $3.4 billion for rural areas. Converting these estimates into 2011 dollars using the Consumer Price Index (105.9 in 2008 and 105.4 in 2011),17 the healthcare cost of active smoking for rural China would be$3.38 billion. Therefore, our estimated healthcare cost attributable to SHS exposure was more than one-third of the cost of active smoking in rural areas. The ratio of SHS costs to active smoking costs in our study was higher than a study conducted in Hong Kong, which reported a ratio of 29%.25 Our study demonstrates that the true impact of smoking on healthcare costs in rural areas would be one-third higher than the estimates by Yang et al24 when the burden of SHS exposure is included. Given that the majority of women and children are non-smokers and exposed to SHS in rural China, tobacco control interventions that reduce SHS exposure are needed.

This research serves as a starting point for a comprehensive assessment of the health and economic impact of SHS exposure in China. The findings provide useful evidence for policymakers who are developing interventions to reduce SHS exposure and increase public awareness in China.

### What this paper adds

What is already known on this subject

• The secondhand smoke (SHS) exposure rate increased substantially in the past decade in China. SHS exposure has been linked to several illnesses.

• No study to date has analysed the healthcare costs attributable to SHS exposure in rural China

What important gaps in knowledge exist on this topic

• This study is the first to estimate the healthcare costs of SHS exposure in rural China.

What this study adds

• We found that the healthcare costs of SHS exposure constitute a large economic burden to China.

• Our findings demonstrate the importance of implementation of tobacco control policies that reduce SHS exposure in rural China as a means of reducing healthcare costs.

## Acknowledgments

The authors would like to thank the China National Health Development Research Center for their collection of data.

## Footnotes

• Contributors T-wH and ZM obtained funding to conduct the National Rural Household Survey and helped collect data. TY conducted all the data analyses and wrote the first draft of the manuscript. WM and H-YS helped design the analytical approach and the data analysis and contributed to the interpretation of the analytical results. All authors provided comments, assisted in revising the drafts and approved the final manuscript.

• Funding This study was funded by grants from the US National Institutes of Health Fogarty International Center (grant R01 TW009295) and the National Cancer Institute; the Australia Government's Overseas Aid Program (HSS080020), the China-Australia Health and HIV/AIDS Facility, the Operational Research on Integration of NCMS and MA Rural Health Financing Schemes; the California Tobacco-Related Disease Research Program, Cornelius Hopper Diversity Award Supplement (20CA-0102); the University of California, San Francisco Dorothy Pechman Rice Postdoctoral Fellowship; and the US National Cancer Institute (grant CA-113710).

• Competing interests None.

• Patient consent Obtained.

• Ethics approval Ethics approval was obtained from the Office of Research Ethics at the China National Health Development Center.

• Provenance and peer review Not commissioned; externally peer reviewed.