The economic burden of hypertension in rural south-west China


Corresponding Author Cai Le, School of Public Health, Kunming Medical University, 1168 Yu Hua Street Chun Rong Road, Cheng Gong New City, Kunming 650500, China. Tel.: +86 871 5922915; Fax: +86 871 5922911; E-mail:


Objective  To estimate the economic burden of hypertension in a given year in rural Yunnan Province of China, including direct, indirect and intangible costs.

Methods  A prevalence-based cost-of-illness method was used to estimate the economic burden of hypertension. Data on participants’ demographic characteristics, inpatient hospitalisation expenditures, outpatient visit expenditures, self-medication costs and indirect costs related to hypertension were collected from a cross-sectional health examination and questionnaire survey, involving 9396 consenting individuals aged ≥18 years and 3500 households. Blood pressure (BP) levels were determined from the average of three BP measurements. Years of life lost (YLL) because of hypertension was estimated using medical death certificates.

Results  The overall prevalence of and YLL/1000 population because of hypertension was 24.8% and 1.5 years for the survey population, respectively. Mean unit direct medical costs, direct non-medical costs, morbidity costs, mortality costs, intangible costs and cost of illness were estimated to be $467.2, $20.1, $23.5, $8265.1, $417.4 and $9393.3, respectively. The total cost of hypertension was estimated to be $231.7 million. Direct costs represented the largest component of economic cost of hypertension. On average, males had higher overall direct, indirect and intangible costs of hypertension than females. Both indirect and intangible costs decreased with age, whereas direct costs increased with age. The incidence of household catastrophic health payment and household impoverishment because of hypertension was 8.9% and 4.1%, respectively.

Conclusions  Hypertension inflicts a considerable economic burden upon individual households and society as a whole in Yunnan Province, China.


Objectif:  Estimer la charge économique de l’hypertension pour une année donnée, dans les régions rurales de la province du Yunnan en Chine, comprenant les coûts directs, indirects et intangibles.

Méthodes:  Une méthode du coût de la maladie basée sur la prévalence a été utilisée pour estimer la charge économique de l’hypertension. Les données sur les caractéristiques démographiques des participants, les dépenses d’hospitalisation des patients hospitalisés, les dépenses des visites ambulatoires, les coûts de l’automédication et les coûts indirects liés à l’hypertension ont été recueillies à partir d’une analyse transversale de la santé et une enquête par questionnaire, impliquant 9.396 personnes consentantes âgées de ≥18 ans et 3500 ménages. La tension artérielle (TA) a été déterminée à partir de la moyenne de trois mesures de la TA. Les années de vie perdues (AVP) à cause de l’hypertension ont été estimées à l’aide des certificats médicaux de décès.

Résultats:  La prévalence globale de l’hypertension et les AVP/1000 habitants due à l’hypertension étaient respectivement de 24.8% et de 1.5 ans pour la population étudiée. Les coûts unitaires moyens, médicaux directs et non médicaux directs, les coûts de la morbidité, les coûts de la mortalité, les coûts intangibles et coûts de la maladie ont été estimés respectivement à 467,2 USD; 20,1 USD; 23,5 USD; 8265,1 USD; 417,4 USD et 9393,3 USD. Le coût total de l’hypertension a été estiméà 231,7 millions de dollars. Les coûts directs représentaient la principale composante du coût économique de l’hypertension. En moyenne, les hommes avaient des coûts globaux directs, indirects et intangibles de l’hypertension plus élevés que les femmes. Les coûts indirects et intangibles diminuaient tous les deux avec l’âge, tandis que les coûts directs augmentaient avec l’âge. L’incidence de paiements catastrophiques des ménages pour la santé et de l’appauvrissement des ménages dus à l’hypertension était de 8,9% et 4,1%, respectivement.

Conclusions:  L’hypertension inflige une charge économique considérable sur les ménages et sur la société dans son ensemble dans la province du Yunnan, en Chine.


Objetivo:  Estimar la carga económica de la hipertensión en un año específico en la provincia rural de Yunnan, China, incluyendo costes directos, indirectos e intangibles.

Métodos:  Se utilizó un método basado en la prevalencia de coste-de-la-enfermedad para calcular la carga económica de la hipertensión. Se recolectaron datos sobre las características demográficas de los participantes, gastos de hospitalización para el paciente, gastos por consultas externas, costes de automedicación y costes indirectos relacionados con la hipertensión de un estudio croseccional de salud con examen clínico y encuesta, involucrando a 9,396 individuos con consentimiento informado y edades ≥18 años, y 3,500 hogares. Los niveles de presión sanguínea (PS) se determinaron del promedio de tres mediciones de la PS. Los años de vida perdidos (AVP) debido a la hipertensión se calcularon utilizando certificados médicos de defunción.

Resultados:  La prevalencia total de y AVP/1,000 habitantes debido a la hipertensión era del 24.8% y 1.5 años para el estudio poblacional, respectivamente. La media unitaria de los costes médicos directos, los costes directos no-médicos, los costes de morbilidad, costes de mortalidad, costes intangibles y costes de enfermedad eran respectivamente $467.2, $20.1, $23.5, $8,265.1, $417.4, y $9,393.3. El coste total de la hipertensión se calculó en $231.7 millones. Los costes directos representaban el mayor componente del coste económico de la hipertensión. En promedio, los hombres tenían unos mayores costes directos, indirectos e intangibles de hipertensión que las mujeres. Tanto los costes indirectos como intangibles disminuyeron con la edad, mientras que los costes directos aumentaron con la edad. La incidencia de los pagos catastróficos para los hogares y el empobrecimiento de los hogares debido a la hipertensión eran de 8.9% y 4.1%, respectivamente.

Conclusiones:  La hipertensión inflinge una carga económica considerable sobre los hogares individuales y la sociedad en general en la provincia de Yunnan, China.


Hypertension is a major health problem; approximately one billion people are affected worldwide (Hajjar et al. 2006). WHO (2002) estimates that half of cardiovascular diseases are attributable to elevated blood pressure. Accompanying economic growth and lifestyle changes, China has experienced an increase in hypertension over the last decade: the prevalence increased 2.4 times between 1991 and 2001 (Gu et al. 2002). Furthermore, hypertension is a growing cause of morbidity and mortality among rural Chinese, and the prevalence of hypertension varies widely among China’s ethnic groups (Yang et al. 2011).

Hypertension may affect more than 90% of hypertensive individuals during their lifetime (Vasan et al. 2002), and morbidity related to hypertension inflicts an enormous economic burden upon patients and societies as a whole (Veronesi et al. 2007). Hypertension is becoming one of the most costly health conditions worldwide (Vasan et al. 2002), and this cost is likely to continue to grow because of a rising prevalence and an ageing society. To optimise the use of limited healthcare resources, and better meet future demands for health services, it is critical to have an understanding of the economics of hypertension, including assessment of its economic impact.

Traditionally, economic burden of disease has been estimated in terms of direct costs, indirect costs and intangible costs. Numerous cost-of-illness studies over the years have been conducted to estimate the global economic burden of hypertension (Fletcher 1994; Druss et al. 2001; Berto et al. 2002). However, the majority of studies conducted in developing countries have largely focused on direct medical costs, while indirect economic consequences related to productivity loss and premature death are rarely calculated (Rizzo et al. 1996; Yang & Huang 1999). Intangible costs related to the psychosocial aspects of hypertension are often omitted from studies because of difficulty of measurement.

A Chinese study indicates that expenditure of drugs takes a large proportion of the total expenditure of hypertensive outpatients (Chen et al. 2003). However, data on the financial impact and economic burden of hypertension in China are still scarce, especially for rural areas. Previous studies indicate that more than 500 million rural Chinese citizens lack adequate access to basic healthcare services and are at risk of impoverishment in the case that a family member becomes afflicted by a chronic illness. Poverty because of illness has become a significant social problem in rural China (Wang et al. 2005; Sun et al. 2009). To fill the gap in the literature, this study used a standard prevalence-based cost-of-illness method from a societal perspective to estimate the direct, indirect and intangible costs of hypertension in a given year (2010) and assessed the economic impact of hypertension on households in rural Yunnan Province of China.


Study area and population

This study was conducted in two rural areas of Yunnan Province. Yunnan is located in south-west China and had a recorded population of 45.9 million people in 2010. Data from the year 2011 China Population Census indicate that people aged 15–59 years accounted for 68.20% of the Yunnan’s population, whereas 11.07% of Yunnan population were aged 60 years and over. Yunnan is a relatively undeveloped province and is home to 25 of China’s ethnic minority groups. In 2010, Yunnan’s per capita GDP was US$2379, and the crude death rate was 487.2 per 100 000 population (Yunnan Statistical Yearbook 2010). Hypertension was responsible for 3.4% of the total deaths.

Study design

A cross-sectional health interview and examination survey were conducted and combined with a review of death certificates.

Death certificate study

A calculation of mortality because of hypertensive disease in 2010 was based on medical death certificate information obtained from the Yunnan Center for Disease Control (CDC). All death reports were grouped by underlying cause of death as defined in the World Bank and World Health Organization’s Global Burden of Disease (GBD) study (Murray & Lopez 1996) and coded using the International Classification of Diseases, 10th revision (ICD-10) coding system.

Community survey

Sampling technique

A multi-stage stratified random sampling method was used to select the study sample. In the first stage, a computer database was used to divide Yunnan Province’s counties into two categories – economically advantaged and economically disadvantaged – according to wealth distribution (per capita GDP), with the median value as the cut-off point. One county was then randomly selected from each of the two categories. In the second stage, the probability proportional to size (PPS) method was used to select 20 villages from each of the two selected counties. In the final stage, a list of individuals aged ≥18 years was obtained from the village committee of every selected village, and then simple random sampling was applied to select the sample subjects.

Guan Du and Luo Ping counties were selected as the study areas. Guan Du is an economically advantaged rural area with a population of 462 246 (229 444 males and 232 802 females) containing nine townships. Luo Ping County is an economically disadvantaged rural area with a population of 560 848 (296 592 males and 264 256 females) containing 12 townships. In 2010, Guan Du and Luo Ping’s per capita GDP was US$ 3175 and US$ 2068 (Yunnan Statistical Yearbook 2010), respectively.

Data collection and measurement

Researchers were trained to take blood pressure measurements and conduct the questionnaire. Each consenting participant was individually interviewed face to face by trained interviewers who used a pre-tested structured questionnaire. Information on demographic characteristics and history of diagnosis and treatment of hypertension, as well as results from anthropometric measurements and blood pressure was obtained. Information on annual inpatient hospitalisation expenditures, outpatient visit expenditures, self-medication costs, costs of transportation and accommodation during hospital visits, and wages lost because of absence from work from hypertension was also collected for subjects who reported a previous diagnosis and/or treatment of hypertension by a medical doctor and/or use of antihypertensive medications during the course of the survey.

Three consecutive blood pressure measurements were made in accordance with American Heart Association recommendations (Perloff et al. 1993). After at least 5 min of rest in a sitting position, systolic and diastolic pressures were taken from the participant’s right arm using standardised mercury sphygmomanometers. Final recorded BP levels were determined from an average of three BP readings.

Ethical approval

This study was approved by the Ethics Committee of Kunming Medical University, before carrying out the research.

Cost estimation

The economic burden of hypertension was calculated from the sum of direct costs, indirect costs and intangible costs.

Direct cost was divided into two subcategories: direct medical costs and direct non-medical costs. Direct medical costs were those healthcare expenditures used for hospitalisation fees, outpatient visits and self-purchased medications. The costs associated with outpatient/inpatient visits were estimated by multiplying the numbers of outpatient visits/inpatient hospital admissions related to hypertension by the outpatient/inpatient unit costs per year. The patient’s self-estimate was recorded for self-medication costs. Direct non-medical costs included costs of transportation to and accommodation for the hypertensive patient and relatives while visiting health providers, as well as the costs of hiring caregivers.

Estimation of indirect costs

Indirect costs were estimated using the human capital model (Pritchard & Sculpher 2000). This model is comprised of two categories: morbidity costs and mortality costs, as elaborated upon below.

Morbidity costs

Morbidity cost was defined as the amount of lost income because of work disability and absence as a result of hypertension. It was calculated by multiplying the total number of days taken off for both the patient and relevant family members, by the average gross earning per day per person in 2010. Income data were taken from the 2010 Yunnan Statistical Yearbook (Yunnan Statistical Yearbook 2010), where the reported annual per capita net income of rural households was 3952 RMB (US$597).

Mortality costs

Mortality cost was defined as the amount of lost income attributable to premature mortality associated with hypertension. Mortality costs were estimated by multiplying YLL by per capita annual net income of rural households.

Calculation of years of life lost (YLL)

Calculation of YLL is one of several methods for estimating and quantifying life lost because of premature death; it is the mortality component of disability-adjusted life years (DALY) (Murray 1994). We followed the methods used in the GBD study (Murray & Lopez 1996). The formula of YLL is as follows:


where K = age-weighting modulation factor; C = the age-weighting correction constant; r = discounting rate; a = age of death; β = parameter from the age-weighting function; L = standard expectation of life at age a from the national life table.

Consistent with standard GBD approach, C was assigned a value of 0.1658, and β was assigned a value of 0.04. YLL was calculated with a 3% discounting rate and a standard age-weighting function.

Estimation of intangible costs

Intangible cost was defined as the cost related to earning loss because of psychological symptoms associated with hypertension. Intangible costs were estimated using a contingent valuation willingness-to-pay approach (Mitchell & Carson 1989). This method involved asking hypertensive patients how much they would have been willing to spend to avoid developing the illness.

The total community cost was estimated by multiplying the overall cost calculated for the sample group by the ratio of entire community population divided by the sample population.

All costs presented in US dollars were calculated using the exchange rate at the time of data collection ($1 = 6.62 RMB).


A patient was defined as having hypertension when they were recorded to have a mean systolic blood pressure measurement of ≥140 mm Hg, a mean diastolic blood pressure measurement of ≥90 mm Hg and/or self-reported current use of antihypertensive medications.

Catastrophic health payment was defined as occurring when health expenditures exceed 40% of a household’s non-food expenditures or capacity to pay for services, in the 12 months prior to the interview (Xu et al. 2003).

Households were defined as impoverished because of medical expenses when per capita expenditure within the household decreased to below that of the established poverty line (1300 yuan per year per capita in 2008 in China) after health payment (Yingchun et al. 2005; O’Donnell et al. 2007).

Statistical analysis

Descriptive analysis methods were used in the statistical analysis of this study. Prevalence of hypertension for the total sample population stratified by age and sex was calculated. Categorical variables were presented in terms of frequency and percentages. The incidence of catastrophic health payment was calculated by dividing the total number of households facing financial catastrophe because of hypertension by the total number of households included in the study. Similarly, the incidence of medical impoverishment because of hypertension was estimated by dividing the total number of households impoverished by related health spending by the total number of households. The direct standardization method was used to calculate the age-adjusted YLL rate. The total population of Yunnan Province was used as the standard population. All statistically significant decisions were based on two-tailed P values. All data analyses were conducted with R 2.9.1 software (R Development Core Team. 2006).


Prevalence and premature death from hypertension

A total of 3500 households involving a total of 10 000 individuals were asked to participate in this study. Of these, 9396 consented, for an overall response rate of 94%. Of the 3500 households, 3052 (87.2%) had more than one family member aged 18 years and over included in the survey.

Table 1 summarises the demographic characteristics of the study population and rates of prevalence, awareness and treatment of hypertension among study participants. The participant population consisted of 4319 males and 5077 females. The overall adult illiteracy rate was 33.1%, and the mean annual household income was $627, varying from $65 to $2993. A total of 2333 people suffered from hypertension among the surveyed population, for an overall prevalence rate of 24.8% (23.8% for males and 25.7% for females).

Table 1. Demographic characteristics of the study population
CharacteristicsMale (= 4319)Female (= 5077)All (= 9396)
Age (%)
 18–44 years1527 (35.4)1722 (33.9)3249 (34.6)
 45–59 years1420 (32.9)1772 (34.9)3192 (34.0)
 ≥60 years1372 (31.8)1583 (31.2)2955 (31.5)
Ethnicity (%)
 Han4020 (93.1)4647 (91.5)8667 (92.2)
 Minorities 299 (6.9) 430 (8.5) 729 (7.8)
Level of education (%)
 Illiterate 890 (20.6)2224 (43.8)3114 (33.1)
 Primary school (grade 1–6)1350 (31.3)1457 (28.7)2807 (29.9)
 Middle school (grade 7–9) or higher2079 (48.1)1396 (27.5)3475 (37.0)
Approximate yearly household income (US$)
 Mean 638 593 627
 Min–Max  87–3025  65–2993  65–2993
 Hypertension prevalence (%)1026 (23.8)1307 (25.7)2333 (24.8)
 Awareness of condition (%) 431 (42.0) 716 (54.8)1147 (49.3)
 Treatment rate of hypertension (%) 383 (37.3) 618 (47.3)1001 (42.9)

Among the 2333 hypertensive patients, 1147 (431 male and 716 female) were aware of suffering from hypertension, for an overall awareness rate of 49.3%, and females had a higher rate of awareness than males (54.8%vs. 42.0%, P < 0.01).

In total, 1001 patients (383 male and 618 female) received antihypertensive treatment, for an overall treatment rate of 42.9%. Females had a higher treatment rate than males (47.3%vs. 37.3%, P < 0.01). Among these patients, the mean age was 63.4 ± 31.7, and the mean household income was $616. The adult illiteracy rate of the patients was 33.5%, with females having a higher illiteracy rate than males (41.9%vs. 19.8%, P < 0.01).

Table 2 illustrates the age-adjusted YLL because of hypertension by sex and age among study participants. There were 143 total deaths attributable to hypertensive disease (75 male, 68 female) among the sample population in 2010. The average age-adjusted YLL/1000 population from the 143 deaths because of hypertension was 1.4 years. Females had slightly higher average age-adjusted YLL/1000 population because of hypertension than males (1.4 vs. 1.3). People aged 60 years and more were responsible for the highest average number of years of life lost in comparison with other age groups.

Table 2. Age-adjusted years of life lost (YLL) because of hypertension by sex and age in Yunnan Province 2010
CharacteristicsNumber of deathsYLLYLL/1000 population (95% CI)Age-adjusted YLL/1000 population (95% CI)
 Male 75 759.11.4 (1.0–1.8)1.3 (0.9–1.7)
 Female 68 778.21.6 (1.1–2.1)1.4 (1.0–1.8)
 18–44 years  6 147.60.1 (0.07–0.13)0.2 (0.09–0.31)
 45–59 years  6 117.80.8 (0.71–0.89)0.7 (0.6–0.8)
 ≥60 years1311272.04.7 (4.1–5.3)4.5 (4.0–5.5)
All1431537.41.5 (0.9–2.1)1.4 (0.8–2.0)

Cost of illness

Table 3 shows the different cost components of hypertension and compares unit cost by age and sex for the surveyed communities. Among the sample population, overall mean unit direct medical costs, direct non-medical costs, morbidity costs, mortality costs, intangible costs and cost of illness were estimated to be $467.2, $20.1, $23.5, $8265.1, $417.4 and $9393.3, respectively. The total cost of hypertension amounted to $231.7 million. Of this, 53.4% was direct costs, 3.1% was indirect costs and 45.7% was intangible costs. Direct costs comprised the largest component of the economic cost of hypertension, whereas hospital cost estimates comprised the largest component of direct medical expenditures.

Table 3. Cost of illness of hypertension (in US$) in rural areas of Yunnan Province 2010
Cost componentsUnit costTotal% of total cost of illness
MaleFemale18–44 years45–59 years≥60 yearsAll
Direct medical costs522.0435.0276.9361.5561.0467.2118 541 254.4 51.2
 Outpatient visits24.030.921.023.131.828.37 180 474.1  3.1
 Hospitalisation367.8295.3202.8232.0400.6322.081 700 094.0 35.3
 Self-medication130.1108.753.1106.4128.5116.929 660 686.3 12.8
Direct non-medical costs20.119.713.41921.120.15 099 912.7  2.2
 Transportation2. 299.1  0.4
 Accommodation14.514.710.214.115.414.63 704 414.2  1.6
 Payment to caregivers3. 199.4  0.2
Morbidity costs25.422.921.628.321.923.55 962 584.5  2.6
Mortality costs8356.68103.015260.912059.46030.68265.11 181 912.2  0.5
Total direct costs542.1454.7290.3380.5582.1487.3123 641 167.1 53.4
Total indirect costs8382.08125.915282.512087.76052.58288.67 144 496.7  3.1
Total intangible costs493.1372.5704.6454.4311.3417.4105 905 780.0 45.7
Total cost of illness9617.29153.116477.413122.67145.99193.3231 651 863.6100.0

Males had higher overall direct, indirect and intangible costs of hypertension than females (P < 0.05). Both indirect and intangible costs decreased with age, whereas direct costs increased with age (P < 0.01). Hypertensive patients aged 45–59 years incurred greater morbidity costs as compared to other age groups, whereas mortality costs were highest among patients aged 18–44 years.

Economic impact of hypertension on households

As shown in Table 4, the incidence of household catastrophic health payment because of hypertension was 8.9% among the surveyed communities. Prior to factoring in hypertension-related health expenditures, 269 households were categorised under the poverty line. Once the health expenditures were factored in, the number of households living under the poverty line increased to 400, thus the incidence rate of household impoverishment because of hypertension was calculated at 4.1%.

Table 4. Incidence of financial catastrophe and household impoverishment because of hypertension in rural areas of Yunnan Province
Variable n/N% (95% CI)
Catastrophe health payment310/35008.9 (8.0–9.9)
Household impoverishment131/32314.1 (3.4–4.8)


This community-based cross-sectional study provided a comprehensive and detailed cost analysis of hypertensive patients using a prevalence-based cost-of-illness methodology. The findings indicate that hypertension is a major public health problem in terms of morbidity and economic burden in rural south-west China.

The study indicated that there are sex differences in the level of awareness and treatment of hypertension in the region under study. Women had higher levels of awareness and treatment of hypertension than men. This finding is in concordance with other studies (Kearney et al. 2004; Pereira et al. 2009). The results suggest that gender should be an important consideration in the management of BP, and programmes geared towards increasing awareness of hypertension should be directed more towards the male gender.

The results show that direct costs comprised the largest component of overall expenditure related to hypertension, whereas hospital cost estimates comprised the largest component of direct medical expenditures (over 50%). This finding is similar to other previous studies conducted in Taiwan and the USA (Yang & Huang 1999; Hodgson & Cai 2001), but differs from a Canadian study, where drug costs accounted for the largest portion of direct medical costs (Maetzel et al. 2004). The high hospital cost in our study possibly due to the fact that 60.3% of the hypertensive patients of the sampled communities were elderly patients aged 60 years and more and older patients had longer average hospitalisation stay than younger patients (15.9 days vs. 12.2 days). Relatively, indirect costs did not have as much impact in the calculation of the economic burden of hypertension, perhaps indicating that hypertension had less impact on the expense of productivity loss because of premature mortality and productivity loss–related morbidity.

In the literature, one previous study suggested that hypertension had minimal adverse effects on the patients’ quality of life (Fletcher 1994). However, our results showed that intangible costs comprised 45.7% of the total cost of illness of hypertension. Comparison of age and sex-specific intangible costs revealed that the amount of willingness to pay was higher among men and decreased with age, suggesting that hypertension imposes a heavy physical and psychological burden to rural south-west Chinese hypertensive patients, especially men and working-age patients.

The present study found that age and sex were strong determinants of the cost of hypertension. Males had higher unit direct medical, morbidity, mortality and intangible costs of hypertension than females. Unit direct cost of hypertension was positively associated with age, whereas unit mortality, indirect and intangible costs were negatively associated with age. The findings suggest that age and sex should be important considerations when determining the appropriating of financial resources for hypertension prevention and intervention programmes.

The economic effect of chronic diseases has adversely and disproportionately affected poor and vulnerable populations in the developing world (Engelgau et al. 2011). Several recent studies indicate that chronic diseases have brought a heavy financial burden to rural communities throughout China and that poverty because of illness has become a significant social problem (Sun et al. 2009; Yu et al. 2010). In our study, the overall incidence of household catastrophic health payment and household impoverishment because of hypertension was 8.9% and 4.1%, respectively. These results are similar to those found in a previous study conducted in rural areas of western and central China (Shi et al. 2010). The findings indicate that medical expenditures have become a significant source of impoverishment in rural China.

It is important to acknowledge methodological differences when comparing the results of cost-of-illness analyses. A comparison of costs of hypertension studies conducted in China is difficult because of the difference in research methodologies. Findings may be heavily influenced by the variety of cost components included and analysed in those studies. Inclusion of more hypertension-related cost categories will result in a higher overall estimated cost. In China, studies on the costs of hypertension are sparse, and the majority of previous studies has focused on direct healthcare costs and overlook indirect and intangible costs. These studies may have underestimated the economic impact of hypertension on healthcare resources. Our findings offer a more complete picture of the economic burden of hypertension, which will be beneficial for policy makers in China.

The large sample size and high response rate enhance the validity of our results, and the measurements being obtained from health examinations, to ensure the accuracy of data. However, the findings in our study may be subject to several limitations. First, we only considered the costs of hypertension directly and did not factor in costs incurred by hypertension-related diseases, which likely affected the accuracy of our estimated overall cost of hypertension. The direct cost of hypertension by itself represented only a small portion of the overall costs of hypertension. Rather, the greatest health services expenditures are because of hypertension-related diseases rather than hypertension itself (Veronesi et al. 2007). It is likely that by disregarding such comorbidities, the true economic burden has been underestimated. Second, the recorded amounts for direct costs related to hypertension were based on patient recall and may therefore be subject to bias. Third, the community’s willingness to pay was significantly affected by household income, which hypertensive patients with higher household income were willing to pay more for avoiding hypertension, and this may have double counted the intangible costs of hypertension. Finally, more than one household member from the same household participated in the survey, this would leave large households weighing more and within household correlations likely occurring and both could bias our results.

In conclusion, this study provides a comprehensive cost analysis of the economic burden of hypertension for a large representative sample of patients. Our findings indicate that hypertension has a great impact on healthcare resource utilisation, and we suggest that more financial resources should be invested towards the prevention and management of hypertension in rural south-west China.


We would like to thank Ms. Alison Borochoff-Porte for reviewing the manuscript. She is a visiting student at the School of Public Health, Kunming Medical University. The study was supported by grants from national natural science funds of China (Grant number: 30960335) and Yunnan provincial natural science funds (Grant number: 2008CD115).