The Analysis Report of Notional Health Services Survey in China 2013 (ARNHSS) provided consultation rates data for 2 weeks, categorized by the diseases and hospitalization rates and constituent ratio of outpatient and inpatient institutions in 2013. The China Health and Family Planning Statistical Yearbook 2014 (CHFPSY) provided the data for the average outpatient costs, hospitalization costs and the length of stay by diseases at different hospital levels and the total national health expenditure in 2013. The China Statistical Yearbook 2014 (CSY) provided the urban and rural population numbers and the Gross Domestic Product (GDP) data in 2013.

Every 5 years, the Chinese government conducts the national health services survey, which is a special sample survey, in order to understand the urban and rural residents' health status, health care utilization, health care costs and burden on the country. In 2013, the fifth national health services survey adopted multi-stage stratified cluster random sampling. The sample covered 31 provinces, 156 counties, 780 towns and 1,560 villages. It collected data from a total of 93,600 families, which is ~30 million people and is a good representative sample of China. The data from CHFPSY and CSY were reported by the administrative agency that is highly credible and known for high quality work. Furthermore, in order to obtain a consistent statistical caliber, each analytical index in CHFPSY was as closely matched as possible with ARNHSS in China.

We selected 30 types of diseases from CHFPSY for this study to estimate their direct and indirect economic burden because the classification of diseases did not precisely match between CHFPSY and ARNHSS in China. Additionally, the CHFPSY classification of diseases was coded according to ICD-10. Therefore, it can be deduced that data integration was in process. Firstly, we listed the 30 types of diseases from CHFPSY 2014, and then we matched the name of the disease with ARNHSS in China 2013. If we could not find the exact same name of the disease, we classified the diseases into related systems according to the systems of diseases or main symptoms. If they still did not match using this method, the disease was classified as missing.

DEBD refers to the total direct costs for the prevention and treatment of diseases, including various costs incurred by the individual, family and society for preventing, diagnosing, treating and rehabilitating the diseases and injuries (20). On the one hand, it includes the costs of health services provided by the health institutions, such as the prevention of capital costs, emergency expenses, outpatient diagnosis and treatment costs, hospital expenses, medical expenses, health technology labor, family bed treatment and care costs, etc. On the other hand, other charges paid by the patients while receiving health services are included, such as nutrition fees, transportation fees, travel expenses, non-prescription costs for purchasing rehabilitation equipment, etc. We estimated the direct economic burden of 30 types of diseases, including outpatient medical expenses and inpatient medical expenses only because we don't have non-medical expenses. The formula is as follows:

DEBDi = N × [2-week consultation rate × 26 × (Σ outpatient facilities constituent ratio × average outpatient medical cost in outpatient institutions at different levels) + hospitalization rate × (Σ inpatient facilities constituent ratio × average hospitalization medical expense in inpatient institutions at different levels)] (1)

In the above formula, DEBDi is the direct economic burden for the class i disease, and N is the number of people in the research population.

IEBD refers to a society and family's loss between the current value and future potential value due to illness that may result in the reduction of effective working time and the ability to work (21). This is also known as indirect costs. Indirect costs represent the value given by society to health and life. While it broadly includes social productivity losses, losses of income, losses of housework, employment costs, training costs, insurance costs, management costs, etc., indirect costs indicate the loss of productivity in this narrow sense. We used the human capital approach (22) to estimate the indirect economic burden of 30 types of diseases and calculate the economic loss of missed working time due to hospitalization. The formula is as follows:

IEBDi = N × hospitalization rate × (Σ inpatient facilities constituent ratio × average hospitalization days in inpatient institutions at different levels) × GDP/365 (2)

In the above formula, IEBDi is the indirect economic burden for the class i disease, and N is the number of people in the research population.

We weighted the average outpatient medical costs that were not caused by disease according to the constituent ratio of first diagnosis mechanism for urban outpatients over 2 weeks. The proportions of average outpatient medical costs in tertiary, secondary and primary hospitals were 14.46, 18.18, and 67.36%, respectively. In 2013 for urban patients, there were 16.04% distributed in tertiary hospitals, 69.28% distributed in secondary hospitals, and 14.68% distributed in primary hospitals. This distribution was used to weigh the average medical costs of hospitalization. China had a population of 731.11 million urban residents in 2013. The result of the outpatient costs, hospitalization costs and DEBD in urban China are shown in Table 1, which is based on Formula (1). The largest DEBD in urban areas was myocardial infarction coronary artery bypass (MICAB), and the second largest was acute myocardial infarction (AMI), while the smallest DEBD was primary nephrotic syndrome (PNS).

Similar to the method used to process urban medical costs, we weighted the average rural outpatient medical costs. As a result, the proportions of average rural outpatient medical costs in tertiary, secondary and primary hospitals for rural outpatients were 2.53, 16.31, and 81.16% in 2013, respectively. In 2013, there were 13.11% of rural inpatients distributed in tertiary hospitals, 56.61% distributed in secondary hospitals, and 30.28% distributed in primary hospitals. China had a rural population of 629.61 million residents in 2013. The result of the outpatient costs, hospitalization costs and DEBD in rural China are shown in Table 2, which is based on Formula (1). The largest DEBD in rural areas was MICAB, and the second largest was acute upper gastrointestinal bleeding (AUGIB), while the smallest DEBD was PNS.

We weighted average hospitalization days by disease based on the constituent ratio for China's urban inpatients in 2013. The GDP per capita was $6,807 in China in 2013. The indirect economic losses of urban patients caused by hospitalization are shown in Table 3, which is based on Formula (2). The largest IEBD in urban areas was MICAB, which was followed by cerebral hemorrhage (CH), and the smallest IEBD was aplastic anemia (AA).

Similar to the method used for dealing with urban patients' data, we weighted average hospitalization days by disease to obtain rural patients' hospitalization days by disease. The indirect economic losses of rural patients due to hospitalization are demonstrated in Table 4, which is based on Formula (2). The largest IEBD in rural areas was CH, which was followed by MICAB, and the smallest IEBD was AA.

The total health expenditure of China in 2013 was $517.64 billion. The top five diseases among 30 types of diseases that had the heaviest TEBD based on the proportion in total health expenditure were MICAB (15.51%), AMI (5.67%), CH (5.11%), AUGIB (5.01%) and acute appendicitis (AAs) (4.45%). The last five diseases that had the smallest TEBD were AA (0.26%), PNS (0.26%), acute leukemia (AL) (0.41%), infiltrative pulmonary tuberculosis (IPT) (0.43%) and viral hepatitis (VH) (0.46%). The average TEBD for cities was $81.39 billion and for the villages was $50.26 billion, respectively. The total economic burden of 30 types of diseases in urban and rural areas in China is shown in Table 5.