METHODS

Sources of data

Data came from current Chinese language sources. The 1999 mortality data by age, sex, rural/urban location, and cause (injury and major disease groups) came from China’s National Center for Health Information and Statistics, Ministry of Public Health.³ This source is based on a geographically representative sample of 36 cities and 85 rural counties. The urban sample included large cities with more than one million non-agricultural population, as well as medium and small cities with less than 300 000 non-agricultural population. The rural sample consisted of counties at various levels of economic development.

Morbidity data came from the 1998 report of the Second National Health Services Survey, which was based on a representative sample of 210 000 individuals from 57 000 households in 1997.¹⁸ Self reported work days lost in the previous two weeks from cause specific acute morbidity were used in computing morbidity-attributed PPYLL.

Estimates of the 0–64 population were based on the latest (1996) population projection in the United Nations Demographic Yearbook.¹⁹ We expanded the final 1996 population by 3% to reflect the increase of population from 1996 to 1999.²⁰

Mortality data from the earlier years’ World Health Organization Statistics Annuals were used to calculate PPYLL by age, gender, rural/urban location, and cause in 1987 and 1992 to show time trends.^21,22

Direct costs of injury data came from the Second National Health Services Survey. Data for average annual earnings came from China Statistics Yearbook.²³

Analysis

The PPYLL method assumes that productive ages are from age 15–64. (The working age span can be changed if it is inappropriate for a given country.) For deaths before age 15, it is assumed that if the children had survived, they would have worked the full 50 years (15–64). The loss of potentially productive years of life for children is discounted back to the midpoint of their age group. PPYLL does not take into account the factor of competing causes of death.

Recognizing the economic principle that years in the future are worth less than years in the present, the PPYLL method applies a 3% discount rate, compounded, for years of life in the future lost to premature death. The World Bank and World Health Organization also use a 3% discount rate.²⁴ The method calculates future years lost from deaths in each 10 year age group from the midpoint of that period. Rather than giving a formula, an example is more easily understood. For example, deaths in the 15–24 year age group would result in an average of 45 lost years of productive work. Discounted by a compounded annual 3%, this would be about 25 years lost. Deaths in the 55–65 years age group would result in an average loss of 5 years (discounted = 4.6 years). Table 1 gives the discounted years of productivity loss by age groups. The total loss of productive years is the sum of all age groups. We expanded the years lost from the sample of deaths to years lost by the total Chinese population of 1999 by multiplying the age specific mortality rates of the sample by the total population of each 10 year age group from 0–64 years.

In addition to the years of life lost in the future from premature mortality, there are current morbidity losses. Losses of current productivity were based on the days of work lost per 1000 population from injury or disease applied to the total working-age population of China. To convert days lost to years lost, we divided the days lost by 260, the estimated work days per year. We calculated PPYLL separately for injury and four other disease groups.

To estimate the losses due to different types of injury (falls, drowning, etc), we used only premature mortality, because morbidity data by type of injury were not available.

To measure the economic impact of injury we used direct costs (physician fees, hospitalizations, medicine, etc) reported in the Second National Health Services Survey. We multiplied the average expenditure per ambulatory visit by the annual ambulatory visits to obtain the direct cost of ambulatory care by injury and disease groups. Direct costs of hospitalizations were calculated by multiplying the average expenditure per hospitalization by the annual number of hospitalizations.

Indirect costs were calculated by multiplying the average per capita earnings by the years of productivity lost from morbidity and premature mortality every year. According to China’s official statistics, rural average earning per capita was approximately 600 US dollars, and urban average earning per capita was 1000 US dollars in 1999.²³

RESULTS

Injury versus disease

Injuries caused an annual 12.6 million years of lost productivity, followed by respiratory diseases (10.2 million years), cardiovascular diseases (9.0 million years), cancer (8.3 million years), and infectious diseases (2.3 million years). The rates of PPYLL/100 000 are shown in fig 1 to permit comparison with other countries.

Of injury-attributed PPYLL, 80% was due to premature mortality, 20% due to morbidity. In contrast, 70% of respiratory disease-attributed PPYLL was due to morbidity (fig 1).

Type of injury

Analysis of PPYLL from premature death by type of injury showed that traffic related injuries were the major source of lost productivity due to injury (fig 2). In 1999, traffic related injury deaths resulted in 250 PPYLL/100 000 population, followed by suicide (230 PPYLL/100 000), drowning (130 PPYLL/100 000), poisoning (45 PPYLL/100 000), and falls (45 PPYLL/100 000). Together, these types of injuries accounted for 80% of all injury death-attributed PPYLL. Motor vehicle crashes caused 25% of total productive years lost from all injuries.

Time trends, rural/urban, and gender differences

Rates of PPYLL/100 000 population permitted comparisons of gender, temporal, and rural/urban differences. PPYLL/100 000 lost from motor vehicle crashes and other transport injury increased from 1987 to 1999 (fig 3). Rural populations had a greater increase (80%) as well as higher PPYLL rates initially. As expected, male rates were higher than female rates for essentially all ages and all causes of injury (fig 3).

Figure 4 displays divergent age specific death rate patterns, with drowning primarily affecting children and motor vehicle injury primarily affecting young adults.

Cost of injury

The total annual cost of injury among 0–64 age groups was calculated to be US$12.5 billion in 1999. Of the total cost, about US$3.5 billion was direct costs for ambulatory and hospitalization care and US$9 billion was indirect economic losses from morbidity and premature mortality.

DISCUSSION

The PPYLL technique, simple in concept, can be modified to suit different countries, and has the advantage of estimating productivity losses. In developing, industrializing countries such as China, it is essential to measure disease and injury losses for the productive working population. Decision makers with interests in increasing the nation’s productivity should find the information persuasive.

The most significant result of our analysis is that PPYLL from injury outweighed the losses due to respiratory diseases, cardiovascular diseases, cancer, or infectious diseases. The ranking is not universal, but may be typical of most countries with rapid economic development²⁵ as well as industrialized countries.

According to the World Health Organization’s estimation using the DALYs approach based on less complete mortality and morbidity data, injury caused 37 million years lost in China, cardiovascular diseases 23 million years lost, respiratory diseases 22 million years lost, cancer 18 million years lost, and infectious diseases 16 million years lost,¹⁶ which are generally consistent with our results.

To place the importance of the injury problem in perspective, a comparison is useful. The annual PPYLL from injury (12.6 million years) equals the loss of 12.6 million laborers in one year. This is more than the total number of new entrants to the labor force each year (9.7 million).²⁶ The cost of injury (US$12.5 billion) was almost four times the total government spending on all public health services (US$3.4 billion) in 1998.²⁰

One limitation of this study is that the mortality, morbidity, and direct costs are drawn from a large, presumably representative sample. But given the tremendous geographic and socioeconomic diversity in China, sampling bias may be inevitable. Due to the lack of detailed disability data by injury and disease groups, we were not able to calculate the PPYLL, direct and indirect costs caused by disability. Therefore, the productivity and economic loss of injury and other disease groups are somewhat under-estimated.

The findings have major implications for relevant government agencies and international organizations such as the World Bank and the World Health Organization. The opportunity for enhanced productivity should give priority for appropriate injury prevention programs. The divergent age patterns of different types of injury indicate that injury control programs should be age appropriate. For instance, drowning prevention should target children and traffic safety programs should emphasize adults.

The excessive PPYLL from motor vehicle crashes in rural areas will worsen the already lower rural health status and hinder economic development in rural areas. Further research is needed to identify risk factors for traffic related injuries among rural residents. It is unclear how much the hazardous road conditions and poorly maintained vehicles have contributed to the high PPYLL rates in rural areas. Farmers who emigrate to cities for seasonal jobs may also be exposed to heightened risk of injury when navigating the often chaotic, unfriendly traffic and unfamiliar road environment of the urban areas.

Research is also needed to develop cost effective intervention programs. Although the magnitude of the injury problem is increasingly recognized by health authorities in China, little research has been done there on injury prevention programs such as use of seatbelts by motor vehicle occupants, and safety helmets by bicyclists, motorcyclists, and moped riders. These programs have been implemented in many developed and some less developed countries and have proved effective.

China started a national poison control center in 2000 to provide technical support for treating poisoning cases. The newly established Chinese Center for Disease Control should consider establishing a center devoted to the control and prevention of injuries. One of the injury control and prevention center’s responsibilities should be to coordinate the health sector and other sectors in formulating and implementing strategic plans for reducing the tremendous burden of injury.