Burden of disease assessment and health system reform: Results of a study in Mexico

of 9

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
9 pages
0 downs
This paper describes in broad terms the first national assessment of the burden of disease using Disability Adjusted Life Years, which formed part of an overall study of health system reform in Mexico. The overall project, ‘Health and the Economy:
  Journal of International Development: Vol 7, No. 3 555-563 1995) BURDEN OF DISEASE ASSESSMENT AND HEALTH SYSTEM REFORM: RESULTS OF A STUDY IN MEXICO R. LOZANO Burden of Disease Unit Harvard Center for Population and Development Studies and National Institute of Public Health Cuernavaca Mexico C. J. L. MURRAY Burden of Disease Unit Harvard Center for Population and Development Studies J. FRENK Fundacidn Mexicana para la Salud Mexico City J.-L. BOBADILLA Population Health and Nutrition Dept. World Bank Washington D. C. Abstract: This paper describes in broad terms the first national assessment of the burden of disease using Disability Adjusted Life Years, which formed part of an overall study of health system reform in Mexico. The overall project, ‘Health and the Economy: proposals for improving the Health System in Mexico’, is one of the first experiences of applying the analytical tools featured in the World Development Reporl 1993: Investing in Health. The Mexican National Burden of Disease (NBD) study was a critical component of the analysis of health system reform in Mexico. It provided a strong analytical basis for the identification of epidemiological priorities. The results of the study document the heterogeneous nature of the epidemiological transition in different parts of Mexico, and included both premature mortality and disability in the analysis. The results of this study were employed along with information on the cost- effectiveness of major health interventions in Mexico to design a package of essential health services for all Mexicans. Not only was the NBD study useful in identifying health needs and interventions that may have a major health impact, but the exercise itself had a number of other benefits. Through the careful review of all available sources of information on the incidence, prevalence, duration and severity of disability and mortality by age, sex and cause, the strengths and gaps in the current information system were identified. The process by which a large number of Mexican experts were involved in the estimation of the burden of each disease encouraged broad partici- pation of health experts in the discussions about health reform in Mexico. CCC 0954-1748/95/030555-09 1995 by John Wiley & Sons, Ltd.  556 R Lozano et al. INTRODUCTION In recent years, important achievements in global health have brought about notable improvements in the conditions of human life. The decline in mortality rates and the consequent rise in life expectancies is indicative of these achievements World Bank, 1993). Although Mexico certainly has made important progress in this regard, there is still a considerable gap between the levels of mortality and disability achieved in the highest income countries and Mexico on average, and within Mexico there are considerable differences between the best and worst off Bobadilla et al. 1993). In 1993, a major analysis of options for health reform was undertaken in Mexico by the Mexican Health Foundation FUNSALUD), a private non profit organi- zation. The objective of this study was to generate an analytical basis for the ongoing debate on health reform. As the first step in the project, a National Burden of Disease study was initiated as a collaborative effort between FUNSALUD, the National Institute of Public Health, the General Directorate of Statistics of the Ministry of Health, the School of Medicine of the National University of Mexico and the Harvard Center for Population and Development Studies. This study was to be the first national application of the newly developed approach to assessing the Global Burden of Disease outlined in the World Development Report 1993: Investing in Health World Bank, 1993). The objective of the NBD study was to define the main causes of health problems in Mexico. Using Disability- Adjusted Life Years DALYs) as the unit of measure would facilitate the simultaneous use of information on cost-effectiveness generated in terms of cost per DALY with the results of the NBD study. The results of the NBD study and the cost-effectiveness analysis were used to select the services to be included in a proposed package of essential health services Frenk et al. 1994). There were four clear advantages for Mexico in applying the Global Burden of Disease methodology to evaluate health priorities. First, unlike traditional health measures such as the infant mortality rate or life expectancy, DALYs include both years of life lost due to premature mortality YLL) and years of life lived with a disability YLD), thus capturing more fully the importance of non-fatal health outcomes. Second, by using a standard method which emphasizes careful review of all data sources and checks of the internal consistency of estimates of different epidemiological indicators such as incidence, prevalence and relative risk, the best estimates of the burden of each cause could be generated. Third, by using time as the common unit of measure ime lost due to premature mortality and time lived with a disability-a uniform metric could be applied across a complex array of different conditions and risk factors. Fourth, the results of the Global Burden of Disease study for five regions divided in rural and urban areas), five age-groups, two sexes and more than 100 causes were available so that detailed comparisons could be carried out. METHODS The study was divided into two major components: the causes of death analysis and the estimation of incidence, prevalence, duration and severity of major dis-  Burden of Disease Assessment and Health System Reform 557 abling sequelae of the conditions included in the study. While the general approach used was consistent with the methods of the Global Burden of Disease GBD) study Murray, 1994; Murray and Lopez, 1994; Murray et al. 1994) several im- portant modifications were made to make the study more applicable to Mexico. The detailed study methodology is provided elsewhere; here a brief synopsis is Frenk et al. 1994, Lozano et al. 1994). The list of causes to be included in the Mexican NBD study was developed based on the diseases expected to be important causes of burden, diseases for which certain interventions were in widespread use, and the existing list of causes for the GBD study. The final list included 108 causes of which 98 were in the GBD list. It was decided to undertake the analysis of the burden of disease separately for each of the 32 states. Within each state, urban and rural’ areas were analysed separately. In total the study comprised 64 geographical units. Using various demographic techniques and survey results, vital registration data were corrected for under-reporting Gomez de Leon and Virgilio Partida, 1992). This information was obtained from the National Institute of Sta- tistics, Geography and Informatics INEGI, 1991). A detailed analysis of Mexican cause of death coding practices was under- taken including examination of all ICD 9 WHO, 1977) 4 digit categories. Local ‘garbage codes’ such as ICD E928.9 and E980-E989 were identified; based on expert consultation, these deaths were redistributed to substantive causes in the list2. Deaths related to code E928 were redistributed across the other unintentional injury categories. Deaths related to code E980-E989 were redistributed by allocating 50 per cent to specific causes within uninten- tional, and 50 per cent to specific causes within intentional, injuries. The modifications adopted in Mexico in order to allocate the majority of deaths to specific causes enabled classification of 91 per cent of total deaths. Years of life lost due to deaths from each cause in each age-sex group were calculated using the standard assumptions in DALYs. Extensive discussions were held on the value judgments incorporated in DALYs, but ultimately the decision was made to keep assumptions the same. For the estimation of years of life lived with a disability, estimates for each region of the incidence, prevalence, duration and severity of disability were required. To facilitate this exercise, a network of Mexican experts on each disease was identified3. Meetings in small groups and one on one were held to explain the methodology and objectives of the study. The study team and the network of experts identified all available sources of information on incidence, prevalence, duration and severity including: the National Health Surveys DGE various dates), the registers of the National System of Epidemiological Surveillance of the Ministry of Health and of the Mexican Institute of Social Security; the hospital registers of the National Rural areas correspond to locations with less than 15000 inhabitants. ICD E928.9 corresponds to ‘Other and unspecified environmental and accidental causes’ and E980E989 to ‘Injury undetermined whether accidentally or purposely inflicted‘. In 1991 there were 8 130 deaths related to the first one and 3 084 to the second one; 3 per cent of the total deaths. The national experts consulted in this exercise were: 26 clinicians and 28 epidemiologists.  558 R Lozano et al. Health System; and multiple small scale epidemiological studies performed in Mexico4. viii) The available results and expert judgements were then analysed using a computer model to search for internal consistency in the estimates of inci- dence, remission, relative risk of death and prevalence as described in Murray and Lopez 1994) and Murray et al. 1994). Inconsistencies were resolved through a process of consultation with the network of Mexican experts and where necessary, international experts. In the cases in which it was impossible to gather information, even through the above means, we drew on the aggregate estimates by the World Bank and WHO for Latin America. However, it is important to note that the outputs were always confirmed with experts before being published. ix) About 85 per cent of the YLD are derived from estimates supported by Mexican data. These data were collated separately by urban and rural areas but were not disaggregated into 64 geographical units. Years of life lived with a disability were calculated by applying national assumptions to each geographical unit. The regional variation in DALYs came principally from mortality data and to a minor extent from differences in incidence and case fatality. x) The results for YLL and YLD and the resulting estimates of DALYs for each of the 64 regions in the study were then subject to an international review process. Final revisions were based on the reviewers’ comments. RESULTS It is not the purpose of this brief paper to present the detailed results of the Mexican NBD study. Rather we provide a very general overview to indicate how the main findings were relevant to the more general task of designing health reform. During 1991, Mexico lost around 12.8 million DALYs, of which 58 per cent were due to premature death and 42 per cent to disability. However, this overall relationship changes with the level of urbanization see Table 1). In rural areas, YLL account for 62 per cent of DALYs, while in urban areas, they account for 55 per cent. In terms of rates per loo0 population, rural areas lost 182 DALYs while urban areas lost 128. The rural rate is 42 per cent higher than the urban rate: the majority of this difference is attributable to years of life lost as opposed to years of life lived with a disability. The principal disparity between rural areas and urban areas is in communicable diseases in the under-five population. Table 1 presents an international comparison between Mexico and the world regions. Based on DALYs per lo00 population, the Mexican rural area is located between Latin America and China, and the Mexican urban area between the Former Socialist Economies and the Established Market Economies. The burden of disease is borne predominantly by males, and losses due to premature mortality are greater in men than in women. Thirty-three per cent of the burden of disease affects children under five years; the majority of it is attribu- table to communicable diseases and premature death. Thirty-two per cent of the The list of studies consulted in this study appears in Frenk et al. 1994) and Lozano et al. 1994).  Burden of Disease Assessment and Health System Reform 559 Table 1. DALY s lost: international comparisons between Mexico and world regions. Sources: Murray and Lopez (1994); Lozano et al. (1994). Region DALYs Lost per 1,000 Prem. Deaths ( ) Disability ( ) Sub-Saharan Africa India Middle Eastern crescent Other Asia and Islands Latin America and Caribbean Rural Mexico China Former Socialist Economies Mexico Urban Mexico Established market economies 574 344 287 259 232 182 178 168 151 123 117 77 68 68 64 56 62 57 59 58 55 53 23 32 32 36 44 38 43 41 42 45 47 Table 2. DALYs Lost in Mexico by Causes and Region 1991, per 1000 inhabitants. (Source: Lozano et al. 1994).) Causes National Rural Urban R/U Group I 47.7 69.4 31.5 2.2 Group I 71.1 78.3 65.8 1.2 Group 111 32.4 34.4 31.0 1.1 Group I Communicable, Maternal, Perinatal and Nutritional Diseases; Group I Non Communicable di- seases and Group I1 Injuries. national burden of disease occurs in the young adult population (15-44 years), 67 per cent in males and the remainder in females. For males, more than half of the losses are due to injuries, divided evenly between death and disability. Twenty- seven per cent of the burden of disease occurs in the population over 45 years of age, among whom 83 per cent of the DALY loss is attributable to non- communicable diseases. In those age groups, years lived with a disability represents 4 per cent of the DALYs lost in the population over 45 years. The burden of disease in Mexico exhibits a slight predominance of non- communicable diseases over the other two major groups of causes see Table 2). However, in considering the disaggregated causes of disease, the most important are unintentional injuries, in particular motor vehicle accidents (including driver, passenger and pedestrian injuries). At the national level, the number of DALYs lost due to motor vehicle accidents were twice as high as for cirrhosis of the liver and diarrhoea1 diseases, and three times as high as for pneumonia, ischaemic heart disease and diabetes mellitus. Neuropsychiatric diseases in adults and helminths in schoolchildren, associated mainly with disability, are among the significant health needs identified through this exercise, that would not otherwise have been recog- nized. The burden of disease in Mexico does not follow one characteristic profile that may be applied to the country as a whole; rather, Mexico is best seen as a collection of different regions. In fact, this regional heterogeneity is of principal importance
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks