- Research article
- Open Access
- Open Peer Review
Quantifying the burden of disease due to premature mortality in Hong Kong using standard expected years of life lost
© Plass et al.; licensee BioMed Central Ltd. 2013
- Received: 5 February 2013
- Accepted: 16 September 2013
- Published: 18 September 2013
To complement available information on mortality in a population Standard Expected Years of Life Lost (SEYLL), an indicator of premature mortality, is increasingly used to calculate the mortality-associated disease burden. SEYLL consider the age at death and therefore allow a more accurate view on mortality patterns as compared to routinely used measures (e.g. death counts). This study provides a comprehensive assessment of disease and injury SEYLL for Hong Kong in 2010.
To estimate the SEYLL, life-expectancy at birth was set according to the 2004 Global Burden of Disease study at 82.5 and 80 years for females and males, respectively. Cause of death data for 2010 were corrected for misclassification of cardiovascular and cancer causes. In addition to the baseline estimates, scenario analyses were performed using alternative assumptions on life-expectancy (Hong Kong standard life-expectancy), time-discounting and age-weighting. To estimate a trend of premature mortality a time-series analysis from 2001 to 2010 was conducted.
In 2010 524,706.5 years were lost due to premature death in Hong Kong with 58.3% of the SEYLL attributable to male deaths. The three overall leading single causes of SEYLL were “trachea, bronchus and lung cancers”, “ischaemic heart disease” and “lower respiratory infections” together accounting for about 29% of the overall SEYLL. Further, self-inflicted injuries (5.6%; ranked 5) and liver cancer (4.9%; ranked 7) were identified as important causes not adequately captured by classical mortality measures. Scenario analyses highlighted that by using a 3% time-discount rate and non-uniform age-weights the SEYLL dropped by 51.6%. Using Hong Kong’s standard life-expectancy values resulted in an overall increase of SEYLL by 10.8% as compared to the baseline SEYLL. Time-series analysis indicates an overall increase of SEYLL by 6.4%. In particular, group I (communicable, maternal, perinatal and nutritional) conditions showed highest increases with SEYLL-rates per 100,000 in 2010 being 1.4 times higher than 2001.
The study stresses the mortality impact of diseases and injuries that occur in earlier stages of life and thus presents the SEYLL measure as a more sensitive indicator compared to classical mortality indicators. SEYLL provide useful additional information and supplement available death statistics.
- Burden of disease
- Hong Kong
- Standard expected years of life lost
All over the world, and not only since the impact of the global financial crisis in 2008, resources in the health care sector are scarce [1, 2]. Epidemiological indicators such as mortality patterns and life-expectancy values, derived from historically observed mortality trends, have been used extensively to describe and quantify health improvements on population level and to set priorities for resource allocation [3, 4]. Classical mortality indicators are strongly influenced by diseases where death occurs at older ages and thus do not represent the complete mortality-associated disease burden. Indicators that take into account age at death allow a more comprehensive assessment of mortality because they not only count the number of deaths but also measure the years of life lost due to premature mortality [5–7]. Health indicators using lost time as an outcome are increasingly used in public health at global, national and local levels [8–12]. Many of these measures combine complementary epidemiologic information and present the health of a population as a single numerical indicator. Such health indicators are categorised as summary measures of population health . The disability-adjusted life year (DALY) has frequently been used in population health assessments . The DALY combines the impact of mortality and morbidity and uses time as the unit to measure health losses in populations . It is presented as the sum of years lived with disability (YLD, morbidity component) and years of life lost due to premature death (YLL, mortality component). Owing to often poor or even unavailable detailed information on non-fatal outcomes , several investigations have focused on the calculation of single DALY components [5–7, 17–21]. Most of these studies used data from vital statistics where the quality and availability of data have improved in the past to calculate years lost due to premature death. One commonly used measure of premature death is the standard expected years of life lost (SEYLL) indicator. The SEYLL, as a standardised measure, can be used to quantify and compare the impact of different diseases and injuries. SEYLL can supplement the traditional set of mortality indicators and enhance the monitoring of disease patterns in populations.
Currently, no comprehensive assessment of premature mortality effects is available for the Special Administrative Region of Hong Kong. Therefore, the major objective of this study is to quantify the SEYLL due to premature mortality in Hong Kong for the year 2010. Secondary objectives are to estimate the effect of different social value choices and to present the changes of mortality associated disease burden from 2001 to 2010.
Standard expected years of life lost
where N is the number of deaths, r is the discount rate (0.03), C is the age-weighting correction constant (0.1658), e is a constant (≈ 2.718), β is the parameter from the age-weighting function (0.04), a is the age at death and L the remaining life-expectancy at age of death .
The diseases and injuries are classified according to the GBD classification system and ordered by four levels of disaggregation . At the first level, conditions are presented in three broad groups, namely communicable, maternal, perinatal and nutritional conditions (group I), non-communicable conditions (group II) and injuries (group III). For detailed analyses, these broad groups are differentiated into 21 condition groupings and then further into more than 100 single disease entities. A normative health goal was set, defining life-expectancy according to a World Health Organisation (WHO) standard at 82.5 and 80 years for women and men, respectively . SEYLL-rates per 100,000 population are calculated using the age and sex stratified mid-year population for Hong Kong in 2010 . SEYLL results are reported with uniform age-weighting and no time-discount [25–27]. Sensitivity analyses were performed using local standard life-expectancy values for the Hong Kong population and including different age-weighting and time-discounting scenarios.
Population data for 2010 were obtained from the HK Government Department with a mid-year population of 7.07 million inhabitants .
Death counts for 2010 were extracted from a dataset provided by the Hong Kong Census and Statistics Department . This dataset contained information about cause of death, age and sex of each deceased. The causes of death were classified according to the tenth revision of the International Classification of Diseases (ICD-10). Deaths classified as S00 to T98 (Injury, poisoning and certain other consequences of external causes) were separately derived from external causes of death.
The causes of death data were translated to the GBD classification system using the recommended algorithms . In addition, ill-defined causes (R00-R99), cancer (C76, C80, C97) and cardiovascular (I47.2, I49.0, I46, I50, I51.4, I51.5, I51.6, I51.9, I70.9) garbage-codes were redistributed according to correction methods provided by the GBD study [9, 31]. A time series analysis of death registry data from 2001 to 2010 was performed to estimate the past trends of SEYLL in Hong Kong.
The calculations were performed using Microsoft Office Excel version 12 and R Statistical Software version 2.15.1.
SEYLL by main broad cause groups
Standard expected years of life lost (SEYLL) by broad cause groups, Hong Kong, 2010
Rate per 100,000*
Rate per 100,000
Rate per 100,000
The female population showed a higher share for group II conditions, with 80% as compared to the male population (78%). Group I conditions accounted for 12.7% of the SEYLLs for both males and females. Men showed a higher percentage attributable to injuries at 9.3% in comparison to 7.3% for women.
Group II conditions presented the lowest observed SEYLL shares in the age-group 25–29 years, with 37.4%. The SEYLL due to group II conditions appeared to increase with age, reaching a peak at 87.5% (age-group 60–64).
For group III conditions, a constant rise of the SEYLL share is observed starting in age-group 0, reaching a peak of 57.5% in age-group 25–29 years, dropping rapidly at 45–49 years (10.6%), and then flattening out with increasing age.
SEYLL by condition groups
Standard expected years of life lost (SEYLL) by condition groupings, Hong Kong, 2010
Rate per 100,000*
Rate per 100,000
Rate per 100,000
Infectious and parasitic diseases
The SEYLL-rates for the male population were generally higher for all conditions (Figure 3), especially for ischaemic heart disease, trachea, bronchus and lung cancers, and liver cancer, with rates for males being 2.1, 2.2 and 3.5 times higher than the female rates, respectively. An exception of course was breast cancer, a sex-specific malignant neoplasm mainly occurring in the female population.
The overall leading causes include many malignant neoplasms, which constituted five out of ten leading causes. It is noteworthy that the group III condition self-inflicted injuries strongly affected both sexes and was ranked 6th and 7th in the male and female leading causes, respectively.
The majority of single leading causes (except self-inflicted injuries) showed high concentrations of SEYLL in the later stages of life.
Time discounting and age-weighting effects
Major changes in leading causes of premature death were observed for scenario three (0,1). Considering higher age-weights for the productive age groups, conditions occurring in earlier stages of life, and in particular self-inflicted injuries, received an upturn in priority ranking from 5th to 3rd position, even exceeding lower respiratory infections (data not shown).
Local Hong Kong life-expectancy
Sensitivity analyses using local Hong Kong standard life-expectancy indicated considerable changes of SEYLL, especially for the female population (Additional file 2). Using the Hong Kong standard, female remaining life-expectancy at birth was 3.5 years longer than the WHO standard. The remaining Hong Kong life-expectancy for men was nearly identical at birth but slightly higher with increasing age. Altering the life-expectancy resulted in an increase of total SEYLL of 10.8% (15.1% for female; 7.4% for male). In particular, SEYLL due to group I conditions for the female population increased by 18.6%.
SEYLL trends over time
SEYLL rates per 100,000* stratified by sex and broad cause groups (2001–2010)
The results of this study represent the first comprehensive quantification of SEYLL for the population of Hong Kong. In general, Hong Kong shows a burden of disease profile that is comparable with the patterns of other regions with a high income level. Malignant neoplasms and cardiovascular diseases were identified to contribute 60.8% to the overall SEYLL of the Hong Kong population, which is consistent with studies in other high-income countries (e.g. Spain: 57.5%; Germany (North Rhine-Westphalia): 69.4%) [5, 7, 18, 21].
Overall, SEYLL-rates for men exceeded those for women for virtually all of the ten leading causes except breast cancer, which is a sex-specific cancer with major impact on the female population. Except for injury conditions, most SEYLL were located in age-groups above 30 years, which is in line with studies from other high-income countries [7, 21].
However, most of the studies identified ischaemic heart diseases as the leading single cause [5, 21]. For Hong Kong, trachea, bronchus and lung cancers were the leading cause of SEYLL, which is probably associated with the still very high prevalence of smoking in Hong Kong. Although the prevalence of daily smokers had decreased from 39.7% in 1983 to 22% in 2000 for men and from 5.6% to 3.5% for women , the effects of past smoking habits still seem to be obvious. A recent household survey showed that the overall smoking prevalence (2009–2010) remained high, at 20.8% for men and 3.7% for women . Despite the fact that smoking has been banned from many public places since 2007 , it is still considered a threat to the health of the population in Hong Kong. Non-smoking campaigns aiming at preventing smoking at younger ages (15–19 years), where smoking prevalence is currently low (1.8%), are important but more worrying is the much higher smoking prevalence in older age-groups, particularly in the group of 30–39 (15.6%) year-old males . This observation is alarming and calls for further targeted public health campaigns to reduce the future smoking-related disease burden and health care expenditures.
At 12.7%, Hong Kong showed a considerably higher proportion of group I conditions than recent studies e.g. from Spain (6.4%)  or Germany (5.6%) . In contrast to other high-income countries, lower respiratory infections (e.g. influenza, pneumonia) were the leading cause of SEYLL among women (e.g. ranked 9th in Spain ). Infectious diseases still appear to be of importance for the Hong Kong population and should be further tackled by preventive measures (e.g. promotion of influenza vaccination in the elderly population).
At 8.5% (6.9%, scenario 3,0), group III conditions showed a higher share of the total SEYLL as compared to the study from Germany (5.3%)  but was considerably lower than for Spain (12.9%) . However, the injury patterns differed greatly as 67.3% of injury SEYLL in Hong Kong were due to intentional injury deaths. These percentages were considerably lower in the Spanish (30%) and German (35.6%) studies [5, 7]. In particular, self-inflicted injuries among men aged 25–29 years alone accounted for 50.5% of the total SEYLL in this age group in Hong Kong. Studies have indicated that during this period of early adulthood both men and women are highly vulnerable to depressive symptoms due to interpersonal crisis and also have a relatively high risk of attempting or committing suicide .
The priorities set up by SEYLL indicated that self-inflicted injuries were ranked 6th and 7th for males and females, respectively. If the rankings were based solely on death counts, self-inflicted injuries for 2010 would not belong to the ten leading causes for women (ranked 17th) and would only be ranked 9th for men. Using SEYLL highlights the importance of self-inflicted injuries because deaths occur at an earlier stage of life and thus cause a considerable number of life years lost.
Malignant neoplasms also played an important role because this cause group accounted for about 39.1% of the total SEYLL disease burden. Trachea, bronchus and lung cancers were the leading single disease entity. However, priorities set by classical death statistics rank trachea, bronchus and lung cancers only in 3rd place. When not considering age at death, priorities are shifted towards lower respiratory infections (ranked 1st) and ischaemic heart diseases (ranked 2nd), conditions with major impact on the elderly population. The use of SEYLL-rates compared to standard death counts drew greater attention not only to breast cancer for women (whose rank changed from 8th to 5th) but also to liver cancer for men (whose rank changed from 7th to 4th).
As shown in the scenario analyses, the choice of social values (time-discounting and age-weighting) can have a large impact on the resulting SEYLL, with reductions of up to half of the total SEYLL. Previous studies presented SEYLL with a 3% time-discounting and uniform age-weighting and thus especially depleted the health losses of conditions occurring in earlier stages of life [5, 7]. In particular, SEYLL due to group III conditions, including self-inflicted injuries, were reduced by 39.8% when a 3% time-discounting was applied.
The effects of replacing the WHO’s standards with the local Hong Kong life-expectancy values mainly impacted on the SEYLL-rates of women and the biggest overall changes were observed for older age-groups because most of the deaths are condensed in this life-span. For the sex-specific leading causes of SEYLL, however, the rankings were almost identical (Additional file 3).
The time-series analyses highlighted the trend of increasing disease burden, especially for group I conditions. Lower respiratory infections were one of the major drivers of SEYLL in Hong Kong and intervention measures should therefore focus on target groups of the population aged 55 and older.
The present study is founded solely on mortality data, which may be deficient in capturing the full impact of diseases and injuries. Including the morbidity component in further assessments is important because, particularly in high-income countries, the share of chronic diseases is increasing dramatically, as indicated by the GBD study results for the years 2004 and 2010 [12, 36]. Further, possible adjustment of the GBD classification system towards the needs of the disease patterns observed in Hong Kong may lead to more detailed results as e.g. the category of “other malignant neoplasms”, which could further be disaggregated, has high numbers of SEYLL and may mask single important entities.
This study has demonstrated that SEYLL is a valuable measure to quantify the impact of premature deaths on population health because it takes into account the potentially remaining life-expectancy at age of death [18, 21]. Another frequently applied measure of premature mortality, the potential years of life lost (PYLL) (used e.g. by the Organisation for Economic Co-operation and Development), use potential limits to life (e.g. 75 years). This method does not take into account lost years due to deaths occurring at ages above this limit [15, 37–39]. Thus, interventions aiming at avoiding deaths above 75 years would result in no benefit. The SEYLL measure uses remaining life-expectancy values and takes into account deaths in the elderly population and thus is more sensitive to deaths occurring in this life-span . Taking into account the age of death and the remaining life-expectancy, the SEYLL offers a more appropriate measure of premature death.
Using the SEYLL as a measure of premature mortality allows considering the age at death and thus represents a highly qualified measure of mortality as compared to traditionally used mortality indicators such as simple death counts. Especially conditions that occur in earlier stages of life are captured more adequately which results in significant changes in priorities. In our study we identified “trachea, bronchus and lung cancers” as the major cause of SEYLL in Hong Kong in the year 2010 probably associated with still high smoking rates in adults. As SEYLL only represents one component of the overall disease burden as measured by the DALY further analyses should aim at estimating the morbidity component (YLD) to allow a more comprehensive estimation of the disease burden for Hong Kong. Including the morbidity perspective would have a high impact on disease patterns and may lead to changes in priority setting.
This study was supported by the Hong Kong Research Grants Council (RGC) and the German Academic Exchange Service (DAAD) under the Germany/Hong Kong Joint Research Scheme (HK Grant no: G_HK032/10; GER Project-ID 50751245).
We acknowledge support of the publication fee by Deutsche Forschungsgemeinschaft and the Open Access Publication Funds of Bielefeld University.
- Horton R: The global financial crisis: an acute threat to health. Lancet. 2009, 373: 355-356. 10.1016/S0140-6736(09)60116-1.View ArticlePubMedGoogle Scholar
- Marmot MG, Bell R: How will the financial crisis affect health?. BMJ. 2009, 338: 858-860. 10.1136/bmj.b858.View ArticleGoogle Scholar
- Chuang Y-C, Chuang K-Y, Chen Y-R, Shi B-W, Yang T-H: Welfare state regimes, infant mortality and life expectancy: integrating evidence from east Asia. JECH. 2012, 66: e23-Google Scholar
- Cheung KS, Yip PS: Trends in healthy life expectancy in Hong Kong SAR 1996–2008. Eur J Ageing. 2010, 7: 257-269. 10.1007/s10433-010-0171-3.View ArticlePubMedPubMed CentralGoogle Scholar
- Penner D, Pinheiro P, Krämer A: Measuring the burden of disease due to premature mortality using standard expected years of life lost (SEYLL) in North Rhine-Westphalia, a federal state of Germany, in 2005. JPH. 2010, 18: 319-325.Google Scholar
- Marshall RJ: Standard expected years of life lost as a measure of mortality: norms and reference to New Zealand data. Aust N Z J Public Health. 2004, 28: 452-457.View ArticlePubMedGoogle Scholar
- Genova-Maleras R, Catala-Lopez F, de Larrea-Baz N, Alvarez-Martin E, Morant-Ginestar C: The burden of premature mortality in Spain using standard expected years of life lost: a population-based study. BMC Public Health. 2011, 11: 787-10.1186/1471-2458-11-787.View ArticlePubMedPubMed CentralGoogle Scholar
- Michaud C, McKenna M, Begg S, Tomijima N, Majmudar M, Bulzacchelli M, Ebrahim S, Ezzati M, Salomon J, Gaber Kreiser J, et al: The burden of disease and injury in the United States 1996. Popul Health Metr. 2006, 4: 11-10.1186/1478-7954-4-11.View ArticlePubMedPubMed CentralGoogle Scholar
- Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL: Global and regional burden of disease and risk factors, 2001: systematic analysis of population health data. Lancet. 2006, 367: 1747-1757. 10.1016/S0140-6736(06)68770-9.View ArticlePubMedGoogle Scholar
- Dodhia H, Phillips K: Measuring burden of disease in two inner London boroughs using disability adjusted life years. J Public Health. 2008, 30 (3): 313-321. 10.1093/pubmed/fdn015. fdn015View ArticleGoogle Scholar
- Zhou SC, Cai L, Wang J, Cui SG, Chai Y, Liu B, Wan CH: Measuring the burden of disease using disability-adjusted life years in Shilin county of Yunnan province, China. Environ Health Prev Med. 2011, 16: 148-154. 10.1007/s12199-010-0176-8.View ArticlePubMedGoogle Scholar
- WHO: The global burden of disease: 2004 update. 2008, Geneva: World Health OrganizationGoogle Scholar
- Field MJ, Gold MR: Summarizing population health directions for the development and application of population metrics. Summarizing population health directions for the development and application of population metrics. 1998, Washington: National Academy Press, 85-Google Scholar
- Murray CJL, Lopez AD, World Health Organization, World Bank, Harvard School of Public Health: The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020: summary. 1996, Geneva: World Health OrganizationGoogle Scholar
- Murray CJL: Quantifying the burden of disease: the technical basis for disability-adjusted life years. Bull World Health Organ. 1994, 72: 429-445.PubMedPubMed CentralGoogle Scholar
- Boerma JT, Stansfield SK: Health statistics now: are we making the right investments?. Lancet. 2007, 369: 779-786. 10.1016/S0140-6736(07)60364-X.View ArticlePubMedGoogle Scholar
- Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB: Years of life lost due to obesity. JAMA. 2003, 289: 187-193. 10.1001/jama.289.2.187.View ArticlePubMedGoogle Scholar
- Mariotti S, D’Errigo P, Mastroeni S, Freeman K: Years of life lost due to premature mortality in Italy. Eur J Epidemiol. 2003, 18: 513-521.View ArticlePubMedGoogle Scholar
- Allard YE, Wilkins R, Berthelot JM: Premature mortality in health regions with high aboriginal populations. Health Rep. 2004, 15: 51-60.PubMedGoogle Scholar
- Burnet NG, Jefferies SJ, Benson RJ, Hunt DP, Treasure FP: Years of life lost (YLL) from cancer is an important measure of population burden and should be considered when allocating research funds. Br J Cancer. 2005, 92: 241-245.PubMedPubMed CentralGoogle Scholar
- Vlajinac H, Marinkovic J, Kocev N, Sipetic S, Bjegovic V, Jankovic S, Stanisavljevic D, Markovic-Denic L, Maksimovic J: Years of life lost due to premature death in Serbia (excluding Kosovo and Metohia). Public Health. 2008, 122: 277-284. 10.1016/j.puhe.2007.06.010.View ArticlePubMedGoogle Scholar
- Lopez AD: Global and regional burden of disease and risk factors, 2001: systematic analysis of population health data. Lancet. 2006, 367: 1747-10.1016/S0140-6736(06)68770-9.View ArticlePubMedGoogle Scholar
- Mathers CD, Vos T, Salomon J, Ezzati M: National burden of disease studies: a practical guide. 2001, Geneva: World Health OrganizationGoogle Scholar
- Hong Kong Census and Statistics Department: Hong Kong annual digest of statistics. Hong Kong annual digest of statistics. 2011, Hong Kong: , 2011Google Scholar
- Anand S, Hanson K: DALYs: efficiency versus equity. World Dev. 1998, 26: 307-310. 10.1016/S0305-750X(97)10019-5.View ArticleGoogle Scholar
- Anand S, Hanson K: Disability-adjusted life years: a critical review. J Health Econ. 1997, 16: 685-702. 10.1016/S0167-6296(97)00005-2.View ArticlePubMedGoogle Scholar
- Arnesen T, Kapiriri L: Can the value choices in DALYs influence global priority-setting?. Health Policy. 2004, 70: 137-149. 10.1016/j.healthpol.2003.08.004.View ArticlePubMedGoogle Scholar
- HK SAR Government Department: Hong Kong yearbook 2010. 2010, Hong Kong: the Goverment Logistics DepartmentGoogle Scholar
- Hong Kong Census and Statistics Department: Hong Kong SAR: 2001–2010 known death microdata sets, 21 december 2011 edition. Hong Kong SAR: 2001–2010 known death microdata sets. 2011, Hong KongGoogle Scholar
- Lopez AD, Disease Control Priorities Project: Global burden of disease and risk factors. 2006, New York, NY, Washington, DC: Oxford University Press, World BankView ArticleGoogle Scholar
- Lozano R, Murray CJL, Lopez AD, Satoh T, World Health Organization: Miscoding and misclassification of ischaemic heart disease mortality. 2001, Geneva: World Health OrganizationGoogle Scholar
- Au JSK, Mang OWK, Foo W, Law SCK: Time trends of lung cancer incidence by histologic types and smoking prevalence in Hong Kong 1983–2000. Lung Cancer. 2004, 45: 143-152. 10.1016/j.lungcan.2004.01.012.View ArticlePubMedGoogle Scholar
- HK SAR Census and Statistics Department: Thematic houshold survey report No. 48. Thematic houshold survey report No. 48. 2011, Hong Kong: Social Surveys Section Census and Statistics DepartmentGoogle Scholar
- Smoking (public health) ordinance. http://www.legislation.gov.hk/blis_pdf.nsf/6799165D2FEE3FA94825755E0033E532/FC20BFF93D75ADCC482575EE00764F46/$FILE/CAP_371_e_b5.pdf,
- Shah A: The relationship between suicide rates and age: an analysis of multinational data from the world health organization. Int Psychogeriatr. 2007, 19: 1141-1152.PubMedGoogle Scholar
- Murray CJL, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud C, Ezzati M, Shibuya K, Salomon JA, Abdalla S, et al: Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the global burden of disease study 2010. Lancet. 2012, 380: 2197-2223. 10.1016/S0140-6736(12)61689-4.View ArticlePubMedGoogle Scholar
- OECD: Health at a glance 2011: OECD indicators. 2011, OECD Publishing, http://dx.doi.org/10.1787/health_glance-2011-en,Google Scholar
- Werber D, Hille K, Frank C, Dehnert M, Altmann D, Muller-Nordhorn J, Koch J, Stark K: Years of potential life lost for six major enteric pathogens, Germany, 2004–2008. Epidemiol Infect. 2012, 141: 1-8.Google Scholar
- Murray CJL, Mathers CD, Salomon JA, Lopez AD: Health gaps: an overview and critical appraisal. Summary measures of population health - concepts, ethics, measurement and applications. Edited by: Murray CJL, Salomon JA, Mathers CD, Lopez AD. 2002, Geneva: World Health Orgnaization, 233-244.Google Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/13/863/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.