Calculating expected years of life lost for assessing local ethnic disparities in causes of premature death

Background A core function of local health departments is to conduct health assessments. The analysis of death certificates provides information on diseases, conditions, and injuries that are likely to cause death – an important outcome indicator of population health. The expected years of life lost (YLL) measure is a valid, stand-alone measure for identifying and ranking the underlying causes of premature death. The purpose of this study was to rank the leading causes of premature death among San Francisco residents, and to share detailed methods so that these analyses can be used in other local health jurisdictions. Methods Using death registry data and population estimates for San Francisco deaths in 2003–2004, we calculated the number of deaths, YLL, and age-standardized YLL rates (ASYRs). The results were stratified by sex, ethnicity, and underlying cause of death. The YLL values were used to rank the leading causes of premature death for men and women, and by ethnicity. Results In the years 2003–2004, 6312 men died (73,627 years of life lost), and 5726 women died (51,194 years of life lost). The ASYR for men was 65% higher compared to the ASYR for women (8971.1 vs. 5438.6 per 100,000 persons per year). The leading causes of premature deaths are those with the largest average YLLs and are largely preventable. Among men, these were HIV/AIDS, suicide, drug overdose, homicide, and alcohol use disorder; and among women, these were lung cancer, breast cancer, hypertensive heart disease, colon cancer, and diabetes mellitus. A large health disparity exists between African Americans and other ethnic groups: African American age-adjusted overall and cause-specific YLL rates were higher, especially for homicide among men. Except for homicide among Latino men, Latinos and Asians have comparable or lower YLL rates among the leading causes of death compared to whites. Conclusion Local death registry data can be used to measure, rank, and monitor the leading causes of premature death, and to measure and monitor ethnic health disparities.


Background
A core function of local health departments is to conduct public health surveillance, including population health assessments [1,2]. Public health surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of data regarding a health-related event for use in public health action to reduce morbidity and mortality and to improve health [3]. For a local health jurisdiction, primary data collection, such as representative population-based surveys, can be expensive and unsustainable. Therefore, local health departments must analyze existing health data, preferably those that are population-based, comprehensive, readily available, and locally relevant.
Death records are an important data source for assessing population health and health disparities because they cover the whole population and include information on key characteristics of decedents, including age, sex, ethnicity, place of residence and of death, and underlying and contributing causes of death. However, cause-specific mortality is typically reported using traditional epidemiologic measures, especially counts and rates (including ageadjusted rates), that are heavily influenced by deaths among older residents. For most causes, these measures are not very sensitive to deaths occurring at younger ages, which are more likely to be premature, preventable deaths.
To identify and prioritize causes of premature death, the standard expected years of life lost (YLL) measure, as developed by the Global Burden of Disease Study [4], provides a valuable analytic tool which can be applied to local geographic areas. The YLL is based on comparing the age of death to an external standard life expectancy curve, and can incorporate time discounting and age weighting. The YLL, combined with the years lived with disability (YLD) measure, make up the disability-adjusted life year (DALY). Unfortunately, directly measuring YLDs (and therefore DALYs) is cost prohibitive and not practical for most local health jurisdictions. In contrast, YLLs are measurable for a comprehensive set of conditions. YLLs, as opposed to more traditional mortality measures (counts, rates, etc.), highlight premature deaths. Where population estimates are available, age-adjusted YLL rates allow comparisons across groups or over time [5]. These deaths are particularly important from a public health and public policy perspective because they represent preventable loss of life.
Although the YLL is a valid, stand-alone measure for identifying and ranking the causes of premature death for a region [6][7][8], this measure has not been widely adopted for local area mortality analyses. There are several reasons for this. First, detailed methods for calculating YLL are not available in standard epidemiology textbooks or scientific journal articles. In contrast, the years of potential life lost (YPLL) is commonly used: it is easily calculated by subtracting the age of death from a chosen cut-off (e.g., 65, 75, or 85 years) [9,10]; however, the YPLL does not measure deaths after the cut-off age, and it does not incorporate time-based discounting used in cost-effectiveness analysis. The YLL measures every death and can incorporate discounting. Second, with few exceptions [5], sufficient local area YLL analyses have not been published to demonstrate their value in assessing population health. And third, readily available software solutions to make analyses more efficient have not been developed.
The purpose of this paper is to provide detailed methods for calculating YLL for a local geographic area (San Francisco, California, United States), and to demonstrate its value as a population health measure to impact local public health priorities. We illustrate how to use YLL to rank causes of death, and how to use average YLLs to identify the leading causes of premature death for major ethnic groups. Analysis and interpretation of death registry data using YLLs provide objective evidence for public health policymakers, partners, and stakeholders to inform and guide the setting of local public health priorities. This is especially important because of the geographic and demographic variation in health outcomes, major risk factors, and health disparities [5,11].   where n a x is the average age of death, and and are   tively. See Table 4 for use of in spreadsheet calculations.

Calculating expected years of life lost (Y LL)
For a group of deaths that occurred at ages within age interval x to x + n, the crude expected years of life lost is where n D x is the number of deaths between age x and age x + n.
To incorporate discounting and age weighting, one would use Equation 3: where . For this equation, r is the discount rate, and β, C, and K are age weighting constants (see Table 1 for complete definitions). To include age weighting, K (the modulation constant) can be set to 1. For this study, age weighting was not used (K = 0) and r = 0.03.

When the discount rate (r) is 0, Equation 3 simplifies to Equation 4:
First, we calculated the expected of years of life lost, comparing men to women, by summing n Y x for all age intervals (Table 4): Using this approach, we calculated Y LLs for 117 specific causes of death stratified by sex, and stratified by sex and ethnicity.

Calculating age-standardized expected years of life lost rates
Using the direct method [18], we calculated age standardized YLL rates (ASYR). First, we calculated age-specific rates of years of life lost ( n y x ). Then, these rates were reweighted using using the Year 2000 United States standard million population ( n w x in Table 3) [18]. The reweighted rates ( ) were summed to get an ASYR (Equation 6).
See Table 5 for use of Equation 6 in spreadsheet calculations.

Ranking leading causes of premature death
Determining the leading causes of premature death required two steps. First, the leading 15 causes of death
Similarly, an analysis was conducted to rank the leading causes of premature death by ethnicity and sex [Additional file 1] for African Americans (Table A-1), Asians/ Pacific Islanders (Table A-2), Latino/Hispanics (Table A-3), and whites (Table A-3). Similar analyses were done for each ethnic group. For example, among African American men, the leading causes of premature death (largest average YLLs) were homicide (25.9 years), HIV/AIDS (19.7 years), hypertensive heart disease (14.7 years), drug over- Age-standardized YLL rates (ASYRs) allow comparisons of the burden of premature mortality by ethnic group and specific cause of death (Figures 1, 2, 3). For example, for almost every leading cause of premature death in men and women, African Americans had the highest ASYRs compared to other ethnic groups. Among African American men, the disparity in ASYRs was most notable for violent assault (homicide), followed by HIV/AIDS, vascular diseases (ischemic and hypertensive heart, and cerebrovascular disease), accidental drug overdose, and lung cancer.
Among African American women, the disparity in ASYRs was most notable for vascular diseases (ischemic and hypertensive heart, and cerebrovascular diseases), breast cancer, HIV/AIDS, and accidental drug overdose.

Discussion
The key findings of this study are that (1) the leading causes of premature mortality were largely preventable: among men, these were HIV/AIDS, suicide, drug overdose, homicide, and alcohol use disorder; and among women, these were lung cancer, breast cancer, hypertensive heart disease, colon cancer, and diabetes mellitus; (2) leading causes of premature death differed remarkably between ethnic groups (Tables A-1-A-4); (3) a large health disparity was measured between African Americans and other ethnic groups: African American age-adjusted overall and cause-specific YLL rates are notably higher, especially for homicide among men (Figures 1, 2, and 3); and (4) except for homicide among Latino men, Latinos and Asians had comparable or lower YLL rates among the leading causes of premature death compared to whites (Figures 2 and 3). These results illustrate how death registry data can be used to measure, rank, and monitor the leading causes of premature mortality for a local geographic region. Such studies can be used to monitor the local mortality burden of disease and injury over time. For example, our results were compared to our previous San Francisco YLL study for the period 1990-1995 [11]. While the burden of HIV/AIDS deaths decreased remarkably, the ethnic health disparities remained, with African Americans continuing to suffer the largest burden. This was especially striking for homicides among African American men. The generally better health status of Asians and Latinos has persisted.
Several of these findings mirror those from national studies [20]. For example, the U.S. Burden of Disease and Injury Study [21] found many of the same preventable causes of premature death among the leading causes, and that the YLL ranking for each ethnic group was unique. Like our study, there were large disparities, measured as DALYs, between African Americans and other ethnic groups, and they reported better health outcomes among  Asians than whites. The Eight Americas Study [22,23] also found large disparities, measured as life expectancy, between Asian Americans and African Americans. A recent examination of the U.S. black-white disparity in life expectancy during the period 1983-2003 [24] found, like our study, that cardiovascular disease (both males and females), homicide (males), and HIV/AIDS (males) were leading contributors to the gap in recent years.
Three measures were used in this study: YLLs, average YLLs, and ASYRs. The YLL is a stand-alone measure of mortality burden not requiring population estimates. It was used to rank the 15 leading causes of death for men and women (Table 7). However, these 15 leading causes were influenced by the larger number of deaths among older residents. To highlight premature, preventable causes of death, we then ranked these top 15 causes by their average YLLs. Notably, many of the leading causes of death have strong social determinants. Alternatively, the ASYR could have been used to rank the leading causes of death; however, this was not our first choice because it requires population estimates, and the rankings would still be influenced by older deaths. Given our availability of population estimates, ASYRs were used to make comparisons among ethnic groups (  [15,21]. Seventh, our study included Latinos/Hispanics, an important segment of the population that was not included in a similar national study [21]. Eighth, with the availability of ethnic-specific population estimates, we were able to age-standardize the YLLs to measure, compare, and monitor the ethnic health disparities in the burden of premature deaths. And ninth, our study findings are directly relevant and can be adapted to the diverse and unique needs of our communities, and to our local government and policymakers. This study also has several limitations. First, the accuracy of data recorded on death certificates (e.g., underlying cause of death and ethnicity) varies by region and underlying cause [25]. Additionally, analyses using underlying cause of death categories may underestimate the mortality burden for selected contributing causes of death listed on the death certificates (e.g., diabetes mellitus) [26]. Second, the YLL metric does not measure well conditions that cause significant disease and disability, but are difficult to measure (e.g., mental illness) or do not result in death (e.g., osteoarthritis). Third, on average, there may be a 10month or longer delay from the time a calendar year ends and the availability of ICD-10-coded death registry data.
Fourth, the ranking of a specific cause of death depends on its individual YLL magnitude as well as its relative contribution compared to other causes; changes in ranking for a cause over time may be due either to changes in the occurrence of that cause, or to changes in the occurrences of other causes ranked above or below it. Fifth, the average YLL could be large for a specific cause of death but only involve a small number of deaths (small burden). To avoid this problem, we only evaluated the average YLL for the highest ranked causes of death based on YLLs. Sixth, the YLL measure is not age-standardized and cannot be used to compare specific causes of death between groups with different age compositions. (With population estimates, YLL can be age-standardized as described in Methods.) And seventh, because of the uncertainty of population estimates, age-standardized rates must also be interpreted with caution. In spite of these limitations, using YLLs to rank the leading causes of premature death provides community residents, community-based organizations, policy makers, public health authorities, and researchers with local, representative, objective, and informative data to guide and inform public health priorities, and to direct and evaluate public health interventions.
This study has the following key implications: First, we provide the methodological details for calculating YLL to measure the burden of premature mortality for any geographic area that has death registry data. We provide both the ICD-10 causes of death classifications used for this study [Additional file 2] and the computational program code for calculating age-interval-specific expected years of life lost that can incorporate discounting (used in this study) and age weighting (not used in this study) [Additional file 3]. This code can be executed in a freely available, open source program for statistical computing and graphics [19]. And second, we demonstrate how these results can be used to rank the leading cause of premature death for major ethnic groups. The rankings can be use to guide, inform, and monitor public health priorities and programs for each group. These analyses can become part of routine public health surveillance for local health jurisdictions, as we have done in San Francisco.

Conclusion
Population health measures based on YLLs are readily calculated and useful for measuring, ranking, and monitoring the leading causes of premature death for a local geographic area, and for measuring and monitoring the impact of local efforts to reduce premature mortality in ethnic groups for which there are health disparities.