Michael K. Freeman

Smoking prevalence and cigarette consumption in 187 countries, 1980-2012.

IMPORTANCE Tobacco is a leading global disease risk factor. Understanding national trends in prevalence and consumption is critical for prioritizing action and evaluating tobacco control progress. OBJECTIVE To estimate the prevalence of daily smoking by age and sex and the number of cigarettes per smoker per day for 187 countries from 1980 to 2012. DESIGN Nationally representative sources that measured tobacco use (n = 2102 country-years of data) were systematically identified. Survey data that did not report daily tobacco smoking were adjusted using the average relationship between different definitions. Age-sex-country-year observations (n = 38,315) were synthesized using spatial-temporal gaussian process regression to model prevalence estimates by age, sex, country, and year. Data on consumption of cigarettes were used to generate estimates of cigarettes per smoker per day. MAIN OUTCOMES AND MEASURES Modeled age-standardized prevalence of daily tobacco smoking by age, sex, country, and year; cigarettes per smoker per day by country and year. RESULTS Global modeled age-standardized prevalence of daily tobacco smoking in the population older than 15 years decreased from 41.2% (95% uncertainty interval [UI], 40.0%-42.6%) in 1980 to 31.1% (95% UI, 30.2%-32.0%; P < .001) in 2012 for men and from 10.6% (95% UI, 10.2%-11.1%) to 6.2% (95% UI, 6.0%-6.4%; P < .001) for women. Global modeled prevalence declined at a faster rate from 1996 to 2006 (mean annualized rate of decline, 1.7%; 95% UI, 1.5%-1.9%) compared with the subsequent period (mean annualized rate of decline, 0.9%; 95% UI, 0.5%-1.3%; P = .003). Despite the decline in modeled prevalence, the number of daily smokers increased from 721 million (95% UI, 700 million-742 million) in 1980 to 967 million (95% UI, 944 million-989 million; P < .001) in 2012. Modeled prevalence rates exhibited substantial variation across age, sex, and countries, with rates below 5% for women in some African countries to more than 55% for men in Timor-Leste and Indonesia. The number of cigarettes per smoker per day also varied widely across countries and was not correlated with modeled prevalence. CONCLUSIONS AND RELEVANCE Since 1980, large reductions in the estimated prevalence of daily smoking were observed at the global level for both men and women, but because of population growth, the number of smokers increased significantly. As tobacco remains a threat to the health of the worlds population, intensified efforts to control its use are needed.

Healthy life expectancy for 187 countries, 1990–2010: a systematic analysis for the Global Burden Disease Study 2010

BACKGROUND Healthy life expectancy (HALE) summarises mortality and non-fatal outcomes in a single measure of average population health. It has been used to compare health between countries, or to measure changes over time. These comparisons can inform policy questions that depend on how morbidity changes as mortality decreases. We characterise current HALE and changes over the past two decades in 187 countries. METHODS Using inputs from the Global Burden of Disease Study (GBD) 2010, we assessed HALE for 1990 and 2010. We calculated HALE with life table methods, incorporating estimates of average health over each age interval. Inputs from GBD 2010 included age-specific information for mortality rates and prevalence of 1160 sequelae, and disability weights associated with 220 distinct health states relating to these sequelae. We computed estimates of average overall health for each age group, adjusting for comorbidity with a Monte Carlo simulation method to capture how multiple morbidities can combine in an individual. We incorporated these estimates in the life table by the Sullivan method to produce HALE estimates for each population defined by sex, country, and year. We estimated the contributions of changes in child mortality, adult mortality, and disability to overall change in population health between 1990 and 2010. FINDINGS In 2010, global male HALE at birth was 58·3 years (uncertainty interval 56·7-59·8) and global female HALE at birth was 61·8 years (60·1-63·4). HALE increased more slowly than did life expectancy over the past 20 years, with each 1-year increase in life expectancy at birth associated with a 0·8-year increase in HALE. Across countries in 2010, male HALE at birth ranged from 27·9 years (17·3-36·5) in Haiti, to 68·8 years (67·0-70·4) in Japan. Female HALE at birth ranged from 37·1 years (26·9-43·7) in Haiti, to 71·7 years (69·7-73·4) in Japan. Between 1990 and 2010, male HALE increased by 5 years or more in 42 countries compared with 37 countries for female HALE, while male HALE decreased in 21 countries and 11 for female HALE. Between countries and over time, life expectancy was strongly and positively related to number of years lost to disability. This relation was consistent between sexes, in cross-sectional and longitudinal analysis, and when assessed at birth, or at age 50 years. Changes in disability had small effects on changes in HALE compared with changes in mortality. INTERPRETATION HALE differs substantially between countries. As life expectancy has increased, the number of healthy years lost to disability has also increased in most countries, consistent with the expansion of morbidity hypothesis, which has implications for health planning and health-care expenditure. Compared with substantial progress in reduction of mortality over the past two decades, relatively little progress has been made in reduction of the overall effect of non-fatal disease and injury on population health. HALE is an attractive indicator for monitoring health post-2015. FUNDING The Bill & Melinda Gates Foundation.

Using verbal autopsy to measure causes of death: the comparative performance of existing methods

BackgroundMonitoring progress with disease and injury reduction in many populations will require widespread use of verbal autopsy (VA). Multiple methods have been developed for assigning cause of death from a VA but their application is restricted by uncertainty about their reliability.MethodsWe investigated the validity of five automated VA methods for assigning cause of death: InterVA-4, Random Forest (RF), Simplified Symptom Pattern (SSP), Tariff method (Tariff), and King-Lu (KL), in addition to physician review of VA forms (PCVA), based on 12,535 cases from diverse populations for which the true cause of death had been reliably established. For adults, children, neonates and stillbirths, performance was assessed separately for individuals using sensitivity, specificity, Kappa, and chance-corrected concordance (CCC) and for populations using cause specific mortality fraction (CSMF) accuracy, with and without additional diagnostic information from prior contact with health services. A total of 500 train-test splits were used to ensure that results are robust to variation in the underlying cause of death distribution.ResultsThree automated diagnostic methods, Tariff, SSP, and RF, but not InterVA-4, performed better than physician review in all age groups, study sites, and for the majority of causes of death studied. For adults, CSMF accuracy ranged from 0.764 to 0.770, compared with 0.680 for PCVA and 0.625 for InterVA; CCC varied from 49.2% to 54.1%, compared with 42.2% for PCVA, and 23.8% for InterVA. For children, CSMF accuracy was 0.783 for Tariff, 0.678 for PCVA, and 0.520 for InterVA; CCC was 52.5% for Tariff, 44.5% for PCVA, and 30.3% for InterVA. For neonates, CSMF accuracy was 0.817 for Tariff, 0.719 for PCVA, and 0.629 for InterVA; CCC varied from 47.3% to 50.3% for the three automated methods, 29.3% for PCVA, and 19.4% for InterVA. The method with the highest sensitivity for a specific cause varied by cause.ConclusionsPhysician review of verbal autopsy questionnaires is less accurate than automated methods in determining both individual and population causes of death. Overall, Tariff performs as well or better than other methods and should be widely applied in routine mortality surveillance systems with poor cause of death certification practices.

Sources and Focus of Health Development Assistance, 1990–2014

IMPORTANCE The governments of high-income countries and private organizations provide billions of dollars to developing countries for health. This type of development assistance can have a critical role in ensuring that life-saving health interventions reach populations in need. OBJECTIVES To identify the amount of development assistance that countries and organizations provided for health and to determine the health areas that received these funds. EVIDENCE REVIEW Budget, revenue, and expenditure data on the primary agencies and organizations (n = 38) that provided resources to developing countries (n = 146-183, depending on the year) for health from 1990 through 2014 were collected. For each channel (the international agency or organization that directed the resources toward the implementing institution or government), the source and recipient of the development assistance were determined and redundant accounting of the same dollar, which occurs when channels transfer funds among each other, was removed. This research derived the flow of resources from source to intermediary channel to recipient. Development assistance for health (DAH) was divided into 11 mutually exclusive health focus areas, such that every dollar of development assistance was assigned only 1 health focus area. FINDINGS Since 1990,

Simplified Symptom Pattern Method for verbal autopsy analysis: multisite validation study using clinical diagnostic gold standards.

458.0 billion of development assistance has been provided to maintain or improve health in developing countries. The largest source of funding was the US government, which provided

National health accounts data from 1996 to 2010: a systematic review

143.1 billion between 1990 and 2014, including

Concentrating risk: a systematic analysis of the global smoking epidemic

12.4 billion in 2014. Of resources that originated with the US government, 70.6% were provided through US government agencies, and 41.0% were allocated for human immunodeficiency virus (HIV)/AIDS. The second largest source of development assistance for health was private philanthropic donors, including the Bill and Melinda Gates Foundation and other private foundations, which provided

Performance of InterVA for assigning causes of death to verbal autopsies: multisite validation study using clinical diagnostic gold standards

69.9 billion between 1990 and 2014, including

Improving performance of the Tariff Method for assigning causes of death to verbal autopsies

6.2 billion in 2014. These resources were provided primarily through private foundations and nongovernmental organizations and were allocated for a diverse set of health focus areas. Since 1990, 28.0% of all DAH was allocated for maternal health and newborn and child health; 23.2% for HIV/AIDS, 4.3% for malaria, 2.8% for tuberculosis, and 1.5% for noncommunicable diseases. Between 2000 and 2010, DAH increased 11.3% annually. However, since 2010, total DAH has not increased as substantially. CONCLUSIONS AND RELEVANCE Funding for health in developing countries has increased substantially since 1990, with a focus on HIV/AIDS, maternal health, and newborn and child health. Funding from the US government has played a substantial role in this expansion. Funding for noncommunicable diseases has been limited. Understanding how funding patterns have changed across time and the priorities of sources of international funding across distinct channels, recipients, and health focus areas may help identify where funding gaps persist and where cost-effective interventions could save lives.

The use of crowdsourcing and the role of game mechanics in identifying erroneous disease burden estimates

BackgroundVerbal autopsy can be a useful tool for generating cause of death data in data-sparse regions around the world. The Symptom Pattern (SP) Method is one promising approach to analyzing verbal autopsy data, but it has not been tested rigorously with gold standard diagnostic criteria. We propose a simplified version of SP and evaluate its performance using verbal autopsy data with accompanying true cause of death.MethodsWe investigated specific parameters in SPs Bayesian framework that allow for its optimal performance in both assigning individual cause of death and in determining cause-specific mortality fractions. We evaluated these outcomes of the method separately for adult, child, and neonatal verbal autopsies in 500 different population constructs of verbal autopsy data to analyze its ability in various settings.ResultsWe determined that a modified, simpler version of Symptom Pattern (termed Simplified Symptom Pattern, or SSP) performs better than the previously-developed approach. Across 500 samples of verbal autopsy testing data, SSP achieves a median cause-specific mortality fraction accuracy of 0.710 for adults, 0.739 for children, and 0.751 for neonates. In individual cause of death assignment in the same testing environment, SSP achieves 45.8% chance-corrected concordance for adults, 51.5% for children, and 32.5% for neonates.ConclusionsThe Simplified Symptom Pattern Method for verbal autopsy can yield reliable and reasonably accurate results for both individual cause of death assignment and for determining cause-specific mortality fractions. The method demonstrates that verbal autopsies coupled with SSP can be a useful tool for analyzing mortality patterns and determining individual cause of death from verbal autopsy data.