Although the relationship between self-rated health (SRH) and physical and mental health is well documented in developed countries, very few studies have analyzed this association in the developing world, particularly in Africa. In this study, we examine the associations of SRH with measures of physical and mental health (chronic diseases, functional limitations, and depression) among adults in Ouagadougou, Burkina Faso, and how these associations vary by sex, age, and education level.
This study was based on 2195 individuals aged 15 years or older who participated in a cross-sectional interviewer-administered health survey conducted in 2010 in areas of the Ouagadougou Health and Demographic Surveillance System. Logistic regression models were used to analyze the associations of poor SRH with chronic diseases, functional limitations, and depression, first in the whole sample and then stratified by sex, age, and education level.
Poor SRH was strongly correlated with chronic diseases and functional limitations, but not with depression, suggesting that in this context, physical health probably makes up most of people’s perceptions of their health status. The effect of functional limitations on poor SRH increased with age, probably because the ability to circumvent or compensate for a disability diminishes with age. The effect of functional limitations was also stronger among the least educated, probably because physical integrity is more important for people who depend on it for their livelihood. In contrast, the effect of chronic diseases appeared to decrease with age. No variation by sex was observed in the associations of SRH with chronic diseases, functional limitations, or depression.
Our findings suggest that different subpopulations delineated by age and education level weight the components of health differently in their self-rated health in Ouagadougou, Burkina Faso. In-depth studies are needed to understand why and how these groups do so.
Ouagadougou; Burkina Faso; Self-rated health; Chronic diseases; Functional limitations; Depression; Adults
The relationship between health services and population outcomes is an important area of public health research that requires bringing together data on outcomes and the relevant service environment. Linking independent, existing datasets geographically is potentially an efficient approach; however, it raises a number of methodological issues which have not been extensively explored. This sensitivity analysis explores the potential misclassification error introduced when a sample rather than a census of health facilities is used and when household survey clusters are geographically displaced for confidentiality.
Using the 2007 Rwanda Service Provision Assessment (RSPA) of all public health facilities and the 2007–2008 Rwanda Interim Demographic and Health Survey (RIDHS), five health facility samples and five household cluster displacements were created to simulate typical SPA samples and household cluster datasets. Facility datasets were matched with cluster datasets to create 36 paired datasets. Four geographic techniques were employed to link clusters with facilities in each paired dataset. The links between clusters and facilities were operationalized by creating health service variables from the RSPA and attaching them to linked RIDHS clusters. Comparisons between the original facility census and undisplaced clusters dataset with the multiple samples and displaced clusters datasets enabled measurement of error due to sampling and displacement.
Facility sampling produced larger misclassification errors than cluster displacement, underestimating access to services. Distance to the nearest facility was misclassified for over 50% of the clusters when directly linked, while linking to all facilities within an administrative boundary produced the lowest misclassification error. Measuring relative service environment produced equally poor results with over half of the clusters assigned to the incorrect quintile when linked with a sample of facilities and more than one-third misclassified due to displacement.
At low levels of geographic disaggregation, linking independent facility samples and household clusters is not recommended. Linking facility census data with population data at the cluster level is possible, but misclassification errors associated with geographic displacement of clusters will bias estimates of relationships between service environment and health outcomes. The potential need to link facility and population-based data requires consideration when designing a facility survey.
Spatial linkage; DHS; SPA; Misclassification error
Estimates of under-5 mortality at the national level for countries without high-quality vital registration systems are routinely derived from birth history data in censuses and surveys. Subnational or stratified analyses of under-5 mortality could also be valuable, but the usefulness of under-5 mortality estimates derived from birth histories from relatively small samples of women is not known. We aim to assess the magnitude and direction of error that can be expected for estimates derived from birth histories with small samples of women using various analysis methods.
We perform a data-based simulation study using Demographic and Health Surveys. Surveys are treated as populations with known under-5 mortality, and samples of women are drawn from each population to mimic surveys with small sample sizes. A variety of methods for analyzing complete birth histories and one method for analyzing summary birth histories are used on these samples, and the results are compared to corresponding true under-5 mortality. We quantify the expected magnitude and direction of error by calculating the mean error, mean relative error, mean absolute error, and mean absolute relative error.
All methods are prone to high levels of error at the smallest sample size with no method performing better than 73% error on average when the sample contains 10 women. There is a high degree of variation in performance between the methods at each sample size, with methods that contain considerable pooling of information generally performing better overall. Additional stratified analyses suggest that performance varies for most methods according to the true level of mortality and the time prior to survey. This is particularly true of the summary birth history method as well as complete birth history methods that contain considerable pooling of information across time.
Performance of all birth history analysis methods is extremely poor when used on very small samples of women, both in terms of magnitude of expected error and bias in the estimates. Even with larger samples there is no clear best method to choose for analyzing birth history data. The methods that perform best overall are the same methods where performance is noticeably different at different levels of mortality and lengths of time prior to survey. At the same time, methods that perform more uniformly across levels of mortality and lengths of time prior to survey also tend to be among the worst performing overall.
Selection bias is common in clinic-based HIV surveillance. Clinics located in HIV hotspots are often the first to be chosen and monitored, while clinics in less prevalent areas are added to the surveillance system later on. Consequently, the estimated HIV prevalence based on clinic data is substantially distorted, with markedly higher HIV prevalence in the earlier periods and trends that reveal much more dramatic declines than actually occur.
Using simulations, we compare and contrast the performance of the various approaches and models for handling selection bias in clinic-based HIV surveillance. In particular, we compare the application of complete-case analysis and multiple imputation (MI). Several models are considered for each of the approaches. We demonstrate the application of the methods through sentinel surveillance data collected between 2002 and 2008 from India.
Simulations suggested that selection bias, if not handled properly, can lead to biased estimates of HIV prevalence trends and inaccurate evaluation of program impact. Complete-case analysis and MI differed considerably in their ability to handle selection bias. In scenarios where HIV prevalence remained constant over time (i.e. β = 0), the estimated β^1 derived from MI tended to be biased downward. Depending on the imputation model used, the estimated bias ranged from −1.883 to −0.048 in logit prevalence. Furthermore, as the level of selection bias intensified, the extent of bias also increased. In contrast, the estimates yielded by complete-case analysis were relatively unbiased and stable across the various scenarios. The estimated bias ranged from −0.002 to 0.002 in logit prevalence.
Given that selection bias is common in clinic-based HIV surveillance, when analyzing data from such sources appropriate adjustment methods need to be applied. The results in this paper suggest that indiscriminant application of imputation models can lead to biased results.
Selection bias; Simulations; Missing data; Multiple imputation; Complete-case analysis
The Millennium Development Goals (MDGs) have prompted an expansion in approaches to deriving health metrics to measure progress toward their achievement. Accurate measurements should take into account the high degrees of spatial heterogeneity in health risks across countries, and this has prompted the development of sophisticated cartographic techniques for mapping and modeling risks. Conversion of these risks to relevant population-based metrics requires equally detailed information on the spatial distribution and attributes of the denominator populations. However, spatial information on age and sex composition over large areas is lacking, prompting many influential studies that have rigorously accounted for health risk heterogeneities to overlook the substantial demographic variations that exist subnationally and merely apply national-level adjustments.
Here we outline the development of high resolution age- and sex-structured spatial population datasets for Africa in 2000-2015 built from over a million measurements from more than 20,000 subnational units, increasing input data detail from previous studies by over 400-fold. We analyze the large spatial variations seen within countries and across the continent for key MDG indicator groups, focusing on children under 5 and women of childbearing age, and find that substantial differences in health and development indicators can result through using only national level statistics, compared to accounting for subnational variation.
Progress toward meeting the MDGs will be measured through national-level indicators that mask substantial inequalities and heterogeneities across nations. Cartographic approaches are providing opportunities for quantitative assessments of these inequalities and the targeting of interventions, but demographic spatial datasets to support such efforts remain reliant on coarse and outdated input data for accurately locating risk groups. We have shown here that sufficient data exist to map the distribution of key vulnerable groups, and that doing so has substantial impacts on derived metrics through accounting for spatial demographic heterogeneities that exist within nations across Africa.
Population; Demography; Mapping; Millenium development goals
In the current issue of Population Health Metrics, two reports paint a bleak picture of American public health. Both physical inactivity and obesity remain highly prevalent; yet, it is not clear that increased physical activity will reduce the burden of obesity. There continue to be widespread disparities in life expectancy across United States counties. These reports appear against a backdrop of debate regarding how we should allocate our scarce resources for improving health: should we focus more on improving access to high-quality medical care, or should we instead focus on more and better public health interventions? While optimal solutions remain obscure, a look at prior successes suggests that ultimately they will come from the conduct and implementation of rigorous science, and in particular event-driven trials.
Public health; Population science; Obesity; Physical activity; Life-expectancy; Randomized trials
Obesity; Salt; Fat; Sugar; Tobacco; Government; Industry; Regulation
The United States spends more than any other country on health care. The poor relative performance of the US compared to other high-income countries has attracted attention and raised questions about the performance of the US health system. An important dimension to poor national performance is the large disparities in life expectancy.
We applied a mixed effects Poisson statistical model and Gaussian Process Regression to estimate age-specific mortality rates for US counties from 1985 to 2010. We generated uncertainty distributions for life expectancy at each age using standard simulation methods.
Female life expectancy in the United States increased from 78.0 years in 1985 to 80.9 years in 2010, while male life expectancy increased from 71.0 years in 1985 to 76.3 years in 2010. The gap between female and male life expectancy in the United States was 7.0 years in 1985, narrowing to 4.6 years in 2010. For males at the county level, the highest life expectancy steadily increased from 75.5 in 1985 to 81.7 in 2010, while the lowest life expectancy remained under 65. For females at the county level, the highest life expectancy increased from 81.1 to 85.0, and the lowest life expectancy remained around 73. For male life expectancy at the county level, there have been three phases in the evolution of inequality: a period of rising inequality from 1985 to 1993, a period of stable inequality from 1993 to 2002, and rising inequality from 2002 to 2010. For females, in contrast, inequality has steadily increased during the 25-year period. Compared to only 154 counties where male life expectancy remained stagnant or declined, 1,405 out of 3,143 counties (45%) have seen no significant change or a significant decline in female life expectancy from 1985 to 2010. In all time periods, the lowest county-level life expectancies are seen in the South, the Mississippi basin, West Virginia, Kentucky, and selected counties with large Native American populations.
The reduction in the number of counties where female life expectancy at birth is declining in the most recent period is welcome news. However, the widening disparities between counties and the slow rate of increase compared to other countries should be viewed as a call for action. An increased focus on factors affecting health outcomes, morbidity, and mortality such as socioeconomic factors, difficulty of access to and poor quality of health care, and behavioral, environmental, and metabolic risk factors is urgently required.
Obesity and physical inactivity are associated with several chronic conditions, increased medical care costs, and premature death.
We used the Behavioral Risk Factor Surveillance System (BRFSS), a state-based random-digit telephone survey that covers the majority of United States counties, and the National Health and Nutrition Examination Survey (NHANES), a nationally representative sample of the US civilian noninstitutionalized population. About 3.7 million adults aged 20 years or older participated in the BRFSS from 2000 to 2011, and 30,000 adults aged 20 or older participated in NHANES from 1999 to 2010. We calculated body mass index (BMI) from self-reported weight and height in the BRFSS and adjusted for self-reporting bias using NHANES. We calculated self-reported physical activity—both any physical activity and physical activity meeting recommended levels—from self-reported data in the BRFSS. We used validated small area estimation methods to generate estimates of obesity and physical activity prevalence for each county annually for 2001 to 2011.
Our results showed an increase in the prevalence of sufficient physical activity from 2001 to 2009. Levels were generally higher in men than in women, but increases were greater in women than men. Counties in Kentucky, Florida, Georgia, and California reported the largest gains. This increase in level of activity was matched by an increase in obesity in almost all counties during the same time period. There was a low correlation between level of physical activity and obesity in US counties. From 2001 to 2009, controlling for changes in poverty, unemployment, number of doctors per 100,000 population, percent rural, and baseline levels of obesity, for every 1 percentage point increase in physical activity prevalence, obesity prevalence was 0.11 percentage points lower.
Our study showed that increased physical activity alone has a small impact on obesity prevalence at the county level in the US. Indeed, the rise in physical activity levels will have a positive independent impact on the health of Americans as it will reduce the burden of cardiovascular diseases and diabetes. Other changes such as reduction in caloric intake are likely needed to curb the obesity epidemic and its burden.
Physical activity; Obesity; Small area measurement; US counties
Age of onset is an important outcome to characterize a population with a chronic disease. With respect to social, cognitive, and physical aspects for patients and families, dementia is especially burdensome. In Germany, like in many other countries, it is highly prevalent in the older population and imposes enormous efforts for caregivers and society.
We develop an incidence-prevalence-mortality model to derive the mean and variance of the age of onset in chronic diseases. Age- and sex-specific incidence and prevalence of dementia is taken from published values based on health insurance data from 2002. Data about the age distribution in Germany in 2002 comes from the Federal Statistical Office.
Mean age of onset of a chronic disease depends on a) the age-specific incidence of the disease, b) the prevalence of the disease, and c) the age distribution of the population. The resulting age of onset of dementia in Germany in 2002 is 78.8 ± 8.1 years (mean ± standard deviation) for men and 81.9 ± 7.6 years for women.
Although incidence and prevalence of dementia in men are not greater than in women, men contract dementia approximately three years earlier than women. The reason lies in the different age distributions of the male and the female population in Germany.
The objective of these analyses is to document the relationship between biomarker-based indicators of health and socioeconomic status (SES) in a low-income African population where the cumulative effects of exposure to multiple stressors on physiological functions and health in general are expected to be highly detrimental for the well-being of individuals.
Biomarkers were collected subsequent to the 2008 round of the Malawi Longitudinal Study of Families and Health (MLSFH), a population-based study in rural Malawi, including blood lipids (total cholesterol, LDL, HDL, ratio of total cholesterol to HDL), biomarkers of renal and liver organ function (albumin and creatinine) and wide-range C-reactive protein (CRP) as a non-specific biomarker for inflammation. These biomarkers represent widely used indicators of health that are individually or cumulatively recognized as risk factors for age-related diseases among prime-aged and elderly individuals. Quantile regressions are used to estimate the age-gradient and the within-day variation of each biomarker distribution. Differences in biomarker levels by socioeconomic status are investigated using descriptive and multivariate statistics.
Overall, the number of significant associations between the biomarkers and socioeconomic measures is very modest. None of the biomarkers significantly varies with schooling. Except for CRP where being married is weakly associated with lower risk of having an elevated CRP level, marriage is not associated with the biomarkers measured in the MLSFH. Similarly, being Muslim is associated with a lower risk of having elevated CRP but otherwise religion does not predict being in the high-risk quartiles of any of the MLSFH biomarkers. Wealth does not predict being in the high-risk quartile of any of the MLSFH biomarkers, with the exception of a weak effect on creatinine. Being overweight or obese is associated with increased likelihood of being in the high-risk quartile for cholesterol, Chol/HDL ratio, and LDL.
The results provide only weak evidence for variation of the biomarkers by socioeconomic indicators in a poor Malawian context. Our findings underscore the need for further research to understand the determinants of health outcomes in a poor low-income context such as rural Malawi.
Biomarkers; Blood lipids; Creatinine; Albumin; Wide-range CRP; Socioeconomic status; Variation; Malawi
Published estimates of Aboriginal mortality and life expectancy (LE) for the eastern Australian states are derived from demographic modelling techniques to estimate the population and extent of under-recording of Aboriginality in death registration. No reliable empirical information on Aboriginal mortality and LE exists for New South Wales (NSW), the most populous Australian state in which 29% of Aboriginal people reside.
This paper estimates mortality and LE in a large, mainly metropolitan cohort of Aboriginal clients from the Aboriginal Medical Service (AMS) Redfern, Sydney, NSW.
Identifying information from patient records accrued by the AMS Redfern since 1980 of definitely Aboriginal clients, without distinction between Aboriginal and Torres Strait Islander (n=24,035), was extracted and linked to the National Death Index (NDI) at the Australian Institute of Health and Welfare (AIHW). Age-specific mortality rates and LEs for each sex were estimated using the AMS patient population as the denominator, discounted for deaths. Directly age-standardised mortality and LEs were estimated for 1995–1999, 2000–2004 and 2005–2009, along with 95% confidence intervals. Comparisons were made with other estimates of Aboriginal mortality and LE and with the total Australian population.
Mortality declined in the AMS Redfern cohort over 1995–2009, and the decline occurred mostly in the ≤44 year age range. Male LE at birth was estimated to be 64.4 years (95%CI:62.6-66.1) in 1995–1999, 65.6 years (95%CI:64.1-67.1) in 2000–2004, and 67.6 years (95%CI:65.9-69.2) for 2005–2009. In females, these LE estimates were 69.6 (95%CI:68.0-71.2), 71.1 (95%CI:69.9-72.4), and 71.4 (95%CI:70.0-72.8) years. LE in the AMS cohort was 11 years lower for males and 12 years lower for females than corresponding all-Australia LEs for the same periods. These were similar to estimates for Australian Aboriginal people overall for the same period by the Aboriginal Burden of Disease for 2009, using the General Growth Balance (GGB) model approach, and by the Australian Bureau of Statistics (ABS) for 2005–2007. LE in the AMS cohort was somewhat lower than these estimates for NSW Aboriginal people, and higher than ABS 2005–2007 estimates for Aboriginal people from Northern Territory, South Australia, and Western Australia.
The AMS Redfern cohort has provided the first empirically based estimates of mortality and LE trends in a large sample of Aboriginal people from NSW.
The estimated gap in life expectancy (LE) between Indigenous and non-Indigenous Australians was 12 years for men and 10 years for women, whereas the Northern Territory Indigenous LE gap was at least 50% greater than the national figures. This study aims to explain the Indigenous LE gap by common modifiable risk factors.
This study covered the period from 1986 to 2005. Unit record death data from the Northern Territory were used to assess the differences in LE at birth between the Indigenous and non-Indigenous populations by socioeconomic disadvantage, smoking, alcohol abuse, obesity, pollution, and intimate partner violence. The population attributable fractions were applied to estimate the numbers of deaths associated with the selected risks. The standard life table and cause decomposition technique was used to examine the individual and joint effects on health inequality.
The findings from this study indicate that among the selected risk factors, socioeconomic disadvantage was the leading health risk and accounted for one-third to one-half of the Indigenous LE gap. A combination of all six selected risks explained over 60% of the Indigenous LE gap.
Improving socioeconomic status, smoking cessation, and overweight reduction are critical to closing the Indigenous LE gap. This paper presents a useful way to explain the impact of risk factors of health inequalities, and suggests that reducing poverty should be placed squarely at the centre of the strategies to close the Indigenous LE gap.
Health status disparities; Risk factors; Life expectancy; Indigenous population; Socioeconomic factors
Stunting results from decreased food intake, poor diet quality, and a high burden of early childhood infections, and contributes to significant morbidity and mortality worldwide. Although food insecurity is an important determinant of child nutrition, including stunting, development of universal measures has been challenging due to cumbersome nutritional questionnaires and concerns about lack of comparability across populations. We investigate the relationship between household food access, one component of food security, and indicators of nutritional status in early childhood across eight country sites.
We administered a socioeconomic survey to 800 households in research sites in eight countries, including a recently validated nine-item food access insecurity questionnaire, and obtained anthropometric measurements from children aged 24 to 60 months. We used multivariable regression models to assess the relationship between household food access insecurity and anthropometry in children, and we assessed the invariance of that relationship across country sites.
Average age of study children was 41 months. Mean food access insecurity score (range: 0–27) was 5.8, and varied from 2.4 in Nepal to 8.3 in Pakistan. Across sites, the prevalence of stunting (42%) was much higher than the prevalence of wasting (6%). In pooled regression analyses, a 10-point increase in food access insecurity score was associated with a 0.20 SD decrease in height-for-age Z score (95% CI 0.05 to 0.34 SD; p = 0.008). A likelihood ratio test for heterogeneity revealed that this relationship was consistent across countries (p = 0.17).
Our study provides evidence of the validity of using a simple household food access insecurity score to investigate the etiology of childhood growth faltering across diverse geographic settings. Such a measure could be used to direct interventions by identifying children at risk of illness and death related to malnutrition.
Surveillance systems often present data by means of summary measures, like age-standardised rates. In this study, we aimed at comparing information derived from commonly used measures of smoking with that presented in modified population pyramids (PPs), using the example of the diffusion of smoking in Italy over the past two decades.
Data were derived from four National Health Interview Surveys carried out in 1983, 1990 to 1991, 1999 to 2000, and 2004 to 2005. After computing both age-specific and age-standardised rates of current, former, and never smoking, we constructed modified PPs by stratifying the male and female populations according to smoking status and educational level.
Modified PPs showed several features of the smoking epidemic in Italy that were not apparent from conventional surveillance techniques. First, they showed that the population of smokers is aging, with most current smokers in 2005 being males aged 25 to 39 and females aged 40 to 49, whereas in 1983 most smokers belonged to the youngest age groups. Second, they showed that in 2005 most smokers were found among subjects with middle and higher education, whereas two decades earlier most smokers were (male) subjects with the lowest education.
Modified PPs were able to show how absolute numbers of smokers were distributed by age and sex, how these numbers varied between population subgroups, and how they changed over time. PPs may help provide information on past and future trends in the absolute number of smokers and in their sociodemographic characteristics, which may be missed using only traditional surveillance methods.
Smoking; Education; Time trends; Surveillance; Population pyramids
Overweight and obesity prevalence are commonly used for public and policy communication of the extent of the obesity epidemic, yet comparable estimates of trends in overweight and obesity prevalence by country are not available.
We estimated trends between 1980 and 2008 in overweight and obesity prevalence and their uncertainty for adults 20 years of age and older in 199 countries and territories. Data were from a previous study, which used a Bayesian hierarchical model to estimate mean body mass index (BMI) based on published and unpublished health examination surveys and epidemiologic studies. Here, we used the estimated mean BMIs in a regression model to predict overweight and obesity prevalence by age, country, year, and sex. The uncertainty of the estimates included both those of the Bayesian hierarchical model and the uncertainty due to cross-walking from mean BMI to overweight and obesity prevalence.
The global age-standardized prevalence of obesity nearly doubled from 6.4% (95% uncertainty interval 5.7-7.2%) in 1980 to 12.0% (11.5-12.5%) in 2008. Half of this rise occurred in the 20 years between 1980 and 2000, and half occurred in the 8 years between 2000 and 2008. The age-standardized prevalence of overweight increased from 24.6% (22.7-26.7%) to 34.4% (33.2-35.5%) during the same 28-year period. In 2008, female obesity prevalence ranged from 1.4% (0.7-2.2%) in Bangladesh and 1.5% (0.9-2.4%) in Madagascar to 70.4% (61.9-78.9%) in Tonga and 74.8% (66.7-82.1%) in Nauru. Male obesity was below 1% in Bangladesh, Democratic Republic of the Congo, and Ethiopia, and was highest in Cook Islands (60.1%, 52.6-67.6%) and Nauru (67.9%, 60.5-75.0%).
Globally, the prevalence of overweight and obesity has increased since 1980, and the increase has accelerated. Although obesity increased in most countries, levels and trends varied substantially. These data on trends in overweight and obesity may be used to set targets for obesity prevalence as requested at the United Nations high-level meeting on Prevention and Control of NCDs.
Overweight; Obesity; Prevalence; Population health; Risk transition; Global health; Noncommunicable diseases
Self-reported height and weight are commonly collected at the population level; however, they can be subject to measurement error. The impact of this error on predicted risk, discrimination, and calibration of a model that uses body mass index (BMI) to predict risk of diabetes incidence is not known. The objective of this study is to use simulation to quantify and describe the effect of random and systematic error in self-reported height and weight on the performance of a model for predicting diabetes.
Two general categories of error were examined: random (nondirectional) error and systematic (directional) error on an algorithm relating BMI in kg/m2 to probability of developing diabetes. The cohort used to develop the risk algorithm was derived from 23,403 Ontario residents that responded to the 1996/1997 National Population Health Survey linked to a population-based diabetes registry. The data and algorithm were then simulated to allow for estimation of the impact of these errors on predicted risk using the Hosmer-Lemeshow goodness-of-fit χ2 and C-statistic. Simulations were done 500 times with sample sizes of 9,177 for males and 10,618 for females.
Simulation data successfully reproduced discrimination and calibration generated from population data. Increasing levels of random error in height and weight reduced the calibration and discrimination of the model. Random error biased the predicted risk upwards whereas systematic error biased predicted risk in the direction of the bias and reduced calibration; however, it did not affect discrimination.
This study demonstrates that random and systematic errors in self-reported health data have the potential to influence the performance of risk algorithms. Further research that quantifies the amount and direction of error can improve model performance by allowing for adjustments in exposure measurements.
To systematically review the methodology of general burden of disease studies. Three key questions were addressed: 1) what was the quality of the data, 2) which methodological choices were made to calculate disability adjusted life years (DALYs), and 3) were uncertainty and risk factor analyses performed? Furthermore, DALY outcomes of the included studies were compared.
Burden of disease studies (1990 to 2011) in international peer-reviewed journals and in grey literature were identified with main inclusion criteria being multiple-cause studies that quantified the burden of disease as the sum of the burden of all distinct diseases expressed in DALYs. Electronic database searches included Medline (PubMed), EMBASE, and Web of Science. Studies were collated by study population, design, methods used to measure mortality and morbidity, risk factor analyses, and evaluation of results.
Thirty-one studies met the inclusion criteria of our review. Overall, studies followed the Global Burden of Disease (GBD) approach. However, considerable variation existed in disability weights, discounting, age-weighting, and adjustments for uncertainty. Few studies reported whether mortality data were corrected for missing data or underreporting. Comparison with the GBD DALY outcomes by country revealed that for some studies DALY estimates were of similar magnitude; others reported DALY estimates that were two times higher or lower.
Overcoming “error” variation due to the use of different methodologies and low-quality data is a critical priority for advancing burden of disease studies. This can enlarge the detection of true variation in DALY outcomes between populations or over time.
Review; General burden of disease; Disability-adjusted life years; Methodology
When disability-adjusted life years are used to measure the burden of disease on a population in a time interval, they can be calculated in several different ways: from an incidence, pure prevalence, or hybrid perspective. I show that these calculation methods are not equivalent and discuss some of the formal difficulties each method faces. I show that if we don’t discount the value of future health, there is a sense in which the choice of calculation method is a mere question of accounting. Such questions can be important, but they don’t raise deep theoretical concerns. If we do discount, however, choice of calculation method can change the relative burden attributed to different conditions over time. I conclude by recommending that studies involving disability-adjusted life years be explicit in noting what calculation method is being employed and in explaining why that calculation method has been chosen.
Disability-adjusted life years; Incidence perspective; Prevalence perspective; Burden of disease; Discounting
During 2010, a community-based, sentinel site prospective surveillance system measured mortality, acute malnutrition prevalence, and the coverage of a Médecins Sans Frontières (MSF) intervention in four sous-préfectures of Lobaye prefecture in southwestern Central African Republic. We describe this surveillance system and its evaluation.
Within 24 randomly selected sentinel sites, home visitors performed a census, weekly demographic surveillance of births, deaths, and in- or out-migration, and weekly anthropometry on a sample of children. We evaluated the system through various methods including capture-recapture analysis and repeat census.
The system included 18,081 people at baseline. Over 32 weeks, the crude death rate was 1.0 (95% confidence interval [CI]: 0.8-1.2) deaths per 10,000 person-days (35 deaths per 1,000 person-years), with higher values during the rainy season. The under-5 death rate was approximately double. The prevalence of severe acute malnutrition (SAM) was 3.0% (95% CI: 2.3-4.0), almost half featuring kwashiorkor signs. The coverage of SAM treatment was 29.1%. The system detected >90% of deaths, and >90% of death reports appeared valid. However, demographic surveillance yielded discrepancies with the census and an implausible rate of population growth, while the predictive value of SAM classification was around 60%.
We found evidence of a chronic health crisis in this remote region. MSF's intervention coverage improved progressively. Mortality data appeared valid, but inaccuracies in population denominators and anthropometric measurements were noted. Similar systems could be implemented in other remote settings and acute emergencies, but with certain technical improvements.
Mortality; Malnutrition; Central African Republic; Surveillance; Rural; Humanitarian
Infertility is a significant disability, yet there are no reliable estimates of its global prevalence. Studies on infertility prevalence define the condition inconsistently, rendering the comparison of studies or quantitative summaries of the literature difficult. This study analyzed key components of infertility to develop a definition that can be consistently applied to globally available household survey data.
We proposed a standard definition of infertility and used it to generate prevalence estimates using 53 Demographic and Health Surveys (DHS). The analysis was restricted to the subset of DHS that contained detailed fertility information collected through the reproductive health calendar. We performed sensitivity analyses for key components of the definition and used these to inform our recommendations for each element of the definition.
Exposure type (couple status, contraceptive use, and intent), exposure time, and outcomes were key elements of the definition that we proposed. Our definition produced estimates that ranged from 0.6% to 3.4% for primary infertility and 8.7% to 32.6% for secondary infertility. Our sensitivity analyses showed that using an exposure measure of five years is less likely to misclassify fertile unions as infertile. Additionally, using a current, rather than continuous, measure of contraceptive use over five years resulted in a median relative error in secondary infertility of 20.7% (interquartile range of relative error [IQR]: 12.6%-26.9%), while not incorporating intent produced a corresponding error in secondary infertility of 58.2% (IQR: 44.3%-67.9%).
In order to estimate the global burden of infertility, prevalence estimates using a consistent definition need to be generated. Our analysis provided a recommended definition that could be applied to widely available global household data. We also summarized potential biases that should be considered when making estimates of infertility prevalence using household survey data.
Infertility; Demographic and health surveys; Population health; Prevalence
A comprehensive revision of the Global Burden of Disease (GBD) study is expected to be completed in 2012. This study utilizes a broad range of improved methods for assessing burden, including closer attention to empirically derived estimates of disability. The aim of this paper is to describe how GBD health states were derived for schizophrenia and bipolar disorder. These will be used in deriving health state-specific disability estimates. A literature review was first conducted to settle on a parsimonious set of health states for schizophrenia and bipolar disorder. A second review was conducted to investigate the proportion of schizophrenia and bipolar disorder cases experiencing these health states. These were pooled using a quality-effects model to estimate the overall proportion of cases in each state. The two schizophrenia health states were acute (predominantly positive symptoms) and residual (predominantly negative symptoms). The three bipolar disorder health states were depressive, manic, and residual. Based on estimates from six studies, 63% (38%-82%) of schizophrenia cases were in an acute state and 37% (18%-62%) were in a residual state. Another six studies were identified from which 23% (10%-39%) of bipolar disorder cases were in a manic state, 27% (11%-47%) were in a depressive state, and 50% (30%-70%) were in a residual state. This literature review revealed salient gaps in the literature that need to be addressed in future research. The pooled estimates are indicative only and more data are required to generate more definitive estimates. That said, rather than deriving burden estimates that fail to capture the changes in disability within schizophrenia and bipolar disorder, the derived proportions and their wide uncertainty intervals will be used in deriving disability estimates.
Global Burden of Disease; Health States; Schizophrenia; Bipolar Disorder
Although obesity is a risk factor for many chronic diseases, we have only limited knowledge of the magnitude of these associations in young adults. A multiethnic cohort of young adults was established to close current knowledge gaps; cohort demographics, cohort retention, and the potential influence of migration bias were investigated.
For this population-based cross-sectional study, demographics, and measured weight and height were extracted from electronic medical records of 1,929,470 patients aged 20 to 39 years enrolled in two integrated health plans in California from 2007 to 2009.
The cohort included about 84.4% of Kaiser Permanente California members in this age group who had a medical encounter during the study period and represented about 18.2% of the underlying population in the same age group in California. The age distribution of the cohort was relatively comparable to the underlying population in California Census 2010 population, but the proportion of women and ethnic/racial minorities was slightly higher. The three-year retention rate was 68.4%.
These data suggest the feasibility of our study for medium-term follow-up based on sufficient membership retention rates. While nationwide 6% of young adults are extremely obese, we know little to adequately quantify the health burden attributable to obesity, especially extreme obesity, in this age group. This cohort of young adults provides a unique opportunity to investigate associations of obesity-related factors and risk of cancer in a large multiethnic population.
Young adults; Obesity; Diabetes; Metabolic syndrome; Cancer; Epidemiology; Cohort study
Accurate measures of mortality level by age group, gender, and region are critical for health planning and evaluation. These are especially required for a country like Tonga, which has limited resources and works extensively with international donors. Mortality levels in Tonga were examined through an assessment of available published information and data available from the four routine death reporting systems currently in operation.
Available published data on infant mortality rate (IMR) and life expectancy (LE) in Tonga were sought through direct contact with the Government of Tonga and relevant international and regional organizations. Data sources were assessed for reliability and plausibility of estimates on the basis of method of estimation, original source of data, and data consistency. Unreliable sources were censored from further analysis and remaining data analysed for trends.
Mortality data for 2001 to 2009 were obtained from both the Health Information System (based on medical certificates of death) and the Civil Registry. Data from 2005 to 2009 were also obtained from the Reproductive Health System of the Ministry of Health (MoH) (based on community nursing reports), and for 2005–2008, data were also obtained from the Prime Minister’s office. Records were reconciled to create a single list of unique deaths and IMR and life tables calculated. Completeness of the reconciled data was examined using the Brass growth-balance method and capture-recapture analysis using two and three sources.
Published IMR estimates varied significantly through to the late 1990s when most estimates converge to a narrower range between 10 and 20 deaths per 1,000 live births. Findings from reconciled data were consistent with this range, and did not demonstrate any significant trend over 2001 to 2009.
Published estimates of LE from 2000 onwards varied from 65 to 75 years for males and 68 to 74 years for females, with most clustered around 70 to 71 for males and 72 to 73 for females. Reconciled empirical data for 2005 to 2009 produce an estimate of LE of 65.2 years (95% confidence interval [CI]: 64.6 - 65.8) for males and 69.6 years (95% CI: 69.0 – 70.2) for females, which are several years lower than published MoH and census estimates. Adult mortality (15 to 59 years) is estimated at 26.7% for males and 19.8% for females. Analysis of reporting completeness suggests that even reconciled data are under enumerated, and these estimates place the plausible range of LE between 60.4 to 64.2 years for males and 65.4 to 69.0 years for females, with adult mortality at 28.6% to 36.3% and 20.9% to 27.7%, respectively.
The level of LE at a relatively low IMR and high adult mortality suggests that non-communicable diseases are having a profound limiting effect on health status in Tonga. There has been a sustained history of incomplete and erroneous mortality estimates for Tonga. The findings highlight the critical need to reconcile existing data sources and integrate reporting systems more fully to ensure all deaths in Tonga are captured and the importance of local empirical data in monitoring trends in mortality.
Mortality; Cause of death; Noncommunicable diseases; Medical record review; Death certification; Tonga; Pacific Islands
This essay asks whether the global burden of diseases, injuries, and risk factors (GBD) should be measured in terms of their consequences for health, as maintained by most of those who are attempting to measure the GBD, or in terms of their consequences for well-being, as argued by John Broome. It answers that the burden of disease should be understood in terms of the consequences of disease for health, and it defends the wider efforts to measure health by those who are in other ways skeptical of the project of measuring the GBD.
Burden of disease; Well-being; Health measurement