Dataset Information

Estimating population cause-specific mortality fractions from in-hospital mortality: validation of a new method.

ABSTRACT:

Background

Cause-of-death data for many developing countries are not available. Information on deaths in hospital by cause is available in many low- and middle-income countries but is not a representative sample of deaths in the population. We propose a method to estimate population cause-specific mortality fractions (CSMFs) using data already collected in many middle-income and some low-income developing nations, yet rarely used: in-hospital death records.

Methods and findings

For a given cause of death, a community's hospital deaths are equal to total community deaths multiplied by the proportion of deaths occurring in hospital. If we can estimate the proportion dying in hospital, we can estimate the proportion dying in the population using deaths in hospital. We propose to estimate the proportion of deaths for an age, sex, and cause group that die in hospital from the subset of the population where vital registration systems function or from another population. We evaluated our method using nearly complete vital registration (VR) data from Mexico 1998-2005, which records whether a death occurred in a hospital. In this validation test, we used 45 disease categories. We validated our method in two ways: nationally and between communities. First, we investigated how the method's accuracy changes as we decrease the amount of Mexican VR used to estimate the proportion of each age, sex, and cause group dying in hospital. Decreasing VR data used for this first step from 100% to 9% produces only a 12% maximum relative error between estimated and true CSMFs. Even if Mexico collected full VR information only in its capital city with 9% of its population, our estimation method would produce an average relative error in CSMFs across the 45 causes of just over 10%. Second, we used VR data for the capital zone (Distrito Federal and Estado de Mexico) and estimated CSMFs for the three lowest-development states. Our estimation method gave an average relative error of 20%, 23%, and 31% for Guerrero, Chiapas, and Oaxaca, respectively.

Conclusions

Where accurate International Classification of Diseases (ICD)-coded cause-of-death data are available for deaths in hospital and for VR covering a subset of the population, we demonstrated that population CSMFs can be estimated with low average error. In addition, we showed in the case of Mexico that this method can substantially reduce error from biased hospital data, even when applied to areas with widely different levels of development. For countries with ICD-coded deaths in hospital, this method potentially allows the use of existing data to inform health policy.

SUBMITTER: Murray CJ

PROVIDER: S-EPMC2080647 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND:Almost all countries without complete vital registration systems have data on deaths collected by hospitals. However, these data have not been widely used to estimate cause of death (COD) patterns in populations because only a non-representative fraction of people in these countries die in health facilities. Methods that can exploit hospital mortality statistics to reliably estimate community COD patterns are required to strengthen the evidence base for disease and injury control programs. We propose a method that weights hospital-certified causes by the probability of death to estimate population cause-specific mortality fractions (CSMFs). METHODS:We used an established verbal autopsy instrument (VAI) to collect data from hospital catchment areas in Chandpur and Comilla Districts, Bangladesh, and Bohol province, the Philippines, between 2011 and 2014, along with demographic covariates for each death. Hospital medical certificates of cause of death (death certificates) were collected and mapped to the corresponding cause categories of the VAI. Tariff 2.0 was used to assign a COD for community deaths. Logistic regression models were created for broad causes in each country to calculate the probability of in-hospital death, given a set of covariate values. The reweighted CSMFs for deaths in the hospital catchment population, represented by each hospital death, were calculated from the corresponding regression models. RESULTS:We collected data on 4228 adult deaths in the Philippines and 3725 deaths in Bangladesh. Short time to hospital and education were consistently associated with in-hospital death in the Philippines and absence of a disability was consistently associated with in-hospital death in Bangladesh. Non-communicable diseases (excluding stroke) and stroke were the leading causes of death in both the Philippines (33.9%, 19.1%) and Bangladesh (46.1%, 21.1%) according to the reweighted method. The reweighted method generally estimated CSMFs that fell between those derived from hospitals and those diagnosed by Tariff 2.0. CONCLUSIONS:Statistical methods can be used to derive estimates of cause-specific probability of death in-hospital for Bangladesh and the Philippines to generate population CSMFs. In regions where hospital death certification is of reasonable quality and routine verbal autopsy is not applied, these estimates could be applied to generate cost-effective and robust CSMFs for the population.

Project description:BackgroundVerbal autopsy (VA) is used to estimate the causes of death in areas with incomplete vital registration systems. The King and Lu method (KL) for direct estimation of cause-specific mortality fractions (CSMFs) from VA studies is an analysis technique that estimates CSMFs in a population without predicting individual-level cause of death as an intermediate step. In previous studies, KL has shown promise as an alternative to physician-certified verbal autopsy (PCVA). However, it has previously been impossible to validate KL with a large dataset of VAs for which the underlying cause of death is known to meet rigorous clinical diagnostic criteria.MethodsWe applied the KL method to adult, child, and neonatal VA datasets from the Population Health Metrics Research Consortium gold standard verbal autopsy validation study, a multisite sample of 12,542 VAs where gold standard cause of death was established using strict clinical diagnostic criteria. To emulate real-world populations with varying CSMFs, we evaluated the KL estimations for 500 different test datasets of varying cause distribution. We assessed the quality of these estimates in terms of CSMF accuracy as well as linear regression and compared this with the results of PCVA.ResultsKL performance is similar to PCVA in terms of CSMF accuracy, attaining values of 0.669, 0.698, and 0.795 for adult, child, and neonatal age groups, respectively, when health care experience (HCE) items were included. We found that the length of the cause list has a dramatic effect on KL estimation quality, with CSMF accuracy decreasing substantially as the length of the cause list increases. We found that KL is not reliant on HCE the way PCVA is, and without HCE, KL outperforms PCVA for all age groups.ConclusionsLike all computer methods for VA analysis, KL is faster and cheaper than PCVA. Since it is a direct estimation technique, though, it does not produce individual-level predictions. KL estimates are of similar quality to PCVA and slightly better in most cases. Compared to other recently developed methods, however, KL would only be the preferred technique when the cause list is short and individual-level predictions are not needed.

Project description:BackgroundIn recent years, national and global mortality data have been characterized in terms of well-established risk factors. In this regard, alcohol consumption has been called the third leading "actual cause of death" (modifiable behavioral risk factor) in the United States, after tobacco use and the combination of poor diet and physical inactivity. Globally and in various regions of the world, alcohol use has been established as a leading contributor to the overall burden of disease and as a major determinant of health disparities, but, to our knowledge, no one has characterized alcohol-related harm in such broad terms at the local level. We asked how alcohol-related premature mortality in San Francisco, measured in years of life lost (YLLs), compares with other well-known causes of premature mortality, such as ischemic heart disease or HIV/AIDS.MethodsWe applied sex- and cause-specific population-attributable fractions (PAFs) of years of life lost (YLLs) from the Global Burden of Disease Study to 17 comparable outcomes among San Francisco males and females during 2004-2007. We did this in three ways: Method 1 assumed that all San Franciscans drink like populations in developed economies. These estimates were limited to alcohol-related harm. Method 2 modified these estimates by including several beneficial effects. Method 3 assumed that Latino and Asian San Franciscans drink alcohol like populations in the global regions related to their ethnicity.ResultsBy any of these three methods, alcohol-related premature mortality accounts for roughly a tenth of all YLLs among males. Alcohol-related YLLs among males are comparable to YLLs for leading causes such as ischemic heart disease and HIV/AIDS, in some instances exceeding them. Latino and black males bear a disproportionate burden of harm. Among females, for whom estimates differed more by method and were smaller than those for males, alcohol-related YLLs are comparable to leading causes which rank somewhere between fifth and fourteenth.ConclusionsAlcohol consumption is a major contributor to premature mortality in San Francisco, especially among males. Interventions to avert alcohol-related harm in San Francisco should be taken at the population level and deserve the same attention that is given to other major risk factors, such as smoking or obesity.

Project description:AimsCardiorespiratory fitness (CRF) is a key predictor of chronic disease, particularly cardiovascular disease (CVD), but its assessment usually requires exercise testing which is impractical and costly in most health-care settings. Non-exercise testing cardiorespiratory fitness (NET-F)-estimating methods are a less resource-demanding alternative, but their predictive capacity for CVD and total mortality has yet to be tested. The objective of this study is to examine the association of a validated NET-F algorithm with all-cause and CVD mortality.Methods and resultsThe participants were 32,319 adults (14,650 men) aged 35-70 years who took part in eight Health Survey for England and Scottish Health Survey studies between 1994 and 2003. Non-exercise testing cardiorespiratory fitness (a metabolic equivalent of VO2max) was calculated using age, sex, body mass index (BMI), resting heart rate, and self-reported physical activity. We followed participants for mortality until 2008. Two thousand one hundred and sixty-five participants died (460 cardiovascular deaths) during a mean 9.0 [standard deviation (SD) = 3.6] year follow-up. After adjusting for potential confounders including diabetes, hypertension, smoking, social class, alcohol, and depression, a higher fitness score according to the NET-F was associated with a lower risk of mortality from all-causes (hazard ratio per SD increase in NET-F 0.85, 95% confidence interval: 0.78-0.93 in men; 0.88, 0.80-0.98 in women) and CVD (men: 0.75, 0.63-0.90; women: 0.73, 0.60-0.92). Non-exercise testing cardiorespiratory fitness had a better discriminative ability than any of its components (CVD mortality c-statistic: NET-F = 0.70-0.74; BMI = 0.45-0.59; physical activity = 0.60-0.64; resting heart rate = 0.57-0.61). The sensitivity of the NET-F algorithm to predict events occurring in the highest risk quintile was better for CVD (0.49 in both sexes) than all-cause mortality (0.44 and 0.40 for men and women, respectively). The specificity for all-cause and CVD mortality ranged between 0.80 and 0.82. The net reclassification improvement of CVD mortality risk (vs. a standardized aggregate score of the modifiable components of NET-F) was 27.2 and 21.0% for men and women, respectively.ConclusionThe CRF-estimating method NET-F that does not involve exercise testing showed consistent associations with all-cause and cardiovascular mortality, and it had good discrimination and excellent risk reclassification improvement. As such, it merits further attention as a practical and potentially and useful risk prediction tool.

Project description:BackgroundTrends in cause-specific mortality in most African countries are currently estimated from epidemiological models because the coverage of the civil registration system is low and national statistics on causes of death are unreliable at the national level. We aim to evaluate the performance of the death notification system in Antananarivo, the capital city of Madagascar, to inform cause-of-death statistics.MethodsInformation on the sex of the deceased, dates of birth and death, and underlying cause of death were transcribed from death registers maintained in Antananarivo. Causes of death were coded in ICD-9 and mapped to cause categories from the Global Burden of Disease 2016 Study (GBD). The performance of the notification system was assessed based on the Vital Statistics Performance Index, including six dimensions: completeness of death registration, quality of cause of death reporting, quality of age and sex reporting, internal consistency, level of cause-specific detail, and data availability and timeliness. We redistributed garbage codes and compared cause-specific mortality fractions in death records and estimates from the GBD with concordance correlation coefficients.ResultsThe death notification system in Antananarivo performed well on most dimensions, although 31% of all deaths registered over the period 1976-2015 were assigned to ICD codes considered as "major garbage codes" in the GBD 2016. The completeness of death notification, estimated with indirect demographic techniques, was higher than 90% in the period 1975-1993, and recent under-five mortality rates were consistent with estimates from Demographic and Health Surveys referring to the capital city. After redistributing garbage codes, cause-specific mortality fractions derived from death notification data were consistent with GBD 2016 for the whole country in the 1990s, with concordance correlation coefficients higher than 90%. There were larger deviations in recent years, with concordance correlation coefficients in 2015 at 0.74 (95% CI 0.66-0.81) for men and 0.81 (95% CI 0.74-0.86) for women.ConclusionsDeath notification in Antananarivo is a low-cost data source allowing real-time mortality monitoring, with a potential to improve disease burden estimates. Further efforts should be directed towards evaluating data quality in urban centers in Madagascar and other African countries to fill important data gaps on causes of death.