Population Health Metrics Research Consortium gold standard verbal autopsy validation study: design, implementation, and development of analysis datasets

Risk factors for incident cardiovascular events and their population attributable fractions in rural India: The Rishi Valley Prospective Cohort Study

OBJECTIVE

We prospectively determined incident cardiovascular events and their association with risk factors in rural India.

METHODS

We followed up with 7935 adults from the Rishi Valley Prospective Cohort Study to identify incident cardiovascular events. Using Cox proportional hazards regression, we estimated hazard ratios (HRs) with 95% confidence intervals (95% CI) for associations between potential risk factors and cardiovascular events. Population attributable fractions (PAFs) for risk factors were estimated using R ('averisk' package).

RESULTS

Of the 4809 participants without prior cardiovascular disease, 57.7% were women and baseline mean age was 45.3 years. At follow-up (median of 4.9 years, 23,180 person-years [PYs]), 202 participants developed cardiovascular events, equating to an incidence of 8.7 cardiovascular events/1000 PYs. Incidence was greater in those with hypertension (hazard ratio [HR] [95% CI] 1.73 [1.21-2.49], adjusted PAF 18%), diabetes (1.96 [1.15-3.36], 4%) or central obesity (1.77 [1.23, 2.54], 9%) which together accounted for 31% of the PAF. Non-traditional risk factors such as night sleeping hours and number of children accounted for 16% of the PAF.

CONCLUSIONS

Both traditional and non-traditional cardiovascular risk factors are important contributors to incident cardiovascular events in rural India. Interventions targeted to these factors could assist in reducing the incidence of cardiovascular events.

Tree-informed Bayesian multi-source domain adaptation: cross-population probabilistic cause-of-death assignment using verbal autopsy

Determining causes of deaths (CODs) occurred outside of civil registration and vital statistics systems is challenging. A technique called verbal autopsy (VA) is widely adopted to gather information on deaths in practice. A VA consists of interviewing relatives of a deceased person about symptoms of the deceased in the period leading to the death, often resulting in multivariate binary responses. While statistical methods have been devised for estimating the cause-specific mortality fractions (CSMFs) for a study population, continued expansion of VA to new populations (or "domains") necessitates approaches that recognize between-domain differences while capitalizing on potential similarities. In this article, we propose such a domain-adaptive method that integrates external between-domain similarity information encoded by a prespecified rooted weighted tree. Given a cause, we use latent class models to characterize the conditional distributions of the responses that may vary by domain. We specify a logistic stick-breaking Gaussian diffusion process prior along the tree for class mixing weights with node-specific spike-and-slab priors to pool information between the domains in a data-driven way. The posterior inference is conducted via a scalable variational Bayes algorithm. Simulation studies show that the domain adaptation enabled by the proposed method improves CSMF estimation and individual COD assignment. We also illustrate and evaluate the method using a validation dataset. The article concludes with a discussion of limitations and future directions.

CKD is the Major Cause of Death in Uddanam: A Population-Representative Study Using Smart Verbal Autopsy

Introduction

Uddanam is an agricultural area with a high burden of chronic kidney disease of unknown etiology (CKDu). Despite reports of many deaths due to CKD in the lay press, the exact contribution of CKD to deaths remains uncertain because most deaths occur outside medical care.

Methods

We used SmartVA automated verbal autopsy tool to ascertain the cause-specific mortality fractions among a 2419 subject-strong general population cohort of adult subjects in Uddanam between 2018 and 2022. Verbal autopsy interviews were conducted twice with the family members of the deceased.

Results

A total of 133 deaths were recorded, giving a crude death rate of 5.5%, 10 times higher than that recorded in national surveys. CKD was responsible for 45% of all deaths, followed by ischemic heart disease (15%) and respiratory disease (6%).

Conclusion

This study confirms CKD as the leading cause of mortality in this high CKD burden area and provides crucial data for public health decision-making and resource allocation.

Agreement between cause of death assignment by computer-coded verbal autopsy methods and physician coding of verbal autopsy interviews in South Africa

BACKGROUND

The South African national cause of death validation (NCODV 2017/18) project collected a national sample of verbal autopsies (VA) with cause of death (COD) assignment by physician-coded VA (PCVA) and computer-coded VA (CCVA).

OBJECTIVE

The performance of three CCVA algorithms (InterVA-5, InSilicoVA and Tariff 2.0) in assigning a COD was compared with PCVA (reference standard).

METHODS

Seven performance metrics assessed individual and population level agreement of COD assignment by age, sex and place of death subgroups. Positive predictive value (PPV), sensitivity, overall agreement, kappa, and chance corrected concordance (CCC) assessed individual level agreement. Cause-specific mortality fraction (CSMF) accuracy and Spearman's rank correlation assessed population level agreement.

RESULTS

A total of 5386 VA records were analysed. PCVA and CCVAs all identified HIV/AIDS as the leading COD. CCVA PPV and sensitivity, based on confidence intervals, were comparable except for HIV/AIDS, TB, maternal, diabetes mellitus, other cancers, and some injuries. CCVAs performed well for identifying perinatal deaths, road traffic accidents, suicide and homicide but poorly for pneumonia, other infectious diseases and renal failure. Overall agreement between CCVAs and PCVA for the top single cause (48.2-51.6) indicated comparable weak agreement between methods. Overall agreement, for the top three causes showed moderate agreement for InterVA (70.9) and InSilicoVA (73.8). Agreement based on kappa (-0.05-0.49)and CCC (0.06-0.43) was weak to none for all algorithms and groups. CCVAs had moderate to strong agreement for CSMF accuracy, with InterVA-5 highest for neonates (0.90), Tariff 2.0 highest for adults (0.89) and males (0.84), and InSilicoVA highest for females (0.88), elders (0.83) and out-of-facility deaths (0.85). Rank correlation indicated moderate agreement for adults (0.75-0.79).

CONCLUSIONS

Whilst CCVAs identified HIV/AIDS as the leading COD, consistent with PCVA, there is scope for improving the algorithms for use in South Africa.

Estimating causes of community death of adults in Myanmar from a nationwide population sample: Application of verbal autopsy

In Myanmar 84% of deaths occur in the community, of which half are unregistered and none have a reliable cause of death (COD) recorded. Since 2018, Myanmar has introduced improved registration practices and verbal autopsy (VA) to assess whether such methods can produce policy relevant information on community COD. Community health midwives and public health supervisors grade II collected VAs on over 80,000 deaths which occurred between January 2018 and December 2019 in a nationwide sample of 42 townships in Myanmar. Electronic methods were used to collect and consolidate data. The most probable COD was assigned using the SmartVA Analyze 2.0 computer algorithm. Completeness of VA death reporting increased to 71% in 2019. Most adult (12+ years) deaths (82%) were due to non-communicable diseases, primarily stroke, ischemic heart disease and chronic respiratory disease, for both men and women. VA results were consistent with Global Burden of Disease (GBD) Study estimates, except for cirrhosis in men, which was more common, and had a younger age distribution of death than the GBD. Large scale implementation of improved death registration practices and COD diagnosis using VA is feasible and provides plausible, timely, disaggregated and policy relevant information on the leading causes of community death. Addressing the burden of non-communicable diseases, particularly cirrhosis in young men, is an important public health priority in Myanmar. Improving completeness of VA death reporting in poorly performing townships and in neonates, children and women will further improve the policy utility of the VA data.

Cause of Death estimation from Verbal Autopsies: Is the Open Response redundant or synergistic?

Civil registration and vital statistics systems capture birth and death events to compile vital statistics and to provide legal rights to citizens. Vital statistics are a key factor in promoting public health policies and the health of the population. Medical certification of cause of death is the preferred source of cause of death information. However, two thirds of all deaths worldwide are not captured in routine mortality information systems and their cause of death is unknown. Verbal autopsy is an interim solution for estimating the cause of death distribution at the population level in the absence of medical certification. A Verbal Autopsy (VA) consists of an interview with the relative or the caregiver of the deceased. The VA includes both Closed Questions (CQs) with structured answer options, and an Open Response (OR) consisting of a free narrative of the events expressed in natural language and without any pre-determined structure. There are a number of automated systems to analyze the CQs to obtain cause specific mortality fractions with limited performance. We hypothesize that the incorporation of the text provided by the OR might convey relevant information to discern the CoD. The experimental layout compares existing Computer Coding Verbal Autopsy methods such as Tariff 2.0 with other approaches well suited to the processing of structured inputs as is the case of the CQs. Next, alternative approaches based on language models are employed to analyze the OR. Finally, we propose a new method with a bi-modal input that combines the CQs and the OR. Empirical results corroborated that the CoD prediction capability of the Tariff 2.0 algorithm is outperformed by our method taking into account the valuable information conveyed by the OR. As an added value, with this work we made available the software to enable the reproducibility of the results attained with a version implemented in R to make the comparison with Tariff 2.0 evident.

Verbal autopsy to assess causes of mortality among the economically productive age group in the tribal region of Melghat, central India

Background & Objectives

Verbal autopsy (VA) is the systematic and retrospective inquiry (from relatives) about the symptoms of an illness prior to death. In tribal India, 67-75 per cent of deaths occur at home with an unknown cause of death (CoD). Hence, the aim of this study was to determine the CoD in the 16-60 yr age group utilizing VA.

Methods

A prospective, community based longitudinal study was conducted in 32 tribal villages in the Melghat region of Maharashtra, between 2004 and 2020. Number of deaths and VAs in 16-60 yr age group were collected by village health workers (VHWs) and supervisors, verified by five different persons (internal-external) and cross-checked by three VA interpretation trained physicians. A modified version of WHO VA was used. Cause-specific mortality fractions were calculated.

Results

Of the 1011 deaths recorded, mortality in males was significantly higher than females (P<0.001). A total of 763 VAs were conducted which revealed that tuberculosis was the leading CoD, followed by jaundice, heart diseases, diarrhoea, central nervous system infections and suicide. Suicides were significantly more common among males than in females (P=0.046). Significantly, more deaths occurred during the monsoon (P=0.002), especially diarrhoeal deaths (P=0.024).

Interpretation & conclusions

The findings of this study suggest that, in Indian tribal areas, infectious diseases are the leading causes of morbidity and one of the major causes of deaths in economically productive age group. Intensified VHW-mediated interventions are required to reduce the premature deaths.

Evaluation of methods for assigning causes of death from verbal autopsies in India

Background

Physician-coded verbal autopsy (PCVA) is the most widely used method to determine causes of death (COD) in countries where medical certification of death is low. Computer-coded verbal autopsy (CCVA), an alternative method to PCVA for assigning the COD is considered to be efficient and cost-effective. However, the performance of CCVA as compared to PCVA is yet to be established in the Indian context.

Methods

We evaluated the performance of PCVA and three CCVA methods i.e., InterVA 5, InSilico, and Tariff 2.0 on verbal autopsies done using the WHO 2016 VA tool on 2,120 reference standard cases developed from five tertiary care hospitals of Delhi. PCVA methodology involved dual independent review with adjudication, where required. Metrics to assess performance were Cause Specific Mortality Fraction (CSMF), sensitivity, positive predictive value (PPV), CSMF Accuracy, and Kappa statistic.

Results

In terms of the measures of the overall performance of COD assignment methods, for CSMF Accuracy, the PCVA method achieved the highest score of 0.79, followed by 0.67 for Tariff_2.0, 0.66 for Inter-VA and 0.62 for InSilicoVA. The PCVA method also achieved the highest agreement (57%) and Kappa scores (0.54). The PCVA method showed the highest sensitivity for 15 out of 20 causes of death.

Conclusion

Our study found that the PCVA method had the best performance out of all the four COD assignment methods that were tested in our study sample. In order to improve the performance of CCVA methods, multicentric studies with larger sample sizes need to be conducted using the WHO VA tool.

Assessment of the underlying causes of adult deaths using a short version of verbal autopsy in Xaiyabouli Province, Lao People's Democratic Republic

BACKGROUND

In developing countries, it is difficult to collect the data of the underlying cause of death (UCOD), especially when a death does not occur in a health facility. This study aimed to develop a short version of verbal autopsy (VA) and identify the UCOD of adults in Lao People's Democratic Republic (Lao PDR).

METHODS

A short version of VA for deaths outside health facilities was developed. This study included all deaths of people aged 15 years old or older in Xaiyabouli Province in 2020. Socio-demographic factors, place of death, and UCOD of the deceased were collected from health facilities or from family members using a questionnaire including the short VA form. UCOD was compared between home deaths and hospital deaths, between the age group of 15-59 years old and the age group ≥ 60 years old, and between males and females.

RESULTS

Of all the 1,235 deaths included in this study, 1,012 deaths (81.9%) occured at home and 223 deaths (18.1%) at hospitals. The most common UCOD was senility (13.3%), followed by heart/renal failure (10.5%), pneumonia (9.6%) and traffic accident (7.1%). Compared to hospital deaths, home deaths had more people who were females, 75 years old or older, and Lao-Tai. Home deaths had more deaths than hospital deaths due to accident/injury (16.0% vs. 8.1%), tumor (4.7% vs. 1.8%), and senility (16.2% vs. 0%); fewer deaths due to heart/renal disease (15.1% vs. 32.3%), respiratory disease (12.2% vs. 18.8%), liver/gastro-intestine disease (5.3% vs. 9.0%), and infection (3.1% vs. 14.3%). The age group of 15-59 years had more deaths in the categories of accident/injury (28.1% vs. 4.4%), liver/gastro-intestine disease (8.1% vs. 4.4%), infection (7.2% vs. 3.5%), and tumor (6.0% vs. 2.8%). Males had more deaths due to tumor (5.2% vs. 3.0%) and fewer natural deaths (11.2% vs. 15.9%) than females.

CONCLUSIONS

The major UCOD category was heart/renal disease in the adult generation in Xaiyabouli Province. Cost-effective interventions based on the multisectoral noncommunicable disease prevention plan should be appropriately implemented. Mortality surveillance using the short VA tool should be conducted for all home deaths in Lao PDR.

Leading causes of deaths in the mortality transition in Papua New Guinea: evidence from the Comprehensive Health and Epidemiological Surveillance System

Background

Changing causes of deaths in the mortality transition in Papua New Guinea (PNG) are poorly understood. This study analysed community-level data to identify leading causes of death in the population and variations across age groups and sexes, urban-rural sectors and provinces.

Method

Mortality surveillance data were collected from 2018–20 as part of the Comprehensive Health and Epidemiological Surveillance System (CHESS), using the World Health Organization 2016 verbal autopsy (VA) instrument. Data from 926 VA interviews were analysed, using the InterVA-5 cause of death analytical tool to assign specific causes of death among children (0–14 years), those of working age (15–64 years) and the elderly (65+ years).

Result

Nearly 50% of the total deaths were attributed to non-communicable diseases (NCDs), followed by infectious and parasitic diseases (35%), injuries and external causes (11%) and maternal and neonatal deaths (4%). Leading causes of death among children were acute respiratory tract infections (ARTIs) and diarrhoeal diseases, each contributing to 13% of total deaths. Among the working population, tuberculosis (TB) contributed to 12% of total deaths, followed by HIV/AIDS (11%). TB- and HIV/AIDS-attributed deaths were highest in the age group 25–34 years, at 20% and 18%, respectively. These diseases killed more females of working age (n = 79, 15%) than males (n = 52, 8%). Among the elderly, the leading causes of death were ARTIs (13%) followed by digestive neoplasms (10%) and acute cardiac diseases (9%).

Conclusion

The variations in leading causes of death across the populations in PNG suggest diversity in mortality transition. This requires different strategies to address specific causes of death in particular populations.

Cohort profile: Research on Obesity and Diabetes among African Migrants in Europe and Africa Prospective (RODAM-Pros) cohort study

PURPOSE

The Research on Obesity and Diabetes among African Migrants (RODAM) prospective (RODAM-Pros) cohort study was established to identify key changes in environmental exposures and epigenetic modifications driving the high burden of cardiovascular disease (CVD) risk among sub-Saharan African migrants.

PARTICIPANTS

All the participants in the RODAM cross-sectional study that completed the baseline assessment (n=5114) were eligible for the follow-up of which 2165 participants (n=638 from rural-Ghana, n=608 from urban-Ghana, and n=919 Ghanaian migrants in Amsterdam, the Netherlands) were included in the RODAM-Pros cohort study. Additionally, we included a subsample of European-Dutch (n=2098) to enable a comparison to be made between Ghanaian migrants living in the Netherlands and the European-Dutch host population.

FINDINGS TO DATE

Follow-up data have been collected on demographics, socioeconomic status, medical history, psychosocial environment, lifestyle factors, nutrition, anthropometrics, blood pressure, fasting blood, urine and stool samples. Biochemical analyses included glucose metabolism, lipid profile, electrolytes and renal function, liver metabolism and inflammation. In a subsample, we assessed DNA methylation patterns using Infinium 850K DNA Methylation BeadChip. Baseline results indicated that migrants have higher prevalence of CVD risk factors than non-migrants. Epigenome-wide association studies suggest important differences in DNA methylation between migrants and non-migrants. The follow-up study will shed further light on key-specific environmental exposures and epigenetic modifications contributing to the high burden of CVD risk among sub-Saharan African migrants.

FUTURE PLANS

Follow-up is planned at 5-year intervals, baseline completed in 2015 and first follow-up completed in 2021.

The openVA Toolkit for Verbal Autopsies

Verbal autopsy (VA) is a survey-based tool widely used to infer cause of death (COD) in regions without complete-coverage civil registration and vital statistics systems. In such settings, many deaths happen outside of medical facilities and are not officially documented by a medical professional. VA surveys, consisting of signs and symptoms reported by a person close to the decedent, are used to infer the COD for an individual, and to estimate and monitor the COD distribution in the population. Several classification algorithms have been developed and widely used to assign causes of death using VA data. However, the incompatibility between different idiosyncratic model implementations and required data structure makes it difficult to systematically apply and compare different methods. The openVA package provides the first standardized framework for analyzing VA data that is compatible with all openly available methods and data structure. It provides an open-source, R implementation of several most widely used VA methods. It supports different data input and output formats, and customizable information about the associations between causes and symptoms. The paper discusses the relevant algorithms, their implementations in R packages under the openVA suite, and demonstrates the pipeline of model fitting, summary, comparison, and visualization in the R environment.

Global, regional, and national burden and attributable risk factors of neurological disorders: The Global Burden of Disease study 1990-2019

Background

Neurological disorders are a major and increasing global health challenge, which accounts for a substantial portion of the disease burden worldwide. The aim of this systematic analysis is to present the most comprehensive and up-to-date estimates of disease burden, epidemiological trends, and attributable risk factors of neurological disorders at global, regional, and national levels.

Methods

We extracted data of 18 neurological disorders from the Global Burden of Disease 2019 study database. The burden of neurological disorders was measured using the incidence, prevalence, mortality, and disability-adjusted life years (DALYs), and further described according to age, sex, year, geographical location and socio-demographic Index (SDI). All estimates were presented with corresponding 95% uncertainty intervals (UIs).

Findings

Globally, in 2019, there were nearly 10 million deaths and 349 million DALYs due to neurological disorders. Among the 18 neurological disorders, stroke was the biggest contributor to DALYs (143232.18 [95%UI 133095.81-153241.82] in thousands) and deaths (6552.72 [95%UI 5995.20-7015.14] in thousands), followed by neonatal encephalopathy due to birth asphyxia and trauma. From 1990 to 2019, the DALYs of neurological diseases belonging to the communicable, maternal, neonatal and nutritional categories showed a sharp decrease, while Alzheimer's disease and other dementias and Parkinson's disease showed a large increase. Neurological disorders exhibited different profiles in different regions and age groups. A significant correlation between the SDI and the age-standardized DALY rates was also found except for Alzheimer's disease and other dementias. In addition, risk factors such as high systolic blood pressure, low birth weight and short gestation period, and metabolic risk contribute significantly to neurological disorders.

Interpretation

The overall burden of neurological disorders has increased from 1990 to 2019, especially for non-communicable neurological disorders. The substantial variations of burden across regions emphasize the need for region-specific interventional strategies and allocation of resources based on priorities.

Performance evaluation of machine learning and Computer Coded Verbal Autopsy (CCVA) algorithms for cause of death determination: A comparative analysis of data from rural South Africa

Computer Coded Verbal Autopsy (CCVA) algorithms are commonly used to determine the cause of death (CoD) from questionnaire responses extracted from verbal autopsies (VAs). However, they can only operate on structured data and cannot effectively harness information from unstructured VA narratives. Machine Learning (ML) algorithms have also been applied successfully in determining the CoD from VA narratives, allowing the use of auxiliary information that CCVA algorithms cannot directly utilize. However, most ML-based studies only use responses from the structured questionnaire, and the results lack generalisability and comparability across studies. We present a comparative performance evaluation of ML methods and CCVA algorithms on South African VA narratives data, using data from Agincourt Health and Demographic Surveillance Site (HDSS) with physicians' classifications as the gold standard. The data were collected from 1993 to 2015 and have 16,338 cases. The random forest and extreme gradient boosting classifiers outperformed the other classifiers on the combined dataset, attaining accuracy of 96% respectively, with significant statistical differences in algorithmic performance (p < 0.0001). All our models attained Area Under Receiver Operating Characteristics (AUROC) of greater than 0.884. The InterVA CCVA attained 83% Cause Specific Mortality Fraction accuracy and an Overall Chance-Corrected Concordance of 0.36. We demonstrate that ML models could accurately determine the cause of death from VA narratives. Additionally, through mortality trends and pattern analysis, we discovered that in the first decade of the civil registration system in South Africa, the average life expectancy was approximately 50 years. However, in the second decade, life expectancy significantly dropped, and the population was dying at a much younger average age of 40 years, mostly from the leading HIV related causes. Interestingly, in the third decade, we see a gradual improvement in life expectancy, possibly attributed to effective health intervention programmes. Through a structure and semantic analysis of narratives where experts disagree, we also demonstrate the most frequent terms of traditional healer consultations and visits. The comparative approach also makes this study a baseline that can be used for future research enforcing generalization and comparability. Future study will entail exploring deep learning models for CoD classification.

Assessing the Diagnostic Accuracy of Physicians for Home Death Certification in Shanghai: Application of SmartVA

Approximately 30% of deaths in Shanghai either occur at home or are not medically attended. The recorded cause of death (COD) in these cases may not be reliable. We applied the Smart Verbal Autopsy (VA) tool to assign the COD for a representative sample of home deaths certified by 16 community health centers (CHCs) from three districts in Shanghai, from December 2017 to June 2018. The results were compared with diagnoses from routine practice to ascertain the added value of using SmartVA. Overall, cause-specific mortality fraction (CSMF) accuracy improved from 0.93 (93%) to 0.96 after the application of SmartVA. A comparison with a “gold standard (GS)” diagnoses obtained from a parallel medical record review investigation found that 86.3% of the initial diagnoses made by the CHCs were assigned the correct COD, increasing to 90.5% after the application of SmartVA. We conclude that routine application of SmartVA is not indicated for general use in CHCs, although the tool did improve diagnostic accuracy for residual causes, such as other or ill-defined cancers and non-communicable diseases.

Validation of physician certified verbal autopsy using conventional autopsy: a large study of adult non-external causes of death in a metropolitan area in Brazil

Background

Reliable mortality data are essential for the development of public health policies. In Brazil, although there is a well-consolidated universal system for mortality data, the quality of information on causes of death (CoD) is not even among Brazilian regions, with a high proportion of ill-defined CoD. Verbal autopsy (VA) is an alternative to improve mortality data. This study aimed to evaluate the performance of an adapted and reduced version of VA in identifying the underlying causes of non-forensic deaths, in São Paulo, Brazil. This is the first time that a version of the questionnaire has been validated considering the autopsy as the gold standard.

Methods

The performance of a physician-certified verbal autopsy (PCVA) was evaluated considering conventional autopsy (macroscopy plus microscopy) as gold standard, based on a sample of 2060 decedents that were sent to the Post-Mortem Verification Service (SVOC-USP). All CoD, from the underlying to the immediate, were listed by both parties, and ICD-10 attributed by a senior coder. For each cause, sensitivity and chance corrected concordance (CCC) were computed considering first the underlying causes attributed by the pathologist and PCVA, and then any CoD listed in the death certificate given by PCVA. Cause specific mortality fraction accuracy (CSMF-accuracy) and chance corrected CSMF-accuracy were computed to evaluate the PCVA performance at the populational level.

Results

There was substantial variability of the sensitivities and CCC across the causes. Well-known chronic diseases with accurate diagnoses that had been informed by physicians to family members, such as various cancers, had sensitivities above 40% or 50%. However, PCVA was not effective in attributing Pneumonia, Cardiomyopathy and Leukemia/Lymphoma as underlying CoD. At populational level, the PCVA estimated cause specific mortality fractions (CSMF) may be considered close to the fractions pointed by the gold standard. The CSMF-accuracy was 0.81 and the chance corrected CSMF-accuracy was 0.49.

Conclusions

The PCVA was efficient in attributing some causes individually and proved effective in estimating the CSMF, which indicates that the method is useful to establish public health priorities.

Are cause of death data for Shanghai fit for purpose? A retrospective study of medical records

OBJECTIVES

To assess the quality of cause of death reporting in Shanghai for both hospital and home deaths.

DESIGN AND SETTING

Medical records review (MRR) to independently establish a reference data set against which to compare original and adjusted diagnoses from a sample of three tertiary hospitals, one secondary level hospital and nine community health centres in Shanghai.

PARTICIPANTS

1757 medical records (61% males, 39% females) of deaths that occurred in these sample sites in 2017 were reviewed using established diagnostic standards.

INTERVENTIONS

None.

PRIMARY OUTCOME

Original underlying cause of death (UCOD) from medical facilities.

SECONDARY OUTCOME

Routine UCOD assigned from the Shanghai Civil Registration and Vital Statistics (CRVS) system and MRR UCODs from MRR.

RESULTS

The original UCODs as assigned by doctors in the study facilities were of relatively low quality, reduced to 31% of deaths assigned to garbage codes, reduced to 2.3% following data quality and follow back procedures routinely applied by the Shanghai CRVS system. The original UCOD had lower chance-corrected concordance and cause-specific mortality fraction accuracy of 0.57 (0.44, 0.70) and 0.66, respectively, compared with 0.75 (0.66, 0.85) and 0.96, respectively, after routine data checking procedures had been applied.

CONCLUSIONS

Training in correct death certification for clinical doctors, especially tertiary hospital doctors, is essential to improve UCOD quality in Shanghai. A routine quality control system should be established to actively track diagnostic performance and provide feedback to individual doctors or facilities as needed.

Assessment of the quality of the vital registration system for under-5 mortality in Yucatan, Mexico

Introduction

Vital registration is an important element in health information systems which can inform policy and strengthen health systems. Mexico has a well-functioning vital registration system; however, there is still room for improvement, especially for deaths of children under 5. This study assesses the quality of the vital registration system in capturing deaths and evaluates the quality of cause of death certification in under-5 deaths in Yucatan, Mexico.

Methods

We collected information on under-5 deaths that occurred in 2015 and 2016 in Yucatan, Mexico. We calculated the Vital Statistics Performance Index (VSPI) to have a general assessment of the vital registration performance. We examined the agreement between vital registration records and medical records at the individual and population levels using the chance-corrected concordance (CCC) and cause-specific mortality fraction (CSMF) accuracy as quality metrics.

Results

We identified 966 records from the vital registry for all under-5 deaths, and 390 were linked to medical records of deaths occurring at public hospitals. The Yucatan vital registration system captured 94.8% of the expected under-5 deaths, with an overall VSPI score of 87.2%. Concordance between underlying cause of death listed in the vital registry and the cause determined by the medical record review varied substantially across causes, with a mean overall chance-corrected concordance across causes of 6.9% for neonates and 46.9% for children. Children had the highest concordance for digestive diseases, and neonates had the highest concordance for meningitis/sepsis. At the population level, the CSMF accuracy for identifying the underlying cause listed was 35.3% for neonates and 67.7% for children.

Conclusions

Although the vital registration system has overall good performance, there are still problems in information about causes of death for children under 5 that are related mostly to certification of the causes of death. The accuracy of information can vary substantially across age groups and causes, with causes reported for neonates being generally less reliable than those for older children. Results highlight the need to implement strategies to improve the certification of causes of death in this population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12963-022-00284-5.

Validation of SmartVA using conventional autopsy: A study of adult deaths in Brazil

BACKGROUND

Accurate cause of death data are essential to guide health policy. However, mortality surveillance is limited in many low-income countries. In such settings, verbal autopsy (VA) is increasingly used to provide population-level cause of death data. VAs are now widely interpreted using the automated algorithms SmartVA and InterVA. Here we use conventional autopsy as the gold standard to validate SmartVA methodology.

METHODS

This study included adult deaths from natural causes in São Paulo and Recife for which conventional autopsy was indicated. VA was conducted with a relative of the deceased using an amended version of the SmartVA instrument to suit the local context. Causes of death from VA were produced using the SmartVA-Analyze program. Physician coded verbal autopsy (PCVA), conducted on the same questionnaires, and Global Burden of Disease Study data were used as additional comparators. Cause of death data were grouped into 10 broad causes for the validation due to the real-world utility of VA lying in identifying broad population cause of death patterns.

FINDINGS

The study included 2,060 deaths in São Paulo and 1,079 in Recife. The cause specific mortality fractions (CSMFs) estimated using SmartVA were broadly similar to conventional autopsy for: cardiovascular diseases (46.8% vs 54.0%, respectively), cancers (10.6% vs 11.4%), infections (7.0% vs 10.4%) and chronic respiratory disease (4.1% vs 3.7%), causes accounting for 76.1% of the autopsy dataset. The SmartVA CSMF estimates were lower than autopsy for "Other NCDs" (7.8% vs 14.6%) and higher for diabetes (13.0% vs 6.6%). CSMF accuracy of SmartVA compared to autopsy was 84.5%. CSMF accuracy for PCVA was 93.0%.

INTERPRETATION

The results suggest that SmartVA can, with reasonable accuracy, predict the broad cause of death groups important to assess a population's epidemiological transition. VA remains a useful tool for understanding causes of death where medical certification is not possible.

Evaluation of a two-way SMS messaging strategy to reduce neonatal mortality: rationale, design and methods of the Mobile WACh NEO randomised controlled trial in Kenya

INTRODUCTION

Globally, approximately half of the estimated 6.3 million under-5 deaths occur in the neonatal period (within the first 28 days of life). Kenya ranks among countries with the highest number of neonatal deaths, at 20 per 1000 live births. Improved identification and management of neonates with potentially life-threatening illness is critical to meet the WHO's target of ≤12 neonatal deaths per 1000 live births by 2035. We developed an interactive (two-way) short messaging service (SMS) communication intervention, Mobile Solutions for Neonatal Health (Mobile women's and children's health (WACh) NEO), focused on the perinatal period. Mobile WACh NEO sends automated tailored SMS messages to mothers during pregnancy and up to 6 weeks post partum. Messages employ the Information-Motivation-Behaviour Skills framework to promote (1) maternal implementation of essential newborn care (ENC, including early, exclusive breast feeding, cord care and thermal care), (2) maternal identification of neonatal danger signs and care-seeking, and (3) maternal social support and self-efficacy. Participants can also send SMS to the study nurse, enabling on-demand remote support.

METHODS AND ANALYSIS

We describe a two-arm unblinded randomised controlled trial of the Mobile WACh NEO intervention. We will enrol 5000 pregnant women in the third trimester of pregnancy at 4 facilities in Kenya and randomise them 1:1 to receive interactive SMS or no SMS (control), and conduct follow-up visits at 2 and 6 weeks post partum. Neonatal mortality will be compared between arms as the primary outcome. Secondary outcomes include care-seeking, practice of ENC and psychosocial health. Exploratory analysis will investigate associations between maternal mental health, practice of ENC, care-seeking and SMS engagement.

ETHICS AND DISSEMINATION

This study received ethical approval from the University of Washington (STUDY00006395), Women and Infants Hospital (1755292-1) and Kenyatta National Hospital/University of Nairobi (P310/04/2019). All participants will provide written informed consent. Findings will be published in peer-reviewed journals and international conferences.

TRIAL REGISTRATION NUMBER

NCT04598165.

Improving methods to measure comparable mortality cause (IMMCMC) gold standard verbal autopsy dataset

Objectives

Gold standard cause of death data is critically important to improve verbal autopsy (VA) methods in diagnosing cause of death where civil and vital registration systems are inadequate or poor. As part of a three-country research study—Improving Methods to Measure Comparable Mortality by Cause (IMMCMC) study—data were collected on clinicopathological criteria-based gold standard cause of death from hospital record reviews with matched VAs. The purpose of this data note is to make accessible a de-identified format of these gold standard VAs for interested researchers to improve the diagnostic accuracy of VA methods.

Data description

The study was conducted between 2011 and 2014 in the Philippines, Bangladesh, and Papua New Guinea. Gold standard diagnoses of underlying causes of death for deaths occurring in hospital were matched to VAs conducted using a standardized VA questionnaire developed by the Population Health Metrics Consortium. 3512 deaths were collected in total, comprised of 2491 adults (12 years and older), 320 children (28 days to 12 years), and 702 neonates (0–27 days).

Verbal autopsy models in determining causes of death

OBJECTIVES

To systematically review current practices, strengths and limitations of existing VA approaches to increase understanding of health system stakeholders and researchers.

METHODS

The review was conducted and reported based on the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines, in which articles were systematically obtained from the PubMed and SCOPUS online databases. The search was limited to English language journal articles published between 2010 and 2020. The review identified 5602 articles and after thorough scrutiny, 25 articles related to VA approaches were included.

RESULTS

(1) InterVA and Tariff are widely used VA models; (2) Bayes rule is the most common and successful algorithm; (3) the lack of standardised datasets and metrics to evaluate models creates bias in determining VA model performance; (4) performance of the models trained using in-hospital data cannot be replicated in community death; (5) the performance of models among physicians and computer-coded algorithms differs with variation in settings.

CONCLUSION

The physician-certified verbal autopsy (PCVA) approaches are more effective in determining community CoD while computerised coding of verbal autopsy (CCVA) models perform well when the underlying CoD are reliably established using hospital data where data are trained in a similar environment to the target population. Our study recommends the use of hybrid models that combine strengths from various models and using an open standards dataset that includes death from different settings.

Verbal/social autopsy analysis of causes and determinants of under-5 mortality in Tanzania from 2010 to 2016

Background

Tanzania has decreased its child mortality rate by more than 70 percent in the last three decades and is striving to develop a nationally-representative sample registration system with verbal autopsy to help focus health policies and programs toward further reduction. As an interim measure, a verbal and social autopsy study was conducted to provide vital information on the causes and social determinants of neonatal and child deaths.

Methods

Causes of neonatal and 1-59 month-old deaths identified by the 2015-16 Tanzania Demographic and Health Survey were assessed using the expert algorithm verbal autopsy method. The social autopsy examined prevalence of key household, community and health system indicators of preventive and curative care provided along the continuum of care and Pathway to Survival models. Careseeking for neonates and 1-59 month-olds was compared, and tests of associations of age and cause of death to careseeking indicators and place of death were conducted.

Results

The most common causes of death of 228 neonates and 351 1-59 month-olds, respectively, were severe infection, intrapartum related events and preterm delivery, and pneumonia, diarrhea and malaria. Coverage of early initiation of breastfeeding (24%), hygienic cord care (29%), and full immunization of 12-59 month-olds (33%) was problematic. Most (88.8%) neonates died in the first week, including 44.3% in their birth facility before leaving. Formal care was sought for just 41.9% of newborns whose illness started at home and was delayed by 5.3 days for 1-59 month-olds who sought informal care. Care was less likely to be sought for the youngest neonates and infants and severely ill children. Although 70.3% of 233 under-5 year-olds were moderately or severely ill on discharge from their first provider, only 29.0%-31.2% were referred.

Conclusions

The study highlights needed actions to complete Tanzania’s child survival agenda. Low levels of some preventive interventions need to be addressed. The high rate of facility births and neonatal deaths requires strengthening of institutionally-based interventions targeting maternal labor and delivery complications and neonatal causes of death. Scale-up of Integrated Community Case Management should be considered to strengthen careseeking for the youngest newborns, infants and severely ill children and referral practices at first level facilities.

Bayesian hierarchical factor regression models to infer cause of death from verbal autopsy data

In low-resource settings where vital registration of death is not routine it is often of critical interest to determine and study the cause of death (COD) for individuals and the cause-specific mortality fraction (CSMF) for populations. Post-mortem autopsies, considered the gold standard for COD assignment, are often difficult or impossible to implement due to deaths occurring outside the hospital, expense, and/or cultural norms. For this reason, Verbal Autopsies (VAs) are commonly conducted, consisting of a questionnaire administered to next of kin recording demographic information, known medical conditions, symptoms, and other factors for the decedent. This article proposes a novel class of hierarchical factor regression models that avoid restrictive assumptions of standard methods, allow both the mean and covariance to vary with COD category, and can include covariate information on the decedent, region, or events surrounding death. Taking a Bayesian approach to inference, this work develops an MCMC algorithm and validates the FActor Regression for Verbal Autopsy (FARVA) model in simulation experiments. An application of FARVA to real VA data shows improved goodness-of-fit and better predictive performance in inferring COD and CSMF over competing methods. Code and a user manual are made available at https://github.com/kelrenmor/farva.

Diversity of epidemiological transition in the Pacific: Findings from the application of verbal autopsy in Papua New Guinea and the Solomon Islands

BACKGROUND

Cause of death data are essential for rational health planning yet are not routinely available in Papua New Guinea (PNG) and Solomon Islands. Indirect estimation of cause of death patterns suggests these populations are epidemiologically similar, but such assessments are not based on direct evidence.

METHODS

Verbal autopsy (VA) interviews were conducted at three sites in PNG and nationwide in Solomon Islands. Training courses were also facilitated to improve data from medical certificates of cause of death (MCCODs) in both countries. Data were categorised into broad groups of endemic and emerging conditions to aid assessment of the epidemiological transition.

FINDINGS

Between 2017 and 2020, VAs were collected for 1,814 adult deaths in PNG and 819 adult deaths in Solomon Islands. MCCODs were analysed for 662 deaths in PNG and 1,408 deaths in Solomon Islands. The VA data suggest lower NCD mortality (48.8% versus 70.3%); higher infectious mortality (27.0% versus 18.3%) and higher injury mortality (24.5% versus 11.4%) in PNG compared to Solomon Islands. Higher infectious mortality in PNG was evident for both endemic and emerging infections. Higher NCD mortality in Solomon Islands reflected much higher emerging NCDs (43.6% vs 21.4% in PNG). A similar pattern was evident from the MCCOD data.

INTERPRETATION

The cause of death patterns suggested by VA and MCCOD indicate that PNG is earlier in its epidemiological transition than Solomon Islands, with relatively higher infectious mortality and lower NCD mortality. Injury mortality was also particularly high in PNG.This study was funded by Bloomberg Philanthropies.

Limitations to current methods to estimate cause of death: a validation study of a verbal autopsy model

Background: Accurate information on causes of death (CoD) is essential to estimate burden of disease, track global progress, prioritize cost-effective interventions, and inform policies to reduce mortality. In low-income settings, where a significant proportion of deaths take place at home or in poorly-resourced peripheral health facilities, data on CoD often relies on verbal autopsies (VAs). Validations of VAs have been performed against clinical diagnosis, but never before against an acceptable gold standard: the complete diagnostic autopsy (CDA). Methods: We have validated a computer-coded verbal autopsy method -the InterVA- using individual and population metrics to determine CoD against the CDA, in 316 deceased patients of different age groups who died in a tertiary-level hospital in Maputo, Mozambique between 2013 and 2015.   Results: We found a low agreement of the model across all age groups at the individual (kappa statistic ranging from -0.030 to 0.232, lowest in stillbirths and highest in adults) and population levels (chance-corrected cause-specific mortality fraction accuracy ranging from -1.00 to 0.62, lowest in stillbirths, highest in children). The sensitivity in identifying infectious diseases was low (0% for tuberculosis, diarrhea, and disseminated infections, 32% for HIV-related infections, 33% for malaria and 36% for pneumonia). Of maternal deaths, 26 were assigned to eclampsia but only four patients actually died of eclampsia. Conclusions: These findings do not lead to building confidence in current estimates of CoD. They also call to the need to implement autopsy methods where they may be feasible, and to improve the quality and performance of current VA techniques.

Limitations to current methods to estimate cause of death: a validation study of a verbal autopsy model

Background: Accurate information on causes of death (CoD) is essential to estimate burden of disease, track global progress, prioritize cost-effective interventions, and inform policies to reduce mortality. In low-income settings, where a significant proportion of deaths take place at home or in poorly-resourced peripheral health facilities, data on CoD often relies on verbal autopsies (VAs). Validations of VAs have been performed against clinical diagnosis, but never before against an acceptable gold standard: the complete diagnostic autopsy (CDA). Methods: We have validated a computer-coded verbal autopsy method -the InterVA- using individual and population metrics to determine CoD against the CDA, in 316 deceased patients of different age groups who died in a tertiary-level hospital in Maputo, Mozambique between 2013 and 2015.   Results: We found a low agreement of the model across all age groups at the individual (kappa statistic ranging from -0.030 to 0.232, lowest in stillbirths and highest in adults) and population levels (chance-corrected cause-specific mortality fraction accuracy ranging from -1.00 to 0.62, lowest in stillbirths, highest in children). The sensitivity in identifying infectious diseases was low (0% for tuberculosis, diarrhea, and disseminated infections, 32% for HIV-related infections, 33% for malaria and 36% for pneumonia). Of maternal deaths, 26 were assigned to eclampsia but only four patients actually died of eclampsia. Conclusions: These findings do not lead to building confidence in current estimates of CoD. They also call to the need to implement autopsy methods where they may be feasible, and to improve the quality and performance of current VA techniques.

Limitations to current methods to estimate cause of death: a validation study of a verbal autopsy model

Background: Accurate information on causes of death (CoD) is essential to estimate burden of disease, track global progress, prioritize cost-effective interventions, and inform policies to reduce mortality. In low-income settings, where a significant proportion of deaths take place at home or in poorly-resourced peripheral health facilities, data on CoD often relies on verbal autopsies (VAs). Validations of VAs have been performed against clinical diagnosis, but never before against an acceptable gold standard: the complete diagnostic autopsy (CDA).

Methods: We have validated a computer-coded verbal autopsy method –the InterVA- using individual and population metrics to determine CoD against the CDA, in 316 deceased patients of different age groups who died in a tertiary-level hospital in Maputo, Mozambique between 2013 and 2015.  

Results: We found a low agreement of the model across all age groups at the individual (kappa statistic ranging from -0.030 to 0.232, lowest in stillbirths and highest in adults) and population levels (chance-corrected cause-specific mortality fraction accuracy ranging from -1.00 to 0.62, lowest in stillbirths, highest in children). The sensitivity in identifying infectious diseases was low (0% for tuberculosis, diarrhea, and disseminated infections, 32% for HIV-related infections, 33% for malaria and 36% for pneumonia). Of maternal deaths, 26 were assigned to eclampsia but only four patients actually died of eclampsia.

Conclusions: These findings do not lead to building confidence in current estimates of CoD. They also call to the need to implement autopsy methods where they may be feasible, and to improve the quality and performance of current VA techniques.

Application of verbal autopsy in routine civil registration in Lusaka District of Zambia

Background

Ascertaining the causes for deaths occurring outside health facilities is a significant problem in many developing countries where civil registration systems are not well developed or non-functional. Standardized and rigorous verbal autopsy methods is a potential solution to determine the cause of death. We conducted a demonstration project in Lusaka District of Zambia where verbal autopsy (VA) method was implemented in routine civil registration system.

Methods

About 3400 VA interviews were conducted for bodies “brought-in-dead” at Lusaka’s two major teaching hospital mortuaries using a SmartVA questionnaire between October 2017 and September 2018. Probable underlying causes of deaths using VA and cause-specific mortality fractions were determined.. Demographic characteristics were analyzed for each VA-ascertained cause of death.

Results

Opportunistic infections (OIs) associated with HIV/AIDS such as pneumonia and tuberculosis, and malaria were among leading causes of deaths among bodies “brought-in-dead”. Over 21.6 and 26.9% of deaths were attributable to external causes and non-communicable diseases (NCDs), respectively. The VA-ascertained causes of death varied by age-group and sex. External causes were more prevalent among males in middle ages (put an age range like 30–54 years old) and NCDs highly prevalent among those aged 55 years and older.

Conclusions

VA application in civil registration system can provide the much-needed cause of death information for non-facility deaths in countries with under-developed or non-functional civil registration systems.

Extracting Cause of Death From Verbal Autopsy With Deep Learning Interpretable Methods

The international standard to ascertain the cause of death is medical certification. However, in many low and middle-income countries, the majority of deaths occur outside of health facilities. In these cases, Verbal Autopsy (VA), the narrative provided by a family member or friend together with a questionnaire is designed by the World Health Organization as the main information source. Until now technology allowed us to automatically analyze the responses of the VA questionnaire with the narrative captured by the interviewer excluded. Our work addresses this gap by developing a set of models for automatic Cause of Death (CoD) ascertainment in VAs with a focus on the textual information. Empirical results show that the open response conveys valuable information towards the ascertainment of the Cause of Death, and the combination of the closed-ended questions and the open response lead to the best results. Model interpretation capabilities position the Deep Learning models as the most encouraging choice.

Improving cause of death certification in the Philippines: implementation of an electronic verbal autopsy decision support tool (SmartVA auto-analyse) to aid physician diagnoses of out-of-facility deaths

Background

The majority of deaths in the Philippines occur out-of-facility and require a medical certificate of cause of death by Municipal Health Officers (MHOs) for burial. MHOs lack a standardised certification process for out-of-facility deaths and when no medical records are available, certify a high proportion of ill-defined causes of death. We aimed to develop and introduce SmartVA Auto-Analyse, a verbal autopsy (VA) based electronic decision support tool in order to assist the MHOs in certifying out-of-facility deaths.

Method

We conducted a stakeholder consultation, process mapping and a pre-test to assess feasibility and acceptability of SmartVA Auto-Analyse. MHOs were first asked to conduct an open-ended interview from the family members of the deceased, and if they were not able to arrive at a diagnosis, continue the interview using the standardised SmartVA questionnaire. Auto-Analyse then presented the MHO with the three most likely causes of death. For the pilot, the intervention was scaled-up to 91 municipalities. We performed a mixed-methods evaluation using the cause of death data and group discussions with the MHOs.

Results

Of the 5649 deaths registered, Auto-Analyse was used to certify 4586 (81%). For the remaining 19%, doctors believed they could assign a cause of death based on the availability of medical records and the VA open narrative. When used, physicians used the Auto-Analyse diagnosis in 85% of cases to certify the cause of death. Only 13% of the deaths under the intervention had an undetermined cause of death. Group discussions identified two themes: Auto-Analyse standardized the certification of home deaths and assisted the MHOs to improve the quality of death certification.

Conclusion

Standardized VA combined with physician diagnosis using the SmartVA Auto-Analyse support tool was readily used by MHOs in the Philippines and can improve the quality of death certification of home deaths.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-10542-0.

Routine mortality surveillance to identify the cause of death pattern for out-of-hospital adult (aged 12+ years) deaths in Bangladesh: introduction of automated verbal autopsy

BACKGROUND

In Bangladesh, a poorly functioning national system of registering deaths and determining their causes leaves the country without important information on which to inform health programming, particularly for the 85% of deaths that occur in the community. In 2017, an improved death registration system and automated verbal autopsy (VA) were introduced to 13 upazilas to assess the utility of VA as a routine source of policy-relevant information and to identify leading causes of deaths (COD) in rural Bangladesh.

METHODS

Data from 22,535 VAs, collected in 12 upazilas between October 2017 and August 2019, were assigned a COD using the SmartVA Analyze 2.0 computer algorithm. The plausibility of the VA results was assessed using a series of demographic and epidemiological checks in the Verbal Autopsy Interpretation, Performance and Evaluation Resource (VIPER) software tool.

RESULTS

Completeness of community death reporting was 65%. The vast majority (85%) of adult deaths were due to non-communicable diseases, with ischemic heart disease, stroke and chronic respiratory disease comprising about 60% alone. Leading COD were broadly consistent with Global Burden of Disease study estimates.

CONCLUSIONS

Routine VA collection using automated methods is feasible, can produce plausible results and provides critical information on community COD in Bangladesh. Routine VA and VIPER have potential application to countries with weak death registration systems.

Does recall time matter in verbal autopsies? Evidence from urban informal settlements in Nairobi, Kenya

Background: To assign a cause of death to non-medically certified deaths, verbal autopsies (VAs) are widely used to determine the cause of death. The time difference between the death and the VA interview, also referred to as recall time, varies depending on social and operational factors surrounding the death. We investigated the effect of recall time on the assignment of causes of death by VA. Methods: This is a secondary analysis of 2002-2015 survey data of the Nairobi Urban Health Demographic Surveillance System (NUHDSS). The independent variable recall time was derived from the date of death and the date when the VA was conducted. Univariate and multivariate logistic regression methods were used to calculate odds ratios of assigning a cause of death in defined categories of recall time. Results: There were 6218 deaths followed up between 2002 and 2016, out of which 5495 (88.3%) had VAs done. Recall time varied from 1-3001 days (median  92 days, IQR 44-169 days). Majority of the VAs (45.7%) were conducted between 1-3 months after death. The effect of recall time varied for different diseases. Compared to VAs conducted between 1-3 months, there was a 24% higher likelihood of identifying HIV/AIDS as the cause of death for VAs conducted 4-6 months after death (AOR 1.24; 95% CI 1.01-1.54; p-value = 0.043) and a 40% increased chance of identifying other infectious diseases as the cause of death for VAs conducted <1 month after death (AOR 1.4; 95% CI 1.02-1.92, p-value = 0.024). Conclusions: Recall time affected the assignment of VA cause of death for HIV/AIDS, other infectious diseases,maternal/neonatal and indeterminate causes. Our analysis indicates that in the urban informal setting, VAs should be conducted from one month up to 6 months after the death to improve the probability of accurately assigning the cause of death.

Effect of Mass Treatment with Azithromycin on Causes of Death in Children in Malawi: Secondary Analysis from the MORDOR Trial

Recent evidence indicates mass drug administration with azithromycin may reduce child mortality. This study uses verbal autopsy (VA) to investigate the causes of individual deaths during the Macrolides Oraux pour Réduire les Décès avec un Oeil sur la Résistance (MORDOR) trial in Malawi. Cluster randomization was performed as part of MORDOR. Biannual household visits were conducted to distribute azithromycin or placebo to children aged 1–59 months and update the census to identify deaths for VA. MORDOR was not powered to investigate mortality effects at individual sites, but the available evidence is presented here for hypothesis generation regarding the mechanism through which azithromycin may reduce child mortality. Automated VA analysis was performed to infer the likely cause of death using two major analysis programs, InterVA and SmartVA. A total of 334 communities were randomized to azithromycin or placebo, with more than 130,000 person-years of follow-up. During the study, there were 1,184 deaths, of which 1,131 were followed up with VA. Mortality was 9% lower in azithromycin-treated communities than in placebo communities (rate ratio 0.91 [95% CI: 0.79–1.05]; P = 0.20). The intention-to-treat analysis by cause using InterVA suggested fewer HIV/AIDS deaths in azithromycin-treated communities (rate ratio 0.70 [95% CI: 0.50–0.97]; P = 0.03) and fewer pneumonia deaths (rate ratio 0.82 [95% CI: 0.60–1.12]; P = 0.22). The use of the SmartVA algorithm suggested fewer diarrhea deaths (rate ratio 0.71 [95% CI: 0.51–1.00]; P = 0.05) and fewer pneumonia deaths (rate ratio 0.58 [95% CI: 0.33–1.00]; P = 0.05). Although this study is not able to provide strong evidence, the data suggest that the mortality reduction during MORDOR in Malawi may have been due to effects on pneumonia and diarrhea or HIV/AIDS mortality.

Does recall time matter in verbal autopsies? Evidence from urban informal settlements in Nairobi, Kenya

Background: To assign a cause of death to non-medically certified deaths, verbal autopsies (VAs) are widely used to determine the cause of death. The time difference between the death and the VA interview, also referred to as recall time, varies depending on social and operational factors surrounding the death. We investigated the effect of recall time on the assignment of causes of death by VA. Methods: This is a secondary analysis of 2002-2015 survey data of the Nairobi Urban Health Demographic Surveillance System (NUHDSS). The independent variable recall time was derived from the date of death and the date when the VA was conducted. Univariate and multivariate logistic regression methods were used to calculate odds ratios of assigning a cause of death in defined categories of recall time. Results: There were 6218 deaths followed up between 2002 and 2016, out of which 5495 (88.3%) had VAs done. Recall time varied from 1-3001 days (mean 115.5 days, sd216.8). Majority of the VAs (45.7%) were conducted between 1-3 months after death. The effect of recall time varied for different diseases. Compared to VAs conducted between 1-3 months, there was a 24% higher likelihood of identifying HIV/AIDS as the cause of death for VAs conducted 4-6 months after death (AOR 1.24; 95% CI 1.01-1.54; p-value = 0.043) and a 40% increased chance of identifying other infectious diseases as the cause of death for VAs conducted <1 month after death (AOR 1.4; 95% CI 1.02-1.92, p-value = 0.024). Conclusions: Recall time affected the assignment of VA cause of death for HIV/AIDS, other infectious diseases and maternal/neonatal causes. Our analysis indicates that in the urban informal setting, VAs should be conducted from one month up to 6 months after the death to improve the probability of accurately assigning the cause of death.

Using Bayesian Latent Gaussian Graphical Models to Infer Symptom Associations in Verbal Autopsies

Learning dependence relationships among variables of mixed types provides insights in a variety of scientific settings and is a well-studied problem in statistics. Existing methods, however, typically rely on copious, high quality data to accurately learn associations. In this paper, we develop a method for scientific settings where learning dependence structure is essential, but data are sparse and have a high fraction of missing values. Specifically, our work is motivated by survey-based cause of death assessments known as verbal autopsies (VAs). We propose a Bayesian approach to characterize dependence relationships using a latent Gaussian graphical model that incorporates informative priors on the marginal distributions of the variables. We demonstrate such information can improve estimation of the dependence structure, especially in settings with little training data. We show that our method can be integrated into existing probabilistic cause-of-death assignment algorithms and improves model performance while recovering dependence patterns between symptoms that can inform efficient questionnaire design in future data collection.

How to deal with the suffering: Utility of an emotional containment strategy to collect data for verbal autopsies in Mexico

Verbal autopsies are often used to establish cause of death but can be emotionally taxing on the interviewers. We conducted focus groups with interviewers (N = 15) who collected data for verbal autopsies in Mexico in order to explore the utility of an emotional containment strategy designed to boost self-confidence and resilience. The interviewers reflected on broader cultural perspectives on illness and death and described the strategy as helpful in developing strategies to manage emotionally stressful situations and develop their confidence in their work performance. This type of intervention may be useful for field personnel who perform verbal autopsy interviews.KEY MESSAGESIn low- and middle-income countries with less reliable statistics systems, a significant proportion of deaths is not certified by a professional doctor. This complicates the registration of causes of death, which is a crucial issue for health systems. In the absence of reliable vital statistics systems, verbal autopsies (VA) offer an alternative for establishing cause of death.In response to emotional crises leading to resignations among the interviewers while testing an instrument for collecting VA, we designed an emotional containment strategy (ECS). It was specifically crafted to boost the self-confidence and resilience of participants in addition to enhancing their capacity for emotional recovery and to regain a functional state. In order to explore ECS results we conducted a qualitative cross-sectional study with four focus groups of interviewers who collected VA.The results obtained were positive, the interviewers were able to perform their work better by overcoming the emotional crisis that occurred both in them and in the people they interviewed.We recommend developing this type of intervention with all field staff performing verbal autopsy interviews, not only as a resource for emotional health, but also as a means of achieving better-quality data collection.

Analysis of causes of death among brought-in-dead cases in a third-level Hospital in Lusaka, Republic of Zambia, using the tariff method 2.0 for verbal autopsy: a cross-sectional study

BACKGROUND

Over one third of deaths in Zambian health facilities involve someone who has already died before arrival (i.e., Brough in Dead), and in most BiD cases, the CoD have not been fully analyzed. Therefore, this study was designed to evaluate the function of automated VA based on the Tariff Method 2.0 to identify the CoD among the BiD cases and the usefulness by comparing the data on the death notification form.

METHODS

The target site was one third-level hospital in the Republic of Zambia's capital city. All BiD cases who reached the target health facility from January to August 2017 were included. The deceased's closest relatives were interviewed using a structured VA questionnaire and the data were analyzed using the SmartVA to determine the CoD at the individual and population level. The CoD were compared with description on the death notification forms by using t-test and Cohen's kappa coefficient.

RESULTS

One thousand three hundred seventy-eight and 209 cases were included for persons aged 13 years and older (Adult) and those aged 1 month to 13 years old (Child), respectively. The top CoD for Adults were infectious diseases followed by non-communicable diseases and that for Child were infectious diseases, followed by accidents. The proportion of cases with a determined CoD was significantly higher when using the SmartVA (75% for Adult and 67% for Child) than the death notification form (61%). A proportion (42.7% for Adult and 46% for Child) of the CoD-determined cases matched in both sources, with a low concordance rate for Adult (kappa coefficient = 0.1385) and a good for Child(kappa coefficient = 0.635).

CONCLUSIONS

The CoD of the BiD cases were successfully analyzed using the SmartVA for the first time in Zambia. While there many erroneous descriptions on the death notification form, the SmartVA could determine the CoD among more BiD cases. Since the information on the death notification form is reflected in the national vital statistics, more accurate and complete CoD data are required. In order to strengthen the death registration system with accurate CoD, it will be useful to embed the SmartVA in Zambia's health information system.

The epidemiological transition in Papua New Guinea: new evidence from verbal autopsy studies

BACKGROUND

Recent economic growth in Papua New Guinea (PNG) would suggest that the country may be experiencing an epidemiological transition, characterized by a reduction in infectious diseases and a growing burden from non-communicable diseases (NCDs). However, data on cause-specific mortality in PNG are very sparse, and the extent of the transition within the country is poorly understood.

METHODS

Mortality surveillance was established in four small populations across PNG: West Hiri in Central Province, Asaro Valley in Eastern Highlands Province, Hides in Hela Province and Karkar Island in Madang Province. Verbal autopsies (VAs) were conducted on all deaths identified, and causes of death were assigned by SmartVA and classified into five broad disease categories: endemic NCDs; emerging NCDs; endemic infections; emerging infections; and injuries. Results from previous PNG VA studies, using different VA methods and spanning the years 1970 to 2001, are also presented here.

RESULTS

A total of 868 deaths among adolescents and adults were identified and assigned a cause of death. NCDs made up the majority of all deaths (40.4%), with the endemic NCD of chronic respiratory disease responsible for the largest proportion of deaths (10.5%), followed by the emerging NCD of diabetes (6.2%). Emerging infectious diseases outnumbered endemic infectious diseases (11.9% versus 9.5%). The distribution of causes of death differed across the four sites, with emerging NCDs and emerging infections highest at the site that is most socioeconomically developed, West Hiri. Comparing the 1970-2001 VA series with the present study suggests a large decrease in endemic infections.

CONCLUSIONS

Our results indicate immediate priorities for health service planning and for strengthening of vital registration systems, to more usefully serve the needs of health priority setting.

Non-confirming replication of "Performance of InSilicoVA for assigning causes of death to verbal autopsies: multisite validation study using clinical diagnostic gold standards," by Flaxman et al

BACKGROUND

A verbal autopsy (VA) is an interview conducted with the caregivers of someone who has recently died to describe the circumstances of the death. In recent years, several algorithmic methods have been developed to classify cause of death using VA data. The performance of one method-InSilicoVA-was evaluated in a study by Flaxman et al., published in BMC Medicine in 2018. The results of that study are different from those previously published by our group.

METHODS

Based on the description of methods in the Flaxman et al. study, we attempt to replicate the analysis to understand why the published results differ from those of our previous work.

RESULTS

We failed to reproduce the results published in Flaxman et al. Most of the discrepancies we find likely result from undocumented differences in data pre-processing, and/or values assigned to key parameters governing the behavior of the algorithm.

CONCLUSION

This finding highlights the importance of making replication code available along with published results. All code necessary to replicate the work described here is freely available on GitHub.

Automated verbal autopsy: from research to routine use in civil registration and vital statistics systems

BACKGROUND

The majority of low- and middle-income countries (LMICs) do not have adequate civil registration and vital statistics (CRVS) systems to properly support health policy formulation. Verbal autopsy (VA), long used in research, can provide useful information on the cause of death (COD) in populations where physicians are not available to complete medical certificates of COD. Here, we report on the application of the SmartVA tool for the collection and analysis of data in several countries as part of routine CRVS activities.

METHODS

Data from VA interviews conducted in 4 of 12 countries supported by the Bloomberg Philanthropies Data for Health (D4H) Initiative, and at different stages of health statistical development, were analysed and assessed for plausibility: Myanmar, Papua New Guinea (PNG), Bangladesh and the Philippines. Analyses by age- and cause-specific mortality fractions were compared to the Global Burden of Disease (GBD) study data by country. VA interviews were analysed using SmartVA-Analyze-automated software that was designed for use in CRVS systems. The method in the Philippines differed from the other sites in that the VA output was used as a decision support tool for health officers.

RESULTS

Country strategies for VA implementation are described in detail. Comparisons between VA data and country GBD estimates by age and cause revealed generally similar patterns and distributions. The main discrepancy was higher infectious disease mortality and lower non-communicable disease mortality at the PNG VA sites, compared to the GBD country models, which critical appraisal suggests may highlight real differences rather than implausible VA results.

CONCLUSION

Automated VA is the only feasible method for generating COD data for many populations. The results of implementation in four countries, reported here under the D4H Initiative, confirm that these methods are acceptable for wide-scale implementation and can produce reliable COD information on community deaths for which little was previously known.

BAYESIAN FACTOR MODELS FOR PROBABILISTIC CAUSE OF DEATH ASSESSMENT WITH VERBAL AUTOPSIES

The distribution of deaths by cause provides crucial information for public health planning, response and evaluation. About 60% of deaths globally are not registered or given a cause, limiting our ability to understand disease epidemiology. Verbal autopsy (VA) surveys are increasingly used in such settings to collect information on the signs, symptoms and medical history of people who have recently died. This article develops a novel Bayesian method for estimation of population distributions of deaths by cause using verbal autopsy data. The proposed approach is based on a multivariate probit model where associations among items in questionnaires are flexibly induced by latent factors. Using the Population Health Metrics Research Consortium labeled data that include both VA and medically certified causes of death, we assess performance of the proposed method. Further, we estimate important questionnaire items that are highly associated with causes of death. This framework provides insights that will simplify future data.

Regularized Bayesian transfer learning for population-level etiological distributions

Computer-coded verbal autopsy (CCVA) algorithms predict cause of death from high-dimensional family questionnaire data (verbal autopsy) of a deceased individual, which are then aggregated to generate national and regional estimates of cause-specific mortality fractions. These estimates may be inaccurate if CCVA is trained on non-local training data different from the local population of interest. This problem is a special case of transfer learning, i.e., improving classification within a target domain (e.g., a particular population) with the classifier trained in a source-domain. Most transfer learning approaches concern individual-level (e.g., a person’s) classification. Social and health scientists such as epidemiologists are often more interested with understanding etiological distributions at the population-level. The sample sizes of their data sets are typically orders of magnitude smaller than those used for common transfer learning applications like image classification, document identification, etc. We present a parsimonious hierarchical Bayesian transfer learning framework to directly estimate population-level class probabilities in a target domain, using any baseline classifier trained on source-domain, and a small labeled target-domain dataset. To address small sample sizes, we introduce a novel shrinkage prior for the transfer error rates guaranteeing that, in absence of any labeled target-domain data or when the baseline classifier is perfectly accurate, our transfer learning agrees with direct aggregation of predictions from the baseline classifier, thereby subsuming the default practice as a special case. We then extend our approach to use an ensemble of baseline classifiers producing an unified estimate. Theoretical and empirical results demonstrate how the ensemble model favors the most accurate baseline classifier. We present data analyses demonstrating the utility of our approach.

Robustness of the Tariff method for diagnosing verbal autopsies: impact of additional site data on the relationship between symptom and cause

Background

Verbal autopsy (VA) is increasingly being considered as a cost-effective method to improve cause of death information in countries with low quality vital registration. VA algorithms that use empirical data have an advantage over expert derived algorithms in that they use responses to the VA instrument as a reference instead of physician opinion. It is unclear how stable these data driven algorithms, such as the Tariff 2.0 method, are to cultural and epidemiological variations in populations where they might be employed.

Methods

VAs were conducted in three sites as part of the Improving Methods to Measure Comparable Mortality by Cause (IMMCMC) study: Bohol, Philippines; Chandpur and Comila Districts, Bangladesh; and Central and Eastern Highlands Provinces, Papua New Guinea. Similar diagnostic criteria and cause lists as the Population Health Metrics Research Consortium (PHMRC) study were used to identify gold standard (GS) deaths. We assessed changes in Tariffs by examining the proportion of Tariffs that changed significantly after the addition of the IMMCMC dataset to the PHMRC dataset.

Results

The IMMCMC study added 3512 deaths to the GS VA database (2491 adults, 320 children, and 701 neonates). Chance-corrected cause specific mortality fractions for Tariff improved with the addition of the IMMCMC dataset for adults (+ 5.0%), children (+ 5.8%), and neonates (+ 1.5%). 97.2% of Tariffs did not change significantly after the addition of the IMMCMC dataset.

Conclusions

Tariffs generally remained consistent after adding the IMMCMC dataset. Population level performance of the Tariff method for diagnosing VAs improved marginally for all age groups in the combined dataset. These findings suggest that cause-symptom relationships of Tariff 2.0 might well be robust across different population settings in developing countries. Increasing the total number of GS deaths improves the validity of Tariff and provides a foundation for the validation of other empirical algorithms.

Mortality Associated With Acute Respiratory Infections Among Children at Home

Background

Numerous deaths in children aged <5 years in the developing world occur at home. Acute respiratory infections (ARIs) are thought to play an important role in these deaths. Risk factors and pathogens linked to fatal episodes remain unclear.

Methods

A case-control study among low-income children aged <5 years was performed in Buenos Aires, Argentina, to define risk factors and viral pathogens among those who died of ARI at home.

Results

A total of 278 families of children aged <5 years (of whom 104 died and 174 were healthy controls) participated in the study. A total of 87.5% of ARI-associated deaths occurred among infants aged <12 months. The estimated mortality rate due to ARI among infants was 5.02 deaths/1000 live births. Dying at home from ARI was associated with living in a crowded home (odds ratio [OR], 3.73; 95% confidence interval [CI], 1.41-9.88), having an adolescent mother (OR, 4.89; 95% CI, 1.37-17.38), lacking running water in the home (OR, 4.39; 95% CI, 1.11-17.38), incomplete vaccinations for age (OR, 3.39; 95% CI, 1.20-9.62), admission to a neonatal intensive care unit (OR, 7.17; 95% CI, 2.21-23.27), and no emergency department visit during the ARI episode (OR, 72.32; 95% CI, 4.82-1085.6). The at-home death rate due to respiratory syncytial virus infection among infants was 0.26 deaths/100 live births and that due to influenza was 0.07 deaths/1000 live births.

Conclusions

Social vulnerabilities underlie at-home mortality due to ARI. Mortality rates due to RSV and influenza virus infection are high among infants at home and are similar to those reported for hospitalized children.

Ability of verbal autopsy data to detect deaths due to uncontrolled hyperglycaemia: testing existing methods and development and validation of a novel weighted score

OBJECTIVES

Verbal autopsy (VA) is a useful tool to ascertain cause of death where no other mechanisms exist. We aimed to assess the utility of VA data to ascertain deaths due to uncontrolled hyperglycaemia and to develop a weighted score (WS) to specifically identify cases. Cases were identified by a study or site physician with training in diabetes. These diagnoses were also compared with diagnoses produced by a standard computer algorithm (InterVA-4).

SETTING

This study was done using VA data from the Health and Demographic Survey sites in Agincourt in rural South Africa. Validation of the WS was done using VA data from Karonga in Malawi.

PARTICIPANTS

All deaths from ages 1 to 49 years between 1992 and 2015 and between 2002 and 2016 from Agincourt and Karonga, respectively. There were 8699 relevant deaths in Agincourt and 1663 in Karonga.

RESULTS

Of the Agincourt deaths, there were 77 study physician classified cases and 58 computer algorithm classified cases. Agreement between study physician classified cases and computer algorithm classified cases was poor (Cohen's kappa 0.14). Our WS produced a receiver operator curve with area under the curve of 0.952 (95% CI 0.920 to 0.985). However, positive predictive value (PPV) was below 50% when the WS was applied to the development set and the score was dominated by the necessity for a premortem diagnosis of diabetes. Independent validation showed the WS performed reasonably against site physician classified cases with sensitivity of 86%, specificity of 99%, PPV of 60% and negative predictive value of 99%.

CONCLUSION

Our results suggest that widely used VA methodologies may be missing deaths due to uncontrolled hyperglycaemia. Our WS may offer improved ability to detect deaths due to uncontrolled hyperglycaemia in large populations studies where no other means exist.

Overview and Development of the Child Health and Mortality Prevention Surveillance Determination of Cause of Death (DeCoDe) Process and DeCoDe Diagnosis Standards

Mortality surveillance and cause of death data are instrumental in improving health, identifying diseases and conditions that cause a high burden of preventable deaths, and allocating resources to prevent these deaths. The Child Health and Mortality Prevention Surveillance (CHAMPS) network uses a standardized process to define, assign, and code causes of stillbirth and child death (<5 years of age) across the CHAMPS network. A Determination of Cause of Death (DeCoDe) panel composed of experts from a local CHAMPS site analyzes all available individual information, including laboratory, histopathology, abstracted clinical records, and verbal autopsy findings for each case and, if applicable, also for the mother. Using this information, the site panel ascertains the underlying cause (event that precipitated the fatal sequence of events) and other antecedent, immediate, and maternal causes of death in accordance with the International Classification of Diseases, Tenth Revision and the World Health Organization death certificate. Development and use of the CHAMPS diagnosis standards-a framework of required evidence to support cause of death determination-assures a homogenized procedure leading to a more consistent interpretation of complex data across the CHAMPS network. This and other standardizations ensures future comparability with other sources of mortality data produced externally to this project. Early lessons learned from implementation of DeCoDe in 5 CHAMPS sites in sub-Saharan Africa and Bangladesh have been incorporated into the DeCoDe process, and the implementation of DeCoDe has the potential to spur health systems improvements and local public health action.