Automated versus physician assignment of cause of death for verbal autopsies: randomized trial of 9374 deaths in 117 villages in India.

BACKGROUND: Verbal autopsies with physician assignment of cause of death (COD) are commonly used in settings where medical certification of deaths is uncommon. It remains unanswered if automated algorithms can replace physician assignment. METHODS: We randomized verbal autopsy interviews for deaths in 117 villages in rural India to either physician or automated COD assignment. Twenty-four trained lay (non-medical) surveyors applied the allocated method using a laptop-based electronic system. Two of 25 physicians were allocated randomly to independently code the deaths in the physician assignment arm. Six algorithms (Naïve Bayes Classifier (NBC), King-Lu, InSilicoVA, InSilicoVA-NT, InterVA-4, and SmartVA) coded each death in the automated arm. The primary outcome was concordance with the COD distribution in the standard physician-assigned arm. Four thousand six hundred fifty-one (4651) deaths were allocated to physician (standard), and 4723 to automated arms. RESULTS: The two arms were nearly identical in demographics and key symptom patterns. The average concordances of automated algorithms with the standard were 62%, 56%, and 59% for adult, child, and neonatal deaths, respectively. Automated algorithms showed inconsistent results, even for causes that are relatively easy to identify such as road traffic injuries. Automated algorithms underestimated the number of cancer and suicide deaths in adults and overestimated other injuries in adults and children. Across all ages, average weighted concordance with the standard was 62% (range 79-45%) with the best to worst ranking automated algorithms being InterVA-4, InSilicoVA-NT, InSilicoVA, SmartVA, NBC, and King-Lu. Individual-level sensitivity for causes of adult deaths in the automated arm was low between the algorithms but high between two independent physicians in the physician arm. CONCLUSIONS: While desirable, automated algorithms require further development and rigorous evaluation. Lay reporting of deaths paired with physician COD assignment of verbal autopsies, despite some limitations, remains a practicable method to document the patterns of mortality reliably for unattended deaths. TRIAL REGISTRATION: ClinicalTrials.gov , NCT02810366. Submitted on 11 April 2016.

Child, maternal, and adult mortality in rural Ethiopia in 2019: a cross-sectional mortality survey using electronic verbal autopsies.

Background: Ethiopia, with about 10% of Africa's population, has little direct information on causes of death, particularly in rural areas where 80% of Ethiopians live. In 2019-2020, we conducted electronic verbal autopsies (e-VA) to examine causes of death and quantify cause-specific mortality rates in rural Ethiopia. Methods: We examined deaths under 70 years in the three years prior to the survey dates (November 25, 2019-February 29, 2020) among 2% of East Gojjam Zone (Amhara Region) using registered deaths and adding random sampling in this cross-sectional study. Trained surveyors interviewed relatives of the deceased with central dual-physician assignment of causes as the main outcome. We documented details on age, sex and location of death, and derived overall rural death rates using 2007 Census data and the United Nations national estimates for 2019. To these, we applied our sample-weighted causes to derive cause-specific mortality rates. We calculated death risks for the leading causes for major age groups. Findings: We studied 3516 deaths: 55% male, 97% rural, and 68% occurring at home. At ages 5 and older, injuries were notable, accounting for over a third of deaths at 5-14 years, half of the deaths at ages 15-29 years, and a quarter of deaths at ages 30-69 years. Neonatal mortality was high, mostly from prematurity/low birthweight and infections. Among children under 5 (excluding neonates), infections caused nearly two-thirds of deaths. Most maternal deaths (84%) arose from direct causes. After injuries, especially suicide, assaults, and road traffic accidents, vascular disease (15%) and cancer (13%) were the leading causes among adults at 30-69 years. HIV/AIDS and tuberculosis deaths were also important causes among adults. Interpretation: Rural Ethiopia has a high burden of avoidable mortality, particularly injury, including suicide, assaults, and road traffic accidents. Funding: International Development Research Centre, and the Canadian Institutes of Health Research.

Enhancing the value of death registration with verbal autopsy data: a pilot study in the Senegalese urban population in 2019.

BACKGROUND: There is no source of data on causes of death in Senegal that covers both community and hospital deaths. Yet the death registration system in the Dakar region is relatively complete (>80%) and could be expanded to provide information on the diseases and injuries that led to death. METHODS: In this pilot study, we recorded all deaths that occurred over 2 months and were reported in the 72 civil registration offices in the Dakar region. We selected the deaths of residents of the region and administered a verbal autopsy to a relative of the deceased to identify the underlying causes of death. Causes of death were assigned using the InterVA5 model. RESULTS: The age structure of deaths registered at the civil registry differed from that of the census, with a proportion of infant deaths about twice as high as in the census. The main causes of death were prematurity and obstetric asphyxia in newborns. Meningitis and encephalitis, severe malnutrition, and acute respiratory infections were the leading causes from 1 month to 15 years of age. Cardiovascular diseases accounted for 27% of deaths in adults aged 15-64 and 45% of deaths among adults above age 65, while neoplasms accounted for 20% and 12% of deaths in these two age groups, respectively. CONCLUSIONS: This study demonstrates that the epidemiological transition is at an advanced stage in urban areas of Dakar, and underlines the importance of conducting regular studies based on verbal autopsies of deaths reported in civil registration offices.

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.