Deriving causes of child mortality by re-analyzing national verbal autopsy data applying a standardized computer algorithm in Uganda, Rwanda and Ghana

Perinatal death in Northern Uganda: incidence and risk factors in a community-based prospective cohort study

Background: Perinatal mortality in Uganda remains high at 38 deaths/1,000 births, an estimate greater than the every newborn action plan (ENAP) target of ≤24/1,000 births by 2030. To improve perinatal survival, there is a need to understand the persisting risk factors for death. Objective: We determined the incidence, risk factors, and causes of perinatal death in Lira district, Northern Uganda. Methods: This was a community-based prospective cohort study among pregnant women in Lira district, Northern Uganda. Female community volunteers identified pregnant women in each household who were recruited at ≥28 weeks of gestation and followed until 50 days postpartum. Information on perinatal survival was gathered from participants within 24 hours after childbirth and at 7 days postpartum. The cause of death was ascertained using verbal autopsies. We used generalized estimating equations of the Poisson family to determine the risk factors for perinatal death. Results: Of the 1,877 women enrolled, the majority were ≤30 years old (79.8%), married or cohabiting (91.3%), and had attained only a primary education (77.7%). There were 81 perinatal deaths among them, giving a perinatal mortality rate of 43/1,000 births [95% confidence interval (95% CI: 35, 53)], of these 37 were stillbirths (20 deaths/1,000 total births) and 44 were early neonatal deaths (23 deaths/1,000 live births). Birth asphyxia, respiratory failure, infections and intra-partum events were the major probable contributors to perinatal death. The risk factors for perinatal death were nulliparity at enrolment (adjusted IRR 2.7, [95% CI: 1.3, 5.6]) and maternal age >30 years (adjusted IRR 2.5, [95% CI: 1.1, 5.8]). Conclusion: The incidence of perinatal death in this region was higher than had previously been reported in Uganda. Risk factors for perinatal mortality were nulliparity and maternal age >30 years. Pregnant women in this region need improved access to care during pregnancy and childbirth.

Causes of death and predictors of childhood mortality in Rwanda: a matched case-control study using verbal social autopsy

Background

Rwanda has dramatically reduced child mortality, but the causes and sociodemographic drivers for mortality are poorly understood.

Methods

We conducted a matched case-control study of all children who died before 5 years of age in eastern Rwanda between 1st March 2013 and 28th February 2014 to identify causes and risk factors for death. We identified deaths at the facility level and via a community health worker reporting system. We used verbal social autopsy to interview caregivers of deceased children and controls matched by area and age. We used InterVA4 to determine probable causes of death and cause-specific mortality fractions, and utilized conditional logistic regression to identify clinical, family, and household risk factors for death.

Results

We identified 618 deaths including 174 (28.2%) in neonates and 444 (71.8%) in non-neonates. The most commonly identified causes of death were pneumonia, birth asphyxia, and meningitis among neonates and malaria, acute respiratory infections, and HIV/AIDS-related death among non-neonates. Among neonates, 54 (31.0%) deaths occurred at home and for non-neonates 242 (54.5%) deaths occurred at home. Factors associated with neonatal death included home birth (aOR: 2.0; 95% CI: 1.4–2.8), multiple gestation (aOR: 2.1; 95% CI: 1.3–3.5), both parents deceased (aOR: 4.7; 95% CI: 1.5–15.3), mothers non-use of family planning (aOR: 0.8; 95% CI: 0.6–1.0), lack of accompanying person (aOR: 1.6; 95% CI: 1.1–2.1), and a caregiver who assessed the medical services they received as moderate to poor (aOR: 1.5; 95% CI: 1.2–1.9). Factors associated with non-neonatal deaths included multiple gestation (aOR: 2.8; 95% CI: 1.7–4.8), lack of adequate vaccinations (aOR: 1.7; 95% CI: 1.2–2.3), household size (aOR: 1.2; 95% CI: 1.0–1.4), maternal education levels (aOR: 1.9; 95% CI: 1.2–3.1), mothers non-use of family planning (aOR: 1.6; 95% CI: 1.4–1.8), and lack of household electricity (aOR: 1.4; 95% CI: 1.0–1.8).

Conclusion

In the context of rapidly declining childhood mortality in Rwanda and increased access to health care, we found a large proportion of remaining deaths occur at home, with home deliveries still representing a significant risk factor for neonatal death. The major causes of death at a population level remain largely avoidable communicable diseases. Household characteristics associated with death included well-established socioeconomic and care-seeking risk factors.

Integrating community-based verbal autopsy into civil registration and vital statistics (CRVS): system-level considerations

Background: Reliable and representative cause of death (COD) statistics are essential to inform public health policy, respond to emerging health needs, and document progress towards Sustainable Development Goals. However, less than one-third of deaths worldwide are assigned a cause. Civil registration and vital statistics (CRVS) systems in low- and lower-middle-income countries are failing to provide timely, complete and accurate vital statistics, and it will still be some time before they can provide physician-certified COD for every death.

Proposals: Verbal autopsy (VA) is a method to ascertain the probable COD and, although imperfect, it is the best alternative in the absence of medical certification. There is extensive experience with VA in research settings but only a few examples of its use on a large scale. Data collection using electronic questionnaires on mobile devices and computer algorithms to analyse responses and estimate probable COD have increased the potential for VA to be routinely applied in CRVS systems. However, a number of CRVS and health system integration issues should be considered in planning, piloting and implementing a system-wide intervention such as VA. These include addressing the multiplicity of stakeholders and sub-systems involved, integration with existing CRVS work processes and information flows, linking VA results to civil registration records, information technology requirements and data quality assurance.

Conclusions: Integrating VA within CRVS systems is not simply a technical undertaking. It will have profound system-wide effects that should be carefully considered when planning for an effective implementation. This paper identifies and discusses the major system-level issues and emerging practices, provides a planning checklist of system-level considerations and proposes an overview for how VA can be integrated into routine CRVS systems.

Validating hierarchical verbal autopsy expert algorithms in a large data set with known causes of death

BACKGROUND

Physician assessment historically has been the most common method of analyzing verbal autopsy (VA) data. Recently, the World Health Organization endorsed two automated methods, Tariff 2.0 and InterVA-4, which promise greater objectivity and lower cost. A disadvantage of the Tariff method is that it requires a training data set from a prior validation study, while InterVA relies on clinically specified conditional probabilities. We undertook to validate the hierarchical expert algorithm analysis of VA data, an automated, intuitive, deterministic method that does not require a training data set.

METHODS

Using Population Health Metrics Research Consortium study hospital source data, we compared the primary causes of 1629 neonatal and 1456 1-59 month-old child deaths from VA expert algorithms arranged in a hierarchy to their reference standard causes. The expert algorithms were held constant, while five prior and one new "compromise" neonatal hierarchy, and three former child hierarchies were tested. For each comparison, the reference standard data were resampled 1000 times within the range of cause-specific mortality fractions (CSMF) for one of three approximated community scenarios in the 2013 WHO global causes of death, plus one random mortality cause proportions scenario. We utilized CSMF accuracy to assess overall population-level validity, and the absolute difference between VA and reference standard CSMFs to examine particular causes. Chance-corrected concordance (CCC) and Cohen's kappa were used to evaluate individual-level cause assignment.

RESULTS

Overall CSMF accuracy for the best-performing expert algorithm hierarchy was 0.80 (range 0.57-0.96) for neonatal deaths and 0.76 (0.50-0.97) for child deaths. Performance for particular causes of death varied, with fairly flat estimated CSMF over a range of reference values for several causes. Performance at the individual diagnosis level was also less favorable than that for overall CSMF (neonatal: best CCC = 0.23, range 0.16-0.33; best kappa = 0.29, 0.23-0.35; child: best CCC = 0.40, 0.19-0.45; best kappa = 0.29, 0.07-0.35).

CONCLUSIONS

Expert algorithms in a hierarchy offer an accessible, automated method for assigning VA causes of death. Overall population-level accuracy is similar to that of more complex machine learning methods, but without need for a training data set from a prior validation study.