Verbal autopsy interpretation: a comparative analysis of the InterVA model versus physician review in determining causes of death in the Nairobi DSS

Abstract

BackgroundDeveloping countries generally lack complete vital registration systems that can produce cause of death information for health planning in their populations. As an alternative, verbal autopsy (VA) - the process of interviewing family members or caregivers on the circumstances leading to death - is often used by Demographic Surveillance Systems to generate cause of death data. Physician review (PR) is the most common method of interpreting VA, but this method is a time- and resource-intensive process and is liable to produce inconsistent results. The aim of this paper is to explore how a computer-based probabilistic model, InterVA, performs in comparison with PR in interpreting VA data in the Nairobi Urban Health and Demographic Surveillance System (NUHDSS).MethodsBetween August 2002 and December 2008, a total of 1,823 VA interviews were reviewed by physicians in the NUHDSS. Data on these interviews were entered into the InterVA model for interpretation. Cause-specific mortality fractions were then derived from the cause of death data generated by the physicians and by the model. We then estimated the level of agreement between both methods using Kappa statistics.ResultsThe level of agreement between individual causes of death assigned by both methods was only 35% (κ = 0.27, 95% CI: 0.25 - 0.30). However, the patterns of mortality as determined by both methods showed a high burden of infectious diseases, including HIV/AIDS, tuberculosis, and pneumonia, in the study population. These mortality patterns are consistent with existing knowledge on the burden of disease in underdeveloped communities in Africa.ConclusionsThe InterVA model showed promising results as a community-level tool for generating cause of death data from VAs. We recommend further refinement to the model, its adaptation to suit local contexts, and its continued validation with more extensive data from different settings.

Applying the InterVA-4 model to determine causes of death in rural Ethiopia

“…TB was also identified as the leading communicable CoD in other similar studies, but the estimates attributed to TB were higher than in our findings (12.5%); 36% in Ethiopia and 31% in Nairobi, Kenya (2, 15). A lower prevalence of TB than the national estimate was also reported in a recent survey in the region where our study was conducted (24).…”

Pattern of all-causes and cause-specific mortality in an area with progressively declining malaria burden in Korogwe district, north-eastern Tanzania

“…However the proportion of deaths which occurred out of health facilities in this site was lower (by approximately 10%) compared to the proportion previously reported in Nairobi HDSS, Kenya [4]. Similarly, a majority of deaths had occurred at home, as shown by other studies conducted in Muheza district (72.4%) [43] and in Kenya (87.0%) [17].…”

“…Such systems are limited in most developing countries especially in sub-Saharan Africa (SSA) [3]. Unfortunately, in these countries, about 80% of deaths occur outside health facilities [4] and even the few deaths which occur at health facilities are usually not recorded or the records are inconsistent and unreliable [5]. To complement that, some of these countries adopted and introduced a health and demographic surveillance system (HDSS) which is routinely conducted at specific sites as the platform for generating supportive vital events, including mortality data [6].…”

Cause of death during 2009–2012, using a probabilistic model (InterVA-4): an experience from Ballabgarh Health and Demographic Surveillance System in India

“…Level of agreement of this method with what is considered to be a gold standard for assigning COD using VA (medical death certificate) has ranged from 40 to 83% in different settings. The level of agreement was found to be better when the diseases were classified under broader groups rather than when compared to more specific CODs and also agreement varied for different groups (13, 19, 28, 29). Since we had used a broader classification of diseases, we expect a high degree of agreement with the physician CODs.…”