Commentary

Verbal autopsy (VA) is a method of finding the cause of death by trained reviewers based on interviews with the next of kin or caregivers. In countries of the middle and low income group, many people die at home without coming in contact with the health systems, medically certified causes of death are unavailable, and vital registration systems are inadequate or absent. Hence, cause of death for the majority of cases is not known. However, knowledge about the causes of death in populations is essential for public health planning and resource allocation. VA serves an alternative method to determine the cause of death in such resource limited settings.

The paper entitled “Assessing the quality of evidence for verbal autopsy diagnosis of stroke in Vietnam”[1] shows that VA can be used as a tool for correctly ascertaining the cause of death in most cases. In fact, VA has been used extensively and validated time and again by several other studies[2–4] performed worldwide. VA is now an internationally accepted tool for ascertaining the cause of death.[256] It can be utilized for making health policies, for planning and evaluating health interventions, and to identify ways to reduce unnecessary deaths by authorities and governments. There are currently 36 demographic surveillance sites worldwide regularly using VA, mostly in Africa and Asia. The Sample Registration System (SRS) sites in India and the Disease Surveillance Points (DSP) system in China use VA on a large scale.[26] The diagnostic accuracy of the VA, i.e. sensitivity, specificity, predictive value positive and cause-specific mortality fraction, shows some variation according to the disease/condition[6–8]

The reported study[1] reaffirms the efficacy of VA as a diagnostic tool for finding the cause of death with content analysis substantiating its quality. It emphasizes that if standardized/key symptoms are used in the content of VA, not only it is easy for the respondents to recall and reaffirm but also the diagnosis is correct more often than not. The assessment of the medical evidence for stroke brings out that the quality of correctness of stroke diagnosis by the VA method is standard in the majority of cases (91.4%). The analysis of the VA procedure for stroke has been done in detail to bring out some finer points, e.g., age–sex differences, hospital admission to recall of key symptoms and duration, and site of paralysis with regard to hospital admission. A significantly higher positive response from respondents of hospital-admitted patients was seen as expected.

The issues of colloquial language and medical terminology as highlighted in the paper[1] are important in the sense that if questions are asked in the same language as that of the respondents, the accuracy of the diagnosis increases.[9] The analysis of content and quality of medical evidence reiterate the validity of properly trained reviewer physicians for diagnosis of cause of death by the VA procedure. There are several approaches to derive the cause of death from VA reports: physician review, predefined expert algorithms, and data-derived algorithms. The physician review is the most widely used approach for analysis of VA reports.[2]

Such types of further VA studies on different diseases/conditions under diverse settings are required to firmly establish VA as a fairly accurate alternative method for diagnosing cause of death in developing countries. Moreover, standardization needs to be established for obtaining internationally acceptable cause of death data which can be incorporated into national health systems as well as be suitable for international comparisons.