Performance of four computer-coded verbal autopsy methods for cause of death assignment compared with physician coding on 24,000 deaths in low- and middle-income countries

Abstract

Background: Physician-coded verbal autopsy (PCVA) is the most widely used method to determine causes of death (CODs) in countries where medical certification of death is uncommon. Computer-coded verbal autopsy (CCVA) methods have been proposed as a faster and cheaper alternative to PCVA, though they have not been widely compared to PCVA or to each other.

Methods: We compared the performance of open-source random forest, open-source tariff method, InterVA-4, and the King-Lu method to PCVA on five datasets comprising over 24,000 verbal autopsies from low- and middle-income countries. Metrics to assess performance were positive predictive value and partial chance-corrected concordance at the individual level, and cause-specific mortality fraction accuracy and cause-specific mortality fraction error at the population level.

Results: The positive predictive value for the most probable COD predicted by the four CCVA methods averaged about 43% to 44% across the datasets. The average positive predictive value improved for the top three most probable CODs, with greater improvements for open-source random forest (69%) and open-source tariff method (68%) than for InterVA-4 (62%). The average partial chance-corrected concordance for the most probable COD predicted by the open-source random forest, open-source tariff method and InterVA-4 were 41%, 40% and 41%, respectively, with better results for the top three most probable CODs. Performance generally improved with larger datasets. At the population level, the King-Lu method had the highest average cause-specific mortality fraction accuracy across all five datasets (91%), followed by InterVA-4 (72% across three datasets), open-source random forest (71%) and open-source tariff method (54%).

Conclusions: On an individual level, no single method was able to replicate the physician assignment of COD more than about half the time. At the population level, the King-Lu method was the best method to estimate cause-specific mortality fractions, though it does not assign individual CODs. Future testing should focus on combining different computer-coded verbal autopsy tools, paired with PCVA strengths. This includes using open-source tools applied to larger and varied datasets (especially those including a random sample of deaths drawn from the population), so as to establish the performance for age- and sex-specific CODs.

Figure 1

Comparison of open-source random forest to IHME random forest. The IHME random forest was tested on a set of IHME hospital-based data, both with and without health care experience (HCE) variables. HCE variables are binary questions on previous medically diagnosed conditions (including high blood pressure, tuberculosis, cancer), and details transcribed from the respondents’ medical records. Our IHME subset contained some, but not all, HCE variables. The ORF performance was similar to the IHME random forest method on the full hospital-based dataset without HCE variables, but performed less well when HCE variables were included [12]. HCE, health care experience; IHME, Institute for Health Metrics and Evaluation; ORF, open-source random forest.

Figure 2

Comparison of open-source tariff method to IHME tariff method. The IHME random forest was tested on a set of IHME hospital-based data, both with and without health care experience (HCE) variables. The ORF was tested on a subset of the full IHME data, containing some, but not all, HCE variables. The OTM performed almost exactly as the similar IHME method on the full hospital-based dataset without HCE variables (for the top cause), but less well than the same IHME analysis with HCE variables. Note that results for the full IHME dataset without HCE were only available for the top assigned cause [13]. HCE, health care experience; IHME, Institute for Health Metrics and Evaluation; OTM, open-source tariff method.