Tuberculosis (TB) and HIV/AIDS are public health concerns in Papua New Guinea (PNG). This study examines TB and HIV/AIDS mortalities and associated sociodemographic factors in PNG. Method: As part of a longitudinal study, verbal autopsy (VA) interviews were conducted using the WHO 2016 VA Instrument to collect data of 926 deaths occurred in the communities within the catchment areas of the Comprehensive Health and Epidemiological Surveillance System from 2018-2020. InterVA-5 cause of deaths analytic tool was used to assign specific causes of death (COD). Multinomial logistic regression analyses were conducted to identify associated sociodemographic factors, estimate odds ratios (OR), 95% confidential intervals and p-values. Result: TB and HIV/AIDS were the leading CODs from infectious diseases, attributed to 9% and 8% of the total deaths, respectively. Young adults (25-34 years) had the highest proportion of deaths from TB (20%) and the risk of dying from TB among this age group was five times more likely than those aged 75+ years (OR: 5.5 [1.4-21.7]). Urban population were 46% less likely to die from this disease compared rural ones (OR: 0.54 [0.3-1.0]). People from middle household wealth quintile were three times more likely to die from TB than those in the richest quintile (OR: 3.0 [1.3-7.4]). Young adults also had the highest proportion of deaths to HIV/AIDS (18%) and were nearly seven times more likely to die from this disease compared with those aged 75+ years (OR: 6.7 [1.7-25.4]). Males were 48% less likely to die from HIV/AIDS than females (OR: 0.52 [0.3-0.9]). The risk of dying from HIV/AIDS in urban population was 54% less likely than their rural counterparts (OR: 0.46 [0.2-0.9]). Conclusion: TB and HIV/AIDS interventions are needed to target high-risk and vulnerable populations to reduce premature mortality from these diseases in PNG.

Data source

Mortality surveillance data were extracted from the Comprehensive Health and Epidemiological Surveillance System (CHESS), operated since 2018 by Papua New Guinea Institute of Medical Research (PNGIMR). CHESS was based on the integrated Health and Demographic Surveillance System (iHDSS), which was established in PNG in the period 2010-2017. CHESS was designed as a population-based longitudinal follow-up cohort system. The overall purpose of CHESS was to provide a reliable and up-to-date data series for monitoring the implementation of socioeconomic development programmes and healthcare interventions at the sub-national level in PNG. CHESS catchment areas include eight surveillance sites located in six provinces: Eastern Highlands Province (EHP), East New Britain (ENB), East Sepik Province (ESP), Central, Madang, and Port Moresby (POM - the National Capital District). By the end 2022, CHESS will cover a population size of approximately 80,000, equivalent to 1% of the total population of PNG. The system provides population data from rural and urban sectors, with approximately 75% of rural and 25% of urban populations, comparable with the national rural-urban population distribution for the period 2018-2022. The designs and methods of iHDSS and CHESS have been previously published. Sociodemographic characteristics of the surveillance population by sites are presented in Table 1. The distance between urban-rural sites in EHP and ENB is about 50 km. This provides a balance between facilitating access and transportation and ensuring differences in socioeconomic development can be observed and captured in the data.

Mortality surveillance data were collected from the population living in the CHESS sites in the period 2018-2020, using the WHO 2016 verbal autopsy (VA) interview instrument. This tool is based on the consolidation of various existing VA tools and programmed for conducting VA interviews using portable electronic devices. The WHO 2016 tool does not require interviewers to have a health and medical background to conduct VA interviews. The WHO 2016 VA instrument was adapted in 2017 for optimal use in the local context and integrated into CHESS surveillance activities in 2018. An additional data module on identification information of the deceased, including household GPS data and individual ID code was included in the VA instrument for this study, allows linkage between mortality and household socioeconomic demographic data.

The data were collected from March 2018 to September 2020. Mortality data used in this study were collected from the communities. Deaths in the communities were identified by village-based data reporters. VA interviewers were conducted by national scientific officers, who work for CHESS in the demographic team. Among the 1021 deaths identified in the communities, consents were obtained for conducing 1003 VA interviews, resulted in a participation rate of 98%.

Data analyses

The InterVA-5 COD analytic tool was used to analyse cause of death (CODs) of 926 VA interviews. This computer-based programme can assign 64 specific CODs and categories in line with the International Classification of Diseases version 10 (ICD-10). To analyse mortalities from TB and from HIV/AIDS by selected sociodemographic factors, VA data were linked with the household socioeconomic (SES) data from the corresponding period of time using the unique household and individual identification codes. Specifically, the 2018-2020 VA data were linked with the 2018 household SES data. Mortality data from 665 deaths were successfully linked with household SES data and included in the analyses. No household SES data for ESP for 2018 were available as the site was not established until early 2019.

A new variable on household wealth index was constructed for each deceased using the principal component analysis (PCA) method. The application of PCA in the PNGIMR’s CHESS has been previously published. Household SES and demographic variables were included in PCA models. Significant variable remained in the PCA model including housing characteristics, water and sanitation, and household assets. Non significant variables were excluded from the models including education, employment, and occupation of the deceased. Household wealth indices were then divided into quintiles and categorised as poorest, poor, middle, richer and richest.

Multinomial Logistic Regression (MLR) was used to identify socioeconomic demographic factors associated with mortalities from TB and from HIV/AIDS and to predict the risk of morality from TB and HIV/AIDS across sub-populations. Two binary variables were created: (i) TB attributed death (Yes/No); and (ii) HIV/AIDS attributed death (Yes/No). These variables were included in MLR models as dependent variables, and sociodemographic factors were independent variables. Significant sociodemographic variables that remained in the MLR models included age at death, sex of the deceased, household wealth quintile, and urban-rural sector. The ‘province’ variable was excluded in the model because of confounding with the urban-rural sector variable. Main effect was selected to produce estimates of odds ratios (OR) for the risks of dying from TB and HIV/AIDS. Statistical likelihood tests were used to provide 95% confidence intervals (CIs) of the estimated ORs. A p-value of less than 0.05 was considered as significant. All analyses were performed using the Statistical Package for Social Sciences (SPSS-version 20).