INTRODUCTION
Accurate and population-representative cause of death (COD) statistics are vital for informing public health policy, addressing emerging health needs, tracking progress towards Sustainable Development Goals, and achieving national developmental milestones.1–3 However, globally, less than one-third of deaths are assigned a cause, with even fewer in Africa.3 The ideal source of population-representative COD statistics for mortality surveillance is a complete, universal, and continuous civil registration and vital statistics (CRVS) system.4,5
However, in many low- and middle-income countries (LMICs), CRVS systems struggle to provide timely, complete, and accurate vital statistics, with less than 20% of estimated deaths registered annually in Zambia.6,7
The gold standard procedure for determining COD is a complete diagnostic autopsy (CDA).8
For the subset of deaths that occur in health facilities where a CDA is neither available nor affordable, medical certification of COD (MCCOD) is conducted using all patient records and other medically relevant information about the terminal illness.8 However, in settings where CDAs or MCCODs are rare and cultural or infrastructural constraints exist, the WHO recommends the use of verbal autopsy (VA) methods.9 Verbal autopsy involves structured interviews with family members or caregivers of the deceased to ascertain probable causes of death using either physicians or computer algorithms.9 Despite limitations at the individual level, VA remains a crucial method for assigning COD among community deaths in LMICs lacking fully developed CRVS systems.7,10–12
In Zambia, VA has been successfully implemented in research settings, notably in the Sample Vital Registration with Verbal Autopsy (SAVVY) project, demonstrating the feasibility of collecting vital statistics using standardized methods. This was conducted as a pilot study in four provinces of Zambia from 2010 to 2011 and established that collecting vital statistics using standardized SAVVY methods was feasible in Zambia.13 Additionally, the Ministry of Home Affairs and Internal Security Department of National Registration Passports and Citizenry (DNRPC) and the Ministry of Health (MOH) have conducted VAs as part of routine death registration processes. Between January 2017 and September 2020, DNRPC’s implementation aimed to complement civil registration.11,13–15 This was implemented across eight sites in six districts of Zambia. This implementation was coordinated by DNRPC with technical assistance from the Bloomberg Data for Health Initiative (D4H) and the CDC Foundation. From 2019 to date, MOH’s expanded VA initiative across 30 sites in 23 districts supported by the President’s Emergency Plan for AIDS Relief through the United States Centre for Disease Control and Prevention-Zambia (U.S. CDC-Zambia) targeted the monitoring of HIV-associated mortality and improving death registration (Figures 1 and 2; Table 1).
Trendline of the weekly number of verbal autopsies conducted in Zambia, January 2020 to June 2024.
Citation: The American Journal of Tropical Medicine and Hygiene 112, 1; 10.4269/ajtmh.24-0302
Frequency of the top five probable causes of death within each age group, Zambia verbal autopsy, 2020–2024.
Citation: The American Journal of Tropical Medicine and Hygiene 112, 1; 10.4269/ajtmh.24-0302
Demographic characteristics of verbal autopsies done in Zambia, January 2020–June 2024
| Characteristics | N = 74,454 | |
|---|---|---|
| n | % | |
| Sex | ||
|  Female | 30,806 | 41.0 |
|  Male | 43,640 | 59.0 |
|  Unknown | 8 | <0.1 |
| Age, years, median (IQR) | 46 | (26.0–70.0) |
|  Unknown | 25 | – |
| Age group | ||
|  <1 | 6,462 | 8.7 |
|  1–5 | 4,080 | 5.5 |
|  6–18 | 3,821 | 5.1 |
|  19–50 | 26,951 | 36.0 |
|  >51 | 33,115 | 44.0 |
|  Unknown | 25 | – |
| Cause of death* | ||
|  Acute cardiac disease | 8,419 | 11.0 |
|  Other and unspecified cardiac disease | 8,174 | 11.0 |
|  HIV/AIDS-related death | 6,339 | 8.5 |
|  Stroke | 5,313 | 7.1 |
|  Acute respiratory infection, including pneumonia | 4,721 | 6.3 |
|  Diarrheal diseases | 4,615 | 6.2 |
|  Digestive neoplasms | 3,675 | 4.9 |
|  Pulmonary tuberculosis | 3,216 | 4.3 |
|  Indeterminate | 2,967 | 4.0 |
|  Diabetes mellitus | 2,718 | 3.7 |
|  Other†| 24,297 | 33.0 |
IQR = interquartile range.
Top 10 causes of death as ascribed by the InterVA-5.
All other probable causes of death assigned by the algorithm.
In 2022, the Zambia National Public Health Institute (ZNPHI) launched the country’s routine mortality surveillance program through the establishment of a mortality surveillance unit at ZNPHI. This program is designed to enhance mortality surveillance in Zambia by bringing together various stakeholders and scaling up the existing sentinel VA to a broader, population-level application. We apply the framework developed by de Savigny et al.3 to describe current VA implementation in Zambia and to develop recommendations for its improvement. This framework outlines the necessary system-level components for effective VA implementation, which include governance, design, operations, human resources, financing, infrastructure, logistics, information technologies, and data quality assurance (DQA).3
Governance.
The implementation of the VA in Zambia has, to date, always been research protocol based,11,13–15 and as such, there has been no national interagency oversight mechanism. In 2022, Zambia established a National Mortality Surveillance Subcommittee under the Monitoring and Evaluation Technical Working Group within the MOH. The subcommittee was created to address gaps in the coordination of mortality surveillance activities in Zambia, to support the routine use of mortality data for public health decision-making, and to bolster ongoing efforts to improve death registration within the Zambian CRVS. This subcommittee comprises members from key government ministries, departments, and institutions, nongovernmental organizations, and academia (Table 2).
Members of the Zambia Mortality Surveillance Subcommittee and their respective roles
| Stakeholder | Roles and Responsibilities |
|---|---|
| MOH | MOH works towards electronic notification of deaths and facilitates paper-based notification of deaths during the verbal autopsy. |
| DNRPC | DNRPC provides technical support to staff conducting death notification and verbal autopsy and provides/collects death notification forms. DNRPC working with the Ministry of Local Government and Rural Development and the MOH ensures registration of all deaths, including those that take place in rural areas and away from burial sites under local council jurisdiction. |
| Ministry of Local Government and Rural Development | Ensures the coordination of death registration and notifications from Local Councils and from the traditional leadership through the House of Chiefs |
| ZAMSTATS | Is responsible for the analysis and dissemination of information generated from mortality surveillance at the national and subnational levels |
| ZNPHI | Collaborates with DNRPC and ZAMSTATS to analyze and map mortality trends and patterns and shares this information with the other members. Any shifts from normal/expected trends are investigated and reported on by ZNPHI. ZNPHI additionally coordinates all the other stakeholders. |
| The University of Zambia | Leads in ensuring that information generated from the mortality surveillance program is used for research purposes. The university will ensure that results stemming from research are available and can be used for informed policy-making decisions. |
| The World Health Organization County office | Provides technical support |
| U.S. CDC-Zambia | Provides technical and financial support |
| Bloomberg Data for Health Initiative | Provides technical and financial support |
DNRPC = Department of National Registration Passports and Citizenry; MOH = Ministry of Health; ZAMSTATS = Zambia Statistics Agency; ZNPHI = Zambia National Public Health Institute.
Design, operations, information technology (IT), infrastructure, logistics, and human resources.
Purposive sampling was used to select districts for inclusion in both systems. In the DNRPC system, sites were chosen based on districts exhibiting high death registration completeness and well established death notification systems. In the MOH system, district selection was limited to provinces where the U.S. CDC-Zambia was currently implementing activities. Within these provinces, high HIV morbidity and mortality served as key inclusion criteria for selecting districts for VA implementation. Both systems used passive data collection methods, relying on the decedent’s next of kin to report the death while accessing burial services. Potential interviewees were identified among mourners as they brought the deceased to the health facility mortuary. At most sites, VA interviewers have dedicated office space but often share it with mortuary attendants or DNRPC officers stationed at health facilities. Verbal autopsy was promptly conducted after the appropriate respondent was identified at the health facility mortuary as the family accessed burial services. The MOH has been implementing tele-VAs to a limited extent, using phone calls to conduct VA interviews remotely.
During data collection, DNRPC initially piloted the electronic SmartVA Questionnaire system,16 paper based, and then later transitioned to the WHO 2016 Questionnaire17 on the Open Data Kit (ODK)18 platform. The D4H provided technical support in system design and questionnaire translation, whereas DNRPC provided IT support staff and systems for transmission and storage through the government-wide area network. Conversely, the MOH VA system began with ODK and is currently transitioning to a KOBO-based data collection platform.19 Both electronic systems transmitted completed VA forms to an online server, which applied an algorithm to assign a probable cause of death (SmartVA Analyze20 and Inter-VA-421 [DNRPC] and Inter-VA-522 [MOH]). These different algorithms are available on the openVA platform, the first platform to offer a standardized framework for analyzing VA data, ensuring compatibility with all publicly available methods and data structures.23 To a limited extent, DNRPC used physician-coded VAs to assign a cause of death. The VA questionnaire was locally adapted and translated into the three most widely spoken local languages (Bemba, Tonga, and Nyanja) in the sentinel sites.
Regarding human resources, full-time project staff with a minimum qualification of a grade 12 General Certificate of Education were hired to serve as VA interviewers by both the MOH and DNRPC. The DNRPC recruited and trained 25 interviewers and a data manager. Meanwhile, the MOH established several full-time project positions for VA implementation: HIV-associated mortality surveillance advisor (1), mortality surveillance epidemiologist (1), mortality surveillance data manager (1), field mortality surveillance officers (55 VA interviewers), and a driver (1). After recruitment, VA interviewers underwent training before field deployment, with subsequent refresher training as needed. The training curriculum for VA interviewers in both the DNRPC and the MOH included the following topics: an overview of VA, the roles of VA interviewers, the WHO 2016 VA questionnaire (covering neonates, children, and adults), interview techniques, ethics, and VA administration using tablets.
Despite the use of electronic data platforms for collecting, transmitting, and analyzing VA data, there were no connections between these IT systems and other electronic health/civil registration platforms. This lack of linkage has hindered the timely release of VA data for public health surveillance and civil registration purposes. As both systems relied on passive data collection from large referral health facilities, deaths occurring within rural communities were underrepresented in both systems. Furthermore, because study sites were nonrandomly selected, neither system generated population-representative VA data. These findings underscore the necessity for study designs that encompass national as well as subnational representativeness and highlight the importance of community-based surveillance platforms capable of capturing all deaths, regardless of location.24–26
Financing.
The primary source of funding for the implementation of VA in Zambia has been donor support, supplemented by funding from the Government of the Republic of Zambia (GRZ). In the DNRPC implementation, donor support came from D4H, the U.S. CDC-Zambia/CDC Foundation, and the Global Fund, providing financial and technical assistance. In the MOH implementation, U.S. CDC-Zambia provided financial and technical aid through the U.S. President’s Emergency Plan for AIDS Relief, whereas the CDC Foundation, through D4H, offered technical assistance. The GRZ, through the MOH and DNRPC, has supported VA implementation through infrastructure and human resources. We could not calculate the average cost per VA in both systems due to a lack of information. However, the main cost drivers in both systems were training and retraining project staff, procuring IT infrastructure, and remunerating VA interviewers and their supervisors.
Data quality assurance.
During the DNRPC implementation, there was no routine DQA. However, in the MOH system, DQA occurs in two phases, during data collection and through on-site data audits. During data collection, the ODK/KOBO platform incorporates internal checks to ensure the consistency of the entered information. Once collected, these data are transmitted directly to the national server, where both automated and manual data checks are conducted. The on-site data audits are conducted using a structured tool. This tool compares the number of VAs in the Verbal Autopsy Explorer,27 an online dashboard tracking conducted VAs, with the number of community deaths recorded in mortuary registers and the District Health Information System-2. There is currently no monitoring and evaluation framework in place to oversee this implementation.
Summary and recommendations for population-representative VA in Zambia.
The implementation of VA in Zambia has improved mortality surveillance by providing timely information on causes of death among community deaths using electronic data capture and cause-of-death determination tools. Achievements include successful pilot projects such as the SAVVY, expanded VA initiatives across multiple districts (DNRPC and MOH implementation), and substantial technical and financial support from donors and the government. Better coordination of mortality surveillance strengthening activities has been achieved through the National Mortality Surveillance Subcommittee. However, limitations include the underrepresentation of rural areas, lack of electronic integration with CRVS and public health systems, non-population-representative data due to purposive sampling, and a heavy reliance on donor funding.
To better coordinate mortality surveillance activities in Zambia, we propose establishing a dedicated interministerial Mortality Surveillance Coordinating Committee. This committee would oversee the expansion of VA initiatives and the establishment of vital connections with the CRVS with technical support from ZNPHI’s mortality surveillance unit. Given the historical reliance on donor support for funding VA initiatives in Zambia, it is imperative to explore avenues for gradually increasing funding through government mechanisms. Sustainable financing strategies will be essential for maintaining the long-term viability of VA programs and ensuring their continued impact on public health.
To address the need for population-representative VA data, we suggest adopting sample-based systems for VA data collection. This approach will help ensure a more comprehensive understanding of mortality patterns across different regions of Zambia. To facilitate the timely collection and utilization of VA data, we recommend integrating electronic data collection and analysis platforms with other electronic public health response and CRVS systems. Further customization of the VA questionnaires into more local languages is required to promote inclusivity and adapt to the diverse linguistic landscape of Zambia (Table 3).
Comparison of various elements of the verbal autopsy systems implemented by the Ministry of Home Affairs and Internal Security and the Ministry of Health, Zambia 2016–2023
| Area | MOHAIS | MOH | Recommendation for Future State |
|---|---|---|---|
| Governance and finance | Project based (D4H, U.S. CDC-Zambia/CDC Foundation, Global Fund) | Project based (PEPFAR, U.S. CDC-Zambia, CDC Foundation through the D4H) | Use the project base to advocate for greater government funding |
| Design | Non-population-representative sample; purposive sampling in 7 sites | Non-population-representative sample; purposive sample selected for high HIV burden districts; 30 sites | A nationally and subnationally representative population-based sample |
| Operations | Passive system | Passive system | Active and passive system |
| Human resources | Project staff | Project staff | Project staff and Government of the Republic of Zambia employees across key ministries involved in mortality surveillance |
| IT and data quality assurance mechanisms | Electronic data management system | Electronic data management systems | Electronic data management systems and system integration between data producers and data users |
D4H = Bloomberg Data for Health Initiative; IT = information technology; MOH = Ministry of Health; MOHAIS = Ministry of Home Affairs and Internal Security; PEPFAR = President’s Emergency Plan for AIDS Relief; U.S. CDC-Zambia = United States Centre for Disease Control and Prevention-Zambia.
ACKNOWLEDGMENTS
We thank the Zambian Government through the Ministry of Health, Ministry of Home Affairs, Ministry of Local Government and Rural Development, and the Zambia National Public Health Institute. We also thank the U.S. CDC National Centre for Health Statistics.
REFERENCES
- 1.↑
United Nations Department of Economic and Social Affairs, 2024. Sustainable Development. The 17 Goals. Available at: https://sdgs.un.org/goals. Accessed August 13, 2023.
- 2.↑
Ministry of Finance and National Planning, 2022. 8NDP (2022–2026). Available at: https://www.mofnp.gov.zm/?wpdmpro=8ndp-2022-2026. Accessed August 13, 2023.
- 3.↑
de Savigny D, et al., 2017. Integrating community-based verbal autopsy into civil registration and vital statistics (CRVS): System-level considerations. Glob Health Action 10: 1272882.
- 4.↑
World Health Organization, 2024. Civil Registration and Vital Statistics (CRVS). Available at: https://www.who.int/data/data-collection-tools/civil-registration-and-vital-statistics-(crvs). Accessed July 10, 2024.
- 5.↑
Mills S, Lee JK, Rassekh BM, 2019. An introduction to the civil registration and vital statistics systems with applications in low- and middle-income countries. J Health Popul Nutr 38 (Suppl 1 ):23.
- 6.↑
Zambia Statistics Agency, 2022. Quality Statistics for Development. Available at: https://www.zamstats.gov.zm/. Accessed August 15, 2023.
- 7.↑
Firth SM, et al., 2021. Integrating community-based verbal autopsy into civil registration and vital statistics: Lessons learnt from five countries. BMJ Glob Health 6: e006760.
- 8.↑
World Health Organization, 2017. Clinical Criteria for the Determination of Death: WHO Technical Expert Consultation. Geneva, Switzerland: WHO.
- 9.↑
World Health Organization, 2023. Cause of Death. Available at: https://www.who.int/standards/classifications/classification-of-diseases/cause-of-death. Accessed August 15, 2023.
- 10.↑
Rahman MH, Macicame I, Wilson E, Nhachungue S, Amouzou A, 2023. Verbal and social autopsy of adult deaths and adult care-seeking pattern in Mozambique, 2019–2020. Am J Trop Med Hyg 108: 17–28.
- 11.↑
Mapoma CC, Munkombwe B, Mwango C, Bwalya BB, Kalindi A, Gona NP, 2021. Application of verbal autopsy in routine civil registration in Lusaka District of Zambia. BMC Health Serv Res 21: 408.
- 12.↑
World Health Organization, 2020. What’s the Cause? Certifying Deaths in Sub-Saharan Africa. Available at: https://www.afro.who.int/news/whats-cause-certifying-deaths-sub-saharan-africa. Accessed August 15, 2023.
- 13.↑
Mudenda SS, Kamocha S, Mswia R, Conkling M, Sikanyiti P, Potter D, Mayaka WC, Marx MA, 2011. Feasibility of using a World Health Organization-standard methodology for sample vital registration with verbal autopsy (SAVVY) to report leading causes of death in Zambia: Results of a pilot in four provinces, 2010. Popul Health Metr 9: 40.
- 14.↑
Yokobori Y, Matsuura J, Sugiura Y, Mutemba C, Nyahoda M, Mwango C, Kazhumbula L, Yuasa M, Chiluba C, 2020. Analysis of causes of death among brought-in-dead cases in a third-level hospital in Lusaka, Republic of Zambia, using the tariff method 2.0 for verbal autopsy: A cross-sectional study. BMC Public Health 20: 473.
- 15.↑
Chisumpa VH, Odimegwu CO, Saikia N, 2019. Adult mortality in sub-Saharan Africa: Cross-sectional study of causes of death in Zambia. Trop Med Int Health 24: 1208–1220.
- 16.↑
University of Melbourne, 2016. Improving Cause of Death Information: Smart VA: Technical User Manual: Resources and Tools 7. Carlton, Victoria, Australia: University of Melbourne.
- 17.↑
World Health Organization, 2016. Verbal Autopsy Standards: The 2016 WHO Verbal Autopsy Instrument. Available at: https://www.who.int/publications/m/item/verbal-autopsy-standards-the-2016-who-verbal-autopsy-instrument. Accessed May 8, 2024.
- 20.↑
GitHub, 2022 ihmeuw/SmartVA-Analyze: SmartVA-Analyze Desktop Application. Available at: https://github.com/ihmeuw/SmartVA-Analyze. Accessed May 8, 2024.
- 21.↑
RDocumentation, 2019. InterVA4 package (version 1.7.6). Available at: https://www.rdocumentation.org/packages/InterVA4/versions/1.7.6. Accessed May 8, 2024.
- 22.↑
GitHub, 2020 peterbyass/InterVA-5: Files Relating to Evaluation of InterVA-5 Software. Available at: https://github.com/peterbyass/InterVA-5. Accessed May 8, 2024.
- 23.↑
Li ZR, Thomas J, Choi E, Mccormick TH, Clark SJ, 2022. The openVA toolkit for verbal autopsies. R J 14: 316–334.
- 24.↑
Amouzou A, Kante A, Macicame I, Antonio A, Gudo E, Duce P, Black RE, 2020. National sample vital registration system: A sustainable platform for COVID-19 and other infectious diseases surveillance in low and middle-income countries. J Glob Health 10: 020368.
- 25.↑
Setel PW, Sankoh O, Rao C, Velkoff VA, Mathers C, Gonghuan Y, Hemed Y, Jha P, Lopez AD, 2005. Sample registration of vital events with verbal autopsy: A renewed commitment to measuring and monitoring vital statistics. Bull World Health Organ 83: 611–617.
- 26.↑
Nkengasong J, Gudo E, Macicame I, Maunze X, Amouzou A, Banke K, Dowell S, Jani I, 2020. Improving birth and death data for African decision making. Lancet Glob Health 8: e35–e36.
- 27.↑
VA Explorer Docs, 2024. VA Explorer: A Full Service VA Management Prototype. Available at: https://va-explorer.github.io/va_explorer/. Accessed July 24, 2024.