gms | German Medical Science

50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds)
12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie (dae)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie
Deutsche Arbeitsgemeinschaft für Epidemiologie

12. bis 15.09.2005, Freiburg im Breisgau

Cause-specific mortality in North-Western Burkina Faso 1999-2003

Meeting Abstract

  • Gaël P Hammer - Universität Heidelberg, Heidelberg
  • Florend Somé - Centre de Recherche en Santé de Nouna, Nouna, Burkina Faso
  • Olar Müller - Universität Heidelberg, Heidelberg
  • Gisela Kynast-Wolf - Universität Heidelberg, Heidelberg
  • Heiko Becher - Universität Heidelberg, Heidelberg

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. Deutsche Arbeitsgemeinschaft für Epidemiologie. 50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie. Freiburg im Breisgau, 12.-15.09.2005. Düsseldorf, Köln: German Medical Science; 2005. Doc05gmds224

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/gmds2005/05gmds105.shtml

Published: September 8, 2005

© 2005 Hammer et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Introduction

Despite international efforts, Mortality in Africa remains the highest in the world [1]. Information on causes of mortality is necessary to better target health interventions, yet such data remain sparse for Africa.

In many developing countries, Demographic Surveillance Systems (DSS) are currently the only reliable source of longitudinal population data. Like a huge open cohort study, they monitor the entire population of a defined (small) geographical area, and serve as a platform for health intervention trials. They usually collect information on probable cause of death (COD) by the verbal autopsy (VA) method [2].

The Verbal Autopsy method consists of gathering information on disease symptoms by standardized questionnaires completed by trained interviewers. The usual way of assigning probable COD is the independent assessment by two physicians (which allows a certain degree of subjectivity). Despite it’s limited sensitivity and specificity, VA has been shown to be of great usefulness in developing countries.

Neither the infrequent population censuses, nor the MEASURE DHS+ surveys [3], which are nationally representative surveys of 5000 to 30000 households and take place every 3-4 years, collect information on causes of death. They only provide estimates of overall childhood mortality, based on base population numbers and person-years, respectively.

The authors present mortality data from one such DSS and discuss possible problems affecting the quality of results and their generalization.

Materials and Methods

The DSS of Nouna is located in North-West Burkina Faso, a landlocked country in sub-Saharan West Africa. A baseline census took place in 1992, and data collection and updates take place in 3 months intervals. The DSS now covers a population of 60.000 individuals. Collection of COD by VA started in 1998. To avoid the “burn-in” phase of VA collection, and incomplete data for 2003, the analysis is limited to the period 1999 to 2003 in this presentation .

The distribution of missing values for COD (by classes of gender, age group, calendar year, season, urban/rural setting) is tabulated. Logistic regression was used to evaluate if the proportion of missing COD depends on demographic variables. Cause-specific mortality rates by age group are calculated directly based on person-years.

Results

About 75% of all cases have a known COD based on the verbal autopsy method. The proportion of missing COD decreased with calendar year which is an indication for increasing quality of data collection in the observation period, and was elevated for the two age classes with low overall mortality, 5-14 and 15-59 years. All other covariables had no effect. In particular, the season of death had no influence; this was unexpected since cause-specific mortality varies substantially over the year, and villages are more difficult to reach in the rainy season.

In general, the observed mortality rates are similar to those of other sub-Saharan malaria-holoendemic regions. Seasonal patterns are strongest in children under five. Infant and childhood mortality rates are high, mainly due to malaria, which causes 50% of all deaths in the rainy season, and 34% in the dry season. In adults, HIV/AIDS is the leading cause of death (20% in women, 16% in men).

Some findings do not match with data form other West African countries: 1) mortality of acute respiratory and gastrointestinal infections in infants and children is low. Other sources report rates nearly as high as those of malaria [4], and 2) there is an observed high malaria mortality in elderly people

Discussion

DSS are an important source of information on cause-specific mortality in developing countries. The quality of data collection procedures in the Nouna DSS has steadily improved since its start in 1998.

Malaria remains a problem in the area of the Nouna DSS, because of mortality and because of possible misclassification of COD. The sensitivity and specificity of the VA questionnaire used until now has not been assessed by validation studies (using hospital records, for example), and malaria death is well known to be difficult to be diagnosed correctly. Future models may account for sensitivity and specificity once they have been quantified.

Information on causes of morbidity and mortality is essential to health system planning, and the Nouna DSS is one of very few in West Africa that can provide such data. Dissemination and discussion of these results will help to improve quality of data reporting and ultimately the health of the population.

[Tab. 1]


References

1.
World Health organization. The World Health Report 2004
2.
Morris SS, Black RE, Tomaskovic L. Predicting the distribution of under-five deaths by cause in countries without adequate vital registration systems. Int J Epidemiol 2003; 32(6):1041-1051
3.
DHS+. DHS+ Mandate and Mission. http://www.measuredhs.com/aboutdhs/mission.cfm. 1984
4.
Bryce J, Boschi-Pinto C, Shibuya K, Black RE. WHO estimates of the causes of death in children. Lancet 2005; 365(9465):1147-1152.