Menu
Menu
Menu
Menu
Background: Current vital statistics from governmental institutions in Cǒte d’Ivoire are incomplete. This problem is particularly notable for remote rural areas that have limited access to the health system. Objective: To record all deaths from 2009 to 2011 and to identify the leading causes of death in the Taabo health and demographic surveillance system (HDSS) in south-central Cǒte d’Ivoire. Design: Deaths recorded in the first 3 years of operation of the Taabo HDSS were investigated by verbal autopsy (VA), using the InterVA-4 model. InterVA-4 is based on theWorld Health Organization 2012 VA tool in terms of input indicators and categories of causes of death. Results: Overall, 948 deaths were recorded, of which 236 (24.9%) had incomplete VA data. Among the 712 deaths analyzed, communicable diseases represented the leading causes (58.9%), with most deaths attributed to malaria (n=129), acute respiratory tract infections (n=110), HIV/AIDS (n=80), and pulmonary tuberculosis (n=46). Non-communicable diseases accounted for 18.9% of the deaths and included mainly acute abdomen (n=38), unspecified cardiac diseases (n=15), and digestive neoplasms (n=13). Maternal and neonatal conditions accounted for 8.3% of deaths, primarily pneumonia (n=19) and birth asphyxia (n=16) in newborns. Among the 3.8% of deaths linked to trauma and injury, the main causes were assault (n=6), accidental drowning (n=4), contact with venomous plants/animals (n=4), and traffic-related accidents (n=4). No clear causes were determined in 10.0% of the analyzed deaths. Conclusions: Communicable diseases remain the predominant cause of death in rural Cǒte d’Ivoire. Based on these findings, measures are now being implemented in the Taabo HDSS. It will be interesting to monitor patterns of mortality and causes of death in the face of rapid demographic and epidemiological transitions in this part of West Africa.
The Taabo HDSS was established in 2008 and the initial census revealed a population of 37,792. In December 2011 the population was 39,422, and in December 2013 it had reached 42,480 (13). The Taabo HDSS includes one small town (Taabo Cité, 7,514 inhabitants at the end of 2013), 13 main villages, and over 100 hamlets. The establishment of the Taabo HDSS, the longitudinal surveillance of demographic and health data, and the implementation of specific interventions and research projects were approved by the institutional research commissions of the Centre Suisse de Recherches Scientifiques en Côte d’Ivoire (Abidjan, Côte d’Ivoire) and the Swiss Tropical and Public Health Institute (Basel, Switzerland). Ethical clearance was obtained from the ethics committees in Côte d’Ivoire (reference no. 1086 MSHD/CNEF) and Basel (EKBB, reference no. 316/08). All households belonging to the Taabo HDSS are visited three times a year by trained field enumerators. They conduct demographic surveillance, including the registration and monitoring of migration, pregnancies, births, and deaths. Monitoring pregnancies helps enumerators obtain information on stillbirths, abortions, and neonatal deaths. All deaths of permanent residents are recorded and – as with other HDSS sites – whenever possible examined by VA techniques in order to determine the most likely causes of death (14–19). All applied registration and monitoring methods were developed and standardized in order to ensure that the data collected are of high quality and will allow for cross-site comparison (10, 20–23). As a member of the International Network for the Continuous Demographic Evaluation of Populations and Their Health (INDEPTH; http://indepth-network.org), the Taabo HDSS adheres to INDEPTH’s standards. Further details regarding the Taabo HDSS have been published elsewhere (13). The reporting of death is facilitated by key informants in the communities who observe and record any death occurring in the study area. The information is then transmitted to a Taabo HDSS VA supervisor or a field enumerator who informs a VA supervisor. Deaths that might have been missed by the key informants can subsequently be identified during demographic surveillance rounds. Whenever possible, the VA supervisor visits the household in which the death has occurred within 2 weeks and contacts the Taabo HDSS data center for verification of the event. Once it has been verified that the deceased was an HDSS resident, the VA supervisor completes a standardized VA questionnaire with one of the deceased’s close relatives. Initially, completed VA questionnaires were submitted to two physicians who independently determined the direct and underlying cause of death and coded it according to the WHO International Classification of Diseases, version 10 (ICD-10) (24). However, this practice proved relatively slow, somewhat expensive, and there was considerable inter-observer variation regarding the causes of death. Since 2012, the computer automated InterVA-4 model has been validated and admitted for use in research and civil registration, both within already enumerated populations and also as a stand-alone death registration tool (25–28). For the present analysis, the InterVA-4 model was employed. The InterVA-4 tool is a freely available standard computerization model for interpreting VA data and determining causes of death. It has been designed to use the VA input indicators defined in the 2012 WHO VA instrument and to deliver causes of death compatible with the ICD-10. The causes of death are categorized into 62 groups, as defined in the 2012 WHO VA instrument (25–28). The InterVA-4 model was developed on the basis of Bayes’ theorem and is therefore an application of the Bayesian approach for diagnostic help (25–29). If the event of interest (A) depends on different mutually exclusive causes C 1, C 2, …, C m (for instance, causes of death) and other factors S 1, S 2, …, S n (for instance, different signs and symptoms leading to death), then Bayes’ general theorem for each C i and S j can be stated as with P(!Ci) = (1−P(Ci)). For the complete set of causes of death C 1, C 2, …, C m, a set of probabilities for each C i can be computed using a normalizing assumption so that the total conditional probability of all causes sums up to unity: While using an initial set of unconditional probabilities for the causes of death C 1, C 2, …, C m(P(C i∣S 0)) and the matrix of the conditional probabilities P(S j∣C i) for indicators S 1, S 2, …, S n and causes of death C 1, C 2, …, C m, it is possible to apply the same calculation for each S 1, S 2, …, S n that applies to one particular death: In short, with the exception of a minority of interviews where information is contradictory or inconsistent and the cause of death has to be classified as undetermined despite a completed VA interview, usually up to a maximum of three most likely causes of death and their probabilities are estimated per case. Whenever these causes do not add up to 100%, the remaining percentage is classified as undetermined. Because we are working with probabilities, one death can thus contribute to the statistics for several possible causes of death, but never at more than 100%. These idiosyncrasies of InterVA-4 lead to minor rounding errors and explain small differences between the disaggregated data and the summed up totals as presented in this study. Like its predecessor version, InterVA-4 employs special procedures for HIV/AIDS and malaria, because these diseases vary greatly from one setting to another (26). For malaria, unconditional probability is applied to the causes of death, especially due to sickle cell disease as there is a close link between these two conditions (26, 30). For the present analysis, all identified causes of death were aggregated into 14 broad groups, as predefined by INDEPTH, using the statistical software package STATA version 12 (StataCorp, College Station, TX, USA).
N/A
