SEARCH

SEARCH BY CITATION

Keywords:

  • Prevalence;
  • Incidence;
  • Mortality;
  • Follow-up;
  • Epilepsy;
  • Benin

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Purpose

Epilepsy is a major clinical and social issue in Africa. This study was conducted to estimate the prevalence, incidence, mortality, and therapeutic outcome in rural Djidja in Benin.

Methods

This was a two-phase study with a cross-sectional phase and 18 months of follow-up. In the first phase, information was obtained using door-to-door surveys, reports from key informants, and medical sources. People were interviewed using a validated screening questionnaire for epilepsy in tropical regions. The diagnosis of epilepsy was confirmed by a neurologist. We used a capture–recapture method to estimate the number of people with epilepsy (PWE). PWE were followed every month for 18 months after the cross-sectional survey. We asked the health services, the general population, and village leaders in the study area to identify suspected cases of epilepsy occurring during the follow-up. New cases were updated every month after confirmation. Antiepileptic drugs were prescribed to PWE.

Key Findings

We surveyed 11,668 subjects (male-to-female ratio 0.9) and identified 123 PWE, yielding a prevalence of 10.5 per 1,000 (95% confidence interval (CI) 8.8–12.6/1,000). Combining the three sources, we found 148 PWE and a prevalence of 12.7 per 1,000 (95% CI 10.7–14.9/1,000). After application of the capture–recapture method, the prevalence was estimated to be as high as 38.4 per 1,000 (95% CI 34.9–41.9/1,000). The cumulative incidence was 104.2 per 100,000 and the mean annual incidence was 69.4 per 100,000. The mean annual mortality was 20.8 per 1,000. After treatment, 45% of PWE had total seizure remission and 35% had a decrease in the number of seizures.

Significance

This study shows that door-to-door survey findings could be improved by using information from other sources. The follow-up suggests that epilepsy could be controlled. Continuous drug delivery and regular follow-up are key.

Epilepsy is a major and common neurologic disorder that affects around 70 million people worldwide (Ngugi et al., 2010). Ninety percent of people with epilepsy (PWE) live in developing countries in Africa, Asia, or Latin America (WHO, 2012). Health care delivery structures are still in their infancy in many African regions because of economic challenges. The median prevalence of epilepsy in sub-Saharan Africa is 15.0 per 1,000, which is twofold to threefold higher than that observed in industrialized countries (<8.0/1,000 inhabitants; Forsgren et al., 2005; Preux & Druet-Cabanac, 2005). There are fewer data on the incidence of epilepsy and mortality of PWE in sub-Saharan Africa. Population-based studies that used door-to-door methods in Benin have yielded high prevalences: 21.1 per 1,000 in Zinvie (n = 3,134), 15.9 per 1,000 in Savalou, and 10.6 per 1,000 among employees of five companies in South Benin (n = 1,232; Avode et al., 1996; Debrock et al., 2000; Houinato et al., 2007). Epilepsy-related challenges in Benin include the lack of treatment, which is significantly related to the high frequency of seizures, the seroprevalence of human cysticercosis, the social stigma related to epilepsy (still widely seen as a supernatural phenomenon), the high levels of anxiety and depression associated with epilepsy, and the higher prevalence of malnutrition among epilepsy patients than controls (22.1% vs. 9.2%, p = 0.0006) in Djidja (Zoli et al., 2003; Nubukpo et al., 2004; Crepin et al., 2007).

The aim of this study is to assess the main epidemiologic parameters of epilepsy in rural Djidja in central Benin and to evaluate a treatment program.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Study setting

This study was carried out in the Djidja community of the Zou region in Central Benin, comprising 10 villages and 14,500 inhabitants. The population density is 38 inhabitants/km2. The population is primarily rural and is dependent mainly on subsistence farming. This region is also marked by the highest level of poverty in Benin (National Institute of Statistics and Economic Analysis, 2002).

Definition of epilepsy

We used the International League Against Epilepsy (ILAE) definition of epilepsy (ILAE, 1993). A participant was considered to have epilepsy if he or she had had at least two epileptic seizures unprovoked by any immediate identified cause.

Study design

The study was in two phases. The first phase, a cross-sectional study, was conducted in 2005. The second phase consisted of a longitudinal study conducted between January 2006 and June 2007. Ethical agreement was obtained from the ethics committee of the government health department in Benin. All people investigated gave informed verbal consent.

Phase 1: cross-sectional study

A door-to-door survey was performed covering the entire Djidja population. Subjects who had presented with an isolated seizure were excluded. Data were gathered by 17 investigators familiar with the language of the region and accustomed to administration of the questionnaire. Investigators used the validated five-item epilepsy screening questionnaire of Limoges Neuroepidemiology Institute, ILAE, and Pan African Association of Neurological Sciences (PAANS) (Preux, 2000). Anyone who gave at least one positive response to the screening questionnaire was further examined by the neurologist to confirm or reject the diagnosis of epilepsy. To identify all PWE, the cross-sectional survey was supplemented with information from two sources: medical and nonmedical. We interviewed 72 key informants: village heads (11), community leaders (3), teachers (30), traditional healers (14), and spiritual leaders (14). We used a capture–recapture method to estimate the completeness of each source and the number of PWE in Djidja.

Phase 2: longitudinal study

In the second phase, we conducted a longitudinal study involving all PWE who were screened during the first phase and people without epilepsy or any other neurologic disorder who agreed to participate in the follow-up. PWE were followed for 18 months from January 2006. We prescribed antiepileptic drugs to PWE on a monthly basis. During the follow-up, health care services identified and recorded all potential cases of epilepsy that occurred. An active search was undertaken for new cases involving the general population and village leaders and making use of global sensitization of the community. An update of the new cases was made every month. Anyone with suspected epilepsy was further examined by the neurologist to confirm or reject the diagnosis. Those who had a diagnosis of epilepsy were treated and followed monthly to identify any epilepsy-related death(s) and elucidate the posttreatment evolution of epilepsy.

Mortality

We considered all deaths (epilepsy- and non–epilepsy-related) that occurred during the follow-up phase of this study.

Therapeutic outcome

All neurologist-confirmed epilepsy patients were treated daily from the end of the cross-sectional phase to the end of the longitudinal study phase. Carbamazepine at a progressively increasing dose was given to adults who had only partial seizures, and phenobarbitone 100 mg was given to children (1 tablet/day) and adults (2 tablets/day) with generalized or secondarily generalized seizures.

Data management and analysis

Data were recorded and analyzed using Epi-Info (version 6.04d, Centers for Disease Control, Atlanta, U.S.A.). Continuous variables were expressed as means ± standard deviation (SD) and qualitative variables as percentages. Confidence intervals (Cis) were estimated at 95%. The chi-square test was used to compare qualitative variables. A p-value < 0.05 was considered statistically significant.

We used a capture–recapture method to estimate the number of PWE not identified by the three sources (Chapman, 1951; Seber, 1970).

The estimation of the total number of PWE and the evaluation of the dependence of the three sources were determined using log-linear models analysis, with a Bio Medical Data Package (BMDP procedure 4F; Statistical Software Inc, Los Angeles, CA, U.S.A.; Hook & Regal, 1997). The model chosen was the one having less interaction terms, and best goodness of fit with the observed data (likelihood ratio statistic non-significant) and lowest information criteria: Akaike information criterion (AIC), Bayesian information criterion (BIC), and Draper information criterion (DIC; Sakamoto et al., 1986; Hook & Regal, 1997). “N Weighted DIC” was calculated by taking into account all N estimates (Draper, 1995).

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Cross-sectional study phase: characteristics of the study population

The study area population is 14,500 and we were able to survey 11,668 individuals, that is, 80.4% of the total (Fig. 1). Among them, 5,623 (48.2%) were male and 6,045 (51.8%) female. The sex ratio was 0.9 and the mean age was 20.1 ± 18.4 years.

image

Figure 1. Flowchart of participants in the prevalence study of epilepsy in Djidja, Benin, 2005.

Download figure to PowerPoint

Prevalence

Of the 11,668 subjects, 148 were confirmed to have epilepsy. Door-to-door survey found 123 PWE, yielding a prevalence of 10.5 per 1,000 (95% CI 8.8–12.6/1,000). Prevalence of epilepsy in investigated villages ranged from 6.6 to 16.9 per 1,000 (Table 1). Medical sources revealed 20 PWE, and 5 PWE were identified by key informants (Fig. 2). The crude prevalence of epilepsy obtained by combining the three sources was 12.7 per 1,000 (95% CI 10.7–14.9/1,000). The majority of cases were male (sex ratio 1.1). The sex-specific prevalences were therefore higher among male (16.2/1,000) than female (13.4/1,000) individuals, but this difference was not statistically significant (p = 0.21). The prevalence is significantly higher for people older than 20 years than for the younger age group (Table 2).

Table 1. Participants in door-to-door survey and prevalence of epilepsy in each investigated village
VillageNumberPWEPrevalence/1,00095% CI
Madjavi4,095276.64.3–9.6
Sovlegni1,110109.04.3–16.5
Hounvi95999.44.3–17.7
Kome9581111.55.7–20.5
Ye7421013.56.5–24.6
Aligoudo1,6072213.78.6–20.7
Wogbaye559814.36.2–28.0
Agonhohoun536814.96.5–29.2
Sanwlakpa570915.87.2–29.8
Dona532916.97.8–31.9
Total11,66812310.58.8–12.6
Table 2. Age-specific epilepsy prevalence in rural Djidja, Benin 2005
Age (years)NumberPWEPrevalence/1,00095% CI
0–94,150215.13.1–7.7
10–192,724207.34.5–11.3
20–291,8103619.914.0–27.4
30–391,2742620.413.4–29.8
40–496411015.67.5–28.5
≥501,069109.44.5–17.1
Total11,66812310.58.8–12.6
image

Figure 2. Distribution of specific and common cases of epilepsy according to three sources of information in Djidja, Benin, 2005.

Download figure to PowerPoint

Using a capture–recapture method, we evaluated the goodness of fit of the log-linear models and estimated the total number of PWE (Table 3). The total number of PWE in the chosen model was estimated to be as high as 448 (95% CI 407–488) and the prevalence was 38.4 per 1,000 (95% CI 34.9–41.9/1,000). The estimated exhaustivity rates were 27.5% for door-to-door, 5.8% for nonmedical sources, 6.3% for medical sources, and 33.0% for the three sources combined. The “N weighted DIC” was estimated at 544 (95% CI 500–588).

Table 3. Goodness of fit of the log-linear models and estimation of the total number of epilepsy patients in Djidja, Benin, 2005
Modeld.f.G2p-valueAICBICDICN95% CI
  1. S1, door-to-door survey; S2, nonmedical source; S3, medical source; N, estimation of total number of cases; d.f., degrees of freedom; G2, likelihood ratio statistic; AIC, Akaike information criterion; BIC, Bayesian information criterion; DIC, Draper information criterion.

  2. a

    Chosen model with lowest AIC, BIC, and DIC.

S1S2, S1S3, S2S300.531.000.50.50.51,2781,212–1,344
S1S2, S1S312.450.110.5−2.5−0.7663614–712
S1S2, S2S311.370.24−0.6−3.6−1.8534490–578
S1S3, S2S318.480.006.53.55.3175149–201
S1, S2S3221.580.0017.611.615.3241211–271
S2, S1S328.950.015.0−1.02.6177151–203
S3, S1S2a22.720.26−1.3−7.3−3.6448407–488
S1, S2, S3321.580.0015.66.612.1241211–271
Table 4. Characteristics of the PWE followed in Djidja, Benin, 2005
CharacteristicN = 105%
Sex  
Men8154.0
Women6946.0
Age (years)  
<102114.0
10–192013.3
20–293624.0
30–392617.3
40–49106.7
≥50106.7
Marital status  
Unmarried11375.3
Married3724.7
Occupation  
Active7952.7
Not active7147.3
Religion  
Animist8053.3
Christian6644.0
Muslim42.7
Ethnic group  
Fon13791.3
Adja74.7
Other64.0
Antecedents  
Familial epilepsy8154.0
Alcoholism74.7
Head injury32.0
Cerebrovascular disease21.3
Infection2013.3
None declared3724.7
Age at seizure onset  
<2011677.3
20–29149.3
30–3964.0
≥40149.4
Type of seizure  
Generalized9865.3
Partial secondarily generalized4630.7
Partial simple/complex64.0
Chronic complications  
Depression (deep sadness)88.0
Problems of concentration/memory (based on declaration)8053.3

Longitudinal study phase: characteristics of the study population

The longitudinal study included 11,520 subjects who did not have epilepsy at the start of the follow-up and PWE identified during the first phase study.

Incidence rates

After the follow-up of 18 months, 12 new cases of epilepsy (seven male, five female) were identified among the nonepilepsy population of 11,520. Most of new cases (n = 10) were younger than 20 years of age. Seven cases were identified during the first 6 months of follow-up, three during the second 6-month period, and two during the third 6-month period. This yielded a cumulative incidence rate of 104.2 per 100,000 (95% CI 53.8–181.9/100,000) and a mean annual incidence of 69.4 per 100,000 (95% CI 30.0– http://intranet:8080/cmsimple/index.php136.8/100,000).

Characteristics of follow-up cases

Of 160 epilepsy cases identified in this study (148 in the first study phase and 12 new cases during the follow-up), 150 were followed up (Fig. 1). Follow-up was not possible in 10 cases due to inability to participate (5), refusal to participate (2), and death before starting the follow-up (3). Sociodemographic and clinical characteristics of the 150 PWE in Djidja are summarized in Table 4. PWE were predominantly men (54.0%), <40 years old (81.3%), and unmarried (75.3%). A family history of epilepsy in first- and second-degree relatives was reported for 54% PWE (Table 4).

Mortality

During the 18 months of follow-up, five deaths (three due to epilepsy) were observed among the PWE identified in Djidja. Two deaths occurred during the first 6-month period, one during the second 6 months, and two during the third 6 months, corresponding to a mortality of 31.3 per 1,000 (95% CI 10.2–71.4/1,000) during the 18 months of follow-up. The mean annual mortality was 20.8 per 1,000 (95% CI 10.0–38.0/1,000). One of the three deaths due to epilepsy occurred during a fishing excursion on a boat, one during the night when a patient living alone had a seizure, and during a recurrence of epileptic seizures lasting more than 5 min.

Therapeutic follow-up

During the 18 months of follow-up, change was favorable for 121 of 150 PWE followed (80.6%). Among them, 68 achieved total remission of seizures (45%) and 53 achieved a reduction in seizures (35.0%).

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

Prevalence

This study used three sources of information and yielded a prevalence of 12.7 per 1,000—close to the median prevalence of epilepsy in sub-Saharan Africa countries (Preux & Druet-Cabanac, 2005). The prevalence calculated from door-to-door survey findings only was 10.5 per 1,000. Previous studies conducted in Benin show similar prevalences in Zinkanme and Savalou, at 14.8 per 1,000 and 15.2 per 1,000, respectively (Avode et al., 1996; Zohoun, 2004). Compared with studies outside Benin, the prevalence was similar to that in Pikine, Senegal (14.2/1,000), and (12.3/1,000) in Togo, neighboring Benin (Grunitzky et al., 1996; Ndoye et al., 2005). Our figure was lower than that found using the same methodology in Zinvie in rural Benin (21.2/1,000; Debrock et al., 2000).

The validity of the capture–recapture method requires that certain conditions be fulfilled. Case definitions were identical for the three sources, and all suspected cases were confirmed by a neurologist. The population could be considered as closed during the study period because the majority of the villagers were farmers and did not travel. Independence within the sources is the principal condition underlying this method, and was taken into account by the log-linear method; however, the numbers of PWE estimated by log-linear models and the “N weighted DIC” were very high and widely variable. These estimations seem unreliable, possibly because of small numbers in some cells of the contingency table. The differences between treated and untreated patients could also influence the probability of a case being reported by a source. In this study, despite repeated visits to the villages to meet as many villagers as possible, the entire population was not interviewed. The study area population was 14,500, and we surveyed 11,668 (80.4%) of the total. The highest exhaustivity rate (27.5%) was observed for the door-to-door survey. This could be considered surprisingly low and may be due to several causes: concealment due to stigma or underrecognition by household informants. In addition, the capture–recapture method is based on a statistical model, and the low numbers of cases in some combinations of sources could lead to overestimation of the capture–recapture estimate, and consequently underestimation of exhaustiveness. Door-to-door surveys are ideal in developing countries, since other sources of case ascertainment such as medical records or referral from key informants may not be sufficiently complete. Similar observations were made in two studies, wherein a protocol that used key informants as the only source yielded a prevalence of 3.6 per 1,000, whereas using the door-to-door survey the estimated prevalence was 18.2 per 1,000 in the same population (Kaamugisha & Feski, 1988). Several factors, such as lack of awareness of seizure types and poor cooperation among traditional healers may partially explain underestimation of prevalence.

Incidence

To our knowledge, very few previous studies have estimated the incidence of epilepsy in sub-Saharan Africa. Calculation of incidence requires long-term monitoring of large numbers of patients or the possibility of a sufficiently reliable retrospective estimation, which may differ from study to study, thus yielding varied incidence results. The mean annual incidence of epilepsy in our study was 69.4 per 100,000 inhabitants-years, close to that observed in Burkina Faso (83.0/100,000 inhabitants-years; Debouverie et al., 1993) or the median incidence of epilepsy 81.7 per 100,000 (28.0–239.5) in low- and middle-income countries (Ngugi et al., 2011). It is possible that our estimate is low as we relied on health care services to identify PWE, rather than active ascertainment or a second cross-sectional survey. Incidence of epilepsy in Ethiopia (64/100,000 habitants per years) was calculated through a comparison of two cross-sectional studies, the second conducted after an average period of 3.5 years (Tekle-Haimanot et al., 1997).

Our estimate is different from that observed in other African countries, for example, in Uganda where the rate age standardized on the world population was 156 per 100,000 inhabitants per year (Kaiser et al., 1998). Our incidence estimate was also slightly different from the 119 per 100,000 inhabitants per year obtained after a follow-up of the population twice a year in Togo (Grunitzky et al., 1996).

Mortality

Mortality data are very scarce in sub-Saharan Africa. The annual mortality observed in our study was 20.8 per 1,000, which is significantly lower than the crude mortality (31.6/1,000) reported in a 2-year Ethiopian study (316 epilepsy cases; Tekle-Haimanot et al., 1997) or in a 10-year follow-up of 128 PWE in rural Cameroon where 37 PWE died (Kamgno et al., 2003). However, in our study, determination of mortality based on very small numbers of cases yielded a wide confidence interval. Therefore, the differences could be due to the different sample sizes, although the mortality depends upon many factors such as availability of suitable drugs, treatment compliance, drug quality and the state of health care delivery in the region, and the types of epileptic seizures, which may differ between populations or studies. More than half of the deaths were epilepsy related. The circumstances under which the deaths occurred show that the deaths were not caused by the seizures themselves but related to their consequences (drowning, solitude).

Treatment

Our study yielded a high proportion of seizure-free cases (45%) and of cases with a decreased seizure frequency (35%). These figures are close to those reported in studies from Malawi (seizure remission in 56% of PWE) and Kenya (total remission in 53% of PWE, and a reduction in 26%; Watts, 1989; Feksi et al., 1991). Our study also revealed a significant monthly decline in the number of seizures. Regular follow-up of PWE is absolutely necessary to improve the health status of PWE and their care. In a previous study conducted in Zinvie in Benin, 31.7% (n = 63) of PWE were lost 1 year after the start of the follow-up. In our study a lower rate of loss was observed after 18 months of follow-up (n = 10, 6.3%). This lower rate suggests that the follow-up was reliable, continuous, and sufficient. This study showed that the maintenance of regular and timely follow-up is needed to ensure the effectiveness of antiepileptic treatment, as long as there are sufficient resources and commitment. Whatever the intervention, it must be fully integrated in the primary health care system. In addition, health personnel also need to be trained and motivated to treat epilepsy to the same degree as they are for other health problems. In addition, high level commitment is needed from governmental agencies to guarantee that epilepsy continues to be a priority and that the supply of drugs is ensured. This study has provided some promising results. Although mortality is high among PWE, the chances of remission are also high and it is therefore probable that young PWE will be able to actively contribute to society.

Information from door-to-door cross-sectional surveys is usefully complemented by data from other sources in order to identify the total number of PWE in low- and middle-income countries. Our results suggest that the care of patients with epilepsy in Africa requires not only the strengthening of human and material resources, but also an effort to ensure continuous follow-up and thereby improve health care status and quality of life.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this paper is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  • Avode DG, Capo-Chichi OB, Gandaho P, Bouteille B, Dumas M. (1996) Epilepsie provoquée par la cysticercose. A propos d'une enquête sociologique et culturelle réalisée à Savalou au Bénin. Bull Soc Pathol Exot 89:4547.
  • Chapman DG. (1951) Some properties of the hypergeometric distribution with applications to zoological sample censuses. Univ Calif Public Stat 1:131160.
  • Commission on Epidemiology and Prognosis, International League Against Epilepsy. (1993) Guidelines for epidemiologic studies on epilepsy. Epilepsia 34:592596.
  • Crepin S, Houinato D, Nawana B, Avode GD, Preux PM, Desport JC. (2007) Link between epilepsy and malnutrition in a rural area of Benin. Epilepsia 48:19261933.
  • Debouverie M, Kabore J, Dumas M, Weber M, Duboz P, Vaugelade J. (1993) Epidemiology of epilepsy in Burkina Faso. In Dumas M, Giordano C, Gentilini M, Chieze F (Eds) Neurologie tropicale. John Libbey Eurotext, Paris, pp. 5761.
  • Debrock C, Preux PM, Houinato D, Druet-Cabanac M, Kassa F, Adjien C, Avode G, Denis F, Boutros-Toni F, Dumas M. (2000) Estimation of the prevalence of epilepsy in the Benin region of Zinvie using the capture-recapture method. Int J Epidemiol 29:330335.
  • Draper D. (1995) Assessment and propagation of a model uncertainty. J R Stat Soc 57:7879.
  • Feksi AT, Kaamugisha J, Sander JW, Gatiti S, Shorvon SD. (1991) Comprehensive primary health care anti-epileptic drug treatment programme in rural and semi-urban Kenya. ICBERG (International Community-based Epilepsy Research Group). Lancet 337:406409.
  • Forsgren L, Beghi E, Õun A, Sillanpää M. (2005) The epidemiology of epilepsy in Europe – a systematic review. Eur J Neurol 12:245253.
  • Grunitzky KE, Dumas M, Mbella EM, Balogou AM, Hegbe M, Raniandrisoa H. (1996) Epilepsies in Togo. Epilepsia 8:189231.
  • Hook EB, Regal RR. (1997) Validity of methods for model selection, weighting for model uncertainty, and small sample adjustment in capture-recapture estimation. Am J Epidemiol 145:11381144.
  • Houinato D, Tibarbache H, Houeze F, Adjien C, Guedou F, Preux PM, Avodé GD, Dumont D, Druet-Cabanac M. (2007) Epilepsie en milieu professionnel au Bénin. Arch Mal Prof Environ 68:244250.
  • Institut National de la Statistique et de l'Analyse Economique. (2002) Third general census of the population and habitation 2002: some results. INSAE, Cotonou, pp. 10.
  • Kaamugisha J, Feski AT. (1988) Determining the prevalence of epilepsy in the semi-urban population of Nakuru, Kenya, comparing two independent methods not apparently use before in epilepsy studies. Neuroepidemiology 7:115121.
  • Kaiser C, Asaba G, Leichsenring M. (1998) High incidence of epilepsy related to onchocerciasis in West Uganda. Epilepsy Res 30:247251.
  • Kamgno J, Pion SDS, Boussinesq M. (2003) Demographic impact of epilepsy in Africa: result of a 10 year cohort study in a rural area of Cameroon. Epilepsia 44:956963.
  • Ndoye NF, Sow AD, Diop AG, Sessouma B, Sene-Diouf F, Boissy L, Wone I, Touré K, Ndiaye M, Ndiaye P, de Boer H, Engel J, Mandlhate C, Meinardi H, Prilipko L, Sander JW. (2005) Prevalence of epilepsy its treatment gap and knowledge, attitude and practice of its population in sub-urban Senegal. An ILAE/IBE/WHO study. Seizure 14:106111.
  • Ngugi AK, Bottomley C, Kleinschmidt I, Sander JW, Newton CR. (2010) Estimation of the burden of active and life-time epilepsy: a meta-analytic approach. Epilepsia 51:883890.
  • Ngugi AK, Ngugi AK, Kariuki SM, Bottomley C, Kleinschmidt I, Sander JW, Newton CR. (2011) Incidence of epilepsy: a systematic review and meta-analysis. Neurology 77:10051012.
  • Nubukpo P, Houinato D, Preux PM, Avodé G, Clément JP. (2004) Anxiety and depression among the epileptics in general population in Benin (Western Africa). Encephale 30:214219.
  • Preux PM, et le Comité de Recherche sur l'Epilepsie de l'Institut d'Epidémiologie Neurologique et de Neurologie Tropicale de Limoges. (2000) Questionnaire d'investigation de l'épilepsie dans les pays tropicaux. Bull Soc Pathol Exot 93:276278.
  • Preux PM, Druet-Cabanac M. (2005) Epidemiology and aetiology of epilepsy in sub-Saharan Africa. Lancet Neurol 4:2131.
  • Sakamoto Y, Ishiguro M, Kitigawa G. (1986) Akaike information criterion statistics. KTK Scientific, Tokyo, Japan.
  • Seber GAF. (1970) The effect of trap response on tag recapture estimates. Biometrics 26:1322.
  • Tekle-Haimanot R, Forsgren L, Ekstedt J. (1997) Incidence of epilepsy in rural central Ethiopia. Epilepsia 38:541546.
  • Watts AE. (1989) A model for managing epilepsy in a rural community in Africa. Br Med J 298:805807.
  • WHO. (2012) Epilepsy. Fact Sheet 999. WHO, Geneva.
  • Zohoun SH. (2004) Estimation of the prevalence of epilepsy in Zinkanmè, commune of Djidja: Significance of the utilization of multiple sources of information. Medical Thesis, School of Medecine Cotonou Benin.
  • Zoli A, Shey-Njila O, Assana E, Nguekam JP, Dorny P, Brandt J, Geerts S. (2003) Regional status, epidemiology and impact of Taenia solium cysticercosis in Western and Central Africa. Acta Trop 87:3542.