The challenge of epilepsy control in deprived settings: Low compliance and high fatality rates during a community-based phenobarbital program in rural Laos


According to the World Health Organization (WHO), 80% of people with epilepsy (PWE) live in poor resource countries, and never receive treatment. Cost-effectiveness and safety profile of phenobarbital make it the recommended drug for these countries. Lao Peoples' Democratic Republic (Lao PDR) is a small landlocked country in South-East Asia, with a sparse 5.6 million multiethnic rural population, ranking 133rd at the Human Development Index; average annual per capita income is US$491. Epilepsy has an estimated prevalence of 7.7‰, with cysticercosis playing little etiologic role (Tran et al., 2006, 2007a). There are no national guidelines and no control program, and misconceptions and stigma toward PWE are common (Tran et al., 2007b). Despite figuring on the national essential drugs list, phenobarbital may be of limited availability. Only 53% of urban pharmacies can provide it, and diazepam short courses are the most used regimen to treat seizures (Odermatt et al., 2007). The intervention presented in this report took place 120 km north of the capital, Vientiane, in Hinheub, in a typical Lao district hospital, with 17 health personnel including one MD for 23,000 inhabitants.

Medical practitioners pursuing continuous training at the Francophone Institute of Tropical Medicine developed a 2-year community-based phenobarbital program. The target population consisted of PWE identified in a previous prevalence survey, with cases referred by the district health personnel. Epilepsy diagnosis was clinically based on international guidelines (Preux et al., 2000). Phenobarbital (100 mg daily) was offered to all adults with active disease (≥2 generalized seizures in the previous 2-year period). Preliminary information was given to PWE and caregivers on epilepsy, treatment principles, and the importance of compliance and follow-up. We obtained each patient's informed consent, and clearance from the Lao National Ethic Committee. Phenobarbital was provided free by the hospital every 2 months, but transportation costs were at charge of the patient. One assistant pharmacist and two nurses received a short training on the program and data management, including clinical follow-up (seizures occurrence, side effects, compliance monitoring by left over tablets count). Patients missing two appointments, or taking <80% of expected number of phenobarbital tablets, were considered low compliant; those missing three consecutive appointments were considered “dropped out.” The medical team visited the district twice a year to check data and management, and conducted a final home interview with compliant patients—on the program's relevance and the patients/community expectations—using a semistructured questionnaire.

The potential population consisted of 53 people. Only the 46 active cases were retained. The mean age was 31.7 years, the M/F sex ratio was 1.2; 19.5% were mentally retarded. Only four people had ever received antiepileptic drugs before. Eighty percent were farmers. The average daily family income was <US$1, the mean distance to hospital was 18.6 km. Seizures frequency ranged from 2/year to 3/week. Eleven patients did not attend the program at all (24%), 16 rapidly dropped out (35%), and only 10 (22%) showed full compliance.

Premature death was a striking finding. Six of the 53 patients (11%) died within the 2-year study period (2 drowning, 1 burns, 1 fall, 2 sudden unspecified cause). Five of these were mentally retarded young males (median age 18 years), and none had been fully compliant (Fig. 1).

Figure 1.

Follow-up diagram of 53 PWE enrolled in a two-year treatment program with phenobarbital.

The 18 people who completed the program deemed it efficient and reported an improved working capacity and quality of life. Moreover, mean seizure frequency decreased from 3.5 to 0.3/month, 11 patients (69%) were seizure-free, 4 reported somnolence, and none had to withdraw because of side effects.

The main expressed difficulties were distance to the health center, transportation availability and costs (range US$1–3), and poor hospital accessibility (drug delivery failures occurred due to inadequate running hours or absent personnel). Finally, 66% of those interviewed advocated a more convenient village-based drug delivery procedure.

Phenobarbital has proved highly efficient, well tolerated, and manageable at a community level (Mani et al., 2001; Nimaga et al., 2002; Asawavichienjinda et al., 2003). Our population of PWE had low education, low income, and poor access to health care, which may all hinder compliance. The over-prevalence of death in noncompliant mentally retarded young men with epilepsy is striking; however, the possibility that causes other than seizures may intervene, including traditional beliefs and stigma related to mental deficiency should be thoroughly investigated (Malina, 2005). The results of this pilot study suggest that a medical intervention aimed at epilepsy treatment in deprived countries might benefit from an anthropological component.


We are indebt to the personnel of Hinheub district hospital, the Agence Universitaire de la Francophonie (AUF), and the French Ministry of Foreign Affairs (project CORUS number 02-811-052) which both granted the program. We thank Dr. Daniel Reinharz for his help.