The Treatment Gap in Epilepsy: The Current Situation and Ways Forward


  • H. Meinardi,

    1. Department of Physiology, Leyden University Medical Centre, Leiden, The Netherlands;
    Search for more papers by this author
  • R. A. Scott,

    1. The National Hospital for Neurology and Neurosurgery, The National Society for Epilepsy, Chalfont St. Peter, Gerrards Cross, Bucks SL9 ORJ, England;
    Search for more papers by this author
  • R. Reis,

    1. Faculty of Political and Social-Cultural Sciences, University of Amsterdam, Anthropological Sociological Centre, Amsterdam, The Netherlands
    Search for more papers by this author
  • J. W. A. S. Sander On Behalf Of The Ilae Commission on the Developing World

    1. The National Hospital for Neurology and Neurosurgery, The National Society for Epilepsy, Chalfont St. Peter, Gerrards Cross, Bucks SL9 ORJ, England;
    Search for more papers by this author

  • Address reprint requests to Executive Office ILAE, c/o Mrs Irene Kujath, Epilepsie-Zentrum Bethel, Klinik Mara I, Maraweg 21, D-33617 Bielefeld, Germany. E-mail:

Address correspondence to Dr. H. Meinardi at Department of Physiology, Leyden University Medical Centre, P.O. Box 9604, 2300 RC Leiden, The Netherlands. E-mail:


Summary: This article is a summary of a workshop held by the ILAE concerning the issue of the epilepsy treatment gap in developing countries. The gap is defined in terms of those people with epilepsy who are not being appropriately treated and is the result of an array of medical, political, social, economic, and cultural factors. The situation regarding the treatment gap for various countries is reviewed, along with some of its causes. Although the overall gap is estimated to be large, a number of recent projects and interventions have been effective in delivering appropriate treatment to people with epilepsy in underresourced countries of the developing world. It is hoped that these may be transferable elsewhere and that, combined with the ILAE/IBE/WHO Global Campaign against Epilepsy and increased support from the worldwide epilepsy community, the treatment gap will begin to be bridged.

There is generally a perception that there is a large difference between the number of people with epilepsy in developing countries and the number of these people who are actually being treated for their epilepsy.

The ILAE Commission on the Developing World convened a workshop (WOTREGAP I) in Marrakech, Morocco, May 1–2,1999, to discuss the issue of this treatment gap and potential ways of remedying the situation. The workshop was attended by 12 members of the Commission and 15 external experts (see Appendix), and the following four questions were discussed and form the basis for this article:

  • 1What is the treatment gap?
  • 2How is this gap measured?
  • 3What are the causes of the treatment gap?
  • 4How can the treatment gap be reduced?

In discussing these questions, the existing literature on the subject was reviewed. Representatives from Africa, Asia, Latin America, and the former Eastern Block then gave an account of the situation in their region relevant to the treatment gap. (However, the Commission will not in this report provide an overview of estimates of treatment gaps per country as the data available have not been based on uniform criteria nor are the sources of information always identifiable). This was followed by group discussions on the definition, epidemiology, social, psychological, political, and economic issues. These issues were discussed further in a plenary session and a consensus evolved on the recommendations to be made.


According to the definition of the World Health Organization (WHO), health is a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity. A disturbance of health can affect all or some aspects or subsets covered by this definition. Treatment refers to any intervention that is intended to restore health. To ensure a more holistic approach, the Commission recommended identifying the subsets involved whenever treatment is discussed. The Commission also discussed whether diagnosis, i.e., establishment that a person's health has been disturbed, is part of treatment (that is part of the intervention intended to restore health), and it was agreed that this was the case.

In view of these comments, if reference is made to the difference in numbers of people treated and people who (according to international standards of the professional community) should be treated, such differences comprise both diagnostic and therapeutic deficits. The technical term designating the estimated or measured difference is called the “Treatment Gap.” However, whenever reference is made to treatment gap, the modality of health should be mentioned. It was decided that Wotregap I should primarily focus on the pharmacologic or surgical treatment of seizures.

Thus the following definition was accepted:

Seizure treatment gap:

The difference between the number of people with active epilepsy and the number whose seizures are being appropriately treated in a given population at a given point of time, expressed as percentage. This definition includes diagnostic and therapeutic deficits.

This required the following subdefinitions:

Active epilepsy:

Two or more unprovoked epileptic seizures on different days in the prior year that are disabling to the individual.

Appropriate treatment:

Diagnosis and treatment of underlying causes; treatment of recurrent seizures according to international standards.

(It should be noted that were the epilepsy treatment gap to be discussed, then appropriate diagnosis and treatment would also have to include the influence of having epilepsy on mental and social well-being. However, this is beyond the remit of this particular article).


At present, surgical treatment is considered appropriate treatment in ∼10–15% of people with epilepsy. In countries where a treatment gap exists, it is very likely that the treatment gap for surgical treatment will exceed that for pharmacotherapy, and when epilepsy is diagnosed, it will be treated with drugs. Therefore, as a general measure for the seizure treatment gap, the solution proposed by Ellison et al. (7) with a slight modification seems appropriate. This measure contains:

  • • An estimate of the number of people treated for epilepsy [based on dividing the amount of antiepileptic drugs (AEDs) sold in a year by one Defined Daily Dose (DDD) of each type of drug] subtracted from an estimate that at least 0.5% of any population has active epilepsy. This difference is then divided by 1% of the estimated number of people with epilepsy.
  • • Although some patients with types of epilepsy that prove hard to treat may be treated with more than one DDD, this is compensated by patients with easy-to-treat epilepsy who may use less than one DDD. (The DDD is determined by WHO on the basis of literature and reports about the usually effective dose of a drug. The published DDDs are probably on the high side, as the data are mainly derived from Western countries and do not take body weight into account. Conversely, AEDs are also used for other indications than epilepsy, such as trigeminal neuralgia and depression.)


Literature review

One of the problems with the analysis of available literature is the publication delay, which is a concomitant of working in the developing world. Once the research has been accomplished (with great effort and investment of the limited time available for anything besides regular clinical work), the additional time needed to examine the data and prepare a manuscript is difficult to find. For example, a multicenter study on the epidemiology of epilepsy carried out as PL480 funded project of the Indian Council of Medical Research was mainly performed in the early 1970s, whereas the report was finally published in 1989 (1). A study about stigma and psychosocial problems associated with epilepsy in Pakistan published in 1997 was based on information gathered by a door-to-door survey in 1980s (2).

Without pretending to be comprehensive, articles from Tanzania (3), southern Africa (4), and South Asia (5) can be cited on epilepsy in developing countries. Furthermore, the efforts of ICBERG, an informal academic grouping of senior epilepsy research workers, should be mentioned. The ICBERG study introduced the concept of a treatment gap and published data for Pakistan, the Philippines, and Ecuador (6,7). Although not entirely uniform, the literature supports the statement by ILAE/IBE/WHO in their introduction to the Global Campaign against Epilepsy “Epilepsy out of the Shadows,” that globally 85% of people with epilepsy are either inappropriately treated or not treated at all.

Regional overviews

Latin America

(Contributions by: J. Fandiño Franky and S. Kochen)

Health care.

In the major cities, there are big hospitals, many of which are university hospitals with facilities for sophisticated diagnosis involving magnetic resonance imaging (MRI), computed tomography (CT) scan, and laboratory investigations. Every intermediate town has a hospital of the first and second levels and should be able to take care of any case of epilepsy. Most of the small towns have a health center of acceptable quality. However, Latin America is a continent with countries in different stages of development and differences in political stability.

In larger countries, there may also be internal differences and inequalities in health care development. For example, health provision in Brazil is excellent in the South, but in the North, this is deficient for social, economic, and geographic reasons. If a patient living in a remote area has a first seizure, it will take time and effort to gain access to the health care system. In status epilepticus, the person may never make it.

Financial coverage of health costs.

It is estimated that in the near future, on average, 80% of the population in Latin America will have social security coverage of health care. However, this does not imply that all available treatment is covered.


Phenobarbital (PB) and other AEDs are usually covered within government health budgets. However, more modern AEDs, whose price can be up to 190 times that of PB, are not. It should also be noted that the price of drugs in developing countries is often higher than the price of the same drug in the United States or Europe.

Surgical treatment.

The development of epilepsy surgery is hampered by the costs of diagnostic tools, such as MRI and 24-h continuous EEG/video-monitoring. Furthermore, the custom for full-time medical specialists to work part-time in several different hospitals is counterproductive. The determinants of cost coverage are not available. There is an exemplary exception in Carthagena dos Indes, Colombia, where affordable neurosurgical treatment for epilepsy may serve as a model for similar developments elsewhere.

Physician density.

On average, the region (as far as data are available) has one physician per 1,467 people. The distribution between countries and within countries is, however, very varied. For example, Cuba has approximately one physician per 332 people, whereas Haiti has only one per 7,143. Concerning internal disparities, Mexico is illustrative with 8% of the population living in states with a physician density of one per 600, and a further 26% of the population live in states with fewer than one physician per 1,200.

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the GNP/capita = –0.475 (two-tailed significance level, 0.054). This would appear to show that there is not a strong relationship between income and the number of doctors in Latin America. It may be the case that the more socialist states, such as Cuba, confound this relationship through higher health expenditures as a proportion of GDP than do other more capitalist nations.

Nurse density.

On average, the region (as far as known) has one nurse per 4,368 people. Best served is Cuba, with one nurse per 180 people, and again, Haiti is worst off with one nurse per 9,091 people.

The Pearson correlation coefficient for the relation between number of inhabitants per nurse and the GNP/capita = –0.403 (two-tailed significance level, 0.136). There would appear to be an even weaker relationship between incomes and the numbers of nurses in the region.

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the number of inhabitants per nurse is 0.666 (two- tailed significance level, 0.007). However, there does appear to be a fairly strong link between the numbers of doctors and the numbers of nurses.


The prevalence of epilepsy in many parts of Latin America is higher than the average in North America and Europe. As a conservative estimate, a prevalence of twice that of Europe and North America (i.e., 10/1,000) can be assumed. Part of the difference can be explained by a higher risk of brain damage due to parasitic (e.g., cysticercosis) and bacterial (e.g., tuberculosis, meningococci) infections, and due to substandard perinatal care.


(Contributions by: H. Aziz, O. Eisler, D. Gunawan, and Byung-In Lee)

Health care.

An overview of the health care facilities is difficult to obtain because no organization collects these data. Although sophisticated equipment like CT scan and MRI were gradually being made available, the recent economic crisis has prevented their proper maintenance.

Financial coverage of health costs.

In Japan, almost all expenses are covered by health insurance. In China, there is coverage for public servants but not for nongovernment employees. In Indonesia, the government and certain companies have their employees covered by insurance. In this country, rural health care centers provide essential free medical services. Young doctors have to serve the first years after graduation in rural areas. In Korea, national health insurance covers 50% of the costs (except MRI). In Taiwan, all medical and surgical expenses are covered by health insurance.


Most drugs that have been licensed in major European or North American countries are also sold in Asia. However, the essential information concerning whether the total number of DDDs matches the estimated prevalence of epilepsy is not available. Furthermore, assuming that the pattern of therapy resistance is not subject to geographic differences, it would also be necessary to know whether the pattern of sales of each drug corresponds with those in other countries. Finally, the availability of a drug is not a guarantee for its quality. For this report, it was impossible to gauge the thoroughness of control by national authorities of the quality of the AEDs marketed.

Surgical treatment.

Epilepsy surgery is well developed in Japan and Korea. Although a few centers in India do operate for temporal lobe epilepsy, this country was not represented in the Global Survey on Epilepsy Surgery 1980–1990 (8). From Japan, five centers were included, and there was also one from Taiwan. No center in China responded to the questionnaire, but two centers were identified from publications. This was also the case for one center in Vietnam.

Physician density.

On average, as far as data are available, there appears to be one physician per 5,300 people. This masks a wide range in which Vietnam has one doctor per 247 people and Japan has one doctor per 653 persons, whereas, at the other extreme, Bangladesh and Myanmar both have one doctor per 12,500 people, and Nepal has one doctor per 16,667 persons. The following information has been provided about China: one doctor with higher education per 793 people; one doctor with middle education per 2,471 people.

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the GNP/capita = –0.467 (two-tailed significance level, 0.092). As with the figures for Latin America, there does not appear to be a strong relationship between the two variables.

Nurse density.

On average, as far as data are available, there appears to be one nurse for every 6,900 inhabitants. Japan is best placed, with one nurse per 653 persons, followed by Vietnam, with one nurse per 1,150 people. Nepal is again at the other extreme, with one nurse for every 33,333 inhabitants, followed by Bangladesh, with one nurse for 20,000 people. The following information has been provided about China: one nurse with higher education per 1,610 people and one nurse with middle education per 2,533 people.

The Pearson correlation coefficient for the relation between number of inhabitants per nurse and the GNP/capita = –0.274 (two-tailed significance level, 0.443). There seems to be an even weaker link concerning the relationship between nurses and GDP than that for doctors and GDP.

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the number of inhabitants per nurse is 0.896 (two- tailed significance level, <0.001). Again, however, this correlation appears to reveal that the numbers of doctor and nurses are linked together.


Several articles on the epidemiology of epilepsy in Asia have been published. The prevalence rates vary. Early studies in India showed extremely low prevalence rates, fewer than three per 1,000, although more recent studies have shown rates between four and nine per 1,000. Similarly, in Japan in 1964, a prevalence rate of fewer than two per 1,000 was reported, whereas in 1980, the prevalence rate was recorded as eight per 1,000. From China in 1985, a prevalence rate of 4.4/1,000 was given, whereas in Pakistan in 1994, a rate of 9.9/1,000 was stated. According to the surveys, prevalence is clearly higher in rural areas.


(Contributions by: G. Diop, M. Dumas, P.M. Preux, A. Martins da Silva, W.B.P. Matuja, and J. Mielke)

Health care.

To summarize health care provisions for Africa is well-nigh impossible, although certain aspects can be mentioned. In many countries, civil war and cross-border fighting is ongoing, or the country is still suffering from its aftermath. Parasitic disorders, such as malaria, are rampant, especially in rural areas. Acquired immunodeficiency syndrome (AIDS) is rapidly changing the demographic structure. In many countries, expatriate health care personnel are provided by donor countries and religious or secular nongovernmental organizations (NGOs). It is also the case that many national doctors emigrate.

Despite these issues, it is possible to distinguish roughly between three regions: the North, the Inter-Tropical area, and the South. Concerning epilepsy care, in the North, there are at least 175 EEG machines, 120 CT scanners, and 22 MRI machines. In the South, there are 100 EEG machines, 140 CT scanners, and 48 MRI machines. In the Inter-Tropical area, there are no more than 50 EEG machines, 25 CT scanners, and three MRI machines.

Financial coverage of health costs.

In several countries, the government will provide free health care or heavily subsidized care. Health insurance appears to be poorly developed.


In French-speaking areas, phenobarbitone is prescribed in 65–85% of treated cases. In English-speaking countries, phenytoin (PHT) is also frequently prescribed. Carbamazepine (CBZ) is the second drug, but is prescribed in only 5–20% of treated cases. Valproate (VPA) is prescribed in 5–15%, but is less widely available, and costs are also much higher than those of the aforementioned drugs. For status epilepticus, injectable PB or diazepam (DZP) is used when available.

Surgical treatment.

In the ILAE's survey for the years 1980 through1990 (8), no epilepsy surgery programs in Africa were identified. In South Africa, and possibly in Egypt, this treatment modality is gradually being offered.

Physician density.

On average, as far as data are available, there is one doctor for 17,547 persons. Libya and Algeria have a coverage of one doctor per 962 and one doctor per 1,064 inhabitants, respectively. At the other extreme are Niger and Malawi, with one doctor per 50,000 people (interestingly Niger has one nurse for 3,846 people and Malawi only one nurse for 33,333 persons).

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the GNP/capita = –0.531 (two-tailed significance level, 0.001). This correlation, although not strong, is greater than that for both Latin America and Asia.

Nurse density.

On average, the available data indicates that there is one nurse for 6,923 people. Libya and Tunisia have figures of one nurse per 328 and per 407 people, respectively. However, Chad has only one nurse for 50,000 persons and is followed by Malawi with one nurse for 33,333 inhabitants.

The Pearson correlation coefficient for the relation between number of inhabitants per nurse and the GNP/capita = –0.294 (two-tailed significance, 0.155). This appears to show that there is a weak correlation between the two variables.

The Pearson correlation coefficient for the relation between number of inhabitants per doctor and the number of inhabitants per nurse is 0.586 (two-tailed significance level, 0.001). Unlike the figures for Latin America and Asia, there appears to be no strong link between the numbers of doctors and the numbers of nurses.


From Africa, the highest prevalence data worldwide have been reported. In the early 1960s, it was reported that in the Wapogoro tribe in Tanzania, 20/1,000 people had epilepsy. Recent reports about the same area show great differences between districts in that area. In Gran Bassa County in Liberia, approximately one person in 30 had epilepsy. In Nigeria, striking differences in prevalence between fairly closely situated cities has also been reported.

The researchers from Tanzania tried to explain the variance from a correlation with onchocerciasis infestation. The Nigerian explanation for the differences reported was that one town had better public health services and perinatal care, as it was a training area for the University Medical School. Reports from the North of Africa are similar to those from Europe and the United States.

Former Eastern Block

(Contributions by: N. Tatishvili and A.B. Guekht)

This area of the world is still very much in upheaval, and it is impossible to describe a general background. Therefore, as an example, one of the states (Georgia) is discussed. Georgia used to be the most advanced state of the Soviet Union in epilepsy care and epileptology.

Health care.

The first diagnosis of a paroxysmal event is made at the local area, often without the help of an EEG. After referrals, it becomes clear that up to 25% of patients with nonepileptic seizures are mistakenly diagnosed as having epilepsy. Referral to the second level is to the city hospitals, all equipped with EEG facilities, and in the two biggest cities, there is also a CT scanner and MRI. Referral to the third level is to the Epilepsy Centre in the capital Tblisi. Here 24-h ambulatory and video-EEG is possible; the centre is also equipped with a CT scanner and MRI. However, since the introduction of fee-based medicine, the proper functioning of the referral system is corrupted. Wherever a diagnosis of epilepsy is made, proper treatment with a regular supply of medication and appropriate supervision is no longer possible for financial reasons.

Financial coverage of health costs.

The health care system after independence has changed from being state provided to being a fee-based health care system. De jure, a small fraction of medical care is still free of charge, but de facto, this is not the case. For example, in the case of epilepsy de jure, only inpatient treatment of children up to age 1 year and treatment of convulsive status epilepticus is free.


At the first and second levels, PB and CBZ are the most frequently prescribed drugs. At these levels, ∼75% is prescribed as monotherapy. At the third level, VPA and CBZ are most frequently used. About 35% of the prescriptions at this level are polytherapy using combinations of VPA and ethosuximide (ESM), PB and ESM, PB and CBZ, PB and clonazepam (CZP), and CBZ with a benzodiazepine (BZP).

Surgical treatment.

Epilepsy surgery is impossible because the presurgical evaluation is too expensive. (Information on the population densities of physicians and nurses and on epidemiology was not available at Wotregap I nor at the time of going to press.)


To know the size of a seizure treatment gap is a necessary but insufficient tool to achieve optimal seizure control in a population. It is essential to know, or at least to suspect, its causes. Some potential causes could be level of health care development, cultural beliefs, economy, distance from health care facilities and supply of AEDs, and a lack of prioritization. These factors are outlined further later, and a more wide-ranging, holistic summary showing their interconnectedness can be found diagrammatically in Figs. 1–4. (It should also be borne in mind that there is a primary distinction between diagnostic deficits and therapeutic deficits. These deficits may have the same and/or different causes. A distinction also should be made between deficits depending on the health-seeking pattern and those caused by lack of compliance once treatment has been initiated.)

Figure 1.

Overall analysis of factors determining treatment gaps in epilepsy care.

Figure 2.

Analysis of factors at government level.

Figure 3.

Analysis of factors from the health care system perspective.

Figure 4.

Analysis from a community perspective. Double arrows, flow in time; single arrows, influence on.

Level of health care development

The level of development is important because it influences knowledge about the true nature of the problem and its potential solutions. It also influences the means to reduce the treatment gap.

Whether the means available will be sufficient depends on the demand. Thus it might be argued that health care should be primarily directed toward prevention. However, in the first place, the prevalence of symptomatic epilepsies is less than one third of all cases. Not all symptomatic cases are attributable to preventable causes, such as deficient perinatal care or cysticercosis. It should also be stated that before the effects of prevention are felt, incidence cases, which arose before prevention became effective, will have on average a duration of 13 years (the ratio of prevalence over incidence data). Therefore closure of the seizure treatment gap remains a high priority.

Cultural beliefs

Cultural values affect people's health-seeking strategies. If people see epilepsy as caused by something that is not natural or biomedical, then treatment through western medicine may not be sought (6). Such beliefs may mean that people with epilepsy seek treatment from traditional (9) or faith healers (10) or that continued compliance with AEDs is difficult to obtain (11). These decisions may be rational within these peoples' cultures, but they may also lead to epilepsy being untreated or not treated successfully. It should also be mentioned that levels of literacy and knowledge will impinge on cultural beliefs insofar as choices arising from alternative beliefs, or evidence to back up choices, are limited if the means of communication and acquiring background knowledge are restricted (12).


As some of the sections concerning health care and financial coverage for each region indicate, the provision of AEDs through government-funded schemes may in some cases be impossible. It should also be stated that many developing countries' health sectors are hindered by World Bank structural adjustment programs, so that what little income they have is spent on debt repayment rather than on health expenditure (13–17), although this is not the only sector to be adversely affected. For example, in sub-Saharan Africa, incomes, exports, investment, health expenditure, and education expenditure have decreased, whereas debt, ecologic damage, malnutrition, and morbidity have increased (18–20). Although governments may lack resources, it may be equally true that individuals are either unable to pay for AEDs out of their own incomes, or such costs cause financial hardship.

The Pearson correlation coefficients do not generally show a strong relationship between health personnel and national income, but these figures may well hide wide disparities between and within countries, related more to income inequalities than to absolute levels of poverty or wealth. It may also, to a large extent, depend on the political orientation of the government in power and how integrated the country is into the world economy and World Bank policies.

Distance from health facilities

The problem of distance from health care facilities is in fact a problem of the know-how available at the community health care level. It has been shown that a reasonable level of seizure treatment can be achieved by primary health care workers (PHWs) (12,21). However, notwithstanding the objections to vertical programs, sufficient back-up for the PHWs should be available to give epilepsy care its proper place among their many other duties. Some participants argued that epilepsy care would be better off, if undertaken by community-based rehabilitation programs (see later).

Supply of antiepileptic drugs

Distance may also play a role in the lack of a sustained supply of prescribed AEDs. The majority of developing countries also have an extremely limited choice of drugs, with newer AEDs being mostly unavailable. However, the reasons for this are more based on income (both personal and governmental) and inequalities than on geography and distance. The problem is compounded by the evil of counterfeit drugs.

A further hindrance to drug supply and use can be pharmaceutical companies' and distributors' natural desire for high profitability. Profit margins may be too small for a drug to be commercially viable for the companies that produce it or for the pharmacies that would sell it. In this light, PB is an unfortunate paradox. Individuals and governments may find it too expensive, but its low cost relative to other AEDs ensures little profitability for pharmaceutical companies. Drugs that provide these companies with higher profit margins are even more expensive to those who need them most. It would seem then that not the diagnosis of epilepsy but the nonavailability of AEDs in developing countries is both a cause of the treatment gap and probably the single most important obstacle in bridging this gap.

Lack of prioritization

Although epilepsy is an important factor in the burden of neurologic and noncommunicable diseases (22) and an effective treatment often exists, it is not generally recognized as a public health priority (23,24). Epilepsy has to compete with many other conditions and illnesses. Such competition is even harsher where there are low health budgets, so that resources are prioritized for conditions perceived to be more important, such as malaria, HIV/AIDS, tuberculosis, etc. Although not wanting to detract from the scale of such communicable diseases, it may be more cost-effective to treat epilepsy than to treat these other conditions.


The causes described could be placed under political, economic, cultural, and medical headings. Each cause may cross over these headings, but it is helpful to think of them in these terms when it comes to discerning potential ways of overcoming the treatment gap. This section briefly outlines some thoughts on these ways and then describes two practical interventions in which the problem has been addressed.


A commitment to resources for epilepsy treatment must be gained from governments and international health organizations. Political patronage must ensure that epilepsy remains on the agenda and that essential drug supplies are assured. National chapters and the executive of the ILAE must foster links with health power brokers in each country. They will need to raise awareness of the scale of the problem and show how intervention makes economic sense.


Income inequalities and poverty at local, national, and global levels must be addressed for the treatment gap to be truly closed. Although these issues seem too large for epilepsy organizations and programs to address properly, epilepsy interventions can seek to make a difference at a local level. This may mean, for example, that an epilepsy intervention operates within a community development framework where literacy and income-generation programs are also included. These programs may help people with epilepsy to feel more accepted in society while also enabling them to afford treatment based in user fees.

It should also be pointed out that economic issues have been discussed further within the context of the ILAE/IBE/WHO Global Campaign against Epilepsy. It has been thought that these could be addressed at a national level through ILAE chapters providing employment and training facilities. This will be costly and possibly beyond the financial abilities of developing country chapters. In some countries, sufficient persons involved in epilepsy care able to constitute a chapter may even be unavailable. Therefore, the ILAE Executive Committee should propose that developed world chapters pay a percentage of their revenue into a “Developing Country Epilepsy Fund,” which can be disbursed to needier members of the ILAE. Links should also be encouraged between chapters in developed countries with those in developing countries. Such links could model the twinning of towns so that an equal partnership exists.


From her work in Swaziland, Reis (11) argued that cultural aspects should be studied in three dimensions: (a) patients' perceptions, (b) their practices in relation to epilepsy, and (c) the social relations in which these patients are embedded. When these issues have been studied, appropriate information, education, and treatment programs can be adopted that are culturally specific and offer more hope of success. This may also mean that anthropologists be included in any proposed intervention.

A further point for consideration concerns traditional healers. Epilepsy would appear to offer a good basis for cooperation between biologic and traditional medicine, particularly when neither may satisfactorily answer the needs of the patients alone. This cooperation should also help any treatment to be more culturally relevant.

In addition to understanding local cultures, communities must be brought into the planning and implementation processes of any intervention. This should help to ensure that these interventions are sustainable in the long term. In this regard, it may be appropriate to work with local or national NGOs in epilepsy service provision. The use of support groups may also positively affect cultural acceptability and sustainability (25).


Education and training of health workers is essential if people with epilepsy are to be correctly diagnosed and appropriately treated. If this treatment is to consist of AEDs, then drug supplies must be assured. Such assurance could come through the political links with government officials mentioned earlier, but it may also be necessary for chapters to work closely with the pharmaceutical industry to ensure that cheap supplies can be guaranteed.

Any treatment program should be fully integrated within a community health care delivery context. This will help to prevent it from following in the footsteps of many failed vertical interventions. As many cases of epilepsy in developing countries arise from other health problems, such as poor perinatal care and parasitic infection, epilepsy services should be integrated with other public health programs that take place at a government and community level. As action is taken against some of epilepsy's preventable causes, this should directly lead to a reduction in the gap over the long term.

Practical examples

Mumbai, India (N. Bharucha and M. Weiss)

Near Mumbai in 1986, a program was started to enable the basic structure of rural health care in India, the primary health care center, to expand its care to epilepsy. Primary health care workers were trained to identify possible patients with epilepsy, with the diagnosis being confirmed or changed by neurologists who would come to the primary health care center. If it was confirmed, PB was prescribed, and the treatment maintained by the PHWs.

When the primary health care system was first visited, the researchers were told not to waste their time because there was no epilepsy. However, now 188 people are under treatment (122 male and 66 female patients). Before entry into the study, 60 people were receiving allopathic treatment, 21 allopathic plus nonallopathic, 39 nonallopathic, and 68 no treatment.

Essentially, the study was trying to tackle how anthropologic and epidemiologic problems could be solved together. The approach involved using quantitative epidemiologic techniques and anthropologic techniques that focused more on illness experience than on traditional rates of occurrence and risk factors in disease outcomes. This led to the development of a method that investigated both the experience and the meaning of epilepsy with respect to outcomes of practical clinical significance.

This method was initiated by asking about patterns of distress: What is it that is bothering people? This would not only include clinical signs and symptoms that led clinicians toward a diagnosis, but also the perceived causes, which may be different from the established biomedical ideology. After this were questions concerning help seeking and treatment: How have people actually tried to address their problems in the community?

It should be mentioned that it was important for people to realize that the questions asked were not meant to test them in any way. If interviewees did think this, they would try to please the interviewer with the answer rather than providing a personal response. Open-ended queries were essential, for example: What is it that has been bothering you? What do you feel is the cause of this problem? What kind of help have you sought? In response to such questions, people generally provided an answer that was at the front of their minds rather than one aiming to please. The answers given then needed to be probed based on knowledge of the problem and of the community. For example: Which problem do you consider most important? Which provider did you go to first? This helped to build up a picture of how things are prioritized within the community and around the illness experience of epilepsy.

From the patterns of distress section, all but one respondent mentioned seizures. Although not usually expressed in response to the open-ended query concerning the nature of their problems, sadness and depressive symptoms were emphasized above all other aspects of problems when the respondents were asked further questions. Stigma was also an important aspect of the study, especially in terms of comparing it with other aspects of illness experience. For example: Are people who emphasize sadness or anxiety in the account of their illness more likely to associate stigma with their problem than are others? Are particular perceived causes or particular emphases on help-seeking more likely to be related to stigma or not?

Although authors were keen not to put much weight on the findings of the study, they stressed that their approach was essential. It was invaluable in understanding how the health system operated and what the perceptions of epilepsy were within the community. It should help to provide a cultural epidemiologic point of view as well a guide to clinical practice. For the clinician operating within a community, it is important to know the local conceptualization of epilepsy and what people's expectations are. The method presented both analyses and summarizes the situation. However, this analysis then begs the question, how can people with epilepsy receive better social and medical treatment? The service described later provides one way forward in answering this question.

Community Based Rehabilitation (D.K. Pal)

One model of service delivery for health and disability services is community-based rehabilitation (CBR). This is an increasingly used model in developing countries and addresses the problem that most people with disabilities in developing countries live in rural areas and are mostly poor. As a result, disability is usually not the sole priority in their lives.

The three main aspects to CBR are

  • 1Rehabilitation within the community, for that person and within that family;
  • 2Equalization of opportunities, fighting the effects of prejudice and discrimination; and
  • 3Integration within one's own community, within a school, within employment.

One of its main advantages is its low-cost approach. This is because it uses local resources: the disabled people themselves, their families, and those people around them. Why is CBR an appropriate vehicle for epilepsy? First, the theoretical rationale: cerebral impairments are often associated with epilepsy. Epilepsy is also a social handicap, and CBR is a life-long approach. Therefore, as long-term follow-up is vital to the success of treatment programs, CBR seems ideally suited for epilepsy treatment.

This present research used the theoretic framework known as the stress and coping research model, based on the work of Lazarus from the 1970s. This framework looks at stresses that may be short term, such as having a bad day, or a short-term illness, and long-term stresses such as poverty or disability. It looks at what resources are available to that individual, and it looks at what coping strategies are used to mobilize those resources, whether they are financial or psychological. Finally, it looks at outcomes in terms of individual and family and other broad outcomes. The aims of the study also partly included a look at a clinical trial for PB. The other main aim was to look at broad outcomes for intervention in the community setting.

This was an intervention study for children with epilepsy. Behavioral problems, how social integration changed, and how parents adapted to their children were studied. It also involved an epidemiologic study. From a population of almost 120,000, more than 300 were screened positive in the first stage, with 132 people being verified as having epilepsy in the second and final stage. Most of those who were eligible for the treatment trial entered that trail.

To perform the intervention, the study joined forces both with the CBR group and also with the local mother-and-child clinic. Doctors were trained in the MCH clinic to diagnose and treat epilepsy, and the disability workers in the CBR program would visit the families at home. These workers had several roles: to counsel families, to give first aid advice, to give support to the whole family, to make a neurodevelopmental assessment of these children, to design an individual program to promote their rehabilitation, to assess what their social relations were like, to act as advocates within the community to get them back into school, to get them playing with other children, to open up opportunities for them, to act as a medical liaison, and to count tablets and report back to the clinic on compliance.

We found with respect to social integration that 90% had deficits of some degree and that preschool children were not fully accepted by their peer group. The social integration of girls was not so affected, because many of their social responsibilities or normal social functions were within the home. After 1 year of intervention, improvements were found in both boys and girls, in young children and for older children, but more so for girls.

Some parents had some very negative feelings about their children, sometimes wishing that they were dead. Many families had difficulties in setting limits for discipline, and this may have been a factor in their children's behavioral problems. Great numbers had low expectations for social success with their children, and many mothers (78%) reported feelings of guilt, low self-esteem, and depression. Many of them believed that it was their fault that they had given birth to a child with epilepsy. Interestingly, the strength of social support available to parents throughout the 1 year follow-up, and whether the drug worked or not, predicted good parent adjustment. However, for children who had a high level of seizures, parents were less easily adjusted to their diagnosis.

Concerning behavioral problems: sex, age, number of seizures, type of seizures, presence of cerebral impairment, which drug was used, paternal education, and income were studied. It was found that none of these were of significant importance. The strength of social support was the sole independent predictor of behavioral problems in children.

At the end of the year, all the families were revisited, especially those who had dropped out. On average, 4 weeks of clinic attendance during the year had been missed by the 90 or so families in the study. The reasons for defaulting were studied, and it was found that these were first, maternal illness; second, child illness; third, domestic commitments, like marriages and going to visit family and so on; and fourth, problems of access, either geographic during the monsoon season, or financial because of seasonal income during the harvest time.

Drop-out was another interesting finding. Fewer than a third of the families dropped out of treatment by the end of 1 year. There were three main categories. The first was that the diagnosis itself was unacceptable. This reflected the screening and ascertainment methods, which forced the diagnosis of epilepsy on some people who did not want it. The second main reason was access, and this was mostly financial and social. For some people, the social cost of taking half a day away from domestic duties to come to the clinic was too much. In the latter part of the year, about a fifth dropped out because they no longer had seizures and so did not see the point of continuing follow-up.

The researchers concluded that CBR proved to be a good vehicle for epilepsy intervention. It promoted fairly good follow-up, almost 70%, and compliance was also good. It meant that people with the most needs were assessed appropriately and received the input they deserved. This enabled the intervention to be equitable. Intervention was tailored to the family needs by going to the home and deciding what their needs were. The costs were also amazingly low. The complete cost, including drugs, salaries, disability workers, and so on, came to $1 per child per month. This shows that including epilepsy in an existing service can enable the intervention to be very cost-effective.

In summary: PB was found to be a very effective drug; there were pervasive deficits in social integration in all the children of both sexes and all ages; and mother/child relations were affected. Intervention objectively improved both social integration and parental adjustment, and social support influenced both parental and child outcomes. This led the researchers to propose the following implications for service delivery:

  • 1PB was acceptable as a first-line agent.
  • 2A nonmedical intervention was very valuable.
  • 3Social support role is vital to prognosis. Service has to relate to a socioeconomic context, and a CBR is one of the valuable frameworks of control.


Despite advances in medicine generally and epilepsy treatment in particular, people with epilepsy in developing countries are continuing to live undiagnosed or inappropriately treated. Wotregap I has begun to address these issues through producing workable definitions and reviewing the current state of play, or lack of play, regarding epilepsy in developing countries. Although the scale of the seizure treatment gap is large and the problems and obstacles in bridging the gap are numerous, it is hoped that Wotregap I has provided evidence and enthusiasm that the gap is not completely insurmountable. It is now up to individuals, chapters, this Commission, and others like it to move this agenda forward.

The ILAE/IBE/WHO Global Campaign against Epilepsy is one means by which this agenda is being addressed. After 3 years of general awareness raising at national and international levels, demonstration projects are being initiated to see how best the treatment gap can be bridged. The protocols for projects in China, Senegal, and Zimbabwe are currently being finalized and are aiming to take forward many of the points and recommendations made in this article. It is hoped that these projects, and those initiated elsewhere, may become beacons by which epilepsy is brought out of the shadows and people with epilepsy are appropriately treated both medically and socially.


1. When dealing with treatment gaps, the modality of health concerned should be mentioned.

2. The definition for the seizure treatment gap recommended is:

The difference between the number of people withactive epilepsy and the number whose seizures are beingappropriately treated in a given population at a given time, expressed as a percentage. This definition includes diagnostic and therapeutic deficits.


Active epilepsy:

Two or more unprovoked epileptic seizures on different days in the prior year that are disabling to the individual.

Appropriate treatment:

Diagnosis and treatment of underlying causes; treatment of recurrent seizures according to international standards.

3. To estimate the percentage of people with epileptic seizures who are not treated [i.e., the seizure treatment gap (STG)], it is recommended that the following formula be used: [(0.005 multiplied by the population size) minus people treated for epilepsy] divided by 1% of people treated for epilepsy.

4. To estimate the number of people treated for epilepsy, it is recommended that the total amount of each AED dispensed be divided by its DDD, with the results being added together. [The denominator (the DDD) may have to be adjusted according to the average body weight of the population and/or the use of AEDs for other indications than epilepsy.]

5. Causal factors for a seizure treatment gap differ from country to country, but also within countries. It is recommended that the chapters of ILAE chart the situation in their own country. This research should be conducted with the help of both medical and social scientists. Drafts of the analyses of national and local seizure treatment gaps should subsequently be discussed at a regional level to study the feasibility of joint interventions to reduce these gaps.

6. A special analysis is needed to gauge whether investment in surgical treatment, notwithstanding its initial high costs, may be a cost-effective measure, even for countries with a relatively low GNP/capita.

7. It is recommended that the Executive Committee of ILAE invites all chapters and key persons in countries without chapters to report on the treatment gap in their country and its background, using this report as template.

8. ILAE chapters should foster strong links with government health ministers and bureaucrats to promote a higher profile for epilepsy.

9. The ILAE Executive Committee should propose a levy from richer chapters and organizations involved in epilepsy care to be distributed for epilepsy treatment in developing countries.

10. ILAE Chapters should have twinning arrangements across developed/developing country divides.

11. Anthropologists should be involved in any proposed epilepsy interventions to ensure that it is appropriate and relevant.

12. Interventions should piggyback on already existing primary health care services or community-based rehabilitation to ensure a more holistic and cost-effective approach to treatment.

13. Outcome measurements for interventions should be couched in both medical and social terms so that the intervention can be evaluated for its effect on both the treatment gap and the burden of epilepsy.

Acknowledgment: The Workshop was made possible by an unrestricted educational grant of SANOFI-SYNTHELABO, France.


  • * 

    Denotes member of the ILAE Commission on the Developing World

  • (From 19 May 2000, the Chair of the Commission on the Developing World has passed on from Rajendra Kale to Ley Sander and the position of Secretary to the Commission moved to Harry Meinardi.)


Participants in Wotregap I

Terry Aspray (UK), H. Aziz* (Pakistan), Nadir Bharucha* (India), Hanneke de Boer* (Netherlands), Zenebe Gedlie Damtie (Ethiopia), Amadou Gallo Diop (Senegal), Michael Dumas* (France), Pete Engel* (ex officio, USA), Olga Eisler (Hungary), Alla B. Guekht (Russia), Dede Gunawan (Indonesia), Peter Halasz* (Hungary), Pierre Jallon* (Switzerland), Rajendra Kale* (Chairman, India), Sylvia Kochen (Argentina), Byung-In Lee* (South Korea), Antonio Martins da Silva* (Portugal), William Matuja (Tanzania), Harry Meinardi* (Organizer, Netherlands), Jens Mielke (Zimbabwe), Shailen Nandy (UK), Deb Pal (UK), Pierre Marie Preux (France), Ria Reis (Netherlands), Ley Sander* (Secretary, UK), Nana Tatishvili (Georgia), Mitchell Weiss (Switzerland).