Originally published in JAMA.1999;281(4):361–367. doi:10-1001/pubs.JAMA-ISSN-0098-7484-281-4-jsc80337
Drugs offer a simple and cost-effective solution to many health problems in the world, provided they are available, affordable and properly used. African Trypanosomiasis, Shigella dysentery, Leishmaniasis, tuberculosis and bacterial meningitis are just some of the examples of health problems in poor countries for which effective treatment is lacking. Treatment may be precluded because no effective drug exists, or because it is too expensive, or even because it has been withdrawn from the market.
Moreover, reasearch and development in the area of tropical diseases has come to a near standstill. This article focuses on the problems of access to quality drugs for the treatment of diseases, problems that predominantly affect the developing world. The aim is to show that the problems are not independent and unrelated, but are a result of the fundamental nature of the pharmaceutical market and the way in which it is regulated.
The effective treatment of patients with drugs depends on a long chain of factors: research and development of an appropriate pharmaceutical agent, production, quality control, distribution, inventory control, reliable information for health professionals and the general public, diagnosis, prescription, financial accessibility, drug dispensing, observance and pharmacovigilance. At each level, those involved may have conflicting interests, and poor populations turn out to be the first victims of frail links in this long chain. Today, entire populations lack access to essential quality drugs, and the situation appears to be deteriorating, further marginalizing much of the world's population.
Essential drugs are the foundation for nearly every public health program aimed at reducing morbidity and mortality in the developing world, and pharmaceutical expenditure can account for a high proportion of the total health expenditure of a country. Important health programs that rely upon essential drugs include child survival programs, antenatal care, treatment of enteric and respiratory pathogens, control of tuberculosis, and malaria. Other major public health issues exist for which there is no effective pharmaceutical treatment.
This article focuses on four main issues associated with the inaccessibility of drugs for populations in greatest need: 1) the prevalence of poor-quality and counterfeit drugs on the market; 2) the lack of availability of existing essential drugs due to fluctuating production or to prohibitive cost; 3) the need to develop field-based drug research to determine optimum utilization conditions and to re-motivate research and development (R&D) programs for new drugs for the developing world; and 4) the potential consequences of the recent World Trade Organization (WTO) agreements on the availability of old and new drugs. For all these issues, practical recommendations to improve the situation are proposed.
The lack of access to essential drugs or vaccines due to economic reasons raises new human rights issues in a world that remains divided among rich countries, developing countries, and the rest of the world. Yet, financial access to drugs does not necessarily mean correct use. Continuous training for health professionals, disseminating reliable pharmacological data, and improving the management of drugs are all fundamental steps in improving the quality of care for patients which are not developed in this paper.
Identification of the Problems
Example of problems related to development and access to drugs and the magnitude of the public health problems concerned are given in Table 1.
Counterfeit and Substandard Products
Drug products must be manufactured according to strict Good Manufacturing Practices. Unfortunately, many developing countries do not have the technical, financial or human resources required for the application of such standards to their production. As for developed countries, they are sometimes less strict when the product being manufactured is destined for exportation. Today, the quality of drug products and therefore their effectiveness and safety is less and less certain, especially for the poorest populations who are attracted by lower priced drugs sold outside pharmacies.
Recent years have seen an increase in the prevalence of counterfeit and substandard drugs on the market. Counterfeit drugs are those that mimic authentic drugs; substandard drugs are those produced with little or no attention to good manufacturing practices.
During the meningitis epidemic in Niger from February to May 1995 (41, 000 cases reported), the Niger authorities organized an extensive vaccination campaign. In March 1995, Niger received a donation of 88, 000 Pasteur Mérieux and SmithKline Beecham vaccines from neighboring Nigeria. A Médecins Sans Frontières (MSF) team working with local health authorities noticed that the vaccines from Nigeria had an unusual appearance (difficult reconstitution, black filaments in the solution). Inquiries were made and Pasteur Mérieux laboratories confirmed that the batch numbers and the expiration dates did not correspond to their manufacturer records. The drugs supplied by these companies had been substituted with counterfeit drugs. Tests carried out found no traces of active product, confirming they were false. Bottles and labels were copied to perfection1, 2. Pasteur Mérieux subsequently filed a counterfeit suit. Some of the false vaccines (approximately 28, 000) were located by batch number and destroyed. According to estimates, around 60, 000 persons were inoculated with false vaccines out of a total 5 million vaccinated during the campaign. Such a production would have necessitated an industrial-scale production facility, and it is probable that the 88, 000 vaccines identified as false did not account for the entire fraudulent production.
MSF teams have encountered many similar field examples that lead to the following conclusions: organized illegal circuits seem more inclined to manufacture copies with the appearance of known trademark drugs (counterfeit), rather than comparatively less expensive generic products; whereas, non-organized illegal circuits (small production) increasingly manufacture drugs whose composition is substandard or inadequate, including generic drugs.
Poor quality may be accidental, with no intention to deceive, but oversights in manufacturing or neglected controls may sometimes have tragic consequences. Such was the case during recent decades with acetaminophen syrups which contained, by mistake, a lethal ingredient3, 4.
Fluctuating Production of Essential Drugs
Drugs necessary for the treatment of certain tropical diseases have begun to disappear from the market because they are commercially unprofitable. Many of these drugs were discovered in the 1950s and 1960s, or earlier, and are seldom or never used in wealthy countries.
An example is seen in the effort to treat epidemic bacterial meningitis, caused by Neisseria meningitidis, which is rampant in sub-Saharan Africa. Efficacy of treatment with chloramphenicol in oily suspension (1 intramuscular injection repeated after 48 hours) for bacterial meningitis is comparable to the traditional treatment with ampicillin (4 times daily intravenous injections over a period of 10 days)5. The lower cost of the chloramphenicol in oily suspension - only one-tenth of the cost of ampicillin - and its simple administration makes it particularly suitable to the precarious working conditions in developing countries. This is particularly important during epidemics, which can affect thousands of patients in the same region at the same time. In Nigeria in 1996 for example over 100,000 cases were seen6. However, production and availability of chloramphenicol in oily suspension are no longer guaranteed: Roussel-Uclaf Laboratory stopped its production in 1995 and transferred its technology to another laboratory which will only start production this year. In the meantime, temporary solutions have ensured that a certain (but far from sufficient) amount of chloramphenicol is made available.
The circumstances described above also apply to other serious illnesses, such as Leishmaniasis and its treatment with meglumine antimoniate, and African Trypanosomiasis and melarsoprol (Table 1). The trypanocidal activity of eflornithine (or DFMO) was discovered in 19857. It is the only treatment proven effective in cases where African trypanosomiasis shows resistance to melarsoprol, such resistance becoming more and more frequent (20% in Omungo, Uganda)8. This drug was sold at an extremely high price, and is now no longer manufactured. Only through a joint effort of the WHO, non-governmental organizations involved in field work, co-operative bodies, and pharmaceutical companies could this drug become available and affordable again.
The prohibitive cost of antiretrovirals for treatments of people with AIDS is well-known9. There are many other examples of existing drugs that are simply not affordable, most of which have been recently marketed and are therefore still patent protected.
Shigella Sd1 dysentery is extremely contagious and without an effective treatment is lethal in 5-15% of cases. Since 1979, this disease has been the cause of large epidemics in the African continent; for example, in Malawi in 1992-9310 and in Burundi in 199411,12,13. Shigella Sd1 bacteria quickly became resistant to traditional treatments. The only effective antibiotics today are fluoroquinolones (ciprofloxacin, norfloxacin). However, treatment with these new drugs is ten-times more expensive than the traditional treatment using nalidixic acid (US $20 vs. US $2). A special agreement was reached between Bayer Pharma Laboratory and MSF to make available 50,000 treatments with ciprofloxacin for a unit price of US $2 per treatment. This example shows that it is possible to find a short-term ad hoc solution with the pharmaceutical industry, yet no medium-term solution is anticipated.
A recent study of bacterial meningitis caused by Streptococcus pneumoniae in children aged 2 months to 3 years shows that by using ceftriaxone mortality could be reduced from 66% to 32% compared to treatment with chloramphenicol in oily suspension14. Both antibiotics have a sustained action and require very simple protocols (daily intramuscular injection for a short period of time): both are therefore equally easy to use under adverse conditions. However, ceftriaxone is ten-times more expensive than the chloramphenicol treatment. Streptococcus pneumoniae infection is also one of the main causes of severe acute respiratory infections – the primary cause of child mortality in Africa. Ceftriaxone is therefore vital but financially inaccessible to those populations that need it most of all.
Prohibitive pricing also extends to prevention when new vaccines are not available for the population most at risk. For example, Hepatitis B and anti-Haemophilus vaccines, which help to prevent major public health problems, are not accessible because of their steep price. Vaccines for Hepatitis B, a disease predominantly found in East Asia and sub-Saharan Africa15, are approximately ten-times more expensive than other vaccines included in the expanded programme on immunization (EPI) promoted by Unicef16.
Essential Drugs Not Adapted to Field Conditions
Tuberculosis caused the deaths of 3 million people in 1997, but the current treatment regimen, known as 'Directly-Observed therapy, Short-course' (DOTS) is impractical and compliance is poor – only 23% of the world's population have access to the WHO tuberculosis control strategy17. Research to simplify or shorten the DOTS regimen is needed to make the treatment more widely available. Furthermore, the emergence of strains that are resistant to the commonly used antibiotics has potentially devastating world-wide consequences. Current second-line treatments are too expensive, too complex, and too long, and therefore not realistic for actual field conditions. Priority should be given to simpler treatment guidelines that combine several antibiotics, which may not achieve the same level of efficacy of more complex protocols, but which are at least more practical for the field. Today, those suffering from multidrug-resistant tuberculosis in countries with limited financial resources are not receiving treatment, and from an ethical medical and humanitarian perspective this is completely unacceptable.
Access to drugs for poor populations would be greatly improved by research into new forms of existing drugs (eg. sustained action or rectal formulation) and the development of simpler treatment guidelines (one-shot or short treatments). This type of research cannot be developed unless technical and financial resources are made available, and more importantly, unless new efficacy criteria are applied to the treatment under study.
Insufficient Research and Development (R&D) for New Drugs
Growing drug resistance, unacceptable side-effects, and the lack of feasibility of current protocols point to the need for greater research and development into new drugs for diseases found in the developing world. From 1910 to 1970, the pharmaceutical industry's contribution was crucial to the fight against endemic tropical diseases: trypanocides and antiamœbic agents in the 1930s (Bayer, Rhône-Poulenc); chloroquine during World War II (Specia, Winthrop); and later in the 1960s, the discovery of leading anthelmintics (Janssen). Since then pharmaceutical companies have adopted a completely different strategy18.
Among the 1,223 new chemical entities commercialized from 1975 to 1997, 379 (30.9%) are considered as therapeutic innovations, but only 13 (1%) are specifically for tropical diseases (Table 2). A close analysis of these results shows that 2 out of those 13 drugs are actually updated versions of previous products (new formulations of pentamidine and amphotericin B), 2 are the result of military research (halofantrine and mefloquine), 5 come from veterinary research (albendazole, benznidazole, ivermectin, oxamniquine and praziquantel) and only 4 (0.3%) may be considered as direct results of R&D activities of the pharmaceutical industry (artemether, atovaquone, eflornithine and nifurtimox)19.
It can therefore be supposed that pharmaceutical R&D is abandoning tropical diseases. There are four main reasons for this shift:
- The costs and risks of R&D relative to the low purchasing power of consumers in developing countries. A typical R&D programme (from initial results to registration) would cost approximately US $160 million and take between 8 and 12 years to complete20. Moreover, a successful outcome is not guaranteed (as was the case with oltipraz, an antibilharzial agent which was abandoned during clinical trials).
- A shift to more profitable production. To cope with large investments and reduce duplicate spending, pharmaceutical companies started an unprecedented cycle of industrial consolidation and mergers at the end of the 1980s (Glaxo and Wellcome, Sandoz and Ciba-Geigy, Roche and Synthex, etc.). This consolidation focused on the most profitable segments of the market (infectious diseases, cardiovascular conditions, cancer, dermatology, and neurology), leaving tropical medicine largely out of the equation.
- Competition and counterfeiting of drugs. Some drugs patented in the developed world are being 'copied' in developing countries where patent rights of pharmaceutical products are not protected. Such production competes, sometimes fiercely, with the innovating laboratory. For example Bayer Laboratories, the patent holder of praziquantel, was outpriced by Shin Poong, a Korean laboratory, which had developed a less expensive manufacturing process21. In addition to these copies resulting from a different notion of intellectual property rights (see below), there are cases of pure and simple piracy (appropriation of the name and appearance of a trademark drug) which are frequent in countries where informal markets play a significant role22.
- The cost associated with adhering to quality standards. There has been a general trend toward heavier regulations with which companies must comply in order to obtain approval before marketing a drug product, which raise the costs of clinical development. The necessity of minimizing therapeutic risks voiced by consumers in the developed world leads to a reinforcement of various quality standards (good clinical, laboratory and manufacturing practices)23. In practice, when clinical development incidentally identifies a promising product (eflornithine used in the treatment of African trypanosomiasis) or a new indication (atovaquone for malaria, ivermectin for onchocerciasis, albendazole for lymphatic filariasis) for the treatment of tropical diseases, the manufacturer often decides not to market the drug, knowing it would be too expensive. The company generally decides to either make exceptional arrangements (donations in the cases of albendazole, atovaquone and ivermectin) or takes negative action (discontinued production in the case of eflornithine).
Globalization and Drugs: Questions and Concerns
A snapshot of the current landscape in the area of drug availability would not be complete without a consideration of the increasing globalization of the pharmaceutical industry and the potential implications of recent and upcoming world trade agreements.
Drugs: Another Industrial Product?
The final GATT agreement was signed on April 15, 1994 and was replaced by the WTO agreement signed in 199724. This agreement ratifies the world-wide implementation of a free trade economy. Its enforcement with regard to the pharmaceutical sector raises certain doubts and concerns. Two types of provision seem particularly important for pharmaceutical companies in developing countries: those whose purpose is to put an end to protectionist measures; and those which define as mandatory the protection of patents on drugs and their respective manufacturing processes, i.e. the trade related aspects of intellectual property rights (TRIPS Agreement). This is particularly important, as many developing countries do not fully acknowledge patent protection rights in the area of pharmaceuticals.
A Newly Invigorated Tropical Research?
Pharmaceutical companies in the developed world have stated repeatedly that the reason for not conducting research on tropical diseases is the lack of protection for innovations in some developing countries, which would also explain their limited investments in the countries concerned25. The moment the enforcement of patent protection becomes effective (in developing countries no later than January 1, 2006) tropical disease research should logically start again, funded by Western companies or by manufacturers in developing countries.
However, it is unlikely that Western manufacturers will devote much of their effort to non-solvent populations, with or without patents. Manufacturing companies in developing countries may actually be motivated to invest more in research for new drugs, but such investments will essentially respond to the need to shift their innovation capacity (devoted today to finding ways to copy the patented drugs of developed countries) to discovering new drugs26. All things considered, it is to be feared that tropical research will not have a more promising future, even if patents are widely enforced.
Increasingly Prohibitive Prices?
A study financed by American pharmaceutical companies shows that granting drug patents does not tend to increase the price of drugs on the market27. This study, however, does not examine the prices of new innovative drugs and declares that, logically, the price of these new drugs should be higher. Naturally, when the manufacturing company is assured that its product cannot be copied, it holds a stronger position to negotiate prices with public health authorities. Moreover, the liberalization of international pharmaceutical trade entails the development of parallel imports between countries where the same drug is sold at different prices. Pharmaceutical companies, which are consequently less and less inclined to grant significantly lower prices to less developed countries, may tend instead to set unique world-wide prices or delay marketing their drugs in developing countries26. In either case, access to drugs is jeopardized.
The revised drug strategy and the essential drug concept are still key strategies to help improve access to essential drugs and contribute to improve health. The essential drugs concept is evidence based, it is simple, it promotes equity and is rooted in firm public health principles. WHO support to countries and advocacy work to promote the essential drugs concept and to support countries in the formulation and implementation of national drug policies has resulted in change for the better. This strategy is a proven success but it needs to be continued and strengthened, and new ways of implementation have to be explored given the changing context. In this spirit, the following recommendations are made with respect to the four main issues that have been developed in this paper.
Procurement of quality drugs
In order to improve the quality of existing drugs and their procurement, it is important to develop a permanent 'observatory of drug quality', established by WHO in collaboration with organizations involved in the provision of essential drugs (Unicef, World Bank, EU, NGOs), that would oversee the implementation of adequate and effective control procedures.
The practical knowledge acquired by international organizations to assure the quality of generic drugs must be shared with health authorities in developing countries. Invitations to bid, required by big sponsors such as the World Bank, EU, and USAID must combine quality criteria and lower costs. Furthermore, procurement of drugs should be centralized at a national level in order to reinforce the responsibility of governments to make procurement, quality control, stock management, and distribution of essential drugs a priority.
To provide better access to effective treatments for people in the greatest need, several initiatives must be launched now, even if their consequences will not be felt immediately. In the short term, practical solutions – involving the various partners – must be found to maintain the production of essential drugs. By establishing public health priorities, recent but high-priced drugs must be made available to the poor through solutions similar to those implemented for EPI vaccines where the supply is guaranteed by Unicef. These drugs could be made available by creating centralized purchase funds whereby manufacturers would be guaranteed large sales volumes (financed by existing public and private funds). The fund would also set forth, by consensus, compliance with drug indications. Finally, operational research in the field must be promoted and developed in close collaboration with health professionals in developing countries. Such research should produce simple, efficient, and low-cost protocols without losing sight of the risk-benefit factor for the poorest countries.
Restart of R&D
In an attempt to offset this costly structural evolution in the pharmaceutical industry, several public and private initiatives have tried to introduce public health criteria in the choice of R&D strategies. The implementation in 1975 of the UNPD/World Bank/WHO Special Programme for Research and Training in Tropical Diseases has had outstanding results in strategic research (entomology, pathogenesis) and has bolstered research potential (epidemiology, training). However, strategies for product research and development that were actually launched in 1994 have yet to produce any convincing results28. Nevertheless, they have succeeded in raising awareness and have promoted reflection on potentially effective tools, even if most projects focus exclusively on Malaria (e.g. The Multilateral Initiative on Malaria). The Orphan Drug Act implemented by the United States in 1983 has also produced significant results for rare diseases (157 new drugs were commercialized and 837 new indications were developed from 1983 to 1997), but no real impact has been seen with respect to tropical diseases29. We can therefore conclude that while such initiatives may occasionally boost the development of new drugs (e.g. derivatives of artemisinine and pyronaridin), they are unable to significantly redirect the R&D process towards tropical diseases. In the medium term, a legal and fiscal framework must be developed to spur research and development on tropical diseases or related areas, similar to those developed in the United States for orphan drugs used in rare diseases.
Humanizing the WTO agreements
On the whole, it is regretful that WTO agreements contain no specific provisions that would guarantee both funding for ambitious tropical pharmaceutical research and realistic pricing of potential drugs. Some developed countries were however able to protect vulnerable economic and business sectors (textiles, agriculture, culture). One can understand why rich countries demand that developing countries comply with regulations on unfair competition. It is obvious that to meet pressing public health needs we need new essential drugs. To develop them we need innovative research and industry. To fund new research, industry needs commercially viable results. It is therefore vitally important that the pharmaceutical industry collaborates with organizations like WHO, Unicef and the World Bank to map out the challenges and get a clearer view of what they can achieve together in developing sustainable markets for new tropical pharmaceuticals.
It must be remembered that those developed countries that are the main sources of cheap copies of patented drugs (28) are nevertheless relatively poor countries. Enforcing the WTO regulations will remove a source of affordable copies of innovative quality drugs upon which the poorest countries depend. Developing countries, particularly the less advanced, should be encouraged to take advantage of the limited alternatives offered by WTO agreements. Specifically, they should be able to obtain 'compulsory licenses' whereby national authorities allow local manufacturers to circumvent patents rights (under certain conditions and in return for the payment of royalties to the inventor; as allowed for in article 31 of the WTO agreements) (30). Judiciously enforced, such an alternative seems to be the only recourse to find a balance between the interests of the developing and rich world.
WHO is in a unique position to argue the case for health at an international level. Health NGOs and consumer organizations certainly have a supportive role to play, but the WHO is the only intergovernmental organization with a formal international mandate to protect and advance health at an international level. While the WHO's authority in this area has suffered in the last decades, part of WHO's strategy should now be to clearly and unambiguously put health first and to provide leadership in promoting access to essential drugs.
As a medical emergency organization present in 80 countries through 400 medical assistance projects, Médecins Sans Frontières undertakes to speak about the living conditions of those who cannot speak for themselves and to defend their right to vital healthcare. However, it would serve no purpose to demand new public health rights or human rights in a manner that would make one believe that such rights will soon become a reality. The current situation points to the opposite. For a great proportion of mankind, health conditions are worsening and perspectives for improvement are not encouraging.
This paper is mainly based on the field experience of Médecins Sans Frontières and our local partners. The authors wish to thank the many field volunteers who in one way or another have participated in the gathering the information related in this paper. Special thanks are due to Nathan Ford for reviewing the manuscript.
1. Pinel J et al. Médecine et maladies infectieuses. 1997;27 (n° spécial 1-563).
2. WHO. Fake drugs: a scourge on the system. WHO drug information. 1995;9(3): 127-129.
3. O'Brien KL et al. Epidemic of pediatric deaths from acute renal failure caused by diethylene glycol poisoning. JAMA. 1998;279:1175-1180.
4. Shakoor O, Taylor RB and Behrens RH. Assessment of the incidence of substandard drugs in developing countries. Tropical Medicine and International Health. 1997;2(9):839-845.
5. Pécoul B, Varaine F, Keita M, et al. Long-acting chloramphenicol versus intravenous ampicillin for treatment of bacterial meningitis. Lancet. 1991;338:862-866.
6. OMS. Riposte à la méningite épidémique en Afrique, 1997. BEH. 1997;72:313-318.
7. Doua F, Boa FY, Schechter PJ et al. Treatment of human late-stage gambiense trypanosomiasis with (-difluoromethylornithine (eflornithine): efficacity and tolerance in 14 cases in Côte d'Ivoire. Am. J. Trop. Med. Hyg. 1987;37(3):525-533.
8. Legros D, Enyaru JCK, Evans S, Maiso F. An outbreak of relapses among patients treated for T. b. gambiense trypanosomiasis in Arua, Northern Uganda. 24th International Scientific Council for Trypanosomiasis Response and Control (ISCTRC). Maputo; September 1997.
9. Coll-Seck A, Sylla A, Over M et al, Implications of therapies for developing countries, 12th World AIDS Conference. Geneva, June 1998.
10. Paquet C, Perea W, Grimont F et al. Aetiology of haemorragic colitis epidemic in Africa. Lancet. 1993;342:175.
11. Center for Disease Control. Mortality from dysentery in Africa: a follow-up study in Burundi. Internal report. Atlanta: CDC; 1993.
12. Paquet C, Leborgne P, Sasse A, Varaine F. An outbreak of Shigella dysenteriae type 1 dysentery in a refugee camp in Rwanda. Cahier Santé. 1995;5(3):181-184.
13. Paquet C, Van Soest M. Mortality and malnutrition among Rwandan refugees in Zaïre. Lancet. 1994;344:823-824.
14. Varaine F, Keita M, Kaninda AV et al. Long acting chloramphenicol versus ceftriaxone for treatment of bacterial meningitis in children aged 2-35 months. 8th International congress on Infectious Diseases. Boston, USA; May 15-18, 1998.
15. Tandon B N, Acharya A, Tandon A. Epidemiology of Hepatitis B virus infection in India. Epidemiol. Infect. 1996;117(2): 313-325.
16. Kaddar M. La mutation du marché mondial des vaccins. La revue Prescrire. 1995; 157:844-847.
17. Sir Crofton J, Chaulet P, Maher D. Principes pour la prise en charge de la tuberculose à bacilles résistants. Geneva: World Health Organization; 1997. WHO/TB/96.210 (Rev.1).
18. Trouiller P. Recherche et développement pharmaceutiques en matière de maladies transmissibles dans la zone intertropicale. Cahiers Santé. 1996;6:299-307.
19. Trouiller P, Rey JL, Olliaro P. Analysis of drug development patterns of six tropical diseases between 1975 and 1997. 8th International congress on Infectious Diseases. Boston, USA; May 15-18, 1998.
20. Grabowski H. The effect of pharmacoeconomics on company research and development decisions. Pharmacoeconomics. 1997;5:389-97.
21. International strategies for tropical diseases treatments: experiences with praziquantel. Harvard School of Public Health. Boston; 1996.
22. WHO. Counterfeit drugs: report of a WHO/IFPMA workshop. Geneva: World Health Organization; 1992. WHO/DMP/CFD/92.
23. WHO. Recommendations from the ICDRA reinforce the mission of regulatory authorities. WHO Drug Information. 1996;4:182-5.