Essential drugs and registration ofpharmaceuticals: the Sri Lankan experienceK. Weerasuriya1

Many factors influence the regulation of pharmaceuticals in a country. The essential drugs concept, for-mulated by the World Health Organization to assist developing countries in selecting appropriate drugs,also provides a basis for regulation. Sri Lanka has long regulated pharmaceuticals as part of its healthpolicy. Over 70% of 3436 pharmaceutical product registrations were found to be drugs (or alternatives)named in the country's essential drugs list. This is despite the fact that product registrations are mainlyfor the private health care sector, and the list is for the state sector. The essential drugs concept there-fore appears to have influenced the pharmaceuticals registered in Sri Lanka.

IntroductionIn the 1970s, in response to the problems faced bythe developing countries in selecting drugs appro-priate for their health needs, the World Health Orga-nization formulated the essential drugs concept(EDC) and presented the first model list of essentialdrugs (1). The essential drugs list (EDL) used gen-eric names and had only one drug from each group(e.g., promethazine from the antihistamines), butsuitable alternatives were also given. Items withmore than one drug (combination preparations) wereincluded only if there was a clear need for them. Thelist was meant to be a guide for developing countriesto develop their own national drug lists. In the 1980sthe focus of the concept shifted from formulation toimplementation (2). Although the main emphasiswas on selecting and supplying essential drugs, theconcept also provided a basis for the regulation ofpharmaceuticals. This article examines the drugregulatory aspect by comparing the pharmaceuticalproducts registered in Sri Lanka with items named inthe country's essential drugs list.

Drug registrationFactors influencing registration

The number and type of drugs that are available in acountry are controlled by registration. Initially thepurpose of registration was to exclude harmful drugs.Most countries now also assess safety and efficacy,and a few may consider the necessity for a particulardrug. Factors not related to the use of the drug in

I Senior Lecturer, Department of Pharmacology, Faculty ofMedicine, University of Colombo, P.O. Box 271, Colombo, SriLanka.Reprint No. 5378

clinical practice could also influence registration:the country's economy and political philosophy, andthe pharmaceutical industry may exert separate andsometimes contradictory influences on registration(3). For example, the drugs required by the healthneeds of a country may not necessarily be profitablefor the pharmaceutical industry. Nevertheless agovernment may consider a profitable pharmaceu-tical industry, particularly one with strong exportearnings, to be economically desirable. Multinationalpharmaceutical companies often wield considerableclout and may well influence registrations in coun-tries (especially the developing ones) that lack astrong regulatory system. The complex interplay ofthese factors in actual practice contrasts with theapparent simplicity of the stated aims of drug regis-tration.

Drugs (active substances) andpharmaceutical products

The term drug, even within the context of pharma-ceuticals, could have different meanings. In thisarticle the term "drug" is used for an active sub-stance (e.g., propranolol), and the term "pharmaceu-tical product" indicates the different dosage formsand strengths of branded and generic preparations aswell as the combinations (e.g., propranolol with adiuretic) of that drug. Depending on the country'sregulations, a brand with all its dosage forms canbe a single registration, or each strength of a dosageform of a brand would have to be registered separ-ately. Hence the number of registrations is not a goodindicator of the number of drugs registered.

Pharmaceutical registrations

With the exception of the Netherlands and Scandina-vian countries, data on pharmaceuticals registered ina country are often difficult to obtain. In the 1980s

Bulletin of the World Health Organization, 71 (2): 229-235 (1993) © World Health Organization 1993 229

K. Weerasurlya

Finland had about 3500 (4) and Sweden about 2800products (5); Norway in 1991 had 2200 productswith 670 drugs (6). A recent symposium on Inter-national Medicines Regulation had data on annualproduct approvals but no estimates of the numberof products in the USA, United Kingdom and Japan(7). Germany may not have a list of registered prod-ucts (8), though estimates range from 15 000 to123 000 products (9). In developing countries the datamay be scanty or non-existent. Colombia and Kenyahave about 6250 and 3500 pharmaceutical products,respectively, but whether these numbers arebrands only, or strengths of the individual dosageforms of each particular brand is not known (10);Thailand has at least 23 000 products (11) and India60 000 (12). Registration was voluntary in Nigeriaand was not required in Burundi until 1990 (10).

Situation in Sri LankaPharmaceutical supply and regulationFrom the 1940s health care has been providedthrough a state-financed curative and preventiveservice. Reasonable health indices have beenachieved; infant mortality is 24.2 per 1000 live birthsand life expectancy, is 67.8 and 71.7 years, respec-tively, for men and women (13). A smaller privatehealth care sector has continued alongside the statehealth service.

Pharmaceuticals were first regulated in SriLanka as part of its health policy in the late 1950s.This was primarily by controlling imports through asystem of gazette notification. There was little localmanufacture then as now. A formulary for the statehealth sector was prepared in 1959 and updatedregularly; the state imported generic drugs in bulkfrom this list, and distributed them to its institutions(14). Doctors outside the state sector were free to useany pharmaceuticals imported by the private sector.In the 1970s the govemment took over the import ofall pharmaceuticals and supplied both the state andprivate health care sectors (15). But with a change ofgovemment in 1977 the private sector was againallowed to import pharmaceuticals. At present thestate is the sole importer for its institutions, and alsosupplies the private health care sector along with theprivate importers.

Registration of pharmaceuticals, which replacedthe previous system of gazette notification, wasstarted in 1987 and is a prerequisite for importingand marketing a pharmaceutical product.a Safety

a Cosmetics, Devices and Drugs Act No. 27 of 1980. TheGazette of the Democratic Socialist Republic of Sri Lanka, 2December 1985, No 378/3.

and efficacy were sufficient grounds for registration;the necessity for the product was not examined. Asubcommittee on drugs from the Ministry ofHealth, comprising clinical pharmacologists, clini-cians from the state health sector and representativesfrom medical associations, decides on applicationsfor registration.

An essential drugs list was prepared by theMinistry of Health in 1985 (16) and the currentversion (revised in 1988) contains 230 items with210 drugs and 14 combinations (these do not add upbecause the same drug can appear as two items, e.g.,propranolol as an antiarrhythmic and as an anti-hypertensive). The list has been emphasized in thestate health institutions but not in the private healthcare sector. Nor was preference given in registrationto drugs in the essential drugs list.

Pharmaceutical products registered

Pharmaceutical products, as defined by legislationand requiring registration, are medications, vaccinesand diagnostic agents. Each of the different dosageforms and strengths of a product must be registeredseparately. Each registration up to June 1991 wasexamined for the following information: drug(s)contained and whether it was a single or combinedpreparation, an over-the-counter (OTC) preparationor prescription only medicine (POM), as well as themanufacturer and the country of origin. At registra-tion, the products were classified according to theAnatomical Therapeutic Chemical (ATC) classifica-tion (17); this allowed a breakdown of the productsby the anatomical system for which it is intended,and also whether the drug in the product was anamed item or an altemative in the EDL.

A total of 3436 registrations were analysed bythe author in December 1991; 31 were excludedowing to incomplete product information, leaving atotal of 3405. In all, there were 773 drugs.

OTC product registrationsThere were 697 OTC products, 318 (46%) of whichwere from the essential drugs list. The great majorityof these had one drug and the rest were from thesmall number of combinations in the list. Whenaltematives to drugs from the EDL were included,the proportion increased to 516 (74%). In all, theOTC category had more single (399, 57%) than com-bined products. Of the 228 drugs in this category,only 55 (24%) were from the EDL but these werepresent in 437 (63%) of the products. Table 1 showsthe ten commonest drugs, and Table 2 gives a break-down by the anatomical system for drug action.

230 WHO Bulletin OMS. Vol 71 1993

Essential drugs and pharmaceutical registration In Sri Lanka

Table 1: The ten commonest drugs registered In theOTC category

Drug No. registered Single Combined

Paracetamol 67 63 4Vitamin A 47 1 46Aluminium hydroxide 40 7 33Vitamin B12 40 4 36Vitamin B2 26 1 25Vitamin D2 25 0 25Vitamin C 25 12 13Betamethasone 24 11 13Ferrous sulfate 23 4 19Magnesium hydroxide 21 4 17

Table 2: Number of registrations of OTC products, bythe anatomical system for drug action

Alimentary 254 (36) aRespiratory 110 (16)Dermatological 98 (14)Nervous System 96 (14)Blood 55 (8)Othersb 84 (12)

Total 697 (100)a Figures in parentheses are percentages.b Of these, 32 (5%) were antiparasitic drugs.

POM product registrations

There were 2708 prescription only medicines, 1332(49%) of which had drugs in the same formulation asin the EDL; this increased to 1913 (71%) when thealternatives were included; 2308 (85%) were singleproducts. Of the 636 drugs, 188 (30%) were from theEDL and were present in 1484 (55%) of the prod-ucts. Table 3 shown the ten commonest drugs andTable 4 gives a breakdown by anatomical system.

Origin of registered productsThe countries of origin are shown in Table 5. Theten countries with the most number of registrationsaccounted for over 75% of the total. The productsmanufactured in Sri Lanka were mainly OTC; themajority of the products from other countries werePOM. The world's top ten pharmaceutical firms,ranked by sales (18), had registered products throughtheir local subsidiaries or agents.

DiscussionAlmost half the pharmaceutical products registered(POM and OTC) in Sri Lanka had drugs straightfrom the essential drugs list and these were formula-

Table 3: The commonest drugs registered in the POMcategory

Drug No. registered Classification a

Ampicillin 75 EDLAmoxycillin 73 AlternativeCotrimoxazole 68 EDLSalbutamol 45 EDLMetronidazole 42 EDLHydrocortisone 42 EDLNeomycin 40 EDLDexamethasone 38 EDLBetamethasone 33 EDLTheophylline 31 EDLChloramphenicol 31 EDLGentamicin 31 EDLa Drugs given in the essential drugs list (EDL) or mentioned asan alternative.

Table 4: POM products, by anatomical category oraction

Anti-infectives 562 (21) aCardiovascular 290 (11)Nervous system 283 (10)Dermatological 264 (10)Alimentary 264 (10)Respiratory 250 (9)Others 795 (29)a Figures in parentheses are percentages.

ted as specified in the list. This proportion increasesto slightly more than 70% when alternatives todrugs in the EDL are included. The total numberof pharmaceutical products registered in Sri Lanka(3436) is small when compared with other devel-oping countries, e.g., Indonesia (12 600), Thailand(23 000) and India (60 000). Only 773 drugs wereregistered, which is also much less than the 1279 inIndonesia, the other developing country for whichthis information is available.b The Scandinaviancountries have well developed regulatory systemsand comprehensive health care schemes and around700 drugs and 2500 products. Sri Lanka appears tobe closer to these countries than to other developingcountries with regard to pharmaceutical registrations.

OTC products contain drugs with a wide safetymargin and are meant for the treatment of minorailments. Generally, combinations are common andthe proof of efficacy required is less stringent than

b Action Programme on Essential Drugs and Vaccines.Review of the drug programme in Indonesia. Unpublished WHOdocument DAP/90.11, 1990.

WHO Bulletin OMS. Vol 71 1993 231

K. Weerasuriya

Table 5: Number of registered products, by country oforigin

No. registered

India 886 (26) aUnited Kingdom 469 (14)Sri Lanka 305 (9)Switzerland 199 (6)Pakistan 158 (5)West Germany 156 (5)USA 146 (4)Australia 105 (3)Belgium 97 (3)Cyprus 82 (2)Others 802 (23)

Total 3405a Figures in parentheses are percentages.

for the POM category. Although combination prod-ucts were common in the OTC products registeredin Sri Lanka, drugs from the essential drugs listfigured prominently among them. Combinations ofvitamins made up the major portion of the productsfor the alimentary system (Table 2). Paracetamol,which is registered mainly as a single product, wasthe commonest drug in the OTC category. Incontrast, in the United Kingdom, despite the BritishNational Formulary discouraging the use of com-bined analgesics, at least 11 combined preparations ofparacetamol are registered (19).

In the POM category, the 12 most commonlyregistered drugs were all from the essential drugs listexcept for amoxicillin which is an alternative. Theywere also single preparations except for cotrimoxa-zole, a combination that is in the list (Table 3).Among these 12 drugs were six antibiotics; anti-infective products (which included antibiotics and asmall number of vaccines and sera) made up thebiggest category (Table 4). Developing countriesgenerally have a high percentage of prescriptions forantibiotics and in some, antibiotics constitute thesingle largest group of the drug budget (20).

Items given in the essential drugs list of a coun-try are related to its health needs. Although drugsfrom the Sri Lankan EDL dominated the registra-tions, some of the disease patterns in the countrywere not reflected in the type of products registered.For example, parasitic diseases are common in SriLanka, as in other developing countries. There were300 000 slide-positive cases of malaria in 1991C andhelminthiasis is common; however, products for

c Administration report of the Anti-Malaria Campaign 1990.Ministry of Health, Sri Lanka.

parasitic diseases constituted less than 3% of allregistrations. This could have been partly due to thesmall number of active substances available for thetreatment of parasitic diseases, and to the low profitmargins on products based on these substances.

The fact that a drug (or one of the alternatives)is from the essential drugs list may not ensure appro-priate drug utilization. For example, propranolol adrug in the Sri Lankan EDL, is cheap; a similar betablocker would be classified as an alternative. If thisalternative were to be costlier but yet used moreextensively than propranolol, that would be regardedas inappropriate drug use. At present there are noindications of such alternatives being used extensive-ly in Sri Lanka, but drug utilization data are requiredfor a definite conclusion.

Nearly all the smaller developing countriesdepend heavily on imports for their pharmaceuticalrequirements. Sri Lanka is no exception (Table 5)and the major multinational companies continue tomarket their products in the country. Improper adver-tising (21), sale (22) and used of pharmaceuticalshave been reported in Sri Lanka as from other deve-loping countries. In such a context it is curious tonote that registrations are dominated by the drugsfrom the EDL.

As the state purchases drugs named in its formu-lary (see above, page 230) from a small number oflong-standing suppliers, the majority of products areregistered by the pharmaceutical suppliers to the pri-vate sector. Thus the, dominance of drugs from theEDL in the registrations implies a popularity of thesein the private health care sector. This popularity can-not be attributed to any official encouragement, asthe EDL is not imposed on the private sector. As farback as 1982, when only the WHO Model EssentialDrugs List had been developed a study in the privatesector in Sri Lanka demonstrated the popularity ofthe drugs from that list; drugs from the model list,with a few minor exceptions, were found to be suffi-cient in a family practice (23). The drugs from theSri Lankan EDL may be popular in the private healthcare sector simply because of their cheapness, safetyand efficacy, which are precisely the reasons forincluding them in the EDL. It would have been use-ful to compare the concordance of registrations andthe EDLs of other countries with the data from SriLanka. However, although about 100 countries haveadopted an EDL (10) and there are many publica-tions on these lists, this study appears to be the firstcomparison of the list with the pharmaceuticalproducts registered in any country.

d Wolffers, I. Changing traditions in health care in Sri LankaDissertation, University of Leiden, Netherlands, 1987.

232 WHO Bulletin OMS. Vol 71 1993

Essential drugs and pharmaceutical registration in Sri Lanka

Although the EDL is not officially imposed onthe private health care sector, the list could haveinfluenced the drugs used in that sector in subtleways. Clinicians in the subcommittee that considersapplications for registration are mainly from the statesector where they work using the state formulary,with its emphasis on drugs in the EDL. Whenevaluating an application for registration, althoughthe criteria specify only safety and efficacy and donot insist on "need", there would still be a tendencyto evaluate the appropriateness of a product forregistration. The reasons for rejection of applicationswould be a useful indicator of the attitudes of thecommittee. This information is not available but thecommittee in 1991 deregistered antidiarrhoeals(loperamide syrup, kaolin/pectin) that were con-sidered harmful or ineffective (24),e thereby pre-cluding the local manufacture or importation ofthese products. Decisions of the subcommittee haveso far not been challenged.

Drugs used in the private sector could have beeninfluenced in another way by the doctors working forthe state. These doctors are allowed private practiceoutside official working hours and would be inclinedto prescribe drugs they are familiar with. Since thesewould be from the essential drugs list, the demandfor these drugs in the private sector would be enhan-ced. This implicit acceptance of the principle of theessential drugs concept and its use in everyday prac-tice by the doctors, even when an opportunity forusing other drugs exists, suggests a positive attitudeto the concept rather than a negative one of limitingthe freedom to prescribe.

It has been claimed by an authoritative commen-tator on the pharmaceutical industry that, in manydeveloping countries, drug regulation "tums out tobe enthusiastically inefficient" and "often ends up bydepriving the very people it is designed to protectand benefit" (25). However, in Sri Lanka the majorand long involvement of the state in pharmaceuticals,and the influence of WHO's essential drugs concepton the registration of pharmaceutical products in thecountry appear to indicate a very opposite, and posi-tive influence.

It is difficult to measure the effect of the essen-tial drugs concept in achieving the ultimate aim ofbetter health. An exhaustive evaluation of WHO'sAction Programme on Essential Drugs in 1989showed strong indications that the availability ofgood quality, low-cost drugs had increased, but therewas insufficient evidence to directly relate the extentof the changes to the programme (10). A similar

e The Gazette of the Democratic Socialist Republic of Sri Lanka,15 November 1991, No. 688/29.

difficulty exists with regard to the EDL and pharma-ceutical registrations in Sri Lanka; emphasis on thedrugs from the list may have contributed to theirdominance in the pharmaceuticals registered in thecountry, but precise measurement of this contributionis difficult.

ConclusionThe essential drugs concept provides a sound basisfor regulating pharmaceuticals by prioritizing thedrugs needed in a country. Pharmaceutical productsregistered in Sri Lanka are dominated by drugs fromthe EDL, a list devised for the state health care sec-tor. However, since the state imports from a smallnumber of long-standing suppliers, the majority ofregistrations involve products for the private healthcare sector, which in no way is restricted to the EDL.Doctors in private health care are equally free to useany pharmaceutical product that is registered. Thepredominance of drugs in the EDL among the phar-maceutical products registered in Sri Lanka, whichindicates their appropriateness to the health needs ofthe country, is very probably a result of the state'slong and major involvement in health care and phar-maceutical supplies.

AcknowledgementsI thank Ms Chinta Abayawardana and Mr Sajeeva Rana-weera for helping in the data analysis, and Professor Col-vin Goonaratna and Dr Laal Jayakody for valuable discus-sion and criticism of the manuscript. also thank theInternational Organisation for Consumer Unions, RegionalOffice for Asia and the Pacific, Penang, Malaysia, for pro-viding a small grant to initiate the database on pharma-ceuticals. I am grateful to the Director-General of theHealth Services of Sri Lanka for permitting me to use thedata on registrations.

ResumeMedicaments essentiels et homologationdes produits pharmaceutiques:I'exp6rience de Sri LankaDans les annees 70, l'Organisation mondiale de laSant6 a formule le concept des medicamentsessentiels pour aider les pays en developpementa choisir les m6dicaments adaptes a leursbesoins de sante. Les medicaments essentielsaux besoins de sante de la majorite de la popula-tion et, le cas echeant, un certain nombre de pro-duits de substitution appropries, ont ete inscritssur une Liste des Medicaments essentiels (LME).

WHO Bulletin OMS. Vol 71 1993 233

K. Weerasuriya

Cette liste devait 6galement constituer un outil debase pour I'homologation des produits pharma-ceutiques dans chaque pays. Le pr6sent articlecompare les medicaments enregistr6s a Sri Lankaa ceux qui figurent sur la LME nationale.

On dispose de peu de donn6es sur I'homolo-gation des m6dicaments; dans la plupart despays, il semble qu'il existe un nombre important(et probablement excessif) de produits pharma-ceutiques, ce qui tient sans doute aux nombreuxfacteurs qui influent sur I'homologation. A SriLanka, l'Etat est le principal dispensateur de soinsde sante et, depuis plusieurs d6cennies, il aentrepris de r6glementer les medicaments dans lecadre de sa politique de sant6 (cela signifie enpratique qu'il contr6le les importations par le biaisde l'homologation). Un formulaire destin6 auxh6pitaux publics a 6t6 adopt6 dans les annees 50;depuis lors, les m6dicaments generiques sontimport6s en vrac par l'Etat qui les distribue a ses6tablissements. Une liste de 221 m6dicamentsessentiels a et6 etablie dans les ann6es 80 pourle secteur public. Cette liste n'est pas impos6e ausecteur priv6 des services de sant6 et lesdemandes d'homologation emanent principale-ment de celui-ci, car les m6dicaments r6pondantaux besoins du secteur public sont obtenusaupres d'un petit nombre de fournisseurs 6tran-gers.

La LME de Sri Lanka a et6 compar6e a laliste des medicaments effectivement homologu6s;les diff6rents dosages d'une meme pr6parationpharmaceutique sont comptes s6parement. Onarrive ainsi a 3405 homologations pour 773 medi-caments; 697 produits sont en vente libre dont46% figurent sur la LME; cette proportion s'6levea 74% si l'on tient compte des produits de substi-tution pr6vus sur la LME. Parmi les 2708 produitsdisponibles uniquement sur ordonnance, 49% sontinscrits sur la LME (71% si l'on tient compte desproduits de substitution). Les dix m6dicaments lesplus courants dans les deux categories figurentsur la LME, a une exception pres. Soixante-quin-ze pour cent des produits homologu6s provien-nent de laboratoires pharmaceutiques situ6s dansdix pays. Parmi les fournisseurs de produitshomologues on trouve les dix soci6t6s pharma-ceutiques les plus importantes du monde par lechiffre d'affaires. II n'existe pas de donn6es ana-logues pour d'autres pays et il semble que cettecomparaison entre les produits homologu6s dansun pays et la liste nationale des m6dicamentsessentiels soit la premiere du genre.

La position dominante des m6dicaments de laLME parmi les produits homologu6s n'est proba-blement pas due a la politique officielle, car la

LME n'est pas impos6e au secteur priv6. Sil'eventail des produits disponibles semble bienadapt6 aux besoins de sant6 du pays, c'est vrai-semblablement en raison des efforts importants etsoutenus que l'Etat a d6ploy6s dans le domainedes soins de sante et de I'approvisionnement enproduits pharmaceutiques.

References1. WHO Technical Report Series No. 615, 1977. (The

selection of essential drugs: report of a WHO ExpertCommittee).

2. Reich, M.R. Essential drugs: economics and politicsin international health. Health policy, 8: 39-57(1987).

3. Lasagna, L. Setting the scene-the role of regula-tion. In: Walker, S.R. & Griffin, J.P., ed. Internationalmedicines regulation, a forward look to 1992.Lancaster, Kluwer Academic Publishers, 1989,pp. 19-26.

4. The Finnish Committee on Drug Information andStatistics. Finnish statistics on medicines 1990.Helsinki, The National Agency for Welfare andHealth, 1990.

5. Bottiger, L. The Swedish medicines regulatorysystem-with some notes on the Scandinaviansituation. In: Walker, S.R. & Griffin, J.P., ed. Inter-national medicines regulation, a forward look to1992. Lancaster, Kluwer Academic Publishers,1989, pp. 101-110.

6. Norwegian Medicines Control Authority. Annualreport 1991. Oslo, Norwegian Medicines ControlAuthority, 1991.

7. Walker, S.R. & Griffin, J.P., ed. International medi-cines regulation, a forward look to 1992. Lancaster,Kluwer Academic Publishers, 1989.

8. Hartog, R. & Schulte-Sasse, H. German and Swissdrug supplies to the Third World: survey and eva-luation of pharmacological rationality. Amsterdam,HAI Europe, 1989.

9. Dukes, M.N.G. An international comparison of medi-cines regulation. In: Walker, S.R. & Griffin, J.P., ed.International medicines regulation, a forward look to1992. Lancaster, Kluwer Academic Publishers,1989, pp. 27-34.

10. An evaluation of WHO's Action Programme onEssential Drugs, London School of Hygiene andTropical Medicine, London, England, and Royal Tro-pical Institute, Amsterdam, Netherlands, December1989.

11. Thamlikitkul, V. Antibiotic consumption throughdrug stores in Thailand: extent, contributing factorsand impact on public health. Alliance on prudentuse of antibiotics newsletter, 8(3): 1, 6-8 (1990).

12. Das, S.K. An appeal. Drug disease doctor, 5: 34(1992).

13. Annual Health Bulletin, Sri Lanka, 1989. Colombo,Ministry of Health, 1990.

14. Wickremasinghe, S.A. & Bibile, S. The manage-ment of pharmaceuticals in Ceylon. Colombo, Indus-trial Development Board of Ceylon, 1971. Summary

234 WHO Bulletin OMS. Vol 71 1993

Essential drugs and pharmaceutical registration In Sri Lanka

in British medical journal, 3: 757-758 (1971).15. Lail, S. & Bibile, S. The political economy of

controlling transnational: the pharmaceutical industryin Sri Lanka (1972-76). World development, 5:677-698 (1977).

16. List of Essential Drugs, Ministry of Health, Colombo,Sri Lanka. 1985, revised 1988.

17. Oydvin, K. & Alhava, E. Guidelines for ATC classi-fication. Oslo, Nordic Council on Medicines, 1990.

18. The 10 largest pharmaceutical companies world-wide. Scrip-World pharmaceutical news, No. 1570:13 (1990).

19. Drugs acting on the central nervous system. In:British national formulary, No. 22. London, BritishMedical Association and Royal PharmaceuticalSociety of Great Britain, 1991, pp. 125-183.

20. Col, N.F. & O'Connor, R.W. Estimating worldwidecurrent antibiotic usage: report of Task Force 1.Review of infectious diseases, 9 (suppl. 3):S233-S243 (1987).

21. Tomson, G. & Weerasuriya, K. 'Codes' and prac-tice: information in drug advertisements-an exam-ple from Sri Lanka. Social science & medicine, 30:737-741 (1990).

22. Wolffers, I. Drug information and sale practices insome pharmacies of Colombo, Sri Lanka: a researchnote. Social science & medicine, 25: 319-321(1987).

23. Aloysius, D.J. & Lionel, N.D.W. Are "EssentialDrugs" in the WHO list sufficient in general prac-tice? Ceylon medical journal, 27: 116-120 (1982).

24. World Health Organization. The rational use ofdrugs in the management of acute diarrhoea in chil-dren. Geneva, 1990.

25. Lall, S. 'The pharmaceutical industry in India: theeconomic costs of regulation', paper presented at theEleventh IFPMA Assembly, Washington, DC, June1982; quoted by Peretz, S.M. Pharmaceuticals in theThird World: the problems from the suppliers' pointof view. World development, 11: 259-264 (1983).

WHO Bulletin OMS. Vol 71 1993 235