University of Groningen

Improving pharmacovigilance and the role of the pharmacistGrootheest, Adrianus Cornelis van

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2003

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Grootheest, A. C. V. (2003). Improving pharmacovigilance and the role of the pharmacist. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 16-07-2021

2Pharmacovigilance

Chapter 2.1 The dawn of pharmacovigilance- Pharmacovigilance in historical perspective

Drug Safety (accepted)

Kees van GrootheestNetherlands Pharmacovigilance Centre Lareb

31

Abstract

In this chapter an anecdotal view from a Dutch perspective is given on somehistorical aspects of both the awareness and the systematic attention regardingadverse drug reactions, in a period before the word pharmacovigilance even existed.What we can learn form history is illustrated by the study of Lely, who discoveredthat 19 people died from digitalis intoxication, due to a production error of thedigitalis tablets. His observations of the symptoms of digitalis intoxication were inthe 18th century already described by Withering in his study on the foxglove. The Professor of Medicines Van Doeveren is another early example of awareness ofadverse drug effects, when he gave the academic lecture called “Remedio Morbi” in 1789 on diseases caused by treatment. This academic lecture was given attentionby Meyler, who wrote his famous book on adverse drug reaction in 1951. His book“Side effects of drugs” is currently in its 14th edition and is still a standard work inthis field. An early example of systemic collection of adverse drug reactions occurred in thebeginning of the 19th century, by the reporting of these during a smallpoxvaccination campaign in the Netherlands. Systematic attention to adverse drug reactions, including the collection of ADRreports, was boosted by the so called thalidomide disaster in the early sixties of thelast century. Since then, governments started to organize national pharmaco-vigilance centres to prevent such a disaster to happen again. The WHO has played an important role in developing this reporting of adverse drugreactions with its WHO International Drug Monitoring Program and the WHOUppsala Monitoring Centre, maintaining the international database of adverse drugreactions.A comprehensive review of the history of pharmacovigilance is still to be written

32

2.1.1 IntroductionSystematic attention to adverse drug reactions started with the thalidomide disaster inthe early sixties of the twentieth century. Of course, also before that period there wasawareness of possible negative aspects of drug therapy. In this chapter we will look,quite anecdotal and mainly from a Dutch perspective, to some historical aspects ofboth the awareness regarding adverse drug events and the systematic attention thathas been given to it in a period when the word pharmacovigilance did not even exist.

2.1.2 Can we learn from the past?In March 1971 A.H. Lely, MD and specialist in internal medicine, received hisdoctorate from Prof. dr. L. Meyler at the University of Groningen for his thesisentitled: ‘Digitalisintoxicatie, waarnemingen betreffende een massale digitoxine-intoxicatie te Veenendaal’ (Digitalis intoxication, observations on a mass digitoxinintoxication in the Dutch town of Veenendaal).(1) Early February 1969 Lely hasdetected symptoms of digitalis intoxication in a number of his patients. Hissubsequent investigations revealed that since December the sixth of the previousyear the local pharmacist had dispensed tablets that were not composed of therecommended 0.25 mg of digoxin but, due to a production error, contained acombination of 0.20 mg of the more potent digitoxin and 0.05 mg of digoxininstead, which magnified the effect of the heart glycoside, making it 2.2-4.2 morestronger than indicated. Lely concluded that this error was implicated in at least 19deaths. Probably this disaster may have caused far more fatalities that were notidentified, despite the fact that nationwide five times as many tablets with thewrong compound and dose had been dispensed than had been the case inVeenendaal and its immediate surroundings.Several aspects of this tragedy merit particular attention. First, there is the fact that untilshort before the pharmacist involved had produced the tablets himself at the pharmacy.The serious production error happened to coincide with his decision to start purchasingand dispensing industrially produced drugs, causing far greater damage than wouldhave been the case if the error was made with a local produced tablet.A second important aspect is that an observant physician had been able to detect theerror based on the symptoms his patients were exhibiting and informed others ofhis suspicions, thus preventing any further harm. It is very likely that many of hiscolleagues were also confronted with the signs of digitalis intoxication. The reasonwhy they failed to recognise the adverse reactions may be that they were not awareof the main concomitant symptoms, namely serious fatigue (93%), disturbed vision(85%), anorexia nervosa (84%), psychological complaints (64%) and abdominalpains (57%).(1) It is also feasible that they may have realised that there were adversereactions but did not inform others of their observations. These elements are stillkey aspects of pharmacovigilance.

33

Within the context of this chapter another aspect of Lely’s study needs to bementioned. Lely points out that the botanist William Withering from Birmingham,England, had already extensively described the symptoms of a digitalis intoxicationas early as in 1785.(2,3) In his book ‘An Account of the Foxglove and Some of ItsMedical Uses: with Practical Remarks on Dropsy and Other Diseases’ he describesthe symptoms as follows: ‘The Foxglove, when given in very large and quicklyrepeated doses, occasions sickness, vomiting, purging, giddiness, confused vision,objects appearing green or yellow; increased secretion of urine, with frequentmotions to part with it, and sometime inability to retain it; slow pulse, even as slowas 35 in a minute, cold sweats, convulsions, syncope and death’.Although the knowledge about the pharmacological properties of digitalis, thediagnostic possibilities and the therapeutic approaches for the treatment of anoverdose have increased significantly in the two centuries following thispublication, the characteristic symptomatology has not changed and has alreadybeen described 200 years before. Knowledge, a keen sense of observation and athorough anamnesis and diagnosis have not only formed, and still form, the basisof medicine, but also of pharmacovigilance, the science that is concerned with thesafe use of drugs in daily practice.Against the background of this case concerning digitalis, some historical aspects ofthe awareness regarding adverse drug reactions and the systematic collection of itwill be discussed.

2.1.3 Adverse drug reactions in historical perspectiveIn 1789, Wouter van Doeveren, Professor of Medicine at Leiden University, gave apublic lecture with the title ‘Remedio morbo, drug diseases or ailments which oftenaffect people as a result of remedies administered to them for therapeuticpurposes’.(4,5,6) This address, which did not appear in print, but the manuscript ofwhich is kept in the University Library of Amsterdam, may be seen as the firstexpression of an emerging scientific interest in adverse drug reactions.(7) Wouter van Doeveren, born in Philippine, a village in the Dutch province of Zeeland,started his study of medicine in Leiden at the age of seventeen. He received a First (anhonours degree cum omnium applausu) for his thesis on worms ‘De vermibusintestinalibus hominum, praecipue de taenia’. This dissertation would long beregarded as a standard work on worms and was translated into German and French.In 1754 he was appointed to the chair of Medicine at Groningen. Here he engaged inmany activities, such as the promotion of the inoculation with cowpox. In 1770 hewas appointed professor of the faculty of medicine at Leiden where he succeededHerman Boerhaave, whose ideas were still the norm at the faculty. Van Doeveren wasa good clinician, an ‘evidence-based’ researcher avant la lettre, and a critic of themedical practices of that time, opposing many of the prevalent irrational therapies.

34

The systematic cataloguing of illnesses that may occur as a concomitant effect ofpharmacotherapy is not taken up until much later. In 1881 a book is published byL. Lewin: ‘Die Nebenwirkungen der Arzneimittel: pharmakologisch – klinischHandbuch’.(9) The key work providing an overview of the existing literature onadverse drug reactions is written by L. Meyler in Dutch under the title ‘Schadelijkenevenwerkingen van geneesmiddelen’ and is published in 1951. It is directlyfollowed by an English translation: ‘Side effects of drugs’, published in 1952. It wassubsequently translated into many other languages.(10,11) New editions appearregularly under the title ‘Meyler’s Side Effects of Drugs’; in 2002 the 14th editionwas released. Meyler’s work, published ten years prior to the thalidomide tragedy,is soon followed by other publications. In 1953 ‘Maladies médicamenteuses’ by C.Albahary is released, followed in 1954 by G. Duchesnay’s ‘Le risque thérapeutique,prévention et traitement des accidents’.(12,13) In 1955 both the English publication‘Reactions with Drug Therapy’ by H.L. Alexander and the German overview‘Erkrankungen durch Arzneimittel, Diagnostik, Klinik, Pathogenese, Therapie’,edited by R. Heintz, appear.(14,15)In 1961 the thalidomide tragedy causes a surge in the publications on adverse drugeffects, famously led by McBride’s article.(16,17)

2.1.4 Pharmacovigilance in historical perspectivePharmacovigilance concerns the science and activities relating to the detection,assessment, understanding and prevention of adverse drug effects or any other possibledrug-related problems, according to a recent WHO publication.(18) Until thethalidomide tragedy in 1961, when the devastating side effects of the drug came tolight, the systematic collection and analysis of adverse drug events had not receivedmuch attention, although there are several examples of a systematic attention for the(negative) effects of drugs that go further back.Already in the beginning of the 19th century the systematic reporting of adverseeffects, in this case regarding vaccines, is noted and will be described here.Above, Van Doeveren’s effort in Groningen to promote the cowpox vaccination ismentioned. He was one of the protagonists in a long-drawn-out debate, initiated byLady Montagu, relating to the religious, medical-scientific and financial implicationsof such an inoculation. During her travels in Turkey in 1718 Lady Montagu had her 6-year-old son vaccinated and, back in England in 1721, also had her 6-year-old daughterimmunised in the presence of three members of the Royal College of Physicians toprove the safety of inoculation. In the Netherlands, the first time the vaccination wasadministered was in Amsterdam in 1748, where Boerhaave’s student TheodoreTronchin, immunised his eldest son after his second son had barely survived a seriousattack of the smallpox.(19) Van Doeveren administers the vaccination for the first timein Groningen in 1759. And although Boerhaave had already claimed that the

36

Especially during his time at Groningen University Van Doeveren produced a largebody of research in the fields of anatomy, physiology and obstetrics. Whenstepping down as rector magnificus of Leiden University in 1779 he makes hisfamous Sermo academicus de remedio morbo (‘Academic lecture on diseases as aresult of treatment’).(7) In this address he observes that physicians are seldommoderate in their judgments of treatment methods and are rather inclined to regardthem as either essential or hazardous. He points out the hazards of the artificialinducement of fevers and argued against bloodletting and the administration ofperspiration-inducing drugs for acute fevers. He concludes his speech with theobservation that many illnesses are the result of ‘empirical treatments’ administeredwithout a proper diagnosis, thus causing remedio morbi, illnesses that result froma treatment. He warned his audience: ‘Lest you will not be too easily persuaded tohand out medications, which may carry the risk of your adding a second ailment tothe first, or through which you may perhaps even invite death or accelerate it’!

Earlier in this chapter William Withering (1741-1799) and his 1785 work on thefoxglove is mentioned. His work has become known as the first systematic paperon a medicinal drug and gives attention to the adverse effects of the treatmentwith digitalis. The Royal College of Physicians of London commemorate hisachievements with their annual William Withering lecture. In 1994 Michael D.Rawlins was invited to deliver that year’s lecture.(3) His address that has sincebecome a standard reference in pharmacovigilance, was entitled ‘Pharmaco-vigilance: paradise lost, regained or postponed?’. He stated that in our moderntimes Withering’s work could easily serve as an Expert Report when applying formarketing approval. Withering, for example, explains the synthesis of digitalisfrom the foxglove in a standardised way, describes animal tests conducted onturkeys, details the therapeutic effects and the symptoms associated with anoverdose, emphasising the importance of a proper measurement of the dose.(2)

Without question, ahead of Van Doeveren and Withering others will have pointedto the possible harmful effects of medications although this cannot be corroboratedby specific evidence. The basic principle that therapeutic gain needs to outweighany possible damage was already recorded by Hippocrates: primum nil nocere, firstof all do no harm. There is historical evidence available showing that authorities supervised themanufacture of drugs. These activities ultimately resulted in the Pharmacopee. Oneof the aims of founding the FDA in 1906 also was to impose quality criteria for themanufacture of drugs. More stringent criteria were drawn up in 1938 following 107fatalities caused by the use of ethylene glycol as a solvent.(8)

35

prophylactic effect of the immunisation was considered to be reliable and safe, also inGroningen a heated debate took place, culminating in a campaign against VanDoeveren. In Leiden the vaccination was prohibited and Van Doeveren undertooklittle to counteract this decision when he became a professor over there. Nevertheless, his efforts to get the cowpox immunisation accepted in Groningenapparently were successful, possibly aided by the hugely positive response toEdward Jenner’s work, who, in 1798, had published his ‘An enquiry into the causesand effects of the variolae vaccinae’, followed a year later by ‘Further observationson variolae vaccinae’. And once again, the Royal College of Physicians played aprominent role and thanks to their involvement and perhaps also partly due to theFrench occupation, a vigorous campaign to boost the use of the cowpox vaccinationis launched in the Netherlands. A directive, dated 25 August 1808, stipulates anumber of measures to promote the vaccination, as is illustrated by a placard the‘Landdrost’ (local administrator) of the province of Groningen has distributed.(20)The three physicians who administered the largest amount of vaccinations, with aminimum of 100, are to be awarded a gold coin to the value of ten ducats. To be eligiblefor the remuneration, they have to submit a record listing the personal data of theirpatients every three months, but also a statement as to ‘whether the injected cowpoxhave taken and have had a regular effect, also stipulating any peculiarities that mayhave occurred’. Thus, in fact, the spin-off of the debate on the cowpox vaccination wasthat its side effects were monitored systematically! What also strikes is the fact that theregulation also mentions that ‘the beneficial effects of the inoculation […] have notonly been proven in this Realm, but have also been conclusively demonstrated in otherNations’. Apparently, evidence was available and exchanged internationally.

In his article on 150 years of pharmacovigilance Philip Routledge refers to a committeethat was set up by The Lancet to collect reports on mortalities resulting fromanaesthesia in Britain and its colonies. The results of which have been published in TheLancet.(21,22) The committee was a reaction to the death of a 15-year-old girl in 1848who had undergone chloroform anaesthesia in connection with an ingrown toenail! In 1950 reports of incidences of aplastic anaemia associated with the use ofchloramphenicol were received in the US.(23) This led the Council on Drugs of theAmerican Medical Association to set up a Blood Dyscrasia Registry.(24)

2.1.5 Drug safety since 1960In 1961 the Australian physician McBride published a letter in The Lancet in which hesuggests a connection between congenital malformations in newly born infants and thehypnotic thalidomide, which was marketed under various names in many countries. Itwas available both as a prescription and as an over-the-counter (OTC) drug and alsoin many compositions of simple analgesics.(16) Thalidomide was first synthesized in

37

1953 and in 1956 an intensive campaign was launched to promote the drug (undervarious names such as Contergan®, Distavel®, Softenon®) first in Germany, then inEngland (1958) and subsequently in other countries. Although McBride had alreadysubmitted his letter in June 1961, the publication appeared in December 1961 afterresubmission. Earlier, on November the 19th 1961, at a conference on paediatrics theGerman physician Dr. Lenz from Hamburg argued that there was a correlationbetween the many births of children with congenital deformities and the use ofthalidomide, which relationship had incidentally already attracted some attentionbefore that time. On 25 November 1961 the manufacturer Chemie Grünenthalwithdrew thalidomide from the market. The number of children born with seriouscongenital malformations as a result of maternal use of thalidomide is estimatedbetween 6,000 and 12,000, the majority of which were born in Germany.(25,26)Noteworthy in this context is that today toxicology and the surveillance of congenitalabnormalities frequently fall outside the scope of the national pharmacovigilancecentres, and have become the designated field of separate specialised organisations.McBride’s appeal ‘Have any of your readers seen similar abnormalities…?’ by whichhe concluded his letter, had much more impact than he ever could have foreseen.(16)

In 1961 the FDA started with the systematic collection of reports on all types ofadverse drug reactions, chiefly through the Hospital Reporting Program. In variouscountries the thalidomide tragedy had prompted the immediate formulation ofcriteria new drugs needed to meet to receive marketing authorisation, whichstandards, apart from quality, also emphasised the safety and efficacy of new drugs.In addition, the governmental marketing authorisation organisations werecommissioned to establish a post-marketing surveillance system to facilitate theearly detection of adverse reactions in order to prevent a similar tragedy fromoccurring in the future. In 1968 ten countries that supported a spontaneousreporting system for adverse drug reactions took the decision to collaborate andjoined the WHO Pilot Research Project for International Drug Monitoring.(27) In1971 a resolution of the Twentieth World Health Assembly laid the foundations forthe WHO international Drug Monitoring Programme.(28) In 1972 a report waspublished which formed the basis of the current international system of nationalcentres collaborating in the WHO Programme.(29,30) The WHO International Drug Monitoring Programme is supported andcoordinated by the WHO Collaborating Centre for International Drug Monitoring(the Uppsala Monitoring Centre), which maintains and implements theinternational database of adverse drug events. Its activities are of major importanceand include providing active support to the pharmacovigilance centres in lowincome countries, and the evaluation and augmentation of pharmacovigilance ingeneral. In 2000 the ‘Guidelines for setting up and running a Pharmacovigilance

38

Centre’ appeared and in 2002 the ‘The Importance of Pharmacovigilance’ ispublished under the auspices of the WHO, which lays the foundation for thefurther theoretical and practical implementation of pharmacovigilance on aninternational scale.(18,31)Both the national authorities and the pharmaceutical industry have played asignificant role in the development of pharmacovigilance, which has resulted in newlegislation and qualitative requirements the industry and their products need tofulfil. The Council for International Organizations of Medical Sciences (CIOMS)and the International Conference on Harmonization (ICH) have been instrumentalis this process.

The literature, both in the form of scientific handbooks and journal articles, bearswitness to the evolution of pharmacovigilance since the 1960s. A relatively smallgroup of scientists have been responsible for providing pharmacovigilance with afirm scientific basis. In his thesis Meyboom mentions the names of Leopold Meyler,Owen Wade, David Finney and William Inman.(32)What strikes one, when studying the literature on the safety of drugs, is that manyof the topics that have recently been presented as novel have been reported onearlier. Apparently, and possibly nowadays more due to the Internet, the scientificmemory is short. As an illustration: the most frequently cited article in the contextof safety of drug is the meta-analysis by Lazarou et al., in which the authorsconclude that 6.7% of all hospitalized patients experience adverse reactions todrugs.(33) However, already in 1955 Barr published an article in the JAMA thatalso reports that 5% of admitted patients are afflicted by ‘major toxic reactions andaccidents’.(34) Similar percentages are given by Seidl et al. in 1966 in relation tohospital admissions and in 1969 by Hurwitz and Wade for drug-relatedcomplaints.(35,36) Time will tell which contributions have genuinely improved thesafety of drugs and its systematic surveillance.

A comprehensive history of pharmacovigilance is still to be written. The introductorychapters of the works by Davies and Wade provide some information.(37,25)Routledge had a concise overview published in The Lancet and Lindquist has recentlywritten an historical overview of the WHO Programme.(21,27) Several of thesepublications offer information about the progress made in the variouscountries.(38,39,40) The development of the pharmacovigilance system in theNetherlands and the role of the Netherlands Pharmacovigilance Centre Lareb hasbeen described by De Koning, Meyboom and Van Grootheest.(32,41,42).Pharmacovigilance is continuously evolving. Waller and Evans have produced ‘Amodel for the future conduct of Pharmacovigilance’. The key concepts for the nearfuture seem to be ‘proactive safety surveillance’ and ‘risk management’.(43-46)

39

References

1. Lely AH. Digitalisintoxicatie, waarnemingen betreffende een massale digitoxine-intoxicatie teVeenendaal. Thesis. State University of Groningen. Stafleu. Leiden 1971.

2. Withering W. An account of the foxglove and some of its medical uses. CGJ and J Robinson.London 1785.

3. Rawlins MD. Pharmacovigilance: paradise lost, regained or postponed. The William WitheringLecture 1994. J Royal Coll Phys London 1995;29:41-9.

4. Doeveren W van. Sermo academicus de remedio morbo, sive de malis, quae hominibus aremedies, sanandi causa adhibitis, saepenumero accidere solent. Academic speech of February8, 1779. Manuscript in University Library Amsterdam.

5. Meyler L. Woord vooraf. In: Meyler L, Schadelijke nevenwerkingen van geneesmiddelen,Supplement II. Van Gorcum. Assen 1958.

6. Meyler L. Introduction drug-induced diseases. In: Drug-induced diseases. L. Meyler, MHPeck, editors. Van Gorcum. Assen 1962.

7. Zwaag P van der. Wouter van Doeveren, leven en werken van een 18-eeuws hoogleraar in degeneeskunde. Van Gorcum’s Historische Bibliotheek nr. 86. Van Gorcum. Assen 1970.

8. Geiling EMK. Cannon PR. pathogenic effects of elixir of sulfanilamide (diethylene glycol)poisoning. JAMA 1938;111:919.

9. See: Anonymus. Idiosynkrasie. Pharm Weekbl 1902;39:629-30.

10. Meyler L. Schadelijke nevenwerkingen van geneesmiddelen. Elsevier Publishing Company.Amsterdam 1951.

11. Meyler L. Side Effects of Drugs. Excerpta Medica Foundation. Amsterdam 1952.

12. Albahary C. Maladies médicamenteuses d’ordre thérapeutique et accidental. Masson & Cie.Paris 1953.

13. Duchesnay G. Le risque thérapeutique, prévention et traitement des accidents. G. Doin & Cie.Paris 1954.

14. Alexander HL. Reactions with drug therapy. Saunders Company. Philadelphia 1955.

15. R. Heintz (editor). Erkrankungen durch Arzneimittel, Diagnostiek, Kliniek, Pathogenese,Therapie. Georg Thieme Verlag. Stuttgart 1966.

16. McBride WG. Thalidomide and congenital abnormalities. Lancet 1961;11:1358.

40

17. Sjöstrom Hm, Nilsson R. Thalidomide and the Power of the Drug Companies. PenguinBooks. Harmondsworth, Middx.1972.

18. WHO. The Importance of Pharmacovigilance, Safety Monitoring of medical products. WHOGeneva 2002.

19. Lindeboom GA. Theodore Tronchin (1709-1781). Bijdragen tot de geschiedenis dergeneeskunde XXXVI 1956:49.

20. Placard, Groningen 1808. (collection author).

21. Routledge P. 150 years of pharmacovigilance. Lancet 1998;351:1200-1.

22. Commission on Anaesthetics. Lancet 1893;i:629-38.

23. Rich ML. Fatal case of aplastic anaemia following chloramphencol therapy. Ann Intern Med1950;33:1459.

24. Wallerstein RO, Condit PK, Kasper CK, Brown JW, Morrison FR. Statewide study ofchloramphenicol therapy and fatal aplastic anemia. JAMA 1969;208:2045.

25. Wade OL. The dawn of concern. In: Adverse reactions to drugs: 1-10. Acford Ltd. Chichester1970.

26. Randell T. Thalidomide’s back in the news, but in more favorable circumstances. JAMA1990;263:467-8.

27. Lindquist A.M. Seeing and Observing in International Pharmacovigilance. Academic thesis.Katholieke Universiteit Nijmegen. Nijmegen 2003.

28. WHO. Handbook of resolutions and decisions of the World Health Assembly and ExecutiveBoard, 11th ed., 1971. Geneva, World Health Organization, 1972. WHA 20.51.

29. WHO. International Drug Monitoring: The Role of National Centres. Technical ReportSeries. Geneva 1972.

30. Olsson S. The Role of the WHO Programme on International Drug Monitoring inCoordinating Worldwide Drug Safety Efforts. Drug Saf 1998;19:1-10.

31. UMC. Guidelines for setting up and running a Pharmacovigilance Centre. UMC. Uppsala 2000.

32. Meyboom RHB. Detecting adverse drug reactions, Pharmacovigilance in The Netherlands.Thesis. Katholieke Universiteit Nijmegen 1998.

33. Lazarou J. Pomeranz BH, Cory PN. Incidence of adverse drug reactions in hospitalizedpatients. JAMA 1998;279:1200-5.

34. Barr DP. Hazard of modern diagnosis and therapies – the price we pay. Frank BillingsMemorial Lecture. JAMA 1955;159:1452.

41

35. Seidl LG, Thornton GF, Smith JW, Cluff LE. Studies on the epidemiology of adverse drugreactions. III. Reactions in patients on a general medical service. Bull Johns Hopk Hosp1966;119:299.

36. Hurwitz NA, Wade OL. The incidence of adverse drug reactions in relation to the use of drugsin hospitals. BMJ 1970;1:531.

37. Davis DM, editor. Textbook of Adverse Drug Ractions. 4e ed. Oxford University Press.Oxford 1991.

38. Mann R, Andrews E, editors. Pharmacovigilance. Wiley. Chichester 2002.

39. Strom BL, editor. Pharmacoepidemiology, 3e editon. Wiley. Chichester 2000.

40. Inman B. 30 Years in Postmarketing Surveillance. A Personal Perpective. PharmacoepidemiolDrug Saf 1993;2:239-58.

41. Koning GPH de. A regionalized spontaneous surveillance programme for adverse drugreactions as a tool to improve pharmacotherapy. Thesis. Utrecht University 1994.

42. Grootheest AC van, Puijenbroek EP van. Pharmacovigilance in the Netherlands. In:Pharmacovigilance: 309-15. Mann R, Andrews E, editors. Wiley. Chichester 2002.

43. Waller PC, Evans SJW. A model for the future conduct of pharmacovigilance.Pharmacoepidemiol Drug Saf 2003;12:17-29.

44. Bortnichak EA, Wise RP, Salive ME, Tilson HH. Proactive safety surveillance.Pharmacoepidemiol Drug Saf 2001;10:191-6.

45. Perfetto EM, Ellison R, Ackermann S, Sherr M, Zaugg AM. Evidence-Based RiskManagement: How Can we Succeed?: Deliberations from a Risk Management AdviseryCouncil. Drug Inf J 2003;37:127-34.

46. Head of Agencies. Establishing a European Risk Management Strategy. Report of the Headsof Agencies ad hoc Working Group. 2003. (http://heads.medagencies.org/heads/docs).

42

Chapter 2.2 Leopold Meyler (1903-1973): A pioneer in the study of adverse effects of drugs

Ned Tijdschr Geneeskd (accepted)Int J Risk Saf Med (accepted)

Kees van GrootheestGraham Dukes

43

2.2.1 IntroductionIn 1951, more than a decade before the first rumblings of the thalidomidecatastrophe, a remarkable book appeared in the Netherlands under the authorshipof Leopold Meyler, that a year later appeared in English translation as ‘Side effectof drugs’. At that moment the book was unique in the world - and that is still thecase. The concept underlying it was that physician and pharmacist alike should beprovided with an overview of all that was known regarding the possible adverseeffects of medicines. The need to make existing information widely availableremains, even today, a major challenge in the world of pharmacovigilance.(1) Following Meyler’s book others appeared in various languages, several adoptingdifferent approaches to the topic. One after another most of them have vanished;Davis’ Textbook of Adverse Reactions is a creditable exception, providing anapproach to the topic which complements that of the Meyler series. In themeanwhile, Meyler’s volume, greatly expanded and updated, is still with us and haslong been acknowledged as the standard work on the unwanted effects of medicines.The name of the original author has become incorporated into the title: and‘Meyler’s Side Effects of Drugs’ is commonly known to the user simply as ‘Meyler’.

2.2.2 The thalidomide disasterA little over forty years ago, medical interest in drugs - and particularly new drugs- was limited almost entirely to their proven or possible therapeutic benefits. Inretrospect it is fair to say that it was the paper published by McBride in the Lancetof December 16th 1961 which triggered a broadening of that view.(2) His paper,alongside simultaneous observations from Germany, provided the principal cluesleading to the realization that there was a link between the use of the hypnoticthalidomide and the fact that a great many children were coming into the worldwith congenital defects.(3) It had of course been realized for a long time that medicines could exert adverseeffects. As early as 1779, Prof. Wouter van Doeveren of the University of Leiden,The Netherlands, delivered a public oration entitled ‘Remedio morbi’, with assubtitle: ‘Drug diseases, or diseases persons often obtain as a result of medicinesthey received as a treatment’. His conclusion, recorded more than two centuries agowas: ‘Do not give a drug too readily, with the risk that you may add a seconddisease to the first, or accelerate the patient’s’ decease’.(4)

2.2.3 AwarenessNevertheless, and as late as the mid twentieth century, physicians continued to beprimarily concerned with the positive rather than the negative aspects of medicinaltreatment, a unilateral interest promoted ever more with the arrival of new drugswhich opened therapeutic perspectives hitherto out of reach, such as the effective

45

treatment of tuberculosis. Interest in the side effects which might accompany theuse of these new agents remained sporadic.(5) Only with the thalidomide disaster were the medical profession and the publicthroughout the world alerted acutely to the need both to examine the safety of newmedicines critically prior to their admission to the market, and to studysystematically their emergent adverse effects once they had come into use. The UK‘yellow card’ system for the reporting of suspects adverse effects came into being in1964, and in 1968 W.H.O., set up a pilot project for international collaboration inthe monitoring van adverse drug reactions. The latter was to become what is nowthe W.H.O. Drug Monitoring Programme, which in October 2002 held its 25thAnnual Meeting in Amsterdam.

2.2.4 Leopold MeylerWho was Leopold Meyler, and what inspired him to create this book at thebeginning of the fifties? Meyler was born just a century ago in Rotterdam, wherehis father was an oil merchant of Jewish extraction. The young Meyler studiedmedicine at the University of Leiden and went on to specialize in internal medicine.After his qualification he joined the Medical Clinic of the University of Groningenand set up in practice. The ensuing war and the Nazi occupation of TheNetherlands were a difficult period for the family. Meyler himself had to go intohiding but his wife, whose mother was not Jewish, was periodically able to visithim. At the time of the liberation, however, Meyler was found to be suffering frompulmonary tuberculosis which he had acquired already the occupation but whichmay well have been aggravated by the conditions under which he had been obligedto live. He was admitted to a sanatorium and prescribed a prolonged period of rest.However, enforced rest and inactivity were not in the nature of an active andcommitted physician such as Meyler. Then it was that the specialist treating himsuggested a productive task which he could undertake from his bed: let himundertake a literature study of the side effects of drugs.

2.2.5 His illnessThere was a particular reason for the choice of topic, for Meyler himself had alreadyencountered the problem of adverse effects. To quote the first Dutch edition of hisbook in 1951: ‘The notion of bringing together the disagreeable effects which can beassociated with the use of medicines was inspired by personal experience, in part of avery serious nature’. In his book he describes in detail the experiences of patients suchas himself treated with para-aminosaliocylic acid (PAS), including the occurrence offever in individuals with an allergic constitution. It is known that Meyler himself wasallergic and suffered from asthma (6). In a later Supplement to his book he describedthe serious psychic effects which can be exerted by isonicotinic acid hydrazide (INH).

46

2.2.6 The writing of his bookIn order to examine the literature from his hospital bed, Meyler enlisted the help ofthe library of the University of Groningen. Every week a suitcase full of journalsand books was delivered to him, which he was obliged to return within a few days.The arrangement was continued after his return home. And so it was that the firstversion of his book appeared in 1951 in Dutch, followed by the English version ayear later. Not surprisingly the unusual approach manifested in the book elicited criticism insome quarters, and his motives were questioned. There were those who imaginedthat he was in principle opposed to the use of medicines, and some alternativehealers even viewed him as an ally in questioning the orthodox practice of medicine.Others criticised the manner in which he had classified drugs. Meyler howevercontinued assiduously to gather his evidence and relevant case reports. Supplementswere published to bring the original volume up to date, and his introductoryeditorials provided an answer to his critics, stressing that he sought only to promotethe appropriate and safe use of drugs. Behind the scenes, his books were founded inan encyclopaedic venture into the acquisition of hitherto scattered knowledge,carried through in a period where there were no computers or photocopyingmachines. Instead, he relied on his hand-written notes and a voluminous cardsystem.

2.2.7 Meyler as a personMeyler was a remarkable individual, characterized by firm opinions, a criticalapproach to information and an independence of spirit. He was also a man with abroad interest and an open mind. Above all he remained a highly competentpractising physician with a deep devotion to the interests of his patients. Hisindependence was such that he would not attend meetings sponsored by the drugindustry; if invited as a speaker he would insist on paying his own expenses.His work was formally recognized with his appointment in 1969 as the firstprofessor of Clinical Pharmacology at the University of Groningen. His inauguraloration was entitled ‘Why clinical pharmacology?’ and in it he provides anoverview of the tasks which the clinical pharmacologist is called upon to fulfil.(7)Not surprisingly his paper devotes attention to adverse reactions: ‘The study of sideeffects is of the greatest importance. Increasingly the need is felt to understand moreexactly how side effects come about, and to devise wherever possible means bywhich these unwanted complications can be avoided.’ The definition of a field ofmedical science which was new to The Netherlands was a task to which Meyler waswell suited. He published in many journals and became a sought-after speaker atmeetings both at home and abroad. The Boerhaave courses at the University ofLeyden, for which he was in part responsible, led to the publication of a series of

47

complementary volumes under the title Drug Induced Diseases, each dealing witha particular aspect of drug toxicology. Re-reading his publications today one isstruck by his extraordinary knowledge and erudition but also by the fact that somany of his views remain entirely valid in our time.

2.2.8 The importance of MeylerMeyler died suddenly during a vacation in France in 1973, possibly as aconsequence of an overdose of adrenaline which he had taken because of hispulmonary disorder. As Lammers wrote in an In Memoriam in the NederlandsTijdschrift voor Geneeskunde: ‘One can truly say that he was among the very firstto have realized how frequently drugs can be the cause of otherwise unexplaineddisorders. In our time, that realization has become so commonplace and so self-evident that one can hardly imagine how, in the recent past, the subject was barelythe subject of enquiry’.(8)Leopold Meylers opus magnum, Side Effects of Drugs, evolved from the firstedition onwards, as a multi-author volume with a distinguished team ofcontributors. Up to the seventh edition in 1972, Meyler himself was Editor-in-Chief, latterly supported by Prof. Andrew Herxheimer in London. From 1973 to2000 the volume was edited by Graham Dukes and the series is now to becontinued by Jeff Aronson in Oxford. Many tens of authors from all parts of theworld have made their contributions to ‘Meyler’, which continues to providepractising physicians and pharmacists with a critical review of new data from thescientific literature relating to adverse drug reactions. Yet there are today few whorealize that the name ‘Meyler’ on the cover is that of a doctor lying in a sanatorium,looking for something useful to occupy his mind. Meyler laid the foundations for asystematic approach to the problems of side effects. In his own time his was at firsta voice crying in the wilderness. Had that voice been heeded earlier, the extent ofthe thalidomide disaster might well have been much more limited. Even today, witha worldwide system in place for the detection and study of adverse drug reactions,an Editorial in The Lancet has rightly raised the question as to how that process canbe further refined and extended.(9) As Meyler himself wrote in the forward to hisfirst volume: ‘Let us be entirely clear that it is not our purpose to discourage the useof any of the drugs in our therapeutic arsenal. The reverse is the case. One will bein a position to use a medicine better if one is aware not only of its benefits, but alsoof its risks.’

48

References

1. Grootheest K van, Edwards R. Labelling and ‘Dear Doctor’ Letters: are they noncommittal?Drug Saf 2002;25:1051-5.

2. McBride WG. Thalidomide and congenital abnormalities. Lancet 1961;11:1358.

3. Sjöstrom HM, Nilsson R. Thalidomide and the Power of the Drug Companies. PenguinBooks, Harmondsworth, Middx. 1972.

4. Zwaag P van der. Wouter van Doeveren, leven en werken van een 18e-eeuws hoogleraar in degeneeskunde. Van Gorcum. Assen 1970.

5. Routledge P. 150 years of pharmacovigilance. Lancet 1998;351:1200-1.

6. Meyler L. Schadelijke nevenwerkingen van geneesmiddelen.: Elsevier. Amsterdam 1951.

7. Meyler L. Waarom klinische farmacologie?. Ned Tijdschr Geneeskd 1969;113:1275-9.

8. Lammers W. In Memoriam Prof. Dr. L. Meyler. Ned Tijdschr Geneeskd 1973;117:1522-3.

9. Editorial. Improving ADR reporting. Lancet 2002;360:1435.

49

2.3.1 History and organization

HistoryIn the Netherlands, consideration for the surveillance of adverse drug reactionsdeveloped at a relatively early stage. In the early 1950s, at a time when internationalliterature had included only incidental reports of 'side effects', Dr Leo Meyler laidthe basis for more systematic attention to adverse drug reactions. In 1951, hepublished his book (in Dutch) Schadelijke nevenwerkingen van geneesmiddelen(literally, 'Harmful effects of prescription drugs'). The second edition, fully revisedwith a number of supplements, appeared in 1954.In his preface to the first edition, Meyler wrote the following (here in translation):

‘The prescribing of drugs will always entail a greater or lesser degree of risk, and ineach case the physician must ask himself whether the nature of the condition aboutwhich he is being consulted justifies taking such a risk’. Meyler's work wasprompted by his own experiences with tuberculostatic preparations. He warnedalso against the inappropriate use of drugs.Meyler based much of his work on reports in various medical journals, at a timewithout the convenience of the Internet or other conveniences of modern times.The first English edition of Meyler’s seminal work was published as The SideEffects of Drugs: an Encyclopaedia of Reactions and Interactions in 1952. Itsfourteenth edition, edited by Graham Dukes, appeared in 2000. Dukes has been theeditor since the eighth edition, published in 1978, of what has now become thestandard reference work in its field. Dukes' own scientific background was largelygained in the Netherlands.

Organization of pharmacovigilance in the NetherlandsFollowing the thalidomide affair of the late 1950s and early 1960s, the Netherlandsdecided to adopt a more systematic approach to the safety of prescriptionmedicines. The Dutch Medicines Evaluation Board was founded in 1963. Based onthe American model of the Food and Drug Administration, this would assess newpharmaceutical preparations for both effectiveness and safety prior to marketingauthorization. Also in 1963, the Royal Dutch Medical Association (KNMG) joinedthe government in setting up a reporting system for adverse drug reactions. In 1965,the task of processing reports was taken over by the National Drug MonitoringCentre, which was part of the Public Health Supervisory Service and came toacquire an extremely good reputation. With a relatively small staff, the Centreproduced a significant number of publications calling attention to the potentialadverse effects of prescription drugs.(1) Each year, the National Drug MonitoringCentre received approximately thousand reports from interested doctors.

53

In 1986, a number of pharmacists called for greater attention to be devoted to thepotential adverse effects of prescription medicines. These pharmacists wereconvinced that greater awareness of the possibility of adverse effects wouldimprove the quality of pharmacotherapy as a whole. Their initiative led to thecreation of the Lareb Foundation (now: Netherlands Pharmacovigilance CentreLareb), in 1991. Lareb collected reports of adverse effects, mostly made by generalpractitioners and pharmacists. This information was supplementary to the workdone by the National Drug Monitoring Centre and at times created overlap.(2) Anew aspect was that pharmacists felt also their responsibility in the identification ofadverse effects and would consider it their task to call attention to such effects.(3)

In 1995, European legislation having been made more stringent, the Dutchgovernment decided to restructure the system of pharmacovigilance in theNetherlands. Lareb was designated the national centre for all reports of suspectedadverse drug reactions concerning registered drugs. Currently the Health Inspectorate is responsible for monitoring the quality ofpharmacovigilance activities and receives reports relating to preparations that havenot (yet) been given marketing authorization, especially reports of adversereactions that are observed during the statutory clinical trials.

The Medicines Evaluation Board Agency plays a central coordinating role. It receivesreports from Lareb, as well as those made directly by the pharmaceutical industry,and it advises the Medicines Evaluation Board. The Medicines Evaluation Boardmakes the final decision regarding marketing authorization for the Netherlands.Where deemed necessary, it is empowered to require amendments to a drug's'Summary of Product Characteristics', and in serious cases may revoke a drug'smarketing authorization altogether. The Medicines Evaluation Board includes apharmacovigilance department primarily concerned with adverse drug reactions andwith maintaining international contacts in this field. Many decisions are taken atEuropean level by the European Medicines Evaluation Agency (EMEA).

2.3.2 Spontaneous reporting in the Netherlands: the NetherlandsPharmacovigilance Centre Lareb

Direct responsibility of doctors and pharmacistsThe Netherlands Pharmacovigilance Centre Lareb is an organization which wasfounded by doctors and pharmacists and which is still run by doctors andpharmacists today. All large medical and pharmacists associations are representedon its administrative board. Lareb maintains the national 'spontaneous' reportingsystem for the Netherlands. That this task falls to an independent centre rather than

54

the government sets the Netherlands apart from most other countries. Althoughsome (such as Germany, Switzerland, New Zealand and Great Britain with its DrugSafety Research Unit) have organizations investigating adverse drug reactions thatare allied to universities or professional organizations, the role of professionalpractitioners is particularly prominent in the Netherlands. The governmentrestricts itself to a supervisory and coordinating role, while also providing fundingfor the Lareb's activities.The Dutch model has a significant number of advantages and works very well inpractice. It is doctors and pharmacists who see the occurrence of adverse drugreactions in day-to-day practice. If given co-responsibility for the propermonitoring of drug safety, they will be more inclined to contribute. This enhancesthe premise that doctors and pharmacists are themselves responsible for the safe andresponsible use of prescription drugs. The barriers to reporting suspected adverse reactions would be significantlylowered if those reports were made to a peer group organization. After all, theoccurrence of an adverse reaction may cause the doctor or pharmacist to ask himself(or herself) whether he should assume partial responsibility for this reaction. It ispossible that some would be less eager to report an adverse drug reaction to a'higher authority' such as the government.

Regional organisationFor an organization such as Lareb, in which several professions meet, it is relativelyeasy to maintain an extensive network of doctors and pharmacists. This is indeedfacilitated by Lareb's regional organization under which the Netherlands is dividedinto five regions. Lareb's headquarters in ’s-Hertogenbosch acts as one regionaloffice, with the other four in university hospitals throughout the country. Eachregional office has a regional coordinator, responsible for maintaining contact withthe doctors and pharmacists in that region. Such contact is both individual (throughpersonal visits) and collective, involving presentations in hospitals and to groups ofinvited general practitioners and pharmacists. Wherever possible, education and'refresher' courses are offered. The regional coordinator also personally assesses some of the incoming reports inorder to remain involved in the Lareb's 'core business' and will contribute torelevant publications wherever possible. A meeting of all the Lareb's scientific staffis held monthly at the head office, providing an opportunity for consultation andfurther 'in-service' training. Lareb is a small organization, with a staff of only 18. Some work part-time. Thereare four supportive (administrative) staff members, the remainder are all doctors,pharmacists or medical biologists by profession. Details are to be found on thecentre's website at www.lareb.nl.

55

Marked involvement of pharmacistsIn the context of pharmaceutical patient care, pharmacists in the Netherlands arevery much involved in ensuring safe and responsible use of medicines. Pharmacistsplayed an important part in setting up Lareb. Today, pharmacists provide about40% of the reports the centre receives (see Table 1).

Most reports are made by community pharmacists, which is perhaps to be expectedgiven Lareb's background. Hospital pharmacists lag somewhat behind in thisrespect. Accordingly, Lareb has joined forces with the Netherlands Society ofHospital Pharmacists in attempting to encourage greater involvement on the part ofits members. One objective is to establish a protocol in hospitals whereby housepharmacists are not only expected to provide effective pharmacotherapy, but willalso play a coordinating and facilitating role in terms of the collation andforwarding of adverse drug reaction reports. A survey held in early 2001 indicatedthat 97% of hospital pharmacists are eager to report any adverse reactions; theyknow what must be reported and in what way. The complaint that pharmacists arethemselves able to provide little or no clinical information in a report has beenshown to present no great problem in practice. Often, the good cooperationbetween doctors and pharmacists will ensure that adequate information can begiven, particularly if the relevant report is being made from a hospital situation. Ifnecessary, it is possible to contact the prescribing doctor to obtain furtherinformation. That pharmacists are able to provide a complete picture of a patient'sprescription history is a significant advantage.

The generation of signals The primary objective of any reporting system is to generate a 'signal': an earlyindicator or warning of a potential problem. It may be compared to the task of afire-watcher, who looks for smoke and, if he thinks he spots it, must then determine

56

Table 1Sources of reports

Year Pharmacists (%) Doctors (%)1996 46.3 53.71997 44.9 55.11998 44.5 55.51999 41.6 58.52000 41.1 58.9

whether there is indeed a fire and where that fire is located. In pharmacovigilance,it falls to the Medicines Evaluation Board to determine whether there are sufficientarguments to shout 'fire!', whereupon it will take the necessary measures. The reports received by Lareb are first assessed by one of its staff doctors orpharmacists. They examine the probability of a causal link, and will use the currentliterature, previous reports and the description of the drugs' pharmacologicalmechanism to assist them. The results of their assessment are notified to thereporter as well as to the government. A weekly assessment meeting involves all scientific staff. The reports and theirsubsequent assessments are discussed to determine whether further action isnecessary. Such further action may entail more detailed analysis of the relationshipbetween the reported reaction and the suspect drug. Research within Lareb hasrevealed a number of factors that can play a significant role in the decision toconduct further analysis. These include the seriousness of the reported reaction, thenumber of reports related to similar reports on other drugs received by the centre,and whether existing literature has devoted attention to the suspected reaction. During the weekly assessment meeting, all new reports are discussed and furtheraction is scheduled if there seems to be sufficient justification. As a general rule,such action will entail notifying the Medicines Evaluation Board, and in many casesan article will be published. Computer automation now plays an important role in the internal reportassessment process, with all incoming reports undergoing a set sequence ofevents. The information on the report forms themselves, together with that inany other relevant documentation, is stored in digital form. The weeklyassessment meeting also makes use of information obtained through automatedquantitative signal generation. The 'Reporting Odds Ratio' is calculated for allreports, providing a statistical indication of the reporting frequency of each ofthe suspected reactions compared to other reports in Lareb's database. Theresults of the Bayesian Confidence Propagation Neural Network analysis,submitted quarterly by the WHO Monitoring Centre in Uppsala are alsoautomatically linked to each report. Lareb's database contains over 40,000 reports. Besides providing a valuable aidto case-by-case analysis, quantitative information can also be used to distil usefulinformation from a large collection of data. Such information will not beprovided by a single case analysis. Lareb is particularly interested in thepossibilities for identifying specific syndromes and in detecting interactionsbetween drugs.(4,5) Ongoing research is being conducted into whether certainrisk factors for drug reactions can be identified using the information now filedin the database.

57

CommunicationLareb is a small organization in a big world. Its most important contacts are theprofessional organizations, the government and the pharmaceutical industry.

- Doctors and pharmacistsBecause the Lareb is itself an organization of doctors and pharmacists, it has readyaccess to practitioners in the field. Current European legislation does not permitdirect reporting by patients themselves, although Lareb would not be opposed toits future introduction. Partly in view of the fact that doctors and pharmacistsreport suspected adverse drug reactions on a purely voluntary basis, it is importantto inform them of the importance of reporting. In addition to the feedback itprovides, both direct and in the form of publications, the centre offers targetedinformation to potential reporters in the form of mailings and presentations. Thereport form itself has a carefully designed layout and is distributed in various ways,such as inclusion with the regular Drug Bulletin and the annualParmacotherapeutisch Kompas, the pharmacopoeia which forms a standard deskreference book for 90% of Dutch doctors. It is important that the reporter can relyon respect to privacy and confidentiality. Lareb doesn’t receive any informationabout the identity of the patient and no information about the reporter will be givento third parties. The Dutch law is also strict on privacy.An important mean of communication with the reporting parties is the 'feedbackreport'. Not only is receipt of each report acknowledged, but the assessment madeby the Lareb and the conclusions drawn with regard to the reported adverse drugreaction are notified to the person making the report. Besides wishing to encourage reporting, Lareb believes that it is important to raisethe level of awareness among doctors and pharmacists with regard to adverse drugreactions. This will not only lead to a better standard of reporting but will serve toreduce significantly the harmful effects of prescription medicines as well. Doctorswill prescribe more critically and will be more inclined to consider adverse drugreactions as the cause of complaints at an earlier stage in their differential diagnosis,whereupon they will become able to discontinue use of the drug or to adapt thedosage to avoid both unnecessary costs and unnecessary impact in terms of patienthealth.

- The governmentBecause Lareb is an independent organization working on behalf of thegovernment, good communication with that government is very important.Reports are forwarded to the Medicines Evaluation Board Agency weekly. Every 6weeks, a meeting is held between Lareb, the Agency and the Health Inspectorate.Besides possible 'signals', these meetings also discuss international developments.

58

- Marketing Authorization Holders Needless to say, Lareb maintains close contact with the pharmaceutical industry,which also has a vested interest in effective pharmacovigilance. All serious ('15-days') reports are forwarded to the relevant Marketing Authorization Holder, asrequired by international legislation. These reports are anonymous, neither thepatient nor the reporter can be traced. Similarly, all such reports made directly bythe pharmaceutical industry to the government are entered into Lareb's database.In the short term, the centre aims to achieve a free exchange of less serious reportsas well. All articles concerning a specific preparation are submitted for comment tothe relevant Marketing Authorization Holder prior to publication.

ResultsThe 'output' of Lareb can be assessed by looking at both the quantity and qualityof incoming reports, aspects that owe much to the efforts of the centre. Othercriteria include the number of publications for which the centre has beenresponsible and the number of notifications of possible signals it has made.

- Reports: quantityThe number of incoming reports continues to increase each year. The developmentin the number of reports included in the database is shown in Table 2.Lareb sees under-reporting as an inherent characteristic of a spontaneous reportingsystem and not necessarily as a drawback.

- Reports: quality Although an adequate number of reports are necessary to ensure a reliablereporting system, Lareb attaches greater importance to the quality of those reports. The quality of reports can also be seen to have risen each year. Quality iscontinuously assessed according to a number of criteria, one of which is the extentto which the report is documented. In an increasing number of cases, reports areaccompanied by adequate clinical information, including the specialists' clinicalnotes to the patient’s family practitioner. The fact that more complete informationis now available may be attributed in part to the greater number of reports beingmade by hospital practitioners. The increase in the number of reports adjudged tobe of a serious nature is shown in Table 3.Although preparations which have been on the market for some time mayoccasionally reveal new adverse reactions (as in the case of vigabatrine, which hadbeen available for over 10 years before a link with patients' restricted field of visionwas made), Lareb is particularly interested in new medicines. Table 2 shows thepercentage of reports relating to preparations that have been on the market for lessthan 5 years.

59

- Publications and presentationsHaving adopted a scientific and academic level as the basis for its workingmethods, Lareb is regarded as a serious partner by others, particularly theprofessional organizations. The scientific quality of the Lareb's work ismonitored by a Scientific Advisory Board, comprising experts in variousdisciplines. Each year, Lareb publishes over thirty articles in international or national journals,among which the Dutch Drug Bulletin. It also makes more than 30 presentations togroups of doctors and/or pharmacists and is frequently represented at internationalscientific conferences.

2.3.3 Further initiatives in pharmacovigilance in the NetherlandsBesides the spontaneous reporting system and the activities undertaken by, orunder the auspices of, the government, there are various other pharmacovigilance

60

Table 2Total reports and percentage of serious reports according to the criteria of the WHOor CIOMS

Year Reports WHO criteria CIOMS criteria total % reports serious %reports serious

1996 3042 10.6 3.61997 3649 12.1 4.71998 3434 14.5 11.51999 3686 17.1 14.12000 3702 19.9 18.1

Table 3Total reports and percentage of reports on drugs marketed less than five years

Year Reports < 5 years (%)

1996 3042 22.31997 3649 23.71998 3434 29.61999 3686 28.32000 3702 35.6

initiatives in the Netherlands. Of these, the most notable are those undertaken bythe marketing authorization holders and universities.

- Marketing authorization holdersNeedless to say, pharmaceutical companies in the Netherlands must comply withinternational legislation relating to pharmacovigilance. Reports that meet theCIOMS criteria must be made to the Medicines Evaluation Board within 15 daysand will also be included in Lareb's database. In addition, Marketing AuthorizationHolders are required to submit periodic safety update reports, to include allinformation known to them concerning the safety of the preparations for whichthey hold marketing authorization. The Netherlands does not have a tradition ofreports being made directly to the pharmaceutical industry by doctors orpharmacists; the vast majority of reports concerning suspected adverse drugreactions pass through Lareb.

- UniversitiesThree Dutch universities include a department of pharmacoepidemiology. Thesedevote considerable attention to the occurrence of adverse drug reactions at grouplevel. A number of initiatives have been developed whereby these can be studiedmore closely in the context of day-to-day medical and pharmacological practice.The resulting systems are more suitable for the assessment of signals than for theirgeneration. The Department of Pharmacoepidemiology of the University of Utrecht developedthe 'PHARMO' system (which is now operated independently). It is a record-linkage system that uses information provided by a number of pharmacists incombination with hospital clinical records. The department of Epidemiology and Bio-statistics of Rotterdam's ErasmusUniversity is responsible for the IPSI system. It is relying on digital informationrecorded by general practitioners. The Department of Social Pharmacy en Pharmacoepidemiology of the Universityof Groningen, which maintains also the InterActie database, has joined forces withLareb in developing an intensive monitoring system which will use the initialsignals notified by pharmacists as well as responses to surveys conducted amonggeneral practitioners. It is believed that such a system will result in a first impressionof possible adverse reactions in the case of newly authorized preparations.

2.3.4 Summary and future developmentsThe Netherlands can now look back on 50 years of systematic attention for adversedrug reactions. This began with the first edition of the book now popularly knownsimply as Meyler’s, and has developed to a stage in which the emphasis is on

61

effective pharmacovigilance and in which Meyler’s is now the work of severaldifferent authors. On behalf of and in co-operation with the government theNetherlands Pharmacovigilance Centre Lareb maintains the spontaneous reportingsystem for the Netherlands. A notable characteristic of the Dutch situation is thatdoctors and pharmacists are themselves responsible for this system, withpharmacists taking a significant role. Besides continued consideration for both the quantity and quality of reports, thefuture is likely to see further development of automatic signal generation and evengreater concern for good communication with potential reporters, in order toincrease awareness of adverse drug reactions. Developments at the European levelare certain to have a significant influence in this regard.

62

Chapter 2.3Pharmacovigilance in the Netherlands

In: ‘Pharmacovigilance’(Mann RD, Andrews EB, editors)Wiley, Chichester 2002

Kees van GrootheestEugène P. van PuijenbroekNetherlands Pharmacovigilance Centre Lareb’s-Hertogenbosch, the Netherlands.

51

References

1. Koning GPH de. A regionalized spontaneous surveillance programme for adverse drugreactions as a tool to improve pharmacotherapy. Thesis. Utrecht University, Faculty ofPharmaceutical Sciences 1994.

2. Meyboom RHB, Gribnau FWJ, Hekster YA, de Koning GHP, Egberts ACG Characteristicsof Topics in Pharmacovigilance in The Netherlands. Clin. Drug Invest 1996;4:207-219.

3. Koning GPH de, Bakker A, Leufkens HGM. Postmarketing surveillance in pharmacy: anorientation. Pharm Weekbl 1992;127:76-79.

4. Puijenbroek EP van, Egberts ACG, Meyboom RHB, Leufkens HGM. Signalling possibledrug-drug interactions in a spontaneous reporting system: Delay of withdrawal bleedingduring concomitant use of oral contraceptives and itraconazole. Br J Clin Pharmacol1999;47:689-693.

5. Puijenbroek EP van, Egberts ACG, Heerdink ER, Leufkens HGM. Detecting drug-druginteractions using a database for spontaneous adverse drug reactions: An example withdiuretics and non-steroidal anti-inflammatory drugs. Eur J Clin Pharmacol 2000;56:733-738.

63

Chapter 2.4Underreporting in the spontaneous reporting of adverse drug reactions - A pharmacovigilance approach

Submitted

Kees van Grootheest MD1

Ingrid Oosterhuis, student2

Eugène P. van Puijenbroek MD, PhD1

Lolkje T.W. de Jong – van den Berg PharmD, PhD2

1. Netherlands Pharmacovigilance Centre Lareb,’s-Hertogenbosch, the Netherlands.

2. Department of Social Pharmacy and Pharmacoepidemiology, GUIDE, Groningen, the Netherlands

65

Abstract

Spontaneous reporting of adverse drug reactions (ADRs) is the main method fordetecting hitherto unknown adverse effects to drugs. Especially from epidemiologicalcircles spontaneous reporting has been frequently criticised because of the problem ofunderreporting. Underreporting implies that far from all side effects are reported,which makes the spontaneous reporting system (SRS) unsuitable to determine theactual incidence of an ADR. Many attempts have been made to try and establish theextent of underreporting to make the SRS results suitable for epidemiological use.Since its inception, rather than an epidemiological methodology, the SRS hasgradually adopted a pharmacovigilance approach, which development is illustratedon the basis of the phenomenon of underreporting. Pharmacovigilance does notrequire all suspected ADRs to be reported but rather relies on an adequate numberof reports containing a sufficient amount of information to facilitate a causalityassessment that can subsequently help to signal unknown risks of drugs.Underreporting is characteristic of an SRS, which, from a pharmacovigilanceperspective, is a prerequisite and therefore should not a priori be labelled as negative.

66

2.4.1 IntroductionBecause at the time of its introduction the available knowledge about and experiencewith the safety of a new drug is limited, it is essential that the effects of the drug areclosely monitored during its use in clinical practice. This process is commonlyreferred to as Post Marketing Surveillance: the systematic surveillance and scientificevaluation of all intended and unintended effects of medicines on human health, aftertheir release for marketing.(1) The method that is most frequently employed is theso-called spontaneous reporting system (SRS), which system was developed in the1960s and 70s.(2) An SRS operates by requesting health professionals and,increasingly, also the users of medicinal drugs, to report their (negative) experienceswith the drugs to a national reporting centre. Well over 70 countries support such asystem and pass on their reports to the WHO Collaborating Centre for DrugMonitoring (the Uppsala Monitoring Centre).(3,4,5)Spontaneous ADR reporting is the main source of information for the detection ofsignals of drug safety hazards and should remain one of the cornerstones of acomprehensive safety monitoring programme’.(2) The great strength of spon-taneous reporting is that it operates for all drugs throughout their lifetime; it is theonly affordable method of detecting rare ADRs.(6) Nevertheless, since the earlydays of the system’s existence there has been criticism. The issue most frequentlymentioned is that the system of spontaneous reporting is associated withunderreporting: far from all adverse effects are being reported, thus rendering thesystem unsuitable to determine the incidence of a particular ADR.(7,8,9,10)In this paper we will explore underreporting more closely and discuss whether it isindeed justified to consider underreporting as a weakness of the SRS by tracing theoriginal context of the concept of underreporting and by looking at the roleunderreporting plays in today’s practice of pharmacovigilance.

2.4.2 Scientific interest for adverse drug reactionsWhen in the 1960s the world was shocked by the thousands of children that wereborn with serious congenital malformations as a result of the maternal use ofthalidomide, people became awoke of the dangers of possible adverse drugreactions resulting in a raised awareness of ADRs among both the nationalauthorities as well as the scientific community. This paved the way for the newdiscipline of pharmacoepidemiology, which can be regarded as a fusion of clinicalpharmacology and epidemiology.(11) The idea that it should also be possible toderive the incidence of adverse events from the ADR reports that were submittednecessitated the assessment of the extent of underreporting. This idea is alsoreflected in the name most commonly used for the system: spontaneous reportingsystem. In this context the term ‘spontaneous’ denotes basically the at randomgathering of reports, allowing extrapolation to all users. Right from the start of the

67

reporting systems it was clear that far from all adverse events would be reported(and possibly could not be reported) and moreover not all ADRs could beattributed to the suspected drug. In addition, the reporting of ADRs is not random.These facts warranted the conclusion that, although suitable for the detection ofhitherto unknown adverse drug reactions, an SRS was by no means appropriate forthe validation of these adverse effects (signal testing) which requires a differentmethodology, such as clinical trials and case control studies. In the past few yearsthere has been a bipartition in the field of ADR monitoring. The detection of newADRs, which is mainly based on the SRS approach, has now become the domain ofpharmacovigilance (see Figure 1). Pharmacoepidemiology, on the other hand,mainly concerns itself with the validation of ADRs and establishing their incidence,for which it employs its own methods. In other words: pharmacovigilance is chieflyconcerned with the clinical-pharmacological aspects of ADRs andpharmacoepidemiology with their epidemiological context. It needs to be notedthat pharmacoepidemiology has since broadened its scope even further, but this isbeyond the scope of this article. The changes just described are best reflected by thefounding of the International Society of Pharmacoepidemiology (ISPE, 1984) andthe International Society of Pharmacovigilance (originally established as theEuropean Society of Pharmacovigilance in 1992).Against this background it is self-evident that the terms ‘underreporting’ and‘spontaneous’ originally had an epidemiological connotation. However,pharmacovigilance has evolved into a scientific field in its own right, providing thetwo concepts with a different meaning.

68

Figure 1Relation between case reports (spontaneous reporting system) and (pharmaco)epidemiology(modified after Van Puijenbroek (48))

2.4.3 What is underreporting?A review of the literature yielded hundreds of articles on underreporting in relationto the reporting of ADRs. However, we were unable to find an unequivocaldefinition for the concept of underreporting. This lack has led to differentviewpoints and an ambiguous use of the term.The term underreporting is predominantly used to indicate that not all the adverseeffects that occur are reported, although that severe ADRs and adverse effects fornew drugs are more frequently reported than ADRs that are alreadyknown.(9,12,13) It is also used as a reference to the fact that only a minority of thepotential participants contribute to the reporting system; it is known that in generalless than 5% of physicians actually report adverse events.(14,15,16)When we look more closely at the most common interpretation of underreporting,i.e. the fact that only a certain proportion of all possible ADRs is reported, thisraises new questions because in this context underreporting is defined by what isunderstood by the term ADR. The definition that is most frequently used todescribe an ADR is given by the WHO: ‘a response to a medicine which is noxiousand unintended, and which occurs at doses normally used in man’.(17) AnotherWHO publication adds: ‘for the prophylaxis, diagnosis, or therapy of disease, orfor the modification of physiological function’.(18)However, this WHO definition fails to delineate the term ADR adequately, giventhat also adverse events that have not yet been diagnosed are included, for instancea liver dysfunction that has not been identified yet. Although such a disorder fallswithin the category of unintended effects, it cannot be reported. This exampleillustrates that ADR reporting will always be deficient and that for this reason alonethere will always be underreporting. Earlier, Edwards and Aronson already pointedto the unlimited scope of the ADR definition provided by the WHO.(19) The scaleof underreporting is large. In the literature Rawlins is often quoted, who, in hisWilliam Withering Lecture indicated that this was in excess of 90% for seriousreactions and higher for non-serious reactions.(14) Later studies have since shownthat this proportion is in fact considerably higher.(15,16,20) Underreporting is acommon feature of all spontaneous ADR reporting systems.(9)

2.4.4 Underreporting from an epidemiological perspectiveThere have been numerous attempts to try and determine the extent ofunderreporting in order to make the results of the SRS suitable for epidemiologicaluse. If the SRS data are corrected for the ratio of underreporting found, theincidence of an ADR may be determined.(21,22,23) Figure 2 indicates the differencebetween incidence (the proportion of users that have experienced an adverse event)and reporting ratio (the proportion of ADRs that have been reported). The modelof a spontaneous reporting system, however, is not suitable for an incidence

69

assessment. It is evident that the results of an SRS are considerably biased. Apartfrom the fact that not all ADRs are reported, reporting is done on a selective basis.New products are given far greater attention, but also serious ADRs and those thatare widely covered in the media raise the number of incidents that arereported.(24,25)Even though in the 1960s and 70s the contribution and limitations of the SRSreceived a great deal of attention from epidemiological circles, ‘very little that is newhas been written in the past 20 years’.(26) In this context Bortnichak et al. refer toit as the ‘childhood’ of pharmacoepidemiology. Epidemiological methods are notcompatible with a spontaneous reporting system because not all the basicrequirements for proper epidemiological research are being met.(27,28) Alvarez etal. rightly concluded that: ‘It is not possible to estimate incidence rates of adverseeffects using spontaneously reported data’.(9)It is clear that the concept of underreporting should no longer be seen in the lightof its original epidemiological background.

2.4.5 What does pharmacovigilance entail?Pharmacovigilance is the science and the activities relating to the detection,assessment, understanding and prevention of adverse effects or any other possibledrug-related problem.(18) Pharmacovigilance primarily is a clinical science thatemploys various methodologies, of which the spontaneous reporting system is the

70

Figure 2Difference between incidence and reporting ratio (modified after Van Puijenbroek (48))

most important. Other fields of science also contribute, such as various clinicaldisciplines, epidemiology, pharmacy, teratology and toxicology. The essence ofpharmacovigilance is that new information is gathered based on the practicalexperiences with individual patients and that this newly acquired knowledge willbenefit the treatment of the individual patient. The actual dataset of an SRS consistsof a series of case reports. This implies that pharmacovigilance closely resembles thedaily practices of the practitioner’s surgery: although use is made of evidence-baseddata, each new case needs to be assessed separately. This also implies that thereporting of possible ADRs is fraught with uncertainty. It are the cases that arecollected in this way that form the basis for the detection of new signals of adverseevents, just like they constitute the basis for the discovery of new symptomatologyand therapies.(29,30) It is exactly the lack of structure that is characteristic of thiscase-by-case research approach that makes its findings so innovative and makes itso highly sensitive in picking up new phenomena.(31) Serendipity often plays amajor role: a valuable new discovery is made without actually looking for it. It isno coincidence that in response to the increasing interest in evidence-basedmedicine the appreciation of the case report is undergoing a revival.(32)A surveillance system that relies on actual cases is not intended to provide aquantitative account of the risk levels involved. This belongs to the domain ofepidemiologists. But before an idea (for example a signal) can be confirmed orquantified, it first has to be discovered.(29) Although cases and case series rank low in the hierarchy of evidence, thespontaneous reporting system increasingly yields reliable signals, often meetingsufficient criteria for marketing authorities to gear their policy to. Venning et al.demonstrated in their retrospective study that of the 47 evaluated first reports in theliterature the majority (35) had been confirmed 18 years later.(33) In a study intothe 22 drugs that had been withdrawn from the Spanish market in the 1990s Arnaizet al. recently reported that in 18 cases the withdrawal had been based on casereports or case series and in the remaining four cases on randomised clinicaltrials.(34). Both studies were done from an epidemiological viewpoint but underlinethe value of cases. Thanks to the high quality of the reports and accurate causalityassessments the evidence for the signals is constantly gaining in strength and thechance of false positive signals is minimised.

2.4.6 Underreporting is desirableIn pharmacovigilance it is all about collecting those ADR reports that will facilitatethe timely detection of new signals of adverse reactions. Rather than requiring allADRs to be reported, an adequate number of reports containing the right kind ofinformation to allow an accurate causality assessment suffices.From a practical viewpoint alone it would be quite inconvenient if all likely ADRs

71

were reported. None of the national reporting systems is equipped to process suchan amount of data. It would be impossible to find the proverbial needles in thegigantic haystack, given the fact that each report needs to be weighed separately,both to allow an accurate causality assessment and to provide adequate feedback tothe reporting party. Finney already pointed out that it is vital to indicate explicitly which specificadverse events need to be reported.(35)In several countries specifically serious ADRs are invited and those associated withnewly marketed drugs.(36,37,38) Selection is a core task of the SRS and relies on thereporters’ expertise. Underreporting appears to be positively selective, since itinvolves mainly the less severe and the well-known effects, and this consequentlypreserves the most valuable asset of spontaneous reporting: signal generation.(9)Thus, from a pharmacovigilance perspective it is undesirable that all adverse eventsare reported to the SRS. It is those reports that may contribute to the signalling ofhitherto unknown risks of drugs that are indispensable. Table 1 provides an illustration of the recommendations the NetherlandsPharmacovigilance Centre Lareb implements with respect to which type of adverseevents it wishes to be reported.

2.4.7 We do need enough case reports of good qualityThe fact that not all ADRs need to be reported does not imply that reporting

72

Table 1Instructions, used by the Netherlands Pharmacovigilance Centre Lareb to informpotential contributors about what to report

What are you supposed to report to Lareb?It is important that you report any suspicions of adverse drug reactions.You can report anything you deem important, even if the correlation with therelevant drug is as yet uncertain.You should always report the following:- All adverse reactions associated with new drugs (until at least 2 years after their

introduction)- All serious adverse drug reactions- Unknown adverse reactions, i.e. any side-effect that is not included in the SPC- Adverse events in infants, including those associated with off-label use- In addition: adverse reactions to alternative (complementary) medication, OTC

products and vaccines

practices are not subject to specific requirements. Within the framework of GoodPharmacovigilance Practice several authors, most notably Meyboom, haveformulated a number of criteria an SRS needs to fulfil.(39,40,41)Each national SRS needs to receive a sufficient number of ADR reports to allow areliable surveillance of unknown harmful effects of drugs. Evidently, it is paramountthat the quality of the reports meets the highest standards, since it is only the reportsthat are well documented that can contribute to our knowledge about ADRs andwill allow both causality assessments and statistical analyses.(35,41)

We wish to highlight three of the factors that promote high-quality reports.a. The relationship between the national reporting centre and potential reporters

should be characterised by acknowledgement of professionalism andresponsibility.(42)This implies that on the one hand the SRS should operate on a scientifically soundbasis that is appreciated by both the physician and the pharmacist.On the other hand the long-established medical-ethical principle that one does notkeep knowledge that may be relevant for others to oneself should be applied hereas well. The creativity and open-mindedness of the contributors form the basis forqualitatively sound reports. Physicians that fail to report their observations ofadverse events do not comply with the standards of their profession.(7,12)

b. It is essential for the national reporting system to provide the reporting partywith proper feedback.(41,42,43) This way the sender of the report is bothinformed that his or her report has been duly taken account of and receivesinformation about the evaluation the experts at the centre have made. Only whenfeedback is provided on an individual basis will the relationship be fully bilateral.In addition, the various health professions need to be given collective feedbackby means of publications and oral presentations, allowing those involved toappreciate the contributions of the SRS and affording them the opportunity toput the newly acquired knowledge to practice.

c. The literature provides ample information as to which practical requirementsneed to be met, such as proper information to potential reporters, theavailability of a reporting form that is easy to complete, clear instructions aboutwhat needs to be reported, and, for hospitals in particular, facilities that promoteADR reporting. Here, it is recommended to integrate a reporting module intothe computerised systems of general practitioners, pharmacists and hospitals.

In order to increase the number of reports one could also try and raise the numberof those that are allowed to report. The number of studies that have been conductedto gain insight into the attitude of potential reporters and the factors that positivelyor negatively affect the commitment to report is substantial.(12,13,15,42)

73

Pharmacists have long been a valuable source of reports and yet in many countriestheir role in ADR reporting is restricted.(44,45) Nursing staff and patients are alsoincreasingly considered, although the contribution of patient reports is still a pointof debate.(46,47)

2.4.8 Statistical evaluation of the data collectedAs was mentioned above, spontaneous reports are unsuitable for epidemiologicalconclusions as regards incidence assessment. This does not imply that the datasetdoes not allow statistical analyses. From the start, the importance and possibilitiesof storing and analysing the data collected has been recognised, both on a nationaland international basis.(18,35) The fundamental principle of the concept of areporting system was developed by Finney, who also described the criteria foradequate analytical processing of the data.(17)The statistical analysis supplements the case-by-case analysis of the submittedreports.(39,48) The relationships between drugs and ADRs that aredisproportionally present, and that may consequently constitute a signal, are tracedin the dataset, for which several methods are applied.(49,50,51) A recent publicationdescribes the application of such a quantitative signal detection to support the case-by-case analysis in the daily operations of a pharmacovigilance centre.(52)In addition to detecting possible signals of new adverse events these methods canalso be employed to trace drug-drug interactions and drug-related syndromes.(53,54) Repeatedly, it has been shown that underreporting interferes with suchanalyses in certain cases only.(55,56)When applying statistical analyses it is essential to keep in mind that it isspontaneous reports that constitute the basic data. This implies that for eachrelevant report the clinical background needs to be checked.(27,39) As early as in1965 Finney, the architect of statistical analysis of SRS data, wrote: ‘The quality ofresults coming out of a computer cannot be higher than the quality of the recordsput in’.(35) Even when statistical analyses are employed, pharmacovigilance willremain primarily a clinical science.Against this background one needs to be cautious with what is sometimes calledautomated signal selection, where the analysis of the incoming reports is solely leftto software programmes.

2.4.9 Final remarksIt can be concluded that from a pharmacovigilance standpoint it is undesirable thatall adverse reactions are reported to a spontaneous reporting system but only thoseADRs that may contribute to the signal detection of hitherto unknown risksassociated with certain drugs. The adjective spontaneous in this context may be confusing. It should be

74

interpreted as the exploitation of the creativity of health professionals or patients.This can, however, only be done with any success if, as Pasteur taught us, thoseinvolved have a ‘prepared mind’. Providing potential reporters with the properinformation about what constitutes good ADR reporting is paramount. Edwardshas advocated replacing ‘spontaneous reporting’ with ‘concern reporting’.(57)The results of an SRS are unsuitable for epidemiological analyses for severalreasons, despite the valuable contribution epidemiology has made to thedevelopment of the system. This fact has been illustrated by the phenomenon ofunderreporting. If this phenomenon is interpreted from an epidemiologicalperspective, rather than a pharmacovigilance approach, making underreporting aweakness or point of critique, this may easily lead to the wrong conclusions. Bycomparing two WHO publications, one dating from 1972 and the other from theyear 2000, we can neatly illustrate the history of the two different interpretations ofthe concept of underreporting.(17,18) Whereas in 1972 it is stated thatunderreporting makes statistical evaluation more difficult and the reports receivedmay not be representative, in the 2000 publication we read under the headingUnderreporting: ‘…in signal detection not only the quantity but also the relevanceof case reports and the quality of data are important’.Research into underreporting is worthwhile only when it is aimed at thequantitative increase and qualitative enhancement of reports. Specifically, studiesinvestigating what type of reports contribute most to the existing knowledge andwhich methods increase the volume of submission of these types of reports arethought to be most useful. Thus, it may be questioned whether reporting knownadverse reactions, however serious these may be, serve any useful purpose.Underreporting is a characteristic of every SRS and, since this is desirable from apharmacovigilance viewpoint, it should not be aprioristically labelled as a negativeaspect.An SRS is not designed for an exact estimation of incidence rates. Looking forconcrete numbers in the data of a spontaneous reporting system is comparable tolooking for oranges on an apple tree. Not only it is the wrong tree, it also grows inanother country.

75

References

1. Gezondheidsraad (Dutch Health Council): Post Marketing Surveillance in the Netherlands.The Hague 1991.

2. Heads of Agencies. Establishing a European Risk Management Strategy – Report of the Headsof Agencies ad hoc Working Group. Brussels 2003. (http://heads.medagencies.org/heads)

3. Olsson S. The Role of the WHO Programme on International Drug Monitoring inCoordinating Worldwide Drug Safety Effects. Drug Saf 1998;19:1-10.

4. Olsson S. National Pharmacovigilance Systems – Country Profiles and Overview (secondedition). Uppsala Monitoring Centre. Uppsala 1999.

5. Lindquist A.M. Seeing and Observing in International Pharmacovigilance. Thesis. KatholiekeUniversiteit Nijmegen. Nijmegen 2003.

6. Mann RD, Andrews EB. Introduction. In: Pharmacovigilance. Mann R, Andrew E, editors.John Wiley and Sons. Chichester 2002.

7. Finney DJ. The detection of adverse reactions to therapeutic drugs. Stat Med 1982;1:153-61.

8. Rogers AS. The Detection and Identification of Unintented Drug Effects. In:Pharmacoepidemiology. An Introduction (Hartzema AG, Porta M, Tilson HH, editors).Harvey Whitney Books Company. Cincinnati 1998.

9. Alvarez-Requejo A, Carvalja A, Bégaud B, Moride Y, Verga T, Martín Arias LH. Under-reporting of adverse drug reactions – estimate based on a spontaneous reporting schema and asentinel system. Eur J Clin Pharmacol 1998;54:483-8.

10. Moore N, Hall G, Sturkenboom, Mann R, Lagnaoui R, Bégaud B. Biases affecting theproportional reporting ratio (PPR) in spontaneous report pharmacovigilance databases: theexample of sertindole. Pharmacoepidemiol Drug Saf 2003;12:271-81.

11. Strom B L. What is pharmacoepidemiology? In: Pharmacoepidemiology. Strom B L, editor.John Wiley and Sons. Chichester 2000.

12. Eland IA, Belton KJ, Grootheest AC van, Meiners AP, Rawlins MD, Stricker BHCh.Attitudinal survey of voluntary reporting of adverse drug reactions. Br J Clin Pharmacol1999;48:623-7.

13. Moride Y. Haramaburu F, Requejo AA, Bégaud B. Under-reporting of adverse drug reactionsin general practice. Br J Clin Pharamcol 1997;43:177-81.

14. Rawlins MD. Pharmacovigilance: paradise lost, regained or postponed? The WilleamWithering Lecture 1994. J R Coll Physicians Lond 1994;29:41-9.

76

15. Martin RM, Kapoor KV, Wilton LV, Mann RD. Underreporting of suspected adverse drugreactions to newly marketed (‘black triangle’) drugs in general practice: observational study.BMJ 1998;317:119-20.

16. Hasford J, Goettler M, Munter K-H, Müller-Oerlinghausen b. Physicians’ knowledge andattituded regarding the spontaneous reporting system for adverse drug reactions. J ClinEpidemiol 2002;55:945-50.

17. WHO. International Drug Monitoring: The Role of National Centres. Technical Report SeriesNo 498. WHO. Geneva 1972.

18. WHO. The Importance of Pharmacovigilance, Safety Monitoring of medical products. WHO.Geneva 2002.

19. Edwards IR, Aronson JK. Adverse drug reactions: definitions, diagnosis, and management.Lancet 2000;356:1255-9.

20. Bégaud B, Martin K, Haramburu F, Moore N. Rates of Spontaneous Reporting of AdverseDrug Reactions in France. JAMA 2002;288:1588.

21. Tubert-Bitter P, Bégaud B, Moride Y, Chaslerie A, Haramburu F. Comparing the Toxicity ofTwo Drugs in the Framework of Spontaneous Reporting: A Confidence Interval Approach. JClin Epidemiol 1996;49:121-3.

22. Fourrier A, Chaslerie A, Dartigues JF, Bégaud B. Under-Reporting of Adverse Drug Reactionin the Elderly. Pharmacoepidemiol Drug Saf 1995;4:379-80.

23. Stather R. Determining and describing the frequency of ADRs in the SPC. Reactions2002;931:3-4.

24. Wallenstein EJ, Fife D. Temporal Patterns of NSAID Spontaneous Adverse Event Reports.The Weber Effect Revisited. Drug Saf 2001;24:233-7.

25. Griffin JP, Weber JCP. Voluntary systems of adverse drug reaction reporting- Part II. AdvDrug React Pois Rev 1986;1:23-55.

26. Bortnichak EA, Dai WS. Epidemiologists and Adverse Event Data – A Challenge to the Field.Pharmacoepidemiol Drug Saf 1999;8:457-61.

27. Jick H. García Rodriguez LA, Pérez-Gutthann. Principles of epidemiological research onadverse and beneficial drug effects. Lancet 1998;352:1767-70.

28. Skegg DCG. Pitfalls of pharmacoepidemiology. BMJ 2000;321:1171-2.

29. Vandenbroucke JP. Case reports in an evidence-based world. J Roy Soc Med 1999;92:159-63.

30. Goodwin JS. The empirical basis for the discovery of new therapies. Persp Biol Med1991;35:20-36.

77

31. Vandenbroucke JP. Het belang van medische casuïstiek te midden van ‘evidence-based’geneeskunde en moleculaire verklaringen. Ned Tijdschr Geneeskd 2002;36:1699-1703.

32. Charlton BD, Walston F. Individual case studies in clinical research. J Evaluation Clin Pract1998;4:147-55.

33. Venning GR. Validity of anecdotal reports af suspected adverse drug reactions: the problem offalse alarms. BMJ 1982;284:249-52.

34. Arnaiz JA, Carné X. Riba N, Codina C, Ribas J, Trilla A. The use of evidence inpharmacovigilance. Case reports as the reference source for drug withdrawals. Eur J ClinPharmcacol 2001;57:89-91.

35. Finney DJ. The design and logic of a monitor of drug use. J Chron Dis 1965;18:77-98.

36. Grootheest AC van, Puijenbroek EP van. Pharmacovigilance in the Netherlands. In:Pharmacovigilance. Mann R, Andrew E, editors. John Wiley and Sons. Chichester 2002.

37. Kennedy DL, Goldman SA, Lillie RB. Spontaneous Reporting in the United States. In:Pharmacepidemiology. WHW Strom, editor. John Wiley and Sons. Chichester 2000.

38. Davis S, Raine JM. Spontaneous Reporting – UK. In: Pharmacovigilance. Mann R, Andrew E,editors. John Wiley and Sons. Chichester 2002.

39. Meyboom RHB. Detecting adverse drug reactions, Pharmacovigilance in The Netherlands.Thesis. Katholieke Universiteit Nijmegen 1998.

40. Meyboom RHB. The Case for Good Pharmacovigilance Practice. Pharmacoepidemiol DrugSaf 2000;9:335-6.

41. UMC. Guidelines for setting up and running a Pharmacovigilance Centre. Uppsala 2000.

42. Biriell C, Edwards IR. Reasons for Reporting Adverse Drug Reactions – Some ThoughtsBased on an International Review. Pharmacoepidemiol Drug Saf 1997;6:21-6.

43. Grootheest AC van, Puijenbroek EP van. Pharmacovigilance in the Netherlands. In:Pharmacovigilance: 309-15. Mann R, Andrews E, editors. Wiley. Chichester 2002.

44. Grootheest AC van, Puijenbroek EP van, Jong – van den Berg LTW de. Contribution ofpharmacists to the reporting of adverse drug reactions. Pharmacoepeidemiol Drug Saf2002;11:205-10.

45. Grootheest AC van, Olsson S, Couper M, Jong-van den Berg LTW de. The role of pharmacistsin reporting adverse drug reaction in international perspective. Pharmacoepidemiol Drug Saf2003 (in press).

46. Morrison-Griffiths S, Walley TJ, Park BK, Breckenridge AM, Pirmohamed M. Reporting ofadverse drug reactions by nurses. Lancet 2003;361:1347-8.

78

47. Grootheest K van, Graaf L de, Jong – van den Berg LTW de. Consumer Adverse DrugReaction Reporting – A New Step in Pharmacovigilance. Drug Saf 2003;26:211-7.

48. Puijenbroek EP van. Quantitative Signal Detection in Pharmacovigilance. Thesis. Universityof Utrecht 2001.

49. Puijenbroek EP van, Bate A, Leufkens HGM, Lindquist M, Orre R, Egberts ACG. Acomparison of measures of disproportionality for signal detection in spontaneous reportingsystems for adverse adverse drug reactions. Drug Saf 2002;11:3-10.

50. Bate A. The Use of a Bayesian Confidence Propagation Neural Network in Pharmaco-vigilance. Umeå University Medical Dissertations New Series No 846. Umeå 2003.

51. Szarfman A, Machado SG, O’Neill RT. Use of screening algorithms and computer systems toefficiently signal higher-then-expected combinations of drugs and events in the US FDA’sspontaneous reporting data base. Drug Saf 2002;25:381-92.

52. Puijenbroek EP van, Diemont WL, Grootheest AC van. Application of Quantitative SignalDetection in the Dutch Spontaneous Reporting System for Adverse Drug Reactions. Drug Saf2003;26:293-301.

53. Puijenbroek EP van, Egberts AC, Heerding ER, Leufkens HGM. Detecting drug-druginteractions using a database for spontaneous adverse drug reactions: an example withdiuretics and non-steroidal anti-inflammatory drug. Eur J Clin Pharmacol 2000;56:733-8.

54. Puijenbroek EP van, Egberts ACG, Meyboom RHB, Leufkens HGM. Association betweenterbinafine and arthralgia, gever and urticaria: symptoms or syndrome? Drug Saf 2001;10:135-42.

55. Tubert-Bitter P, Bégaud B, Moride Y, Chaslerie A, Haramburu F. Comparing the Toxicity ofTwo Drugs in the Framework of Spontaneous Reporting: A Confidence Interval Approach. JClin Epidemiol 1996;49:121-3.

56. Heijden PGM, Puijenbroek EP van, Buuren S van, Hofstede JW van der. On the assessmentof adverse drug reactions from spontaneous reporting sytems. The influence ofunderreporting on Odds ratios. Stat Med. 2002;21(14):2027-44.

57. Edwards IR. Spontaneous reporting – of what? Clinical concerns about drugs. Br J ClinPharmacol 1999;48:138-41.

79