- Email: [email protected]
Physicians' knowledge and attitudes regarding the spontaneous reporting system for adverse drug reactions
Journal of Clinical Epidemiology 55 (2002) 945–950
PHARMACOEPIDEMIOLOGY REPORT
Physicians’ knowledge and attitudes regarding the spontaneous reporting system for adverse drug reactions J. Hasforda,*, M. Goettlera, K.-H. Munterb, B. Müller-Oerlinghausenb a
Department of Medical Informatics, Biometry and Epidemiology (IBE), Ludwig-Maximilians-Universität, IBE Marchioninistrasse 15, München, Germany b Arzneimittelkommission der Deutschen Ärzteschaft, Aachener Strasse 223-237, Köln, Germany Received 25 August 2000; received in revised form 13 February 2002; accepted 15 May 2002
Abstract The spontaneous reporting system (srs) is the most important early warning system of adverse drug reactions. As there is serious under-reporting we studied the respective knowledge and attitudes of two samples of physicians in Germany. Five hundred randomly sampled physicians and 815 physicians who had actually reported an ADR were included; the response rate to the mail questionnaire was 51.4 and 43.9%, respectively; 61.3% said to have reported at least one case in their life. As many as 75–85% of physicians said never to have sent an ADR report to the governmental or professional reporting systems. Reporting to pharmaceutical companies, on the other hand, has been substantially better. Sixty-eight and two-tenths percent indicated to have suspected an ADR without reporting it. Major reasons for not reporting were: ADR well known (75.6%), too trivial (71.1%), causality uncertain (66.3%). The ADR with the highest probability of being reported were serious unknown adverse reactions of a new drug (81.1%) or an established drug (72.9%) and serious known reactions to a new drug (65.2%). Almost 20% of the physicians admitted to not know the spontaneous reporting system and 30% to not know how to report; 54% would rather report an ADR if therapeutic advice was offered. The results indicate that the traditional ways of advertising the srs and communicating with physicians could be improved. Proactive services and professional marketing of srs are needed to reduce underreporting. © 2002 Elsevier Science Inc. All rights reserved. Keywords: Averse drug reaction monitoring; Pharmacovigilance; Survey; Spontaneous reporting system; Attitude; Under-reporting
1. Introduction The development of highly effective drugs in the last decades has brought remarkable benefits for the patients. The other side of the coin is that adverse drug reactions (ADR) occur. About 6% of all hospital admissions to medical wards have been shown to be due to adverse drug reactions [1], and fatal adverse drug reactions rank among the most common causes of death in the United States [2]. The economic burden of ADRs is considerable also: the estimated direct costs of ADR-related hospital admissions in Germany are well above 1 billion German marks per year [3]. For the United States, a total cost of $47.4 billion annually for 8.7 million drug-related admissions were reported [4]. Thus, there are many reasons—from the well-being of the patient to the economical use of resources—to identify adverse drug reactions as early as possible and to prevent them, if possible. The spontaneous reporting system is, as a signal generating system, currently the most important early * Corresponding author. Tel: 49-89-7095-7480; fax: 49-89-70957482. E-mail address: [email protected] (J. Hasford).
warning system of previously unknown and rare adverse reactions to drugs [5,6]. The spontaneous reporting system relies on vigilant physicians who generate a suspicion that a particular drug has caused an adverse reaction and who report it. Although the value of the spontaneous reporting system has been repeatedly proven, there still is considerable under-reporting [7–9]. Even serious suspected adverse reactions are only reported in 5–15% of all incident cases [10–12]. Thus, the usefulness of the spontaneous reporting system is impaired, as too few physicians report and much less adverse drug reactions than actually occur are notified. The attitudes of the physicians and their knowledge of the spontaneous reporting system are essential determinants of underreporting [7,13–16]. In Germany, the Drug Commission of the German Medical Association (DC) started to run a spontaneous reporting system as early as 1958. Since 1988, the official regulations of conduct oblige physicians to report all adverse drug reactions including suspected cases [17]. Although the spontaneous reporting system in Germany has existed for about 40 years, there were no empirical data about the attitudes and the knowledge of the physicians.
0895-4356/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S0895-4356(02)00 4 5 0 - X
946
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
The primary objectives of our study were to assess the attitudes of the physicians, to evaluate their knowledge of the spontaneous reporting system, and to identify the reasons for under-reporting. In addition, we wanted to find starting points for the improvement of the current system.
2. Methods Two different samples of physicians were studied: (1) a random sample of physicians working in Germany (n 500) from a large professional address data bank, stratified for federal states; (2) all physicians who reported an adverse drug reaction to the DC from mid-June to the end of November 1997 (n 815). The questionnaire used was based on the attitudinal survey II questionnaire of the European Pharmacovigilance Research Group. The self-administered questionnaire sought information regarding (a) demographic factors, (b) personal history of reporting ADR, (c) knowledge of the spontaneous reporting system, (d) attitudes towards reporting, (e) quantitative understanding of various frequency terms, and (f) incentives to improve reporting. A feasibility study was done. The final questionnaire (see www.pharmacoepi.de) was distributed by mail from May to November 1997. The data were analyzed using SPSS Version 7.5. t-Tests and 2-tests were used; the P-values given are two-sided.
3. Results The response rate of the random sample was 51.4% (n 245, evaluable: 230), of the DC sample 43.9% (n 358); 35.1% of the responding physicians worked in hospitals, 62.6% were primarily office-based, 2.3% worked in other fields (Table 1); the most common medical specialities were internal medicine, general practice, and pediatrics. 3.1. Reporting history Seventy-six percent of the physicians of the random sample had already diagnosed an adverse drug reaction at least once in the past. There are five different institutions that collect spontaneous reports in Germany: the Drug Commission of the German Medical Association (DC), the Federal Institute for Drugs and Medical Devices (BfArM—the national drug regulatory authority), the pharmaceutical manufacturer directly or via the sales representative, the hospital pharmacy and the “Arzneitelegramm”—a privately owned, independent monthly drug bulletin that runs a so-called “network for drug safety.” Almost two-thirds of the physicians (61.3%) of the random sample declared to have reported an adverse event/adverse drug reaction at least once in their professional life, 37.4% within the last 5 years, 15.8% at least 5 or more years ago, and 8.3% in both time
Table 1 Demographic profile of the samples
Age, years x, s Sex female male Specialist degree GP Internal medicine Pediatrics (Primarily) office-based Hospital-based Other fields Western states Eastern states Response rate
Random sample (n 230)
Drug Commission (DC) sample (n 358) P-value
48 9.3
41.2 8.7
P .05a
62 (27%) 168 (73%) 172 (74.8%) 12.2% 18.7% 10.8% 157 (71.1%0 62 (27.7%) 5 (2.2%) 187 (81.3%) 43 (18.7%) 51.4%
102 (28.8%) 252 (71.2%) 204 (57%) 25.9% 28.4% 11.9% 203 (56.9%) 140 (39.9%) 8 (2.3%) 291 (89%) 36 (11%) 43.9%)
P .05b P .05b
P .05b P .05b P .05b
a
t-test. chi2-test.
b
periods. Pharmaceutical manufacturers and their sales representatives, who are legally obliged to collect and to forward ADR reports, were the two most important addressees of ADR reports. About 35% of the physicians of both samples said to have reported at least once to sales representatives. Hospital pharmacies share a similar legal obligation, and about 14% of the physicians have used this service; 87.4% of the physicians of the random sample admitted that they never had reported to the drug commission of their own professional board (Table 2). 3.2. Knowledge and attitudes An amount of 68.2% of the physicians indicated to have already suspected an adverse drug reaction without reporting it. The three major reasons for not reporting were (Table 3): adverse drug reaction already well known (75.6%), adverse drug reaction too trivial to report (71.1%), and uncertainty concerning definite causality (66.3%). Almost 20% of the physicians of the random sample admitted to be unaware of the existence of a national reporting scheme, about 30% declared to know neither how to report nor the respective official regulations of conduct. With less than 1%, legal concerns were extremely rarely mentioned as reason for not reporting. An adverse drug reaction that is due to a drug that had been prescribed by another physician or is due to a drug taken in self-medication is also less likely to be reported. To find out whether there are special characteristics of a suspected ADR that make a report more likely, short case scenarios were presented permuting three contrasts: serious ADR—nonserious ADR, new ADR—well-known ADR, and adverse reaction of a new drug—ADR of an established drug. The adverse drug reactions with the highest probability of being reported were serious unknown adverse reac-
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
947
Table 2 Reporting history Random sample (n 230) (primarily) office-based Hospital-based Western states Eastern states Never reported reported last five years rep. more than 5 years ago reported both time periods Drug commission sample (n 358) Office-based Hospital-based Western states Eastern states Never reported reported last 5 years rep. more than 5 years ago reported both time periods
Drug authority
Drug commission
Pharmac. manufact.
Sales reps
Drug bulletin
Hospital pharmacies
21 (77.8%) 6 (22.2%) 22 (78.6%) 6 (21.4%) 202 (87.8%) 13 (5.7%) 13 (5.7%) 2 (0.9%)
19 (67.9%) 9 (32.1%) 29 (100%) 0 (0%) 201 (87.4%) 15 (6.5%) 13 (5.7%) 1 (0.4%)
55 (74.3%) 19 (25.7%) 64 (83.1%) 13 (16.9%) 153 (66.5%) 47 (20.4% 25 (10.9%) 5 (2.2%)
59 (72%) 23 (28%) 68 (78.2%) 19 (21.8%) 143 (62.2%) 66 (28.7%) 15 (6.5%) 6 (2.6%)
8 (72.7%) 3 (27.3%) 11 (100%) 0 (0%) 219 (95.2%) 8 (3.5%) 3 (1.3%) 0 (0%)
16 (53.3%) 14 (46.7%) 26 (83.9%) 5 (16.1%) 199 (86.5%) 18 (7.8%) 13 (5.7%) 0 (0%)
49 (62.8%) 29 (37.2%) 68 (85%) 12 (15%) 273 (76.3%) 69 (19.3%) 6 (1.7%) 10 (2.8%)
182 (60.9%) 117 (39.1%) 251 (88.7%) 32 (11.3%) 0 (0%) 323 (90.2%) 5 (1.4%) 30 (8.4%)
93 (62.8%) 55 (37.2%) 125 (87.4%) 18 (12.6%) 200 (55.9%) 134 (37.4%) 17 (4.7%) 7 (2.0%)
91 (77.1%) 27 (22.9%) 92 (85.2%) 16 (14.8%) 237 (66.2%) 106 (29.6%) 11 (3.1%) 4 (1.1%)
41 (80.4%) 10 (19.6%) 46 (97.9%) 1 (2.1%) 307 (85.8%) 34 (9.5%) 5 (1.4%) 12 (3.4%)
30 (66.7%) 15 (33.3%) 33 (82.5%) 7 (17.5%) 312 (87.2%) 38 (10.6%) 5 (1.4%) 3 (0.8%)
The percentages are calculated per section, e.g., 21 (77.8%) in the first column means that 77.8% of the responders reporting to the drug authority were office based; 202 (87.8%) means that 87.8% of all responders had never reported to the drug authority.
tions of a new drug, which would be reported by 81.1% of the physicians and serious unknown adverse reactions of an established drug (72.9%), followed by a serious known adverse reaction to a new drug (65.2%). Less than 10% of the physicians would report nonserious adverse reactions to established drugs.
4. Further results Textbooks, summaries of product characteristics, and patient information leaflets quite often use verbal terms to describe the incidence of adverse drug reactions. In Germany, the official terms are: common, occasional, rare, and single cases or very rare. The questionnaire asked the physicians to link these verbal terms with incidence estimates. The results show considerable variability regarding physicians’ interpretation of the official terminology: 68.9% of the physicians think that the category very rare means an incidence of one per 100,000, the correct answer for Germany—less than one per 1000—was selected by only 8.2% of the physicians (Table 4). The results indicate that the official terminology differs considerably from the common interpretation of these frequency terms. The final part of the questionnaire looked for starting points to improve the reporting behavior. About 54% asked for the possibility to report by phone to a qualified drug information service, and 43% stated that their motivation to report could be increased if their PC offered the option to print out the official report form. The chance to report by e-mail was positively considered by 12.4%. There were no statistically significant differences between the physicians’ attitudes in West and East Germany.
5. Discussion 5.1. Representativeness, reporting history, and validity Although the spontaneous reporting system in Germany is about 40 years old, this is the first survey concerning the knowledge and attitudes towards this system. Thanks to a couple of reminders a response rate of about 50% could be achieved in the random sample. Most probably one is well advised to assume that the nonresponders were mostly nonreporters. Thus, the generalizability of the results of this survey is limited. According to official data there are about 280,000 physicians in Germany, of whom 40% are office based, 49% are working in a hospital, and 11% in other areas like research and administration [18]. About 36% of the physicians are feTable 3 Percentage of respondents giving the stated reason for not having reported a suspected adverse drug reaction
ADR too well known* ADR too trivial Uncertain causality Reporting too bureaucratic* Lack of time* Resp. Rules of conduct unknown* Susp. drug prescribed by colleague* Reporting process unknown* Lack of financial reimbursement* Susp. drug was self-medication* Reports considered useless* Spont. reporting system unknown* Fear of legal liability *P .05, chi2-test.
Drug Commission sample
Random sample
80.8 70.8 63.5 27.3 33.8 18.8 8.3 4.6 13.1 5.5 7.7 4.6 0.8
65.9 71.7 71.7 39.9 22.5 29.0 28.0 27.5 6.5 20.8 16.7 18.1 0.7
948
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
Table 4 Interpretation of the official terminology for risk quantification by physicians (%) (random and DC sample combined, n 588), e.g., 53.4% of the physicians think that a rare ADR occurs in 1 in 10,000 patients
Official terminology
Common Occasional Rare Very rare ADR ADR ADR ADR 1:10 1:100–1:10 1:1000–1:100 1:1000
Physicians’ incidence estimate 1/10 patients 31.6 1/100 patients 24.0 1/1,000 patients 5.9 1/10,000 patients 2.0 1/100,000 patients 0 35.9 Missinga
1.0 20.7 42.2 14.5 4.8 16.8
0.7 2.0 18.3 53.4 11.2 14.4
0.7 0.2 1.0 8.2 68.9 21.1
a
Refers to missing responses (%) per frequency term.
male, 85% live in the Western and 15% in the Eastern states of Germany. Sixty-six percent of the physicians have a specialist degree. Relating these data to both samples studied, it seems that male office-based specialists are over-represented. The proportion of physicians from the new Eastern states in Germany in the DC sample is fairly low, and indicates that they are less accustomed to report to the Drug Commission than their Western colleagues (P .05, 2-test). Almost 25% of the physicians declared to have never diagnosed an adverse drug reaction. This result indicates that a considerable part of the physicians has major problems to diagnose an ADR. Sixty-one and three-tenths percent state that they have already reported at least one suspected ADR. This percentage is somewhat higher than the average of 52% reported recently for Europe, with a range from 19.4% in Italy up to 74.4% in France [15,19,20]. There are, however, some doubts about the validity of these self-reported activities. If 6.5% of the approximately 250,000 physicians who treat patients in Germany had actually reported at least once to the Drug Commission within the last 5 years (Table 2), about 16,250 reports should have been received. But there were only about 9,000 reports in this period [21]. Such attempts to crossvalidate survey data are missing in previous reports [15,19]. Whereas in most European countries alleged adverse drug reactions are mainly reported to national or regional centers [15], in Germany the pharmaceutical manufacturers and their sales representatives are the most common recipients of reports. We assume that most “reports” to sales representatives were actually not formal written reports, but oral reports during an informal conversation. The German drug law states that sales representatives must record and forward any report of a suspected ADR. Comparably few physicians report directly to the German drug authority, the BfArM. The rather low proportion (0– 8.4%) of physicians of both samples who stated to have reported in both predefined time periods indicates that none of the institutions managed satisfactorily to establish a stable relationship with their reporters.
5.2. Reasons for under-reporting One of the major objectives of our study was to find out the extent of and the reasons for under-reporting. The most common explanation for not reporting was that the ADR was already well known, a finding that confirms results of previous studies [7,8,10,12,14,22]. Apparently it is largely unknown that the spontaneous reporting system could also be used for monitoring changes of the type and/or frequency of serious known ADRs. This can only be achieved, if already known serious ADR are reported as well. From a physician’s point of view, however, it may seem “natural” and cost-efficient to not report even serious ADRs if well known, as there is no financial compensation and a good case report takes some time. About two-thirds of the physicians stated that they did not file a report because they were in doubt about causality. This finding was unexpected, because the official reporting form explicitly states in its heading that suspected cases of adverse drug reactions should also be reported. Twenty percent of the physicians of the random sample did not know the spontaneous reporting system at all, and almost 30% had no idea how to report. In countries such as the United Kingdom, Denmark, and Sweden, only 2.7, 3.4, and 8.6% of the physicians, respectively, were unsure how to report [15]; 86.7% of the physicians stated to not know any criteria concerning ADR/adverse events that should be reported, although the drug commission regularly publishes such criteria in the German Medical Journal, which is mailed weekly to every physician in Germany. These results indicate that the efforts of the DC in the past to make the physicians familiar with the spontaneous reporting system have failed to a relevant extent. New marketing concepts are thus needed and have to be evaluated. 5.3. Starting points for improvements The questionnaire also asked for incentives to improve motivation for reporting and for techniques to make reporting easier. As in other studies, the possibility to report by phone was mentioned by a majority of the physicians [22,23]. By offering the option to report by phone the number of reports could be increased fourfold in a project in Maryland [14]. This option could be combined with a toll-free telephone number. The finding that about 50% of the physicians asked for the possibility to report by fax indicates again that too many physicians do not really know how to report. Reporting by fax has been possible for many years, and the respective fax number is printed on the official report form, which is printed on the last inner page of the Deutsche Ärzteblatt (the weekly German Medical Journal) every second week. An additional starting point for the improvement of the spontaneous reporting system is the instant availability of reporting forms [5]. As most offices in Germany nowadays use PC-based programs for prescribing, a standard tool con-
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
taining the report form with an option for printout was mentioned by more than 40%. The DC has recently developed such a program that copies relevant information like demographic data, underlying diseases, and medication data automatically if these data are available electronically. Finally, a qualified therapy information service might help. Experiences with the French decentralized pharmacovigilance system show that about half of the ADR reports are derived from inquiries for therapeutic advice. Our data suggest that especially physicians, who have never reported up to now, might report in the future if such services are provided. This survey’s focus was to identify reasons why physicians do not report. Biriell and Edwards published an interesting approach insofar as they tried to explore the reasons of physicians and pharmacists who had taken time to report alleged ADRs [24]. This approach is particularly valuable to find starting points for establishing a positive relationship between reporter and center and thus to increase the proportion of physicians who will report more than once in their lifetime. The reasons for reporting mentioned most often in their survey [24]: motivation to contribute to medical knowledge, reaction previously unknown, reaction to new drug, evident association between drug and reaction, and severity of reaction, coincide fairly well with our findings. They propose active personal encouragement, recognition, and feedback from the national center for those submitting reports.
6. Conclusions In conclusion, the survey generated data about the attitudes of physicians towards the spontaneous reporting system and provided starting points to improve the current spontaneous reporting system. Considering that many doctors have with little credit or appreciation reported their personal clinical suspicions, the reporting centers should develop and adopt a service-oriented approach, for example, by providing qualified and confidential therapeutic advice. Certainly, every innovation has to be evaluated. There is no doubt, however, that any acceptable effort is needed to strengthen the current system, as the spontaneous reporting system is the only population-wide approach for signal generation of new, previously unknown serious ADRs of new drugs.
Acknowledgments This study was supported, by a grant to the European Pharmacovigilance Research Group from the Commission of European Communities under the Biomed Program (Contract number: PL 92 1556). We thank K. Belton and M. Rawlins for providing the EPRG Attitudinal Survey II questionnaire and all physicians who were willing to respond. We acknowledge the help of S. Schneeweiss and I. GarcíaGomez in organizing the survey. We appreciate the valuable comments of the reviewers.
949
References [1] Muehlberger N, Schneeweiss S, Hasford J. Adverse drug reaction monitoring—cost and benefit considerations part I: frequency of adverse drug reactions causing hospital admissions. Pharmacoepidemiol Drug Safety 1997;6(Suppl 3):S71–7. [2] Lazarou J, Pomeranz BH, Corey PH. Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. JAMA 1998;279(15):1200–5. [3] Goettler M, Schneeweiss S, Hasford J. Adverse drug reactions monitoring—cost and benefit considerations part II: cost and preventability of adverse drug reactions leading to hospital admission. Pharmacoepidemiol Drug Safety 1997;6(Suppl 3):S79–90. [4] Johnson JA, Bootnman JL. Drug-related morbidity and mortality. A cost-of-illness model. Arch Intern Med 1995;155:1949–56. [5] Scott HD, Thacher-Renshaw A, Rosenbaum SE, Waters WJ Jr, Green M, Andrews LG, Faich GA. Physician reporting of adverse drug reactions: results of the Rhode Island drug reaction reporting project. JAMA 1990;263:1785–8. [6] Beach JE, Faich GA, Bormel FG, Sasinowski FJ. Black box warnings in prescription drug labeling: results of a survey of 206 drugs. Food Drug Law J 1998;53:403–11. [7] Walker SR, Lumley CE. The attitudes of general practitioners to the monitoring and reporting of adverse drug reactions. Pharmaceut Med 1986;1:195–203. [8] Egberts TCG, Smulders M, de Koning FHP, Meyboom RH, Leufkens HG. Can adverse drug reactions be detected earlier? A comparison of reports by patients and professionals. BMJ 1996;313:530–1. [9] Martin RM, Kapoor KV, Wilton LV, Mann RD. Underreporting of suspected adverse drug reactions to newly marketed (“black triangle”) drugs in general practice: observational study. BMJ 1998;317: 119–20. [10] Smith CC, Bennett PM, Pearce HM, Harrison PI, Reynolds DJ, Aronson JK, Grahame-Smith DG. Adverse drug reactions in a hospital general medical unit meriting notification to the Committee on Safety of Medicines. Br J Clin Pharmacol 1996;42(4):423–9. [11] Moride Y, Haramburu F, Requejo AA, Begaud B. Under-reporting of adverse drug reactions in general practice. Br J Clin Pharmacol 1997; 43(2):177–81. [12] Lumley CE, Walker SR, Hall GC, Staunton N, Grob PR. The underreporting of adverse drug reactions seen in general practice. Pharmaceut Med 1986;1:205–12. [13] Koch-Weser J, Sidel VW, Sweet RH, Kanarek P, Eaton AE. Factors determining physician reporting of adverse drug reactions. Comparison of 2000 spontaneous reports with surveillance studies at the Massachusetts General Hospital. N Engl J Med 1969;280(1):20–6. [14] Rogers AS, Israel E, Smith CR, Levine D, McBean AM, Valente C, Faich G. Physician knowledge, attitudes and behavior related to reporting adverse drug events. Arch Intern Med 1988;148:1596–600. [15] Belton KJ, and The European Pharmacovigilance Research Group. Attitude survey of adverse drug-reaction reporting by health care professionals across the European union. Eur J Clin Pharmacol 1997;52: 423–7. [16] McGettigan P, Feely J. Adverse drug reaction reporting: opinions and attitudes of medical practitioners in Ireland. Pharmacoepidemiol Drug Safety 1995;4:355–8. [17] Bundesaerztekammer. Berufsordnung für die deutschen Ärzte. Deutsches Ärzteblatt 1988;85:3601. [18] Der Bundesminister fuer Gesundheit. Daten des Gesundheitswesens. Ausgabe. Baden: Nomos Verlagsgesellschaft Baden; 1997. [19] Pierfitte C, Hanesse B, Royer RJ. Enquete d’opinion sur la pharmacovigilance Européenne: le sentiment des médecins franais. Therapie 1995;50:171–2. [20] Cosentino M, Leoni O, Banfi F, Lecchini S, Frigo G. Attitudes to adverse drug reaction reporting by medical practitioners in a Northern Italian district. Pharmacol Res 1997;35(2):85–8. [21] Munter K-H. Bearbeitung von UAWs in der Arzneimittelkommission
950
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
der deutschen Ärzteschaft. In: Müller-Oerlinghausen B, Lasek R, Düppenbecker H, Munter K-H, editors. Handbuch der unerwünschten Arzneimittelwirkungen. München-Jena: Urban & Fischer; 1999. p. 571–6. [22] Fincham JE. Pilot projects to stimulate adverse drug reaction reporting. J Clin Pharm Ther 1987;12:243–7.
[23] Robins AH, Weir M, Biersteker EM. Attitudes to adverse drug reactions and their reporting among medical practitioners. S Afr Med J 1987;72(2):131–4. [24] Biriell C, Edwards RI. Reasons for reporting adverse drug reactions—some thoughts based on an international review. Pharmacoepidemiol Drug Safety 1997;6:21–6.
PHARMACOEPIDEMIOLOGY REPORT
Physicians’ knowledge and attitudes regarding the spontaneous reporting system for adverse drug reactions J. Hasforda,*, M. Goettlera, K.-H. Munterb, B. Müller-Oerlinghausenb a
Department of Medical Informatics, Biometry and Epidemiology (IBE), Ludwig-Maximilians-Universität, IBE Marchioninistrasse 15, München, Germany b Arzneimittelkommission der Deutschen Ärzteschaft, Aachener Strasse 223-237, Köln, Germany Received 25 August 2000; received in revised form 13 February 2002; accepted 15 May 2002
Abstract The spontaneous reporting system (srs) is the most important early warning system of adverse drug reactions. As there is serious under-reporting we studied the respective knowledge and attitudes of two samples of physicians in Germany. Five hundred randomly sampled physicians and 815 physicians who had actually reported an ADR were included; the response rate to the mail questionnaire was 51.4 and 43.9%, respectively; 61.3% said to have reported at least one case in their life. As many as 75–85% of physicians said never to have sent an ADR report to the governmental or professional reporting systems. Reporting to pharmaceutical companies, on the other hand, has been substantially better. Sixty-eight and two-tenths percent indicated to have suspected an ADR without reporting it. Major reasons for not reporting were: ADR well known (75.6%), too trivial (71.1%), causality uncertain (66.3%). The ADR with the highest probability of being reported were serious unknown adverse reactions of a new drug (81.1%) or an established drug (72.9%) and serious known reactions to a new drug (65.2%). Almost 20% of the physicians admitted to not know the spontaneous reporting system and 30% to not know how to report; 54% would rather report an ADR if therapeutic advice was offered. The results indicate that the traditional ways of advertising the srs and communicating with physicians could be improved. Proactive services and professional marketing of srs are needed to reduce underreporting. © 2002 Elsevier Science Inc. All rights reserved. Keywords: Averse drug reaction monitoring; Pharmacovigilance; Survey; Spontaneous reporting system; Attitude; Under-reporting
1. Introduction The development of highly effective drugs in the last decades has brought remarkable benefits for the patients. The other side of the coin is that adverse drug reactions (ADR) occur. About 6% of all hospital admissions to medical wards have been shown to be due to adverse drug reactions [1], and fatal adverse drug reactions rank among the most common causes of death in the United States [2]. The economic burden of ADRs is considerable also: the estimated direct costs of ADR-related hospital admissions in Germany are well above 1 billion German marks per year [3]. For the United States, a total cost of $47.4 billion annually for 8.7 million drug-related admissions were reported [4]. Thus, there are many reasons—from the well-being of the patient to the economical use of resources—to identify adverse drug reactions as early as possible and to prevent them, if possible. The spontaneous reporting system is, as a signal generating system, currently the most important early * Corresponding author. Tel: 49-89-7095-7480; fax: 49-89-70957482. E-mail address: [email protected] (J. Hasford).
warning system of previously unknown and rare adverse reactions to drugs [5,6]. The spontaneous reporting system relies on vigilant physicians who generate a suspicion that a particular drug has caused an adverse reaction and who report it. Although the value of the spontaneous reporting system has been repeatedly proven, there still is considerable under-reporting [7–9]. Even serious suspected adverse reactions are only reported in 5–15% of all incident cases [10–12]. Thus, the usefulness of the spontaneous reporting system is impaired, as too few physicians report and much less adverse drug reactions than actually occur are notified. The attitudes of the physicians and their knowledge of the spontaneous reporting system are essential determinants of underreporting [7,13–16]. In Germany, the Drug Commission of the German Medical Association (DC) started to run a spontaneous reporting system as early as 1958. Since 1988, the official regulations of conduct oblige physicians to report all adverse drug reactions including suspected cases [17]. Although the spontaneous reporting system in Germany has existed for about 40 years, there were no empirical data about the attitudes and the knowledge of the physicians.
0895-4356/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S0895-4356(02)00 4 5 0 - X
946
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
The primary objectives of our study were to assess the attitudes of the physicians, to evaluate their knowledge of the spontaneous reporting system, and to identify the reasons for under-reporting. In addition, we wanted to find starting points for the improvement of the current system.
2. Methods Two different samples of physicians were studied: (1) a random sample of physicians working in Germany (n 500) from a large professional address data bank, stratified for federal states; (2) all physicians who reported an adverse drug reaction to the DC from mid-June to the end of November 1997 (n 815). The questionnaire used was based on the attitudinal survey II questionnaire of the European Pharmacovigilance Research Group. The self-administered questionnaire sought information regarding (a) demographic factors, (b) personal history of reporting ADR, (c) knowledge of the spontaneous reporting system, (d) attitudes towards reporting, (e) quantitative understanding of various frequency terms, and (f) incentives to improve reporting. A feasibility study was done. The final questionnaire (see www.pharmacoepi.de) was distributed by mail from May to November 1997. The data were analyzed using SPSS Version 7.5. t-Tests and 2-tests were used; the P-values given are two-sided.
3. Results The response rate of the random sample was 51.4% (n 245, evaluable: 230), of the DC sample 43.9% (n 358); 35.1% of the responding physicians worked in hospitals, 62.6% were primarily office-based, 2.3% worked in other fields (Table 1); the most common medical specialities were internal medicine, general practice, and pediatrics. 3.1. Reporting history Seventy-six percent of the physicians of the random sample had already diagnosed an adverse drug reaction at least once in the past. There are five different institutions that collect spontaneous reports in Germany: the Drug Commission of the German Medical Association (DC), the Federal Institute for Drugs and Medical Devices (BfArM—the national drug regulatory authority), the pharmaceutical manufacturer directly or via the sales representative, the hospital pharmacy and the “Arzneitelegramm”—a privately owned, independent monthly drug bulletin that runs a so-called “network for drug safety.” Almost two-thirds of the physicians (61.3%) of the random sample declared to have reported an adverse event/adverse drug reaction at least once in their professional life, 37.4% within the last 5 years, 15.8% at least 5 or more years ago, and 8.3% in both time
Table 1 Demographic profile of the samples
Age, years x, s Sex female male Specialist degree GP Internal medicine Pediatrics (Primarily) office-based Hospital-based Other fields Western states Eastern states Response rate
Random sample (n 230)
Drug Commission (DC) sample (n 358) P-value
48 9.3
41.2 8.7
P .05a
62 (27%) 168 (73%) 172 (74.8%) 12.2% 18.7% 10.8% 157 (71.1%0 62 (27.7%) 5 (2.2%) 187 (81.3%) 43 (18.7%) 51.4%
102 (28.8%) 252 (71.2%) 204 (57%) 25.9% 28.4% 11.9% 203 (56.9%) 140 (39.9%) 8 (2.3%) 291 (89%) 36 (11%) 43.9%)
P .05b P .05b
P .05b P .05b P .05b
a
t-test. chi2-test.
b
periods. Pharmaceutical manufacturers and their sales representatives, who are legally obliged to collect and to forward ADR reports, were the two most important addressees of ADR reports. About 35% of the physicians of both samples said to have reported at least once to sales representatives. Hospital pharmacies share a similar legal obligation, and about 14% of the physicians have used this service; 87.4% of the physicians of the random sample admitted that they never had reported to the drug commission of their own professional board (Table 2). 3.2. Knowledge and attitudes An amount of 68.2% of the physicians indicated to have already suspected an adverse drug reaction without reporting it. The three major reasons for not reporting were (Table 3): adverse drug reaction already well known (75.6%), adverse drug reaction too trivial to report (71.1%), and uncertainty concerning definite causality (66.3%). Almost 20% of the physicians of the random sample admitted to be unaware of the existence of a national reporting scheme, about 30% declared to know neither how to report nor the respective official regulations of conduct. With less than 1%, legal concerns were extremely rarely mentioned as reason for not reporting. An adverse drug reaction that is due to a drug that had been prescribed by another physician or is due to a drug taken in self-medication is also less likely to be reported. To find out whether there are special characteristics of a suspected ADR that make a report more likely, short case scenarios were presented permuting three contrasts: serious ADR—nonserious ADR, new ADR—well-known ADR, and adverse reaction of a new drug—ADR of an established drug. The adverse drug reactions with the highest probability of being reported were serious unknown adverse reac-
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
947
Table 2 Reporting history Random sample (n 230) (primarily) office-based Hospital-based Western states Eastern states Never reported reported last five years rep. more than 5 years ago reported both time periods Drug commission sample (n 358) Office-based Hospital-based Western states Eastern states Never reported reported last 5 years rep. more than 5 years ago reported both time periods
Drug authority
Drug commission
Pharmac. manufact.
Sales reps
Drug bulletin
Hospital pharmacies
21 (77.8%) 6 (22.2%) 22 (78.6%) 6 (21.4%) 202 (87.8%) 13 (5.7%) 13 (5.7%) 2 (0.9%)
19 (67.9%) 9 (32.1%) 29 (100%) 0 (0%) 201 (87.4%) 15 (6.5%) 13 (5.7%) 1 (0.4%)
55 (74.3%) 19 (25.7%) 64 (83.1%) 13 (16.9%) 153 (66.5%) 47 (20.4% 25 (10.9%) 5 (2.2%)
59 (72%) 23 (28%) 68 (78.2%) 19 (21.8%) 143 (62.2%) 66 (28.7%) 15 (6.5%) 6 (2.6%)
8 (72.7%) 3 (27.3%) 11 (100%) 0 (0%) 219 (95.2%) 8 (3.5%) 3 (1.3%) 0 (0%)
16 (53.3%) 14 (46.7%) 26 (83.9%) 5 (16.1%) 199 (86.5%) 18 (7.8%) 13 (5.7%) 0 (0%)
49 (62.8%) 29 (37.2%) 68 (85%) 12 (15%) 273 (76.3%) 69 (19.3%) 6 (1.7%) 10 (2.8%)
182 (60.9%) 117 (39.1%) 251 (88.7%) 32 (11.3%) 0 (0%) 323 (90.2%) 5 (1.4%) 30 (8.4%)
93 (62.8%) 55 (37.2%) 125 (87.4%) 18 (12.6%) 200 (55.9%) 134 (37.4%) 17 (4.7%) 7 (2.0%)
91 (77.1%) 27 (22.9%) 92 (85.2%) 16 (14.8%) 237 (66.2%) 106 (29.6%) 11 (3.1%) 4 (1.1%)
41 (80.4%) 10 (19.6%) 46 (97.9%) 1 (2.1%) 307 (85.8%) 34 (9.5%) 5 (1.4%) 12 (3.4%)
30 (66.7%) 15 (33.3%) 33 (82.5%) 7 (17.5%) 312 (87.2%) 38 (10.6%) 5 (1.4%) 3 (0.8%)
The percentages are calculated per section, e.g., 21 (77.8%) in the first column means that 77.8% of the responders reporting to the drug authority were office based; 202 (87.8%) means that 87.8% of all responders had never reported to the drug authority.
tions of a new drug, which would be reported by 81.1% of the physicians and serious unknown adverse reactions of an established drug (72.9%), followed by a serious known adverse reaction to a new drug (65.2%). Less than 10% of the physicians would report nonserious adverse reactions to established drugs.
4. Further results Textbooks, summaries of product characteristics, and patient information leaflets quite often use verbal terms to describe the incidence of adverse drug reactions. In Germany, the official terms are: common, occasional, rare, and single cases or very rare. The questionnaire asked the physicians to link these verbal terms with incidence estimates. The results show considerable variability regarding physicians’ interpretation of the official terminology: 68.9% of the physicians think that the category very rare means an incidence of one per 100,000, the correct answer for Germany—less than one per 1000—was selected by only 8.2% of the physicians (Table 4). The results indicate that the official terminology differs considerably from the common interpretation of these frequency terms. The final part of the questionnaire looked for starting points to improve the reporting behavior. About 54% asked for the possibility to report by phone to a qualified drug information service, and 43% stated that their motivation to report could be increased if their PC offered the option to print out the official report form. The chance to report by e-mail was positively considered by 12.4%. There were no statistically significant differences between the physicians’ attitudes in West and East Germany.
5. Discussion 5.1. Representativeness, reporting history, and validity Although the spontaneous reporting system in Germany is about 40 years old, this is the first survey concerning the knowledge and attitudes towards this system. Thanks to a couple of reminders a response rate of about 50% could be achieved in the random sample. Most probably one is well advised to assume that the nonresponders were mostly nonreporters. Thus, the generalizability of the results of this survey is limited. According to official data there are about 280,000 physicians in Germany, of whom 40% are office based, 49% are working in a hospital, and 11% in other areas like research and administration [18]. About 36% of the physicians are feTable 3 Percentage of respondents giving the stated reason for not having reported a suspected adverse drug reaction
ADR too well known* ADR too trivial Uncertain causality Reporting too bureaucratic* Lack of time* Resp. Rules of conduct unknown* Susp. drug prescribed by colleague* Reporting process unknown* Lack of financial reimbursement* Susp. drug was self-medication* Reports considered useless* Spont. reporting system unknown* Fear of legal liability *P .05, chi2-test.
Drug Commission sample
Random sample
80.8 70.8 63.5 27.3 33.8 18.8 8.3 4.6 13.1 5.5 7.7 4.6 0.8
65.9 71.7 71.7 39.9 22.5 29.0 28.0 27.5 6.5 20.8 16.7 18.1 0.7
948
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
Table 4 Interpretation of the official terminology for risk quantification by physicians (%) (random and DC sample combined, n 588), e.g., 53.4% of the physicians think that a rare ADR occurs in 1 in 10,000 patients
Official terminology
Common Occasional Rare Very rare ADR ADR ADR ADR 1:10 1:100–1:10 1:1000–1:100 1:1000
Physicians’ incidence estimate 1/10 patients 31.6 1/100 patients 24.0 1/1,000 patients 5.9 1/10,000 patients 2.0 1/100,000 patients 0 35.9 Missinga
1.0 20.7 42.2 14.5 4.8 16.8
0.7 2.0 18.3 53.4 11.2 14.4
0.7 0.2 1.0 8.2 68.9 21.1
a
Refers to missing responses (%) per frequency term.
male, 85% live in the Western and 15% in the Eastern states of Germany. Sixty-six percent of the physicians have a specialist degree. Relating these data to both samples studied, it seems that male office-based specialists are over-represented. The proportion of physicians from the new Eastern states in Germany in the DC sample is fairly low, and indicates that they are less accustomed to report to the Drug Commission than their Western colleagues (P .05, 2-test). Almost 25% of the physicians declared to have never diagnosed an adverse drug reaction. This result indicates that a considerable part of the physicians has major problems to diagnose an ADR. Sixty-one and three-tenths percent state that they have already reported at least one suspected ADR. This percentage is somewhat higher than the average of 52% reported recently for Europe, with a range from 19.4% in Italy up to 74.4% in France [15,19,20]. There are, however, some doubts about the validity of these self-reported activities. If 6.5% of the approximately 250,000 physicians who treat patients in Germany had actually reported at least once to the Drug Commission within the last 5 years (Table 2), about 16,250 reports should have been received. But there were only about 9,000 reports in this period [21]. Such attempts to crossvalidate survey data are missing in previous reports [15,19]. Whereas in most European countries alleged adverse drug reactions are mainly reported to national or regional centers [15], in Germany the pharmaceutical manufacturers and their sales representatives are the most common recipients of reports. We assume that most “reports” to sales representatives were actually not formal written reports, but oral reports during an informal conversation. The German drug law states that sales representatives must record and forward any report of a suspected ADR. Comparably few physicians report directly to the German drug authority, the BfArM. The rather low proportion (0– 8.4%) of physicians of both samples who stated to have reported in both predefined time periods indicates that none of the institutions managed satisfactorily to establish a stable relationship with their reporters.
5.2. Reasons for under-reporting One of the major objectives of our study was to find out the extent of and the reasons for under-reporting. The most common explanation for not reporting was that the ADR was already well known, a finding that confirms results of previous studies [7,8,10,12,14,22]. Apparently it is largely unknown that the spontaneous reporting system could also be used for monitoring changes of the type and/or frequency of serious known ADRs. This can only be achieved, if already known serious ADR are reported as well. From a physician’s point of view, however, it may seem “natural” and cost-efficient to not report even serious ADRs if well known, as there is no financial compensation and a good case report takes some time. About two-thirds of the physicians stated that they did not file a report because they were in doubt about causality. This finding was unexpected, because the official reporting form explicitly states in its heading that suspected cases of adverse drug reactions should also be reported. Twenty percent of the physicians of the random sample did not know the spontaneous reporting system at all, and almost 30% had no idea how to report. In countries such as the United Kingdom, Denmark, and Sweden, only 2.7, 3.4, and 8.6% of the physicians, respectively, were unsure how to report [15]; 86.7% of the physicians stated to not know any criteria concerning ADR/adverse events that should be reported, although the drug commission regularly publishes such criteria in the German Medical Journal, which is mailed weekly to every physician in Germany. These results indicate that the efforts of the DC in the past to make the physicians familiar with the spontaneous reporting system have failed to a relevant extent. New marketing concepts are thus needed and have to be evaluated. 5.3. Starting points for improvements The questionnaire also asked for incentives to improve motivation for reporting and for techniques to make reporting easier. As in other studies, the possibility to report by phone was mentioned by a majority of the physicians [22,23]. By offering the option to report by phone the number of reports could be increased fourfold in a project in Maryland [14]. This option could be combined with a toll-free telephone number. The finding that about 50% of the physicians asked for the possibility to report by fax indicates again that too many physicians do not really know how to report. Reporting by fax has been possible for many years, and the respective fax number is printed on the official report form, which is printed on the last inner page of the Deutsche Ärzteblatt (the weekly German Medical Journal) every second week. An additional starting point for the improvement of the spontaneous reporting system is the instant availability of reporting forms [5]. As most offices in Germany nowadays use PC-based programs for prescribing, a standard tool con-
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
taining the report form with an option for printout was mentioned by more than 40%. The DC has recently developed such a program that copies relevant information like demographic data, underlying diseases, and medication data automatically if these data are available electronically. Finally, a qualified therapy information service might help. Experiences with the French decentralized pharmacovigilance system show that about half of the ADR reports are derived from inquiries for therapeutic advice. Our data suggest that especially physicians, who have never reported up to now, might report in the future if such services are provided. This survey’s focus was to identify reasons why physicians do not report. Biriell and Edwards published an interesting approach insofar as they tried to explore the reasons of physicians and pharmacists who had taken time to report alleged ADRs [24]. This approach is particularly valuable to find starting points for establishing a positive relationship between reporter and center and thus to increase the proportion of physicians who will report more than once in their lifetime. The reasons for reporting mentioned most often in their survey [24]: motivation to contribute to medical knowledge, reaction previously unknown, reaction to new drug, evident association between drug and reaction, and severity of reaction, coincide fairly well with our findings. They propose active personal encouragement, recognition, and feedback from the national center for those submitting reports.
6. Conclusions In conclusion, the survey generated data about the attitudes of physicians towards the spontaneous reporting system and provided starting points to improve the current spontaneous reporting system. Considering that many doctors have with little credit or appreciation reported their personal clinical suspicions, the reporting centers should develop and adopt a service-oriented approach, for example, by providing qualified and confidential therapeutic advice. Certainly, every innovation has to be evaluated. There is no doubt, however, that any acceptable effort is needed to strengthen the current system, as the spontaneous reporting system is the only population-wide approach for signal generation of new, previously unknown serious ADRs of new drugs.
Acknowledgments This study was supported, by a grant to the European Pharmacovigilance Research Group from the Commission of European Communities under the Biomed Program (Contract number: PL 92 1556). We thank K. Belton and M. Rawlins for providing the EPRG Attitudinal Survey II questionnaire and all physicians who were willing to respond. We acknowledge the help of S. Schneeweiss and I. GarcíaGomez in organizing the survey. We appreciate the valuable comments of the reviewers.
949
References [1] Muehlberger N, Schneeweiss S, Hasford J. Adverse drug reaction monitoring—cost and benefit considerations part I: frequency of adverse drug reactions causing hospital admissions. Pharmacoepidemiol Drug Safety 1997;6(Suppl 3):S71–7. [2] Lazarou J, Pomeranz BH, Corey PH. Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. JAMA 1998;279(15):1200–5. [3] Goettler M, Schneeweiss S, Hasford J. Adverse drug reactions monitoring—cost and benefit considerations part II: cost and preventability of adverse drug reactions leading to hospital admission. Pharmacoepidemiol Drug Safety 1997;6(Suppl 3):S79–90. [4] Johnson JA, Bootnman JL. Drug-related morbidity and mortality. A cost-of-illness model. Arch Intern Med 1995;155:1949–56. [5] Scott HD, Thacher-Renshaw A, Rosenbaum SE, Waters WJ Jr, Green M, Andrews LG, Faich GA. Physician reporting of adverse drug reactions: results of the Rhode Island drug reaction reporting project. JAMA 1990;263:1785–8. [6] Beach JE, Faich GA, Bormel FG, Sasinowski FJ. Black box warnings in prescription drug labeling: results of a survey of 206 drugs. Food Drug Law J 1998;53:403–11. [7] Walker SR, Lumley CE. The attitudes of general practitioners to the monitoring and reporting of adverse drug reactions. Pharmaceut Med 1986;1:195–203. [8] Egberts TCG, Smulders M, de Koning FHP, Meyboom RH, Leufkens HG. Can adverse drug reactions be detected earlier? A comparison of reports by patients and professionals. BMJ 1996;313:530–1. [9] Martin RM, Kapoor KV, Wilton LV, Mann RD. Underreporting of suspected adverse drug reactions to newly marketed (“black triangle”) drugs in general practice: observational study. BMJ 1998;317: 119–20. [10] Smith CC, Bennett PM, Pearce HM, Harrison PI, Reynolds DJ, Aronson JK, Grahame-Smith DG. Adverse drug reactions in a hospital general medical unit meriting notification to the Committee on Safety of Medicines. Br J Clin Pharmacol 1996;42(4):423–9. [11] Moride Y, Haramburu F, Requejo AA, Begaud B. Under-reporting of adverse drug reactions in general practice. Br J Clin Pharmacol 1997; 43(2):177–81. [12] Lumley CE, Walker SR, Hall GC, Staunton N, Grob PR. The underreporting of adverse drug reactions seen in general practice. Pharmaceut Med 1986;1:205–12. [13] Koch-Weser J, Sidel VW, Sweet RH, Kanarek P, Eaton AE. Factors determining physician reporting of adverse drug reactions. Comparison of 2000 spontaneous reports with surveillance studies at the Massachusetts General Hospital. N Engl J Med 1969;280(1):20–6. [14] Rogers AS, Israel E, Smith CR, Levine D, McBean AM, Valente C, Faich G. Physician knowledge, attitudes and behavior related to reporting adverse drug events. Arch Intern Med 1988;148:1596–600. [15] Belton KJ, and The European Pharmacovigilance Research Group. Attitude survey of adverse drug-reaction reporting by health care professionals across the European union. Eur J Clin Pharmacol 1997;52: 423–7. [16] McGettigan P, Feely J. Adverse drug reaction reporting: opinions and attitudes of medical practitioners in Ireland. Pharmacoepidemiol Drug Safety 1995;4:355–8. [17] Bundesaerztekammer. Berufsordnung für die deutschen Ärzte. Deutsches Ärzteblatt 1988;85:3601. [18] Der Bundesminister fuer Gesundheit. Daten des Gesundheitswesens. Ausgabe. Baden: Nomos Verlagsgesellschaft Baden; 1997. [19] Pierfitte C, Hanesse B, Royer RJ. Enquete d’opinion sur la pharmacovigilance Européenne: le sentiment des médecins franais. Therapie 1995;50:171–2. [20] Cosentino M, Leoni O, Banfi F, Lecchini S, Frigo G. Attitudes to adverse drug reaction reporting by medical practitioners in a Northern Italian district. Pharmacol Res 1997;35(2):85–8. [21] Munter K-H. Bearbeitung von UAWs in der Arzneimittelkommission
950
J. Hasford et al. / Journal of Clinical Epidemiology 55 (2002) 945–950
der deutschen Ärzteschaft. In: Müller-Oerlinghausen B, Lasek R, Düppenbecker H, Munter K-H, editors. Handbuch der unerwünschten Arzneimittelwirkungen. München-Jena: Urban & Fischer; 1999. p. 571–6. [22] Fincham JE. Pilot projects to stimulate adverse drug reaction reporting. J Clin Pharm Ther 1987;12:243–7.
[23] Robins AH, Weir M, Biersteker EM. Attitudes to adverse drug reactions and their reporting among medical practitioners. S Afr Med J 1987;72(2):131–4. [24] Biriell C, Edwards RI. Reasons for reporting adverse drug reactions—some thoughts based on an international review. Pharmacoepidemiol Drug Safety 1997;6:21–6.