Results: According to selection criteria, 54 546 reports were included in the analysis

(1)

patients with at least one ADR reported under the following WHOART preferred terms: ‘‘parosmia’’, ‘‘taste loss’’, ‘‘taste alterations’’. The non-cases comprised all remaining patients. Index reports included ADR associated with drug classes (third level ATC) with at least 5 reports of taste and/or smell alterations, while all ADR reports not involving index drugs were used as controls.

Results: According to selection criteria, 54 546 reports were included in the analysis. A total of 197 cases were identified (mean age:

56.17– 14.83; females: 66.50%). Taste alterations were reported in 176 patients (89.3%) and smell disturbances in 47 patients (23.9%). Drug classes most frequently reported with taste and/or smell alterations were macrolides (30 patients, 16.5%; ROR: 7.6; 95% CI: 5.1, 11.2), antifungals for dermatologic use (all terbinafine; 17 patients; 9.3%;

ROR: 64.8; 95% CI: 41.2, 101.7), fluoroquinolones (15 patients, 8.2%;

ROR: 1.7; 95% CI: 1.1, 3.0) and protein-kinase inhibitors (10 patients;

5.5%; ROR: 4.1; 95% CI: 2.1, 7.7). The outcomes for these ADR were

‘‘resolved’’ (76 reports; 38.6%), ‘‘not yet resolved’’ (44 reports; 22.3%),

‘‘resolving’’ (20 reports; 10.1%) and ‘‘resolved with sequelae’’ (12 reports; 6.1%), while 45 reports (22.8%) were recorded as ‘‘not available outcome’’.

Conclusions: Significant ROR for taste and/or smell alterations was found for macrolides, antifungals for dermatologic use (terbinafine), fluoroquinolones and protein-kinase inhibitors. Taste impairments were generally expected, while smell disturbances were unexpected or expected only for some drugs within each class. High rates of unresolved outcomes may suggest that some of these ADR may be permanent or associated with long-term recovery. Further investiga- tions are warranted to validate these signals and their impact on patient quality of life.

Reference

1. Doty RL, Bromley SM. Effects of drugs on olfaction and taste.

Otolaryngol Clin North Am 2004; 37: 1229-54

46. Physician’s Awareness of Safety Information on the Risk of Cardiovalvulopathy Related to Pergolide Use in Japan

K. Tsuruta, N. Kikuchi, A. Kokan, A. Kudo, S. Nishiuma and F. Sakakibara

Global Patient Safety Japan, Eli Lilly Japan K.K., Kobe, Japan Background: Several studies have suggested a high frequency of cardiovalvulopathy in Parkinson’s disease (PD) patients treated with ergot derived dopamine agonists, including pergolide. In order to mitigate the risk of valvulopathy in patients treated with pergolide and to optimize the safe use of pergolide in patients with PD, a revised package insert was developed in 2007, warning prescribing physicians of the risk of new-onset valvulopathy with pergolide. The Contra- indications and Important Precautions Section of the package insert in Japan were modified.

Aim: To assess physician’s awareness of the changes made to the package insert of pergolide as well as their understanding of the wording and their adherence to the package insert changes which include:

mandatory echocardiograms for all patients upon initiation of pergolide treatment and regularly throughout the treatment. This survey will determine the effectiveness of the risk minimization activities as im- plemented by neurologists in their daily clinical practice, and also act as a reminder to neurologists that a change in practice may be needed.

Methods: We surveyed neurologists to determine their awareness on safety information on the risk of cardiovalulopathy related to pergolide use in Mar, 2009. Self-administered questionnaires were mailed to a neurologist population (n= 2374) pooled in a commercial database

for physician survey (EPOCA, Osaka, Japan), and collected from them during the survey period of two weeks.

Results: Response rate was 14.3% (n = 339). Overall awareness of the package insert change for pergolide was 98%. Most neurologists performed echocardiogram at baseline for patients who started treatment with pergolide/cabergoline (77%) and during treatment with pergolide/

cabergoline (70%) according to the package insert.

Conclusions: Most neurologists who responded to this survey were aware of the package insert change of pergolide and indicated that they would perform echocardiograms prior to starting treatment, as well as during treatment. Although the response rate was low, this survey showed that neurologists were aware of the changes made to the package insert in relation to the valvulopathy risk of pergolide.

47. Bias in Signal Detection D. Arora

Wockhardt Limited, Mumbai, India

The tendency or preference towards a particular perspective especially when it adversely affects the results is known as bias. Like any other activity, the process of signal detection may be biased and can result in a systematic, non-random deviation of results from truth. Sources of bias in signal detection include the limitations of spontaneous reporting, like incomplete reporting, duplicate reports; incorrect causality assessment and the methodological errors. In addition, several external factors like media attention, privacy and data protection issues, law suits, marketing and financial pressures on the marketing authoriza- tion holders can also influence signal detection. Further, methodological errors like inappropriate selection of the comparator drug, incorrect stratification and inadequate use of MedDRA for database search while developing case series can adversely influence the process of signal detection and the life cycle management of a medicinal product. Detailed descriptions of various types of biases, viz. Notoriety bias, Protopathic bias, Channeling, 2-peak distribution, Weber effect etc are available in literature^[1]and there are case studies describing regulatory actions (and amendments thereafter) resulting from such biases.^[2,3]

Therefore, the possible sources of bias in signal detection, their poten- tial impact on signal detection and the feasible solutions should be considered and appropriately addressed while developing the Risk Management Plan. The presentation will discuss various sources of bias, their effect on life cycle management and handling bias.

References

1. Stang P. Epidemiological Context of Signalling. Drug Saf 2007; 30 (7): 611-3 2. Moore N, Hall G, Sturkenboom M, et al. Biases affecting the proportional reporting ratio (PRR) in spontaneous reports pharmacovigilance databases: the example of sertindole. Pharmacoepidemiol Drug Saf 2003; 12: 271-81

3. Pariente A, Gregoire F, Fourrier-Reglat A, et al. Impact of Safety Alerts on Measures of Disproportionality in Spontaneous Reporting Databases.

The Notoriety Bias. Drug Saf 2007; 30 (10): 891-8

48. Quality Management System as a Tool for

Improvement of the Dutch Pharmacovigilance System P.G.M.A. Zweers, E.P. van Puijenbroek and A.C. van

Grootheest

Netherlands Pharmacovigilance Centre Lareb,

‘s-Hertogenbosch, the Netherlands

Introduction: The Netherlands Pharmacovigilance centre Lareb started as a regional organization in 1985. In 1995 it was appointed by the 896

(2)

Health Authorities as the national centre for collection and analysis for reports of adverse drug reactions. Since then, Lareb has become a solid, professional organization with an efficient way of handling reports of adverse drug reactions and signal detection. Procedures covering these primary processes of pharmacoviglilance in Lareb have been developed during the past years. In 2008 a start was made with the introduction of a quality management system according to the ISO (International Organization for Standardization) standards.^[1,2] In March 2009 Lareb achieved the ISO 9001-2000 certificate.

Aim: The aim is to describe the experience with the implementation of a quality management system and the influence hereupon the pharmacovigilance system in Lareb.

Methods: Description of the process of achieving the ISO 9001–2000 certificate. An analysis of the main aspects influencing the pharmacovigilance processes is performed.

Results: The organisation responsible for certification according to ISO standards in ‘knowledge-driven’ associations, investigated Lareb for the status of main topics in the ISO 9001–2000 requirements, including policy, organization, processes, results, employees and cooperation/automation. Within all these aspects, the basic principle of Plan-Do-Check-Act should be visible. Following this, several actions had to be undertaken to prepare for the final certification in- spection. One of the main issues was to formalize all processes in a quality management manual covering all requested topics. Beside this, a process for internal auditing had to be commenced. The audits of the primary processes resulted in a welcome improvement in the perfor- mance of the pharmacovigilance processes. Finally a (bi)annual management review had to be undertaken, which ended up in a better view of the necessary steps for efficient planning of activities and resources.

Conclusions: The implementation of a quality management system shows to be a valuable and supplementary system to the already existing procedures and instructions regarding the collection and analysis of reports of possible adverse drug reactions. The continuous process of improvement warrants an efficient method for pharmacovigilance activities.

References 1. http://iso.ch

2. Quality management systems. Requirements ISO 9001:2000. Comite´

Europe´en de Normalisation, 2000

49. Report Processing and Signal Management at the Netherlands Pharmacovigilance Centre E.P. van Puijenbroek, L. Ha¨rmark and A.C. van Grootheest Netherlands Pharmacovigilance Centre Lareb,

the Netherlands

Introduction: The fast developments in pharmacovigilance in the past decades have had far-reaching consequences for the way of working in national pharmacovigilance centers. Firstly, processing and distribution of adverse drug reaction reports on both national and international levels became more complicated and required an adequate working flow of the reports and quality control to meet all regulatory requirements and technical standards. Secondly, the application of quantitative approaches in signal detection enabled the detection of large numbers of possible signals, still needed to be evaluated from a clinical perspective. This requires an adequate prioritisation and efficient diffusion of the signals involved. Both developments ask for a more efficient way of report processing and signal management. In the past decades the Netherlands Pharmacovigilance Centre Lareb developed a way of working to improve report processing and signal detection to meet the current needs.

Aim: The aim of this presentation is to describe the various work flows concerning report processing and signal detection. In addition, steps taken to increase the efficiency, reliability and transparency of the system will be discussed.

Description: The infrastructure of the Netherlands Pharmacovigilance centre consists of four main elements. In the first place a computerised working flow for the coding, assessment, filing and distributing the ADR reports in the database is available. Quality checks and systems are in place to ensure an effective processing of the reports. Secondly, information from other sources like literature, other databases and analysis which are previously carried out, are made automatically available for the assessors to reduce the workload. In addition, an automated signal management system ensures a transparent signal detection and handling of the publication process. Finally, a dedicated website presents the output in a transparent and automated way, both in respect to the individual case reports as for signals issued. Handling of the reports and signal detection are closely interrelated. Actions needed for processing reports are also the basis in the first steps of signal detection and diffusion of the signals, thus increasing the efficiency of the system.

Conclusion: The growing number of reports and signals are a challenge for many pharmacovigilance centers and asks for tuning of these various processes in which information technology plays a pivotal role.

In order to ensure a high quality of both reports and signals regulatory requirements should be realistic and easy implementable in the current routines. Transparent processes will enhance the acceptability of our work by the general public.

50. Hepatic Adverse Events Associated with the Use of Herbal Drugs in Norway (‘‘the Fortodol Case’’) C.S. Nerga˚rd and K. Myhr

RELIS Medicines Information Centres, Oslo University Hospital – Ulleva˚ l, N-0407 Oslo, Norway

Introduction: In Norway the majority of herbal products are legally sold without applying for a marketing authorisation, and thus without the products being evaluated for quality, safety and efficacy. The reporting of adverse effects from health care personnel is spontaneous and not mandatory for herbals. Only about 30 adverse events associated with herbal drugs are reported annually in Norway. In spite of the limitations of the reporting system for herbals, we illustrate how it has worked in ‘‘the Fortodol case’’. The company (Bringwell) declared that Fortodol sold in Norway as a ‘‘natural’’ analgesic contained a turmeric extract (Curcuma longa).

Case reports: RELIS received the first report on Fortodol in 2007.

A 47-year old female was diagnosed with hepatitis after use of 1–2 capsules per day for approx. 2–3 weeks. Several weeks after withdrawal her liver enzymes turned to normal. In 2008 RELIS received 4 more reports on Fortodol associated with hepatic adverse events. Symptoms developed from 2 weeks to 6 months after start. In three of four cases the patients also used other drugs associated with increase in liver enzymes. In 2 cases liver tests normalised after withdrawal of Fortodol, one patient was recovering at the time of reporting and the last patient died. In the same period the Swedish Medical Products Agency received 4 reports of liver toxicity, including one fatal case. An analysis of the Fortodol capsules used by the patient who died in Norway showed an undeclared content of 42 mg nimesulid per capsule.

Discussion and conclusion: Seven million capsules of Fortodol has been sold in Norway from 2007 to withdrawal of the product from the Norwegian market on 25 February 2009, and 7 of 9 lots have contained nimesulide.^[1]Nimesulide is not authorised in Norway and is

Abstracts 897