GCP Auditing - Methods and Experiences

Systems Audits in Pharmacovigilance Michaela Rittberger, Michael Stegemann, and Jürgen-Hans Schmidt In the light of the changing legal situation for pharmacovigilance, the con- duct of an audit in this area has an increased impact on assuring and / or improving the quality of the processing of pharmacovigilance data as part of the respective pharmacovigilance system. This is especially true as the pharmacovigilance process becomes increas- ingly globalised and complex, while the necessity to meet timelines and the need for effectiveness remains as important as ever. This chapter describes the performance of a pharmacovigilance audit, from the audit planning to writing an audit report. It gives an overview of several aspects of audit conduct. Moreover, experiences with inspections and com- mon audit and inspection findings are described. However, this chapter should not be understood as a comprehensive working instruction and not be regarded to reflect all possibilities and aspects of a pharmacovigilance audit. 1. Introduction The purpose of pharmacovigilance (PhV) is to provide accurate, complete and timely information about the benefit-risk profile of medicinal products throughout their life cycle to interested parties (e.g. company management, trial subjects, consumers, healthcare professionals, authorities). Therefore, pharmaceutical companies need to maintain a system that en- sures An active, systematic and continuous monitoring of pharmacovigilance cases received. Early detection of any regular safety signals or irregular crisis situations. That appropriate actions are taken to minimise risks and prevent harm to the company and customers. The quick and effective identification and evaluation of pharmacovigilance signals is dependent on early access to complete information. This is essen- tial for the Marketing Authorisation Holders (MAHs) as well as Marketing Authorisation Applicants (MAAs) and competent authorities to be able to protect public health in taking appropriate actions swiftly. Signal detection should identify this problem at an early stage. 164 Principles of signal detection in pharmacovigilance: “... Although pharma- covigilance is especially concerned with adverse effects a signal is more broadly defined as a set of data constituting a hypothesis that is relevant to the rational and safe use of a drug in humans ...” [1]. Procedures for the detection of safety signals, based on pharmacovigilance data, should be complemented by a risk management system. A risk management system should be understood “as a set of pharmacovig- ilance activities and interventions designed to identify, characterise, prevent or minimise risks relating to medicinal products, and the assessment of the effectiveness of those interventions.” [2] At the time of this publication much relevant legislation and many guidance documents had been changed as well as developed, and a lot of work is still in progress to this end. As a consequence of the rapidly changing legal situation, several points need further clarification and not all questions have been answered until now. However, one result of the changes in legislation and guidance is that more EU member states are implementing routine pharmacovigilance inspec- tion programmes. There is, therefore, an increased focus on pharmacovigilance activities, which result in an increased number of PhV inspections. These focus on whether pharmaceutical companies maintain a reliable pharmacovigilance system fit for the purpose mentioned above. To meet the increasing legal and business requirements, systems audits in pharmacovigilance can evaluate whether the system is designed to ensure that activities applied therein are performed in compliance with applicable requirements. Failures, areas of improvement and the need to initiate cor- rective and preventive actions can be identified. In addition, systems audits can encourage a continuous improvement pro- cess as understood by ISO 9004:2000 [3]. This publication focuses on the performance of systems audits in pharma- covigilance for medicinal products in human use concerning the pre- and post-authorisation phases. In this context experience already gained in common inspections and audit findings will be presented. 2. Audit methodology 2.1. Objectives of a pharmacovigilance systems audit The following five audit objectives are proposed as being systematically applicable to pharmacovigilance units to independently confirm that: the organisational unit responsible for pharmacovigilance is capable of performing all related tasks. interfaces, e.g. to product quality, regulatory affairs, marketing, are regulated and functioning. 165 the processes established are performed in accordance with external and internal quality requirements. the safety data received and processed in the database(s) are consistent with the source documents and are evaluated and timely reported to all relevant parties. there is a quality management system in place supporting continual im- provement of the pharmacovigilance system. 2.2. Audit planning In order to be customer-oriented and to set up an optimal operational course for the audit, the planning should be done in close cooperation with the auditees and the responsible line management (called process owner in the following). If the individual audit of a pharmacovigilance unit is part of a systems audit programme, the audit planning would be based on the latter. In this programme, the focus of the individual audit would already be determined. For the audit planning the following basic elements need to be clarified or done first: The audit setting (organisational unit, auditee, process owner, locations, duration of the audit, staff to be contacted, auditor with contact details). Listing objectives, scope and topics of the audit. Defining applicable external and internal quality standards. Selecting elements of the quality management system (e.g. quality man- ual, job descriptions, resource management, training activities, Standard Operating Procedures (SOPs) management, facilities). Selecting processes (specifying how the functionality will be evaluated: either by checking the quality of the process outputs, e.g. the individual cases or safety reports, or rather by reviewing the process flows for ad- herence to the SOPs and for efficiency). Existing quantitative quality parameters for corresponding processes (e.g. a maximum acceptable number for late cases is defined); determin- ing them, if applicable. For the review of case processing, defining an observation period and the sampling approach. If applicable, check if special safety monitoring programmes are imple- mented. Involvement of pharmacovigilance interfaces (e.g., regulatory affairs, pharmaceutical production, marketing, or clinical development; their agreement on audit support is to be obtained first). Outlining the audit report flow, option of a draft report, and timelines for reporting. When selecting an auditor or audit team, adequate qualification of the audit staff should be confirmed. Auditors should be familiar with interna- tional and applicable local regulatory requirements on pharmacovigilance, 166 have been trained on such audits preferably in co-audits, be experienced and familiar with the systems audit approach (e.g. in terms of DIN ISO 9001 norms on quality management system). 2.3. Audit preparation Logistical and content aspects of the audit would be worked out in more detail in this phase. The duration of the audit depends on the scope of the audit, number of audit staff, location(s) to be visited, availability of staff, types and number of safety cases to be reviewed. An audit agenda can structure all this in- formation. It is crucial to obtain all necessary information about the pharmacovigil- ance unit to prepare the audit effectively. The following documents (copies) should be requested: Summary of pharmacovigilance system (SPS). Organisational charts. Further important quality documents, e.g. specific policies, SOP index, selected SOPs, quality monitoring plans, e.g. with Contract Research Or- ganisations (CROs). List of investigational/marketed products included in the safety surveil- lance. Survey on co-development and/or co-marketing (license partners) affect- ing safety reporting procedures. Description database(s) used (requesting read access) during the audit. Status of safety cases processed in the defined audit observation period. Familiarisation with process descriptions is mandatory to understand how the processes are organised in detail and how to derive audit questions. Due to the usually high number of cases processed, the sample size for cases to be audited needs to be calculated. If quantitative acceptance values were, for example, specified for the case processing time lines, these limits − so-called Acceptance Quality Limits − would be used as an audit reference. If possible, an electronic export of relevant data might be requested from the pharmacovigilance database. Such requests can be made quite early to allow for a pre-check of data and of any technical adjustments, if necessary (compatible data format). The preparation phase would be completed by an official announcement letter to the auditees, providing all information on audit timing, location, required availability of staff, documents, database access, work space, etc. The audit agenda would be attached to this letter. 2.4. Audit conduct The conduct of pharmacovigilance audits is very similar to that of other systems or process audits.