Quality of ICSRs
Call for action for marketing authorisation holders in relation to the quality of reports of adverse reactions originating from spontaneous sources and non-interventional studies submitted via EudraVigilance
The National Institute of Pharmacy and Nutrition (OGYÉI) dedicates significant effort to check the quality of adverse reaction reports received via EV from marketing authorisation holders (MAHs). In this document OGYÉI would like to draw attention to the most typical mistakes encountered in reports of the MAHs that make the interpretation of the data difficult and may falsify or thwart signal detection activities.
The most typical mistakes are listed and detailed below. MAHs are asked to always check the compliance of the ICSR with the EV requirements prior to sending, as these mistakes may not be filtered and rejected by the system (not even a warning message may be generated).
1. Differences between the content of structured fields and the narrative
It is a typical mistake that the narrative contains information that is not provided in the appropriate structured field (e.g. information regarding case history, medicinal products previously administered, adverse reactions and test results) or information provided in the narrative and in the structured fields is contradictory.
All information, when possible, should be recorded in a structured form as well, it is not sufficient to refer to it in the narrative, since filtering on adverse reaction reports in EV and signal detection based on PRR/ROR values are carried out with the use of structured fields. The reverse also holds true; narrative should have all the information contained in the structured fields. Furthermore, any information that cannot be provided in the structured fields should be present in the narrative. In the latter case, please, do not attempt to code the information (e.g. populating the medicinal product name with the therapeutic group if the active substance name is not known), since it also leads to incorrect population. Please, make sure before submission that the report does not contain discrepancies between the content of the structured fields and the narrative.
2. Indication of follow-up information
MAHs often update the narrative with new information obtained during follow-up, which are not always indicated at the end of the narrative. This is not the appropriate practice.
In case of receipt of follow-up information, the existing text of the narrative should be updated with the new information so that a detailed, still concise description is obtained that contains the totality of data and facts. Furthermore, please indicate at the end of the text the date when the follow-up information was received and specify the information with which the narrative has been updated. Structured fields need to be updated according to the new information.
3. Submission of reports without an adverse reaction
The Institute does not require reports to be submitted when no adverse reaction was experienced by the user (e.g. overdose, off-label use or medication error without an adverse reaction, drug exposure during pregnancy without any untoward consequences to the mother or foetus, etc.) These cases shall be collected by the MAHs and analysed in periodic safety update reports. Furthermore, please consider submitting cases, where only the consequences and not the exact adverse reactions are known (e.g. surgery, hospital admission, etc.). The validity of these cases can be questioned, since there is no exact information on one (adverse reaction) of the 4 minimum criteria.
4. Nullification of cases
The Institute became aware of several unsubstantiated nullifications of individual case safety reports and furthermore, recreation of such cases with new case identifiers. Both scenarios are considered serious errors and wrong practice.
MAHs are asked to study thoroughly the relevant section of GVP Module VI before nullifying any reports, and exercise due caution. As EV receives many more cases daily (by orders of magnitude), all nullification reports are likely to be accepted without a deeper check of the reason of nullification.
If one of duplicate reports needs to be nullified, the remaining ’master’ case should be updated with all relevant information from the case to be deleted (if applicable). The MAH should submit a follow-up to the remaining case with updated information simultaneously with the nullification request of its duplicate. Alternatively, the MAH should confirm that the remaining case already contains all relevant information, thus, no follow-up report is needed.
5. No causal association is identified between the reaction and the drugs administered to the patient
If neither the reporter nor the MAH suspect a causal relationship between the drugs taken by the patient and the adverse reactions suffered, the case does not qualify for reporting to the authorities. This applies to all types of reports including spontaneous cases, cases originating from non-interventional studies or from the scientific or medical literature. Although causality is usually suspected in spontaneous reports, if the reporter explicitly denies any association, supposing causality automatically without any assessment and justification by the MAH and forwarding the case to the authorities is wrong practice. Considering fatal cases automatically related is also inappropriate. Frequently, death results from the natural course/progression of the disease. MAHs are asked to exercise due caution and consider the data available while deciding on the possible causal association.
6. Reports originating from patient support programs (PSPs)
Several badly documented, poor quality reports are received by the Institute in which significant data are lacking to assess the causal relationship. These reports introduce undesirable „noise” in ADR databases without presenting any relevant information.
When initiating such programs in Hungary with an intention to collect ADRs as well, MAHs are requested well in advance to develop and guarantee all conditions of obtaining detailed adverse reaction descriptions from healthcare professionals or patients in order to submit well-documented, good quality case reports to authorities. Please avoid initial reports with only minimum information; try to obtain follow-up information as soon as possible and forward cases to authorities only thereafter, with complete information. Please ensure prior to initiating a program that possibility of obtaining medical confirmation of patient reports is available.
Furthermore, it is the responsibility of the MAH to always assess the causal relationship between the drug and the adverse reaction. If neither the reporter nor the MAH assumes a causal relationship, the case does not qualify for reporting.
7. Adverse reactions reported as spontaneous reports originating from clinical trials
The Institute is still receiving numerous cases with report type spontaneous that occurred during clinical trials and where investigational drugs can be among the suspected drugs (either blinded or placebo).
Adverse reactions originating from clinical trials falling under the scope of directive 2001/20/EC, should be reported as spontaneous adverse reactions only when a non-investigational medicinal product or other product not included in the protocol is suspected and an interaction with the investigational medicinal product* can be excluded. In these cases, the type of the report should be spontaneous. When only the investigational medicinal product or the investigational product and other medicinal products are suspected, the case should not be reported as spontaneous to EVPM. For reporting requirements in these cases, please refer to directive 2001/20/EC. For further information please refer to the Clinical Trials Division, since this department deals with suspected adverse reaction reports in relation to investigational medicinal products within the Institute.
There are several examples of cases when the investigator does not suspect an association between the investigational medicinal product and the adverse reaction, which is stated in the case narrative or sender’s comment; still, the product is entered as suspect in the structured fields due to the wrong in-house procedure of the MAH. These reports may lead to the detection of false signals. MAHs are requested to correct these errors (either a wrong procedure in place or inappropriate settings of the ADR processing software) as soon as possible, and forward corrected reports to the relevant EV module. As such incorrect reports will not be filtered out during technical validation, when the Pharmacovigilance Division detects repeated non-compliance from MAHs despite of previous call for correction; it will refer the issue to the Inspectorate.
*Investigational medicinal product (IMP): a pharmaceutical form of an active substance or placebo being tested or used as a reference in a clinical trial, including products already with a marketing authorization but used or assembled (formulated or packaged) in a way different from the authorised form, or when used for an unauthorised indication, or when used to gain further information about the authorised form
8. Processing of literature cases
An important source of information on the safety profile of drugs is the scientific and medical literature.
MAHs are requested to monitor regularly the scientific literature, paying special attention to literature published in the territory of Hungary. Literature monitoring is a requirement regarding all active substances for which the MAH holds a valid marketing authorisation in the territory of Hungary, and have de facto been launched. Following the launch of the medical literature monitoring service provided by the European Medicines Agency, starting on 1 July 2015, this monitoring and reporting obligation will not apply to Individual Case Safety Reports related to active substances and published in literature sources covered by the above service (the lists of the active substances and literature are to be found on EMA’s homepage).
No need to report cases from literature, if:
· the suspect medicinal product and the source of the concerned publication are contained in the official lists of active substances and journal/reference databases published by the European Medicines Agency on its homepage in relation to EMA’s medical literature monitoring service, including the literature sources indexed at the followed reference databases,
· the suspect drug is unequivocally identified as being a product of another MAH, e.g. brand name of the medicinal product is clearly mentioned in the article,
· the MAH did not market the product in the period to which the article refers, i.e. no patients could have been treated with the drug,
· the article refers to aggregate drug safety data** and no single patients can clearly be assigned to specific drugs or adverse reactions. Novel, significant drug safety results obtained from such articles should not be reported as individual cases, but need to be collected and discussed in the next PSUR,
· if the author of the article does not suspect any relationship between the drug administered and the occurrence of the adverse reaction.
To our understanding, case descriptions identified in the scientific and medical literature cannot automatically be regarded as spontaneous reports (with an automatic suspicion of an association), since their objective is not necessarily to discuss the adverse events and to establish the relationship with the drug use. Several times, the adverse reaction is only mentioned incidentally, and this is not the primary focus of the article. Therefore, MAHs are requested to provide an assessment of drug-reaction relatedness in all cases that has been found in the literature, if no such assessment of the author is available. If there is no reasonable association of the drug with the reaction, do not submit the case to the authorities.
If there are doubts concerning the relatedness or the article contains contradictory data on the adverse reaction observed, MAHs are recommended to contact the author of the article to clarify the details prior to the submission of the initial report.
** Examples of aggregate cases:
If several patients are mentioned in an article along with several drugs; however, it cannot be established which patient received which drug, and who suffered the adverse reaction; these cases should not be reported. E.g. 10 patients were receiving simvastatin or atorvastatin and one patient developed an adverse reaction.
Similarly, the case should not be reported if only the pharmacological/therapeutic class is known, and no single active substance is mentioned in the article. E.g. one patient out of 10 developed a serious adverse reaction while being treated with fluoroquinolones.
Further aspects of processing a literature report
Case narrative of reports originating from the scientific and medical literature should contain a clear and concise summary of the publication in English.
If more medicinal products/active substances are mentioned in the publication, only those should be structured as suspects for which the author of the publication explicitly presumes an association with the adverse reaction. If a patient took more drugs in the period relevant for the development of the ADR, and other products than that of the reporting MAH could also be associated with the reactions, all drugs should be structured as suspects and discussed in the report. Neglecting those products that might have a causal role, and focusing only on the products of the reporting MAH is wrong practice.
If the patient received more than one drug one after the other, successively, the assessment of the MAH should focus on its particular drug and the adverse reaction this drug may have elicited. Nevertheless, the complete article should be processed, and the case narrative should include all details from prior or subsequent medication, if these are relevant for the assessment of causality.
If more drug-event pairs separated in time are described referring to the same patient, these distinct ’episodes’ should be handled as separate safety reports; nevertheless, they should be linked to one other.
If more than one identifiable patient is described in a publication, separate cases should be generated for each individual patient. These individual safety reports should also be linked to each other. One report should contain only one patient and at least one patient identifier as required by EV, should be available (e.g. age at onset of the reaction, birthdate, age group, gender). The case should include data from the patient as complete as possible and relevant to the medical assessment of the case.
Primary source of a case identified from the literature is the first or corresponding author of the publication.
The appropriate provision of reference of literature cases helps considerably the filtering of duplicates, therefore marketing authorisation holders are required to provide references in a uniform way according to Vancouver style at all times.
Please note that the Institute will continuously place high emphasis on the quality checking of reports submitted, and it will contact the MAH in order to correct the errors detected that make the evaluation and signal detection activity more difficult or impossible. Should the errors persist or occur again despite this notification, the Institute will consider the need of further action against the MAH (e.g. pharmacovigilance inspection).