Use of Visualization Techniques to Augment Traditional Adverse Event Signal Detection and Evaluation

Traditional qualitative signal detection and evaluation is a process that needs to be carefully designed to ensure efficiency and reproducibility of analysis results. It is focused on deciphering a potential signal of a possible unexpected adverse reaction, or change in severity, characteristics or frequency of an unexpected event. Such a potential signal would normally be an emerging safety pattern that is embedded within a dataset.

Multiple data sources need to be screened for possible signals as part of post-market signal detection and evaluation process and may include spontaneous AEs/ADRs (including published literature sources), medical/insurance claims databases, drug usage, and Safety Registries (post-market) databases. For marketed drugs with ongoing clinical trials, data screened should also include post-market clinical trials safety data and epidemiological data where applicable.

HEPATIC

For easier and faster review and signal screening of data, this process can include the use of a visualization application that may or may not be integrated with a validated safety database application from where the analysis datasets will be obtained. Also, trained Signaling Assessors and Medical Reviewers should be part of the core procedure with clearly defined roles and responsibilities, as well as integrated quality control and quality assurance procedures being critical aspects of the overall process.

Use of visualization methods is not meant to replace but augment and facilitate traditional qualitative review of Individual Case Safety Reports (ICSRs) and/or aggregate data.

Example of a Visualization Analysis: Demographics Trends Analysis:

Where data is available, visualization signal screening of spontaneous data should include review and analysis of the following parameters:
Serious adverse events (ICSRs and aggregate data) Non-serious events (Aggregate data analysis) Reporting frequency of events, especially ADRs that are, or may be, causally related to the use of the drug or biologics product Dosage Analysis: Evaluation of the extent of exposure at relevant doses; this could be an estimated exposure rate for spontaneous events Other Dosage Analysis: Dose, plasma level (if available), duration of exposure, etc. Adverse event profile of vulnerable populations such as: pediatric, elderly, pregnant women, etc. Adverse event profile of high-risk populations: for example, patients with impaired liver or kidney functions, consistent, spurious or intermittent spikes or abnormalities of hepatic tests or renal function tests, co-morbid illnesses, metabolic status and genetic characteristics, such as, genetic predispositions, etc. Cardiac events and related events (also included as DMEs) Drug-Drug interactions or potential interactions, other drug-related factors Demographic analyses, e.g., age, gender, ethnicity, race, etc. Concomitant medications burden Comorbidity burden Designated Medical Events (DMEs) Rare events, that is, events that are less likely to occur in the absence of drug exposure

In a visualization analysis, the Signal Assessor/Medical Reviewer considers all of the findings of the traditional qualitative signal detection analysis including the following:

Demographic trends: AE/ADR distribution across demographic groups such as age, gender, disease indication sub-populations (geriatric age grouping, pediatric, women of child bearing age, pregnancy, organ-impaired patients-decreased hepatic and renal function, etc.) Polypharmacy: The confounding effect (impact) of one or more concomitant medications Concurrent/Inter-current Illnesses: Underlying diseases that may predispose patients to developing DMEs, TMEs or other events Rarity of AE/ADR: Some events are rare in the disease indication population or even in the natural population in absence of drug exposure, e.g., Progressive multifocal leukoencephalopathy (PML)/Progressive Multifocal Leukoencephalitis Race and genetics: Race and genetic polymorphisms (where data is available) should be considered; some hereditary factors known to affect bioavailability of drugs and which may contribute to the development of ADRs. Impact of any medication errors and potential for medical errors Impact of off-label use (including off-label pediatric use) and potential for off label use. Impact of any misuse/overdose and potential for misuse/overdose

Use of Visualization Techniques to Augment Traditional Adverse Event Signal Detection and Evaluation

HEPATIC