One of the key roles of pharmacovigilance departments is to understand the benefit-risk profile of a drug. But departments are often bogged down by manual tasks that can absorb time, which might be better spent on value-centric activities. Perhaps the most tedious pharmacovigilance activity is medical literature monitoring (MLM), and as we discussed in our last blog, manual, repetitive tasks are a primary cause of employee dissatisfaction.
Automation has the potential to significantly relieve this burden, reducing cost for the organization by not only taking this task off the plate of busy pharmacovigilance professionals but also reducing reliance on third-party services.
Real-time, comprehensive medical literature review can be conducted through robotic automation, natural language processing, machine learning and data mining, information retrieval, classification and filtering systems. Automation tools can scan all relevant databases for global publications, such as PubMed and Embase and local publications for potential safety issues.
The ability to review local publications is crucial, because although regulatory authorities are taking steps to harmonize the medical literature review process – as is the case with the European Medicines Agency (EMA), which is conducting selected medical literature monitoring – there are limitations. EMA, for example, only covers selected databases, and marketing authorization holders (MAHs) remain responsible for monitoring all other medical literature and reporting suspected adverse drug reactions (ADRs).
Real-time Benefits
The benefit of real-time monitoring is that, as soon as a case is mentioned in the literature, it becomes available to pharmacovigilance departments, enabling them to act promptly – to the advantage of both the patient and the company. Pharmacovigilance isn’t necessarily tied to real-time updates if that would overly burden the department; rather it can allow a department or company to select a desired frequency for medical literature searches.
Once a potential ADR has been identified, the automated tools can detect relevant information and auto-extract it for case creation. This includes information about the person reporting the ADR, the drug used, the ADR experienced, the demographics of the patient (where available), and so on.
An audit trail can be automatically generated for all searches undertaken, and records can be generated to demonstrate which articles have been reviewed for possible ADR identification. Furthermore, duplication issues that often arise in literature searches can be managed with advanced automation by having the tool scan individual case safety reports (ICSRs).
Cognitive intelligence
Automation of medical literature monitoring goes well beyond simply looking for mentions of drugs or cases and shepherding data through pre-determined pathways. Indeed, advanced cognitive automation tools are trained in identifying relevant articles and quickly learn which potential cases might be relevant for the drug in question.
In addition, the tool can incorporate internally developed intelligence with a growing external knowledge base provide meaningful recommendations over time. With experienced cognitive systems can make assist in constructing a search, improve precision of a database search. What that means is not only is an advanced automation tool removing manual work, but it also can increase the quality and sensitivity of detecting cases in the literature.
To learn more about how cognitive computing can alleviate many manual processes in pharmacovigilance and enhance processes, view our recent webcast: Productivity, Compliance and Quality: The Holy Grail of Pharmacovigilance.
[poll id=”3″][poll id=”3″]One of the key roles of pharmacovigilance departments is to understand the benefit-risk profile of a drug. But departments are often bogged down by manual tasks that can absorb time, which might be better spent on value-centric activities. Perhaps the most tedious pharmacovigilance activity is medical literature monitoring (MLM), and as we discussed in our last blog, manual, repetitive tasks are a primary cause of employee dissatisfaction.
Automation has the potential to significantly relieve this burden, reducing cost for the organization by not only taking this task off the plate of busy pharmacovigilance professionals but also reducing reliance on third-party services.
Real-time, comprehensive medical literature review can be conducted through robotic automation, natural language processing, machine learning and data mining, information retrieval, classification and filtering systems. Automation tools can scan all relevant databases for global publications, such as PubMed and Embase and local publications for potential safety issues.
The ability to review local publications is crucial, because although regulatory authorities are taking steps to harmonize the medical literature review process – as is the case with the European Medicines Agency (EMA), which is conducting selected medical literature monitoring – there are limitations. EMA, for example, only covers selected databases, and marketing authorization holders (MAHs) remain responsible for monitoring all other medical literature and reporting suspected adverse drug reactions (ADRs).
Real-time Benefits
The benefit of real-time monitoring is that, as soon as a case is mentioned in the literature, it becomes available to pharmacovigilance departments, enabling them to act promptly – to the advantage of both the patient and the company. Pharmacovigilance isn’t necessarily tied to real-time updates if that would overly burden the department; rather it can allow a department or company to select a desired frequency for medical literature searches.
Once a potential ADR has been identified, the automated tools can detect relevant information and auto-extract it for case creation. This includes information about the person reporting the ADR, the drug used, the ADR experienced, the demographics of the patient (where available), and so on.
An audit trail can be automatically generated for all searches undertaken, and records can be generated to demonstrate which articles have been reviewed for possible ADR identification. Furthermore, duplication issues that often arise in literature searches can be managed with advanced automation by having the tool scan individual case safety reports (ICSRs).
Cognitive intelligence
Automation of medical literature monitoring goes well beyond simply looking for mentions of drugs or cases and shepherding data through pre-determined pathways. Indeed, advanced cognitive automation tools are trained in identifying relevant articles and quickly learn which potential cases might be relevant for the drug in question.
In addition, the tool can incorporate internally developed intelligence with a growing external knowledge base provide meaningful recommendations over time. With experienced cognitive systems can make assist in constructing a search, improve precision of a database search. What that means is not only is an advanced automation tool removing manual work, but it also can increase the quality and sensitivity of detecting cases in the literature.
To learn more about how cognitive computing can alleviate many manual processes in pharmacovigilance and enhance processes, view our recent webcast: Productivity, Compliance and Quality: The Holy Grail of Pharmacovigilance.