Pharmacovigilance (PV) is defined as the science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem. There are many significant approaches were considered while developing the overall process of pharmacovigilance i.e. how these activities like detection, assessment, understanding, and prevention will be carried out in actual PV operations.
The significant approach towards pharmacovigilance planning was initially developed by EMA via proposing the GVP module-VI (i.e. Guideline on good pharmacovigilance practices (GVP) Module VI – Collection, management and submission of reports of suspected adverse reactions to medicinal products) in which EMA had put fourth step by step guidelines of defining & managing the activities of pharmacovigilance and also US-FDA had prescribed the guidelines for pharmacovigilance planning (PVP) for establishing an efficient PV operations.
Although efficient guidelines are at place but there are many hurdles usually occurs along with immense amount of manual efforts in all day to day PV operations. Hence, there is an emergent need of automation to be implemented in certain PV processes which will reduce the manual efforts and will result in maximum productivity and efficiency of operations team. In this blog we will discuss about the pain areas in PV operations and how we can automate them.
Types of PV operations, challenges & their automation aspects:
- PV mail-box handling
One of the most integral part of the PV operations, which required a lot of efforts of manually checking each mail for the presence of unstructured source documents, once the mail is screened for the source document, the downloaded document will be analysed and assessed for validity criteria’s for a valid ICSR and it will be processed and attached in the safety database. A case initiator (case intake personnel) had to take cumbersome manual efforts in order to handle the huge amount of case volume on daily basis which results in more resource expenditure and more time consumption.
However, an automation of mailbox will immensely help the end users who are involved in this activity. This can be achieved by automating the process of initial auto assessment of mails received in mailbox by auto-reading the subject and mail text strings and we can also segregate all the mails based on clients by mapping of their mail domain ID. (e.g. @wipro.uk).
- Structured and Unstructured case intake:
Case initiation (intake) is most crucial aspect of ICSR processing. ICSR initiation happens via two ways structured case intake and unstructured case intake. In case of structured case intake, E2B R3-XML files for ICSRs are downloaded from the EvWEB portal and later imported to the safety database and further ICSR will be processed. In this process a lot of manual steps will be performed like segregation of cases based on MAH/non-MAH, duplicate check, follow up check, selection and ignorance of valid/invalid cases. All these steps were performed manually by analysing each XML file which was downloaded.
Automation for all these processes can be achieved by setting up the algorithms in the system for all these processes by which system will be able to perform all these tasks of structured case intake in automated manner i.e. auto-duplicate check, MAH/non-MAH check, valid/Invalid check etc.
In unstructured case intake, mainly unstructured source documents such as CIOMS, MedWatch and handwritten documents (PDF/word) were analysed and processed. This is also one of the cumbersome process in PV operations which can also be automized by implementing by using artificial intelligence algorithms like natural language processing and OCR reading of the unstructured documents for capturing the events, suspect drugs, reporter details, and patient details which are exactly required for creating an ICSR.
- Case allocation or reallocation:
Once the case intake will be performed, next important aspect which comes in the picture is case allocation within the team by the respective leads, which is also currently happening manually at most of the PV operations. Conventional method is pulling over the excel listings of active workflow cases and assigning the cases to respective workflow users by manually prioritizing them. This whole process is cumbersome for any lead in the PV operations on daily basis they have to perform this activity of assigning the cases and sending mailers of assignments to respective users which is also time consuming.
The automation in this case allocation process can be easily achieved by developing an automated case allocation module in the safety database itself in which the admin (lead) will able to visualize all the active the cases at one platform and will be able to select the users according to the workflow and assign them from the same platform without sending an individual mailers or alerts. Case reallocation can also be done via same module to avoid the confusion of duplicate assignments.
- ICSR data entry & Narrative writing
This is the heart of ICSR processing, at the initiation level minimal data entry is done, while the case processor has to perform full data entry in this step by adding all the case related information starting from awareness date, source of the report, reporter and patient details, medical history, laboratory details, event details, event assessments (causality and labelling), narrative, and ICSR comments. All these processes can easily be automated in case of structured case intake. When the E2B-XML ICSR will be imported in the safety database all the above-mentioned information can be automatically be imported in the safety database in their respective fields due to algorithm set behind the R3-XML schema.
Automation in reading and assessment of unstructured document can be automated with the help of natural language processing and OCR readings and all the requisite details can be auto captured in data entry stage in the safety database. Automation in the narrative can also be achieved, auto narrative can be mapped with all the requisite fields which are required as per compliance and client narrative format (customizable template).
- Quality review process:
This process is at most required to maintain desired quality level of ICSRs processed in the safety database. There is a separate SOP, which is client specific for the process of quality review. In most of the PV operation currently a manual process of recording the errors of the ICSRs in the excel sheets is practiced which is so much cumbersome and it becomes very difficult when the sampling percentage will become 100% in quality review. Along with keeping a manual case specific track record in excel sheet the quality review team lead also has to pull the monthly quality data of each user and manually calculate the case level and field level quality in order to have a quality check for audit compliance.
Automation for QC review process will be extremely helpful if it will be implemented for auto-tracking of user specific quality of each user. It can be customizable as per client SOP or client QC check list. Overall team’s quality can be easily traceable with one click for audit compliance.
- Event assessment:
This is the most important part of ICSR case processing from PV operations, which can be taken care by physicians (Drug safety physicians) which involves assessment of company causality of the suspect product with the reported events. Most of the times considerations of confounding factors like de-challenge and re-challenge, medical history, event outcome after action taken, concomitant medications, and lab details can be taken to assess the causality.
Latest techniques from artificial intelligence by using algorithms like neural networks, the event assessments and labelling process can be automized. For which immense amount retrospective data of the event is required for iterations in the algorithm.
- ICSR submissions and acknowledgement receipt
This is the end stage of the of the ICSR processing, in which we need to manually schedule the corresponding reports (either in form of R2/R3 XML or CIOMS or MedWatch) from the safety database on the basis of the source and submission country submission rules (domestic and foreign submissions). Country specific submissions rules were followed in the reporting of ICSRs.
Automation can be achieved by auto-scheduling of the reports by implementing the country specific submission logic in the back end of the safety database. With one click we can schedule the country specific reports for that INN/active pharmaceutical ingredient who has marketing authorization in that territory. Further, we can have an integrated data interchange gateway with our safety database through which we can transmit those ICSRs XML files to the respective regulatory authority with ease.
We at Sarjen system came up with many automated solutions in our safety database PvEdge as listed below, which will be of immense help in automating all above mentioned processes which will in turn result in increased efficiency and productivity of PV operations team.
- Front end masters for clients, projects, products, formulation, scopes, role assignment, agency, causality assessment, reporter, sender & receiver masters
- Efficient auto-scheduling module for automatic report generation for ICSR submission
- Efficient auto narrative generation from the fields filled in the database
- Literature automation module for medical literature monitoring (MLM)
- PvEdge-Integrated user quality monitoring module (Daily, weekly, monthly & yearly analysis from database itself on one click)
- Integrated case allocation module for daily case assignments for each workflow
- Integrated alert management module: For sending an important alert to specific users regarding priority case
- Integrated follow up management
- Front end importing facility of XMLs to PvEdge for Case creation (no need to have drop-box)
- Integrated notice management tool for latest client updates on notice board
- Front end case copy option for quick creation of a copy case
- Accurate audit trail compliance reports (for various masters, users, listings, import log, project specific audit trail)
- Front end facility of MedDRA upload and assignment to client (to avoid time consuming downtimes)
- Front end master for clinical study set up for CT cases (Protocol configuration wizard)
- Integrated user productivity monitoring via workflow efficiency module (Daily, weekly, monthly & yearly analysis from database itself on one click)