Pharmacovigilance Analytics

Your best resource for PV analytics news, content and innovation!

Opinion

Pharmacovigilance Audits/Inspections and PV Analytics!

by Leave a Comment

Disclaimer: This article is written by a Safety Physician to provide pharmacovigilance (PV) operational perspective to experts from automation, artificial intelligence (AI), and other analytical domains.

Why do we need PV analytics in Inspections/ Audits processes of PV?

Pharmaceutical industry, and more particularly pharmacovigilance, is tightly regulated ship in terms of having strict laws and regulations set up by health authorities with stringent timelines to follow them.

To ensure that pharma companies follow the rules and regulations, it is mandatory for their PV systems to be audited by internal and external auditors and then again also by inspectors from various health authorities.

However, preparing for audit/inspections is a tremendous time consuming and costly process. This is one area of PV, that can obtain tremendous help from analytics systems. Also, with innovations in technology and automation processes, many times, processes are handled by robots, and audits and inspections are carried out for these robotic tools too. The data handled by robotic processes is varied, large and complex, and takes time for inspectors/auditors to manually understand and then identify issues in it.

An audit/inspection is carried out according to the following steps:

1.            Communication regarding the audit/inspections

2.            Introductory meeting with all stakeholders participating in audit/inspections

3.            Conduct of audit/inspection

4.            Interviews with relevant personnel involved in personnel

5.            Process demonstration of PV systems

6.            Document review conducted of documents involved in PV process

7.            Exit meeting to discuss with involved stakeholder in audits/inspections

8.            Follow up request to request additional details related to PV systems

9.            Report preparation of audits/inspection to summarise the audit/inspection

10.          Response to be provided by pharma companies that was audited/inspected

11.          Follow up to ensure the root cause analysis (RCA), corrective and preventive actions (CAPA) are identified for findings and actions are taken to resolve the finding.

As we can see, the amount of data gathered, discussed, and identified is complex and huge during the entire process of audits/inspection and has to be checked if it is in compliance with the required regulatory guidelines. One additional challenge is multiple regulations and evolution of regulation with more requirements from health authorities for a better understanding of the safety of pharmaceutical products and, ultimately, better patient care.

An example

To give an example, I will quote paper published by MHRA. MHRA conducted 22 inspections in 1 year between April 2017 and March 2018, which took tremendous amount of time, money and resources of both, pharma companies and MHRA. During these inspections, MHRA identified 89 major findings in risk management plans, noncompliance in quality systems, analysis of safety data, and management of adverse drug reactions. However, if these inspection findings took time to be identified, inspectors had to spend hours with pharma companies, prepare these findings, involve multiple stakeholders, to finally release those findings to pharma companies. On the other hand, after receiving these findings, it again involved additional time at pharma companies end, to track these findings, to analyze them, find their root cause, and issue corrective and preventive actions. Lot of time, efforts and money is spent before the actual correction is implemented. These times is critical, as any delay is directly or indirectly affecting the safety of patients. E.g. if a finding is about not being able to identify a signal in relation to a product a product due to weak process of signal detection, by the time we strengthen the process by writing a new SOP, training people etc, we might be possibly lagging to identify signal for other products too.

What is PV analytics?

The concept of PV analytics is explained in previous articles on this blog. To clarify in short is that it’s a process of using different data techniques to analyze large and varied datasets to make informed decisions in effective manner.

When we talk about audits and inspection, PV Analytics data techniques will be used to identify cross functional PV systems to identify any weak links in these systems. Any company has interlinked SOPs such as PV SOPs are linked with clinical, regulatory, quality, legal departments, and others. Despite trainings and years of experience, the process set up in PV Systems is still not followed 100% and this weak links in the process are not immediately identified until an audit or inspection takes place. Because the data is so varied for each department and process is so complex, its almost impossible to do the identification of weak links on regular basis by operations team themselves. PV analytics can help in doing so, by being a watchdog of these data processes, and triggering an alert for a weak link to us before audit or inspections.

How?

Due to the above reasons, a good solution would be to look at application of analytical systems to automate, streamline or to identify tools that can make this process a less time consuming and more solid and cost effective. Hence, collaboration with IT solution providers selected according to their technical performance, but also to their level of expertise in quality and regulatory compliance is essential. We can streamline the SOPs interlinked with PV from other departments and apply tools that can identify a possible weak link or trigger a warning, whenever a process is not followed by other department which can have impact on PV audit or inspections. E.g. as a part of GCP compliance, Investigators should file the SUSARs in Site Trial Master File (TMF), PV professional send the SUSAR as part of their due diligence within timelines to Investigators, but if they are included in TMF or not, is not under PV purview. To check if TMF has SUSAR or not, we employ CRA, Site co-Ordinator’s etc, and yet one of the frequent finding in GCP inspection is SUSAR are not present in TMF.

If we put tools in place, that can identify a SUSAR submission from PV database until its presence in TMF. Many pharma companies do have such automated tools in place, but they are costly for small and medium pharma companies and yet have not been 100% effective.

Another aspect to keep in mind is the reconciliation between PV databases and clinical database. The reconciliation is performed on periodic bases between the two databases, finding potential discrepancies that otherwise would remain unnoticed. As an example, one of the frequent finding in audits/inspections is AE/SAE missing in PV database which is present in clinical database.

PV analytics can design cost efficient tools that can trigger such findings of discrepancies well within time, before an audit or inspection.

Another interesting aspect would be post audit/inspection follow up. Multiple audits/inspections happen every year for PV systems. It is very difficult to track findings of each audit/inspection manually, and many times findings from one audit are not rectified in timely manner, which then becomes a finding in next inspection or audit. But if we can track the findings in parallel manner, and do analysis thoroughly in time, we can avoid such issues.

The data audited/inspected is tracked by multiple different stakeholders, reports are made by different teams, and corrective and preventive actions are implemented by separate teams, which again is costly and time consuming. Often this leads to ineffective implementation of corrective actions. Preparing the findings, drafting a response to audit/inspection report requires a huge effort from various stakeholders, often by those who are not involved in the system audited/inspected. This leads to communication delay, where the corrective and preventive action proposed is not clear, does not reach on time to actual operations team responsible to implement it. If we have tools and analytics that can immediately identify any CAPA initiated by any department that affects the PV operations, we will have their agreement in terms of implementing the CAPA. The tools and analytics will help PV operations team to see the logic of RCA and CAPA and as it reached in time to them, they will be open to accept it.

Summary

For this purpose, numerous of current providers of pharmacovigilance solutions should also look into analytical solutions that will reduce time, cost and risk associated with Audits/Inspection process. The analytics that will help with compliance by pharma companies to Health Authorities to patient safety. But these solutions should be based on thorough knowledge of the actual reality of inspection or audit findings, and preferably be done internally or in close collaboration with quality and regulatory PV experts.

References

Pharmacovigilance Inspection Metrics April 2017 to March 2018

https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/union-guidance-record-keeping-archiving-documents-obtained-resulting-pharmacovigilance-inspections_en.pdf

EU Legislation a regular audit pharmacovigilance system (Directive 2010 / 84 article 104)

https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E6/E6_R1_Guideline.pdfhttps://www.longdom.org/proceedings/minimizing-reconciliation-between-safety-and-clinical-databases-14571.html

Artificial Intelligence in pharmacovigilance? What a challenge!

by 3 Comments

Abstract

Pharmaceutical industry, and more particularly pharmacovigilance, has seen the amount of data from individual safety reports grow exponentially due to the evolution of regulation with more requirements from health authorities for a better understanding of the safety of pharmaceutical products and ultimately better patient care. This has led companies to hire more Health Care Professionals to deal with this increased workload.  At the same time a fast development, as in any other field of the industry, of IT solutions or more tendentiously Artificial Intelligence has occurred with many suppliers offering these services and whose selection must be made successfully to reorganize pharmacovigilance activities in order to reduce the time and costs of their activities as a result.  Hence, collaboration with IT solution providers selected according to their technical performance but also their level of expertise in quality and regulatory compliance are essential.

The constant evolution of pharmacovigilance regulations towards Manufacturing Authorization Holders (MAH) but also Health Care Professionals and the awareness of patients and patient associations in the era of social media has contributed to the increase in the workload and cost of pharmacovigilance activities.

This has created an urgent need for organization and solutions to the pharma industry and also to the outsourcing companies to reach compliance and then offer the highest quality of patient care.

These technology solutions will not be without passing through the Artificial Intelligence through which all other fields of industry pass, and in particular the pharmaceutical one.

Since then, IT solutions providers have been offering and continuing to develop products to organize PV at all stages from data entry to data analysis.

But have we really succeeded at the present time in meeting this challenge and optimising pharmacovigilance tasks and costs or what is the remaining path to be taken if not?

It is clear that many pharmaceutical companies have set up PV databases and have transformed their workload and obviously, many of them  implemented advanced PV platforms to transform their entire pharmacovigilance workflow which reduces time and costs, while accelerating information processing.

This intelligent automation allows predictive analysis but also the capture and translation of adverse event data to identify significant safety trends. It also helps PV professionals to better manage future adverse events and better understand safety issues. These additional benefits include reduced manual data entry errors and strict patient data confidentiality controls, which minimize pharmacovigilance risks to the company.

For this purpose, numerous of current providers of pharmacovigilance solutions offer a wide range of offers. The most interesting are those of web and mobile automation functions to ensure the processing, analysis and follow-up of adverse events from different sources (patient files, social media, literature…etc.) and their capture from different document formats.

The most efficient of these IT solutions reduce time, cost and risk associated with manual processing and generate compliant information for Health Authorities and pharmaceutical companies to enable evidence-based decisions on product safety.

But these solutions will only be effective if they are developed on the basis of a thorough knowledge of the reality and needs of the pharmaceutical industry. Hence, to avoid biasing results, the design of such technological solutions should preferably be done internally or in close collaboration with an expert provider to avoid exposure to a high risk of failure often due in particular to a lack of understanding of the rendering of its technological solutions and their impact on their process. This can be also due to collaboration with suppliers who are unaware of the strategies and regulations governing pharmacovigilance.

To reduce this risk, collaboration with IT solution providers selected according to their technical performance but also their level of expertise in quality and regulatory compliance are essential.

These IT solutions are evolving very quickly and with them the way in which pharmacovigilance activities are managed, which is a permanent challenge for companies to meet this evolution and achieve the targeted objectives in order to optimize the expected productivity and efficiency as per the investments they have committed.

As an example, in my previous experience I have succeeded such projects with suppliers who are not only familiar with the current regulation and pharmacovigilance issues but also their history and the likely trends in their future development. Suppliers whose personnel have also worked internally in the industry with a very good knowledge of its quality and compliance requirements. this can be one of the keys to success and a positive impact on your organization.

References

  1. Danysz K, Cicirello S, Mingle E. Artificial Intelligence and the Future of the Drug Safety Professional. Drug Safety (2019) 42:491–497.
  2. Pitts PJ, Le Louet H, Advancing Drug Safety Through Prospective Pharmacovigilance. Ther Innov Regul Sci. (2018) 52:400-402.
  3. https://www.iqvia.com/blogs/2018/11/pharmacovigilance-automation-has-arrived

Real World Evidence (RWE): Predictive Analytics to Impact Patient Safety

by Leave a Comment

Real World Evidence and SafetyThe use of new analytical tools applied to large, diverse, complex data sets, so called “big data”, the development of devices to track and gather real-time healthcare data and information and the use of digital media are on the increase in the healthcare environment and have the potential to be of great value if harnessed and utilized appropriately.

The current main system for keeping track of dangerous side effects of prescription drugs is deeply flawed according to the Institute for Safe Medications Practices (ISMP). The study conducted by the ISMP found that only about half of reports of serious side effects submitted by manufacturers met basic standards for completeness.

No longer just a sideline used to fill knowledge gaps, the evidence generated from real-world data (RWD) is rapidly becoming an integral component of product evidence strategies. However, the growing volumes and heterogeneity of real-world data sources are creating increasingly inefficient and chaotic analytic environments and as a result, new approaches for database analyses are needed.

Real-world evidence (RWE) allows companies to make more informed and reliable strategic decisions earlier when it comes to protect patient safety. Another value added is converting RWD into valuable RWE that offers great scientific and patient benefit trials as well as shortening phase III to accelerate the approval procedures. These benefits include improving the ability to positively impact patient outcomes through understanding of disease characteristics and treatment patterns, enhancing medicines compliance and aiding in interpreting treatment outcomes for individual patients. It also enables organizations to demonstrate health outcomes and support the case for the value of their products to health authorities, payers, health care providers and patients.

The opportunity to use these technologies and derive their potential benefits to assess the efficacy and effectiveness of therapeutic options is in its infancy. Although they could be the key to establishing a credible new generation of fit-for-purpose RWE, pharmaceutical companies have been cautiously investigating the use of the various technologies that contribute to big data collection (e.g., social media, electronic health records, insurance data claim databases) and the application of analytics to these data sets including pharmacovigilance to assess side effects. The case for accessing and evaluating the plethora of different data sets is clear. However, while the opportunity is large, exploiting the value requires the appropriate governance, knowledge and analytics capabilities to stay within the acceptable tolerance levels for compliance and reputation risk management. Pharmacovigilance departments will become a key element within the organizations.

Questions have been raised as to how best to deploy innovative collection and analytic technologies to maximize their effectiveness. Approaches such as the Advancing Medical Innovation initiative encourage the FDA to identify opportunities to use big data to streamline and support pre- and post-approval activities. In Europe, collaborative projects in the area of post-authorization efficacy studies have identified the need for companies and agencies to be able to measure safety and effectiveness in the real-world use of new medicines.

This is mirrored by the need for pharmaceutical and biotech companies to quickly adapt their pharmacovigilance departments into cross-functional teams to strategically make decisions regarding how new medicines will be used in the real world and to confirm the expected benefit and value, often derived largely from controlled clinical studies. In a rapidly changing regulatory environment, with a diverse set of data sources, contractual mechanisms and data privacy requirements, the capabilities needed to extract the value from the data become strategic in their own right.

Over the last years the potential of real-world data and analytics has been discussed as opportunities to enhance patient engagement, reduce uncertainty in the development and approval space, as well to serve as a natural process for the collection of benefit and risk data post-authorization. Collecting data from a mix of evidentiary experiences would support novel flexible regulatory pathways that accelerate reviews and access to medicines, and therefore, will likely play a key role in transforming medicine development and access over the next decade.

Are we ready for this?