We Are DIA

A Career in Pharmacovigilance: Automation with a Human Touch

Lilai Teum
Laura Anne MacEntee

harmacovigilance, monitoring the safety of therapeutic products approved for patient or consumer use, stepped into the world’s center stage during the urgent global development of vaccines and therapies to combat COVID-19. In the following Q&A, an experienced PV professional shares her original motivation for choosing a career in PV and safety, and the vision that keeps inspiring her work in this field, with a young professional who recently started this vocation.

Lilai Teum: When did you realize you wanted to work in pharmacovigilance (PV) and safety, and what specifically attracted you to the field?

Laura Anne MacEntee: I’m a retail pharmacist by training; I spent my days counseling patients about side effects, how to take medications and counter adverse events. What attracted me to pharmacovigilance is its focus on monitoring adverse event profiles and being able to communicate the benefit-risk profile of a drug so that patients, physicians, and other healthcare providers can ensure that the prescribed medication is most appropriate. I realized that PV was the next step in my progression as a pharmacist: Instead of impacting numerous individuals who came to my counter one at a time, I could impact much larger patient populations by supporting PV activities.

I started as a case processor in what I would call the PV foundation: performing analysis on the information that was reported, writing a structured case narrative around it, and forwarding the case for medical review so that the report could be assessed and distributed to different health authorities if it required expedited reporting, or included into an aggregate analysis that might be submitted to the authorities on a routine basis later. Following case processing, I worked my way toward automation by working on safety databases, because I started to see that this was the foundation of what was coming next.

LT: What do you like the most and what do you like the least about pharmacovigilance?

LAM: I don’t like how often PV is overlooked. Patients will take note of side effects listed on their medications, but they don’t understand where that information comes from or that there’s a team of professionals available to protect them.

What I like most about PV is that you are protecting people on a global scale in this profession. A group of collaborative professionals analyze real-life data and align the story of the patient with the story of the product to benefit more and more patients.

Some people say that PV professionals are like policemen and look at everything from the perspective of what went “wrong” with the product. That’s not our role. Yes, we take what could be perceived as a negative drug effect but we’re communicating that risk to positively impact (i.e., protect) others. We’re providing knowledge that people can use to make better decisions about how they want to manage the use of their medicines.

LT: What degree of automation exists in PV today? Will it increase due to machine learning? Could PV be totally automated—and is that even desirable?

LAM: The PV industry trend is to automate, not just because of resource constraints but to manage the increased volume of reported drug adverse events to pharmaceutical companies. For example, from 2016 to 2021, the FDA received an increase of approximately 649,000 cases (approximately 6% growth year on year). Regarding case intake, information may come in readable format, and natural language processing (NLP) and translation technology can transcribe it directly into the database. Similar tools can be used elsewhere within the organization to scan other information into systems. We’ve had electronic case exchange within industry, a very impactful automation, for close to 20 years.

I do see the advancement of machine learning (ML) in PV, particularly to aid in identifying signals and target populations. But people must understand that ML will require human intervention, specifically to feed information into the machine so the machine can learn.

Could PV be totally automated? The word “totally” is very absolute, so: No, I don’t think you can completely remove humans from PV. We can automate a large portion of our processes. But that still leaves me to wonder: Computers provide logic—logic is their baseline and that’s how computers work. People can provide logic, too. But people also have intuition, based on experience, that can contribute to more accurate evaluation and judgment.

LT: What is the greatest challenge in the PV field today? And what do you think it will be in five to 10 years?

LAM: The greatest challenge right now is the volume of data being reported and being able to drill down to its most salient points. Social media contributes to adverse event (AE) reporting but must be filtered to make sure it meets appropriate reporting standards. This may still be our greatest challenge for the next five to 10 years. In addition, increased use of wearable devices will require companies to create reliable toolsets to derive meaningful insights from disparate data.

LT: What is your vision for PV in the year 2030?

LAM: The 2030 vision is more patient interaction and engagement through outreach and/or support programs that focus on patients’ insight on their preferred mechanism for reporting their adverse events. We currently spend much more time with the data rather than upfront, direct interaction and engagement with patients and healthcare providers as they report adverse events. Advanced digital tools, artificial intelligence, or process optimization should foster a more practical patient approach to obtaining upfront and complete adverse event information including applicable medical history, concomitant medication, test results, etc., versus an iterative communication process with patients and healthcare professionals to obtain the patient’s story. We can also engage patients and healthcare providers to provide perspectives on communication preferences for how they wish to receive benefit-risk information and remove the many stage-gates that information flows through to provide the right information in a timely manner to either the patient or the healthcare provider overseeing the patient.

Part of my vision is to use technology to enable more patient-preferred technical interfaces, and by 2030 we should be able to achieve it. COVID-19 accelerated the concept of telemedicine for the global public. By 2030, we could offer “tele-PV”: The evolution of technology could make it happen. Every pharmaceutical product comes with a risk; through patient engagement, we can optimize communication and education to mitigate that risk and optimize the benefit.

LT: How would you like to see the PV field evolve?

LAM: PV will evolve into being more proactive and less reactive. We receive reports after events occur and process that information through many channels. But companies also proactively distribute timely safety information so that patients can make better decisions about their products. We cannot entirely eliminate risk. Even tools like targeted therapy and genetic testing can only decrease, but not completely remove, the potential for risk. Risk is really a decision; it’s also an outcome. PV professionals just need to continue to effectively and efficiently gather, analyze, and provide safety information and make benefit-risk assessments from the patients’ and healthcare providers’ perspectives by utilizing scalable processes and agile technology.

LT: What advice would you give to your younger self about a PV career?

LAM: Borrowing from a Nike® ad: Just do it. As much as I treasured the patient interactions as a community pharmacist, I have no regrets about transitioning to an industry role in PV. Trust yourself, put up your hand when you need help, and say yes to that PV career. Say yes to the opportunities because every opportunity in PV will further your professional and personal journey.

Disclaimer: The opinions expressed in this article are solely the authors’ and do not represent the views of their organizations.