Transform safety surveillance with data analytics and automation

Transform safety surveillance with data analytics and automation

The pharmacovigilance (PV) process to capture, evaluate and report adverse drug reactions is heavily dependent on manual interactions during each step of the process. The current process is labor intensive, reliant on patients to voluntarily provide information, time consuming, and prone to human error. As the volume of adverse drug events (ADEs) increases, this historical way of operating is becoming increasingly unreliable and costly.

Pharmaceutical companies need their PV systems to be more than just a way to capture and report adverse events.

To meet the market demands for increased patient safety and satisfy regulatory requirements, companies must have a safety surveillance system that quickly produces reliable, actionable, patient-centric information.

This white paper discusses the challenges of today’s PV systems to manage adverse events. It also covers how this work can be transformed from a passive to an active surveillance system using data analytics and automation.

The rise in prescription drug use

Prescription drug use in the U.S. has dramatically increased in recent years due to an aging population managing chronic conditions along with the availability of new therapies and generic drugs. Consumers today are accustomed to having access to a wide range of brand-name therapies to treat ailments such as high blood pressure, high cholesterol, anxiety and depression. In fact, prescription drug use reached a record 194 billion daily doses in 20211.

Similarly, health care spending in the U.S. is also on the rise, especially spending for prescription drugs. In 2016, the U.S. spent $3,337 billion, or 17.9 percent of the gross domestic product (GDP), on national health expenditures, of which $329 billion was spent on prescription drugs2. Additionally, the Centers for Medicare and Medicaid (CMS) projects that in the next 10 years spending for retail prescription drugs will be the fastest-growing healthcare category.3

The high cost of adverse events

With more people taking a combination of prescription drugs daily, the healthcare industry is now facing a significant increase in ADEs, which is defined as any negative medical occurrence associated with the use of a medication. While a drug’s safety and efficacy must be demonstrated during clinical trials prior to full approval by the Food and Drug Administration (FDA), many ADEs are detected only after a drug has been on the market and used by a larger and more diverse population, sometimes in combination with other various drugs.

According to a 2018 FDA report4 , ADEs are one of the leading causes of morbidity and mortality in health care. The agency quotes an Institute of Medicine report that 44,000 to 98,000 deaths occur annually from medical errors. Of this total, an estimated 7,000 deaths occur due to Adverse Drug Reactions. Other studies of hospitalized patient populations have found much higher estimates of the overall incidence of serious ADEs. These studies estimate that 6.7% of hospitalized patients have had a serious adverse drug reaction causing over 106,000 deaths annually.

A large pharmaceutical company typically receives 500K Adverse Events in a year and estimates report that individual case safety reports (ICSR) volume is increasing by 15% every year. This translates to a doubling volume every five years, meaning that a cluster of 20 pharmaceutical companies will have approximately eight million ICSRs processed every year.

Typical growth in ICSR volume of a large pharma company

 

The failings of today’s PV systems

Pharmaceutical companies rely on manual processes to detect, research, report and mitigate adverse events. With a typical drug manufacturer receiving a half a million adverse events per year, it makes the work of managing and reporting this critical drug safety information a huge operational expense.

In a traditional drug safety surveillance program, adverse events are collected and validated through a manual process. After verification, the information is recorded in an adverse experience database. Then an analyst conducts a literature search to find causalities for the reported drug interaction. Once a conclusion is reached that the adverse event was indeed caused by the drug and not by any other factors, a safety report is developed and sent to the FDA.

An overview of the manual processes of managing adverse events in a traditional drug safety surveillance program.

Figure 1: An overview of the manual processes of managing adverse events in a traditional drug safety surveillance program.

This manual and labor-intensive process also is encumbered by operational and technical issues. Companies often lack standardized processes, reducing efficiency in data collection and analysis. Analysts must rely on data using legacy IT systems that either cannot fully accommodate an information request or cannot produce the necessary results in a timely manner.

Additionally, the entire drug safety ecosystem is becoming increasingly complex and difficult to navigate, combined with a greater volume of reporting and therapy combinations. Plus, the current system is reliant upon spontaneous reporting. While this reporting is mandatory for drug companies, it is voluntary for patients. Studies have shown that as many as 90% of serious ADEs go unreported.5

An outline of the numerous operational issues, technical challenges, and market demands affecting and functionality of safety surveillance programs today.

Figure 2: An outline of the numerous operational issues, technical challenges, and market demands affecting and functionality of safety surveillance programs today.

Traditional drug safety surveillance programs are plagued by the limitations of passive reporting, long periods of time to consolidate information, and high implementation costs. To meet this increased demand and complexity, corporate spending on PV management has tripled in recent years. And yet drug safety has not significantly improved.

Today’s PV systems remain in a passive, reactive mode rather than serving as an active surveillance system to improve drug safety and health outcomes for the U.S. consumer. Pharmaceutical companies need to re-examine their legacy systems to add data analytics and automation to achieve a new level of active engagement that is more productive and cost effective.

Transformation through automation

Implementing advanced technologies for automation offers new capabilities to add scale, consistency, and speed to the entire PV process from intake, triage, evaluation and reporting, to signal detection and management. Pharmaceutical companies can transform their entire PV process by utilizing technologies that can store massive quantities of data from multiple data sources related to adverse drug reactions. Automation allows companies to reduce costs by capturing better information more quickly, with the ultimate goal of improving patient outcomes.

At a high level, AI-driven automation is the only solution for a large volume of diverse, dynamic, and distributed structured or unstructured data. By using advanced data mining, the system can more easily and quickly determine findings such as drug event pairing, adverse event tracking, and hypothesis testing. With more real-time, evidence-based data, companies now can feel more confident with their decision-making and communications regarding the safety of their drug products. This translates increased consumer confidence in their brand, creating stronger brand awareness and loyalty.

Looking at the day-to-day activities of managing ADEs, an automated system allows analysts to spend less time conducting information gathering and more time doing higher quality work. An analyst can capture information in real-time by using a combination of existing data from various sources such as the FDA’s adverse experience database, a company’s own internal adverse experience database, along with patient electronic records.

Electronic health records are now a viable option for AI-driven data analysis including both structured or coded data and unstructured clinical narratives that could provide more information about adverse drug reactions. Finding this information in electronic records can be achieved by utilizing advanced natural language processing (NLP) techniques. What once would have taken days or weeks to accomplish now can be find in minutes.

Preparing for future needs

Using data analytics and AI automation, a company can more quickly identify possible adverse event scenarios and develop a proactive intelligence system. While some companies have started the journey towards automation, few have yet to fully implement intelligent automation technologies to transform their safety surveillance program.

With timeliness and quality of safety operations becoming more critical and as caseloads continue to increase, pharmaceutical companies will need to rely on automation to reduce effort and costs. This is especially true for companies working in markets with heightened regulatory obligations and price caps that will limit retail consumer prices. Making the change to automation is the answer to improving security surveillance while maintaining profitability.

To transition from a legacy system to an AI-driven system, companies should rely on a technology partner that not only knows the PV environment but also understands its larger impact on patient safety. Companies should seek a partner with both technical capabilities and deep expertise in the healthcare industry who understands that ultimately the most important customer is the patient.

For more information about EXL Health’s offering for Pharmacovigilance and safety surveillance: Life sciences | EXL (exlservice.com)

References:

1. https://www.iqvia.com/insights/the-iqvia-institute/reports/the-use-of-medicines-in-the-us-2022

2. https://www.healthaffairs.org/doi/pdf/10.1377/hlthaff.2017.1299

3. https://www.actuary.org/content/prescription-drug-spending-us-health-care-system

4. https://www.fda.gov/drugs/drug-interactions-labeling/preventable-adverse-drug-reactions-focus-drug-interactions#ADR%20Reporting

5. https://link.springer.com/article/10.1007/s40264-018-0766-8