Contributed Commentary by Benzi Alex Mathew

March 17, 2021 | The biggest US vaccine initiative is currently underway, with 32.3 million people having received at least one dose of either Pfizer-BioNTech or Moderna’s COVID-19 vaccine—including 37 million who have been fully vaccinated—as of the second week of March. 

This welcomed news comes as the U.S. death toll from the novel coronavirus exceeds 534,000 but is tempered by research indicating that 39% (4 in 10) of Americans said they definitely or probably won’t take the vaccine. Much of that concern can be attributed to worries over adverse events (AEs) and safety. However, approximately half of the people in that group say it is possible they would change their mind as more people get immunized and more information about the vaccine becomes available. 

Manufacturers of COVID-19 vaccines have committed to conducting long term follow-up of clinical trial patients, and such data will be extremely valuable in understanding the safety effects of these vaccines. However standard passive/spontaneous surveillance approaches do not offer the insight needed to boost public confidence in the vaccines.

The use of Active Safety Analytics is the way to close the gap between those willing to be vaccinated and those who are hesitant. This state-of-the-art technology compliments current passive/spontaneous safety surveillance while also overcoming inherent limitations. 

Current Surveillance Challenges 

While spontaneous reports benefit from being submitted due to a suspected causal association with the medicinal product, they shed light on only a short period of a patient's life. Cases may be under-reported as reporting is voluntary, or over-reported if the adverse event is in the news or social media. Relying on passive surveillance alone may not be sufficient to identify all potential adverse events, protect and inform the patient population, and ultimately boost public confidence in therapy. 

Advantages of Using Active Safety Analytics for Vaccines

Active surveillance uses real-world data encompassing a patient's longitudinal history of medications, diagnoses, procedures, and labs to identify potential adverse events associated with a specific drug. Real-world data negates the influence of over or under-reporting, and provides consistent patient health information for all patients included in an analysis. The FDA Sentinel Initiative, launched in 2008, leverages active safety analytics to assess the safety of approved medical products including drugs, vaccines, and medical devices. The life sciences industry now has the ability to deploy similar innovative technology, including the FDA-developed TreeScan methodology for safety signal detection using real-world data, and can leverage it to more firmly establish public confidence in and acceptance of COVID-19 vaccines.

TreeScan is a data mining method that involves hierarchically applying organized diagnostic codes for scanning deidentified patient data to detect any statistically higher-than-expected clustering of cases within the post-exposure follow-up period while adjusting for multiple testing. Beyond identifying specific adverse events, the Tree Temporal TreeScan methodology identifies risk windows when the events are more likely to be diagnosed. 

The use of the TreeScan methodology for safety signal detection has been extensively studied in many vaccines as a part of the Sentinel Initiative. The mass administration of COVID-19 vaccines implies the exposure of vaccines to a vast population. Hence, adding active surveillance to regular drug safety efforts can increase the effectiveness of the process exponentially. Availability of real-world data for a large pool of patients also provides an excellent statistical power to detect rare events.

The FDA announced its plans for the active surveillance of COVID-19 vaccines during the 12th Annual Sentinel Initiative Public Workshop in October 2020. The FDA will use near-real-time surveillance or rapid cycle analysis (RCA) to monitor about 10-20 specific outcomes of interest that have been pre-identified for monitoring. To leverage active safety analytics most efficiently, leading pharma companies need to have their own active safety surveillance initiatives and not rely solely on federal agencies to perform this function.

With the US federal and state governments taking up managing the vaccination program, multiple individual state Immunization Information Systems have become sources of vaccination data. This patient-level vaccination data then flows through the Immunization (IZ) Data Clearing House and COVID-19 Data Lake of the Centers for Disease Control and Prevention to Tiberius, the nerve center of the data infrastructure created by Operation Warp Speed to coordinate the distribution of COVID-19 vaccines across the US. 

While the FDA has access to several data sources, including the CMS, vaccination data need to be made available to real-world data providers. Connecting COVID-19 vaccination data to claims and electronic health records and making them accessible to data providers will enable pharmaceutical manufacturers to better understand their vaccines' safety and effectiveness in real-world settings. Active surveillance will become an essential tool and enormous advantage for successfully combating the COVID-19 pandemic. 

Pharma’s Adoption of Active Safety Analytics

Pharma organizations of all sizes can now leverage cloud-based analytics technology for their own COVID-19 active safety analytics capability. Besides making analytics-ready, real-world data available for researchers and data scientists, innovative solutions are available with built-in TreeScan algorithms for safety signal detection during routine safety surveillance.   

The last time an epidemic caused so much devastation in the U.S. was in 1918, when the Spanish Flu claimed more than 650,000 lives. The life sciences responded to the COVID-19 pandemic with unprecedented speed and agility, developing vaccines in months versus years to prevent the novel coronavirus. But more remains to be done if we are to ensure public confidence in these therapeutics. Active Safety Analytics complements existing spontaneous/passive safety analytics capabilities and can go a long way toward increasing the public comfort level with COVID-19 vaccination programs, leading to faster herd immunity and more lives saved. 

Benzi Alex Mathew, Vice President and Partner at Saama Technologies, is committed to enabling Pharma companies to use Real World Data (RWD) to regularly monitor, evaluate and act on the safety signals from drug and vaccine usage to meaningfully impact human lives. Benzi has been engaged with the Life Sciences industry for over two decades, running innovation labs, deploying various technologies at a global scale, crafting strategies, leveraging data and advanced analytics, and implementing business process transformation. He can be reached at benzi.mathew@saama.com.