An adverse event refers to an unexpected problem that happens when a new drug or therapy is being investigated. According to the National Cancer Institute, it can range from mild to severe and may be caused by the drug itself or something else.
In this post, we explore why reporting adverse events remains low and how the process can be improved through technology and better systems.
What’s not often spoken about is how reporting adverse events is not necessarily an objective practice. It differs from most scientific endeavors, which are based on objective techniques, and allows subjective bias to enter the judgement process, explains Dr. Robert Jeanfreau, owner and principal investigator at clinical research consortium MedPharmics.
“In clinical trials, we don’t take a scientific, objective view of AEs because we are starting from the perspective that if something happens, it is an adverse event, suggesting something negative. That is already a biased view. It would be more objective to observe the event as a ‘change in health,’” he says.
For example, Jeanfreau notes that a clinical investigator could regard a minor decrease in the ratio of volumes of red blood cells to the total volume of blood as an adverse event. A similar size increase in ratio, however, might not be viewed in the same way.
Not only is reporting of AEs prone to biases, it also tends to be underreported. This is particularly so for serious adverse events (SAEs), says life sciences writer Ann Neuer.
Clinical trial regulations stipulate that researchers must report SAEs immediately to the sponsor, which informs the FDA within 15 days. But Neuer says reporting remains problematic, often not appearing in published articles at all.
Reporting AEs is valuable to pharma companies as they can collect post-marketing safety data. And healthcare professionals are well placed to make these reports. Yet very few do.
Researchers investigated the reasons why HCPs tend not to report AEs, and published their findings in The Journal of Drug Assessment. They used a questionnaire of 43 questions to assess pharmacovigilance knowledge and activities. Respondents, comprised mainly physicians and nurses, reported that 69.6 percent had encountered an AE at least once, but only 40.5 percent reported it.
Reasons included that only 34.7 percent of respondents knew where to look for an AE form and even fewer people, 25.5 percent, had completed or read an AE reporting form. Even the term “pharmacovigilance” was new to many respondents, with 35.5 percent encountering it for the first time when completing the survey.
It is this limited knowledge that is keeping healthcare professionals from reporting AEs, the researchers conclude. Proper training is required, along with check-ins from regulatory authorities.
Given poor AE reporting rates, proper interventions and training is required. The important factor is that interventions and encouragement to report AEs need to be frequent, say researchers in their report published in the Frontiers in Pharmacology.
They tested their theory at a pediatric facility in Israel where they created an interventional program and compared it with a control period to determine which drugs lead to AEs.
Over three months, the intervention study gathered data. Then they compared it with data gathered for 12 months before and after the study. The researchers found that the 12 month period before intervention showed an AE report rate of zero. It increased to 46 reports during the intervention but decreased significantly in the year after the intervention.
Hence, the conclusion that interventions and training need to be frequent.
The FDA is working on an electronic submission process to streamline AE reports for investigational new drug (IND) applications. The Technology Modernization Action Plan (TMAP), aims to modernize the ways the agency can process data for better regulatory governance, explains FDA principal deputy commissioner Dr. Amy Abernethy.
The agency is designing interfaces and tools to streamline the submission and review of data for improved safety of therapies, Abernethy adds. The IND safety reports should be submitted in accordance with data standard ICH E2B, which is consistent with the reporting format in the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines.
Because adverse events are not often reported or the process of reporting tends to be untimely, innovators have looked to use new technology to make improvements. The FDA MedWatcher Social system, for instance, uses automated tools to monitor reporting of AEs on social media and via crowdsourcing initiatives, according to an unpublished preprint at BioRxiv.
The system, which was launched in 2012, uses an indicator score approach to identify AEs. But researchers say this score can be improved and should be updated. Doing so will enhance automatic pharmacovigilance systems. The researchers regard this simply as best practice and consider any algorithmic designs for computational epidemiology should be regularly assessed and improved.
The days of paper-based AE reporting are on their way out. The team at the Patient Safety Network list web-based systems that integrate with electronic health records and other specialized systems for specific clinical settings. But as much as the tech matters, so too do the people. Key to an effective AE reporting system is a supportive environment that protects staff that report adverse events, encourages event reporting from various personnel and shares these reports within a suitable time frame.
After patients undergo cancer surgeries, many require post-acute care to treat adverse events. But patients also tend to be released from hospital and sent home sooner these days due to improved treatment provision. This makes it harder for medical teams to report AEs, which is why a system of monitoring is needed, explain researchers at the University of Bristol.
An effective means of achieving AE monitoring is through routine electronic patient-reported outcome (ePRO) data capture, they say. Using ePRO systems can help to manage patients by automatically notifying clinicians of any patient reported symptoms that may be related to AEs.
Patient data is more available to researchers than ever before. It can be accessed from patient EHRs, wearables and from their opinions and discussions on social media and in online groups. All of these sources provide information which can be used to track a new drug’s safety.
Real-world evidence and big data provide valuable information but pharmacologist Bruce Donzanti, senior group director of global pharmacovigilance innovation policy at Genentech, says all of this data brings new challenges.
Much time can be spent trawling the data for insightful analysis and trying to bridge the gaps in information that can arise. It’s debatable whether the time spent is worth it, says Donzanti. Efficiencies in safety operations and safety signal assessments are not guaranteed and there’s a risk of adding more data to safety database, which Donzanti notes could conceal true safety signals and rare events.
A better solution, then, is AI technologies, with machine learning and deep learning capabilities, he suggests. This is already being used to support pharmacovigilance activities in single case assessments for identifying AEs. The FDA and pharma industry is already investigating how algorithms can be used to improve the identification of adverse events that arise from more than a million individual reports a year.
Not all adverse events warrant reporting and it is necessary for sponsors to understand the difference between those that do and don’t.
In a special article at the Journal of Clinical Oncology, researchers say that sponsors that have improved reporting by focusing only on actionable AEs have done so by creating a company culture that eschews uninformative reporting, following consistent reporting standards, using a monitoring system to encourage company-wide AE reporting, and providing relevant AE reporting training to staff.
Reporting on actionable AEs is not only good for pharma organizations’ efficiency and resource allocation, but is in line with the FDA goal of reducing the number of uninformative reports. Indeed, the agency has been focused on reducing reports since 2010, when it published a final rule (2010 Final Rule) on expedited IND safety reporting.
Sponsors were submitting AE reports without evidence to show the drug had caused the event when the reality was that adverse events were caused by an underlying disease. Reporting adverse events accurately is essential for patient safety and improved drug development. Currently, much needs to be done to improve the reporting process and employing suitable technology will help streamline the reporting process.
Images by: antonioguillem/©123RF Stock Photo, adiruch/©123RF Stock Photo, dobledphoto/©123RF Stock Photo