AABB News: The Future of Hemovigilance

June 17, 2021

Note: This article originally appeared in the June 2021 issue of AABB News, a member benefit of AABB.

Since the introduction of blood transfusion into routine clinical practice about 100 years ago, practitioners have been aware of the risk for transfusion-associated adverse events.

“The spectrum of these events has changed over the decades, but accrediting bodies, regulatory agencies and professional societies all have expectations that transfusion services are monitoring for these events and doing due diligence in the work-up of the cases,” explained Chester Andrzejewski, MD, PhD, FCAP, medical director of Baystate Health’s System Blood Banking and Transfusion/Apheresis Medicine Services, and assistant professor of pathology at University of Massachusetts Medical School-Baystate in Springfield, MA.

The United States has required mandatory reporting of transfusion and blood collection deaths since 1976.1 However, it was only more recently — in 2010 — that the Centers for Disease Control and Prevention (CDC) created the National Healthcare Safety Network (NHSN), which houses the U.S.’s national hemovigilance module (NHSN HM). The NHSN HM is a voluntary, passive system available for reporting of transfusion-related adverse reactions.

In the first year of the NHSN HM, 82 facilities enrolled; by 2016 that number had increased to 277, a minority of the more than 4,600 acute care facilities in the U.S. One-quarter of enrolled facilities are located in Massachusetts, where participation is mandatory. Reporting on total transfused components by enrolled facilities is estimated to represent about 10% of units transfused in the U.S.2

Without greater numbers of U.S. hospitals participating and contributing data to this effort, however, opportunities to improve transfusion safety via national hemovigilance reporting and monitoring in the U.S. remain to be fully realized, Andrzejewski said, including efforts directed toward more in-depth routine analyses of large amounts of data that can further inform reaction taxonomy standardization, the identification and capture of additional data elements of interest pertinent to hemovigilance, and the potential for the introduction of automation involving hospital computer systems to help streamline data entry in the reporting process.

The Benefits of Tracking
The aim of any hemovigilance reporting system is to improve transfusion safety by analyzing reports of adverse events and errors in transfusion procedures and by making recommendations for changes in practice at the national level, explained Michael Murphy, MD, FRCP, FRCPath, FFPath, professor of transfusion medicine at the University of Oxford, and consultant hematologist for NHS Blood & Transplant and the Oxford University Hospitals NHS Foundation Trust. 

In the United Kingdom, an increased focus on the transfusion process was prompted in part by a small study published in the early 1990s. The study surveyed 400 hematology laboratories in Great Britain and found that one-third of responding laboratories reported incidents in which patients received the wrong blood.3

“This finding showed these incidents resulting from errors in transfusion procedures were more frequent than expected,” Murphy said. “We began to think about interventions to stop them.”

Indeed, hemovigilance efforts in many countries began to attempt to track some of the commonest and deadliest transfusion-related reactions. 

“When I begin a discussion on transfusion reactions, I start with the commonest [1 in 10 risk range] events,” said Christine Cserti-Gazdewich, MD, FRCPC, assistant professor in the division of hematopathology and transfusion medicine at the University of Toronto, Canada. 

The likeliest acute (same-day) disturbances to be on watch for include transfusion-associated circulatory overload (TACO), minor allergic transfusion reactions and febrile non-hemolytic transfusion reactions, according to Cserti-Gazdewich. 

“These reactions happen in real-time, during, within hours of, or by the end of the day of the transfusion,” she said. “There is a lot of promise for how technologies can screen for these top three most common transfusion disturbances.”

The more daunting — less common but more serious — tier of reactions or adverse events are the ones that occur in the 1 in 1,000 to 1 in 100,000 risk range. These include transfusion-related acute lung injury (TRALI), higher-grade allergic reactions like anaphylaxis, bacterial contamination associated with transfusion (BaCon) or errors in the blood product received.

Tracking these adverse reactions is important, Cserti-Gazdewich said, especially given that even tracking the numbers likely underestimates the associated morbidity and mortality for each reaction.

“If a reaction increases the length of stay or throws a patient into the ICU, these could be associated with bad outcomes,” she said. “If a reaction occurs, even in the 1 in 1,000 range, if you multiply that number by the number of units administered in each country, that adds up to
a tremendous number of lives and a lot of traumatizing events for patients.”

Barriers To Reporting
“The more we know and the better our numbers are, the more likely they will reflect the actual scope of the problem,” Cserti-Gazdewich said. 

However, one of the biggest barriers to accurate reporting, at least in the U.S., is voluntary participation in programs like the NHSN HM. 

In stark contrast is reporting in the U.K. to its national program, Serious Hazards of Transfusion (SHOT), which tracks not just errors in transfusion but all adverse events related to transfusion. 

“Through this collection of data around all hospitals in the U.K., we can analyze the data, present findings in an annual report, and get people thinking about interventions to reduce these events,” Murphy explained. “For example, the use of only male fresh frozen plasma to reduce the risk of TRALI came out of an analysis of SHOT data.”4 

Each year SHOT releases an annual report and summary. In 2019 — among other important information — it reported 17 transfusion-related deaths including five that could have been prevented and confirmed that TACO and delays in providing blood are the most prevalent causes of transfusion-related deaths year over the year.5 

Reports on these adverse reactions in the U.S. are more challenging, according to Aaron Hettinger, MD, medical director and director of cognitive informatics at MedStar Health Research Institute. 

“There are challenges related to not only detecting the cases, but getting the right information, formatting it and submitting it,” he said. “There might be inconsistencies across health systems, across team members or even across departments. There need to be efforts for widespread standardization.” 

This lack of standardization and the burdens associated with reporting in the U.S. leaves a very incomplete picture of safety events. 

Some of the burdens include finding the right information in the electronic health record (EHR), transferring the data in a reporting format, and securely transmitting the data. According to Hettinger, these can be a challenge for those who report infrequently and are less familiar with the process, or time consuming for those that oversee large programs that may have frequent reporting needs. 

“These [hemovigilance] programs have the potential to be helpful and may be free to join, but practices are then responsible for the equipment, personnel to input data, and other associated costs,” said Jason E. Crane, DO, Medical Director of Vitalant-Illinois. “Anytime you are requiring something like this, it will come down to whether practices have the money and staff available to participate.” 

In addition, reporting hemovigilance data to the CDC does not replace mandatory or regulatory reporting requirements to FDA, or the practices’ state agencies, or any other agency with required reporting. 

Automation
Although the U.K.’s SHOT program in considered to be a successful example of a hemovigilance program, it does not collect data electronically, Murphy said. 

 “That idea is of interest,” he said. “Whether trying to collect the data that is currently reported to the electronic health records will be effective in picking up errors and adverse events has yet to be proven. It may be that some aspects will and some won’t be well picked-up. We just don’t know yet.” 

One of the ways that the U.S. is attempting to streamline hemovigilance reporting is with FDA’s Biologics Effectiveness and Safety (BEST) Initiative. The vision for BEST is for it to be “the pre-eminent resource for evaluating biologic product safety and effectiveness that leverages high-quality data, analytics and innovation to enhance surveillance, real-world evidence generation, and clinical practice that benefits patients.” 

According to Barbee I. Whitaker, PhD, lead general health scientist, Office of Biostatistics and Epidemiology (OBE) within FDA’s Center for Biologics Evaluation and Research (CBER), the FDA is cognizant of the burden on transfusion services for hemovigilance reporting. 

“It is our goal to make it as easy as possible to conduct safety reporting so that we are able to ensure blood and blood products are safe and effective and to identify emerging risks,” Whitaker said. 

Hussein Ezzeldin, PhD, senior staff fellow, co-lead of BEST Innovative Methods (IM) initiative OBE/CBER, said that the hope for BEST is that it will reduce burden on clinical transfusion staff, enhance detection of transfusion events, and provide a standardized review of the cases by the transfusion medicine clinicians. 

Currently, the BEST program is contracting with companies to develop methods and tools, BEST prototype, for using electronic health records to establish semi-automated adverse events reporting for therapeutics including blood and blood products. 

“Electronic health records are a wealth of information and this semi-automated program is hoping to pull out pertinent information and have it in front of the clinician to review,” Hettinger explained. 

The program is currently in a pilot phase with data being reviewed and validated to make sure it is meaningful in order to fine tune the artificial intelligence (AI) algorithms that aim to assist clinicians and practitioners. 

Hettinger is involved in projects designed to help understand the workflow involved in gathering hemovigilance data and reporting it to the right places. 

“We are trying to understand the gap between work as imagined and as performed,” he said. “The FDA is trying to understand that workflow, cut down on excess burden and make it easier to report.” 

At the 2019 AABB Annual Meeting, Whitaker reported information on the BEST initiative. In the study, the program’s algorithms were trained using 727 transfusion adverse events documented in NHSN hemovigilance reporting from 2014 to 2018. Subsequent active learning-based development and validation will allow the system to use automated transfusion adverse events detection from electronic health record data.6 

The auto-detection will flag and rank cases to be reviewed by the clinicians. This automation will attempt to modernize the current manual process clinicians undergo by automatically populating supporting data for a case and facilitating semi-automated verification and reporting of adverse events. The process can be complemented with computational phenotypes, which are EHR-based algorithms to flag possible adverse events that can then be selectively reviewed in more detail, optimizing reviewer time. 

“The more we can do to get consistent, accurate and timely reporting across the country, the better off we will be,” Hettinger said. 

Additionally, Ezzeldin said, after this review and clinician assessment, with a click of a button, the software will automatically fill out the clinical information to create an Individual Case Safety Report (ICSR). After the reporter reviews the ICSR and the software runs a last check to ensure it completeness, it produces an ICSR xml file to send to the FDA. 

“This will reduce a lot of time and effort by bypassing the information collection and re-entry steps into creating the ICSR,” Ezzeldin said. 

The system remains in its pilot phase and has some additional hurdles to clear before it can be extended further and is ready to scale on a national level, Ezzeldin said. 

Health Benefit
“Anytime we can use data to help determine the best practices for transfusion, we may be able reduce the incidence and cost of adverse reactions,” Crane said. “These efforts should be looked at as both a cost-savings and a lifesaving effort.” 

Cserti-Gazdewich added that although these efforts would likely have a larger public health benefit, it is also important to remember the benefit that any avoided reactions would have on each individual patient. She described seeing a patient chart recently that described the patient having post-traumatic stress after a transfusion-related adverse reaction. 

“Even though I walk among my patients every day, I haven’t imagined how often flashbacks and fears affect them,” she said. 

Cserti-Gazdewich said she also has a greater understanding for how one bad experience can negatively bias future encounters for patients who are transfusion-dependent, describing the worry of recurrences as “heartbreaking.” 

Despite the work left to be done, there is hope for the further improvement of hemovigilance and transfusion safety, Cserti-Gazdewich said. TRALI is one of the best examples of how studying the problem led to a great solution. 

“We are getting better at all of this, and I maintain hope that we will continue to identify areas for potential improvement and will act quickly and effectively to do things better,” Cserti- Gazdewich said.
Endnotes
1. Vamvakas EC, Blajchman MA. Transfusion-related mortality: the ongoing risks of allogeneic blood transfusion and the available strategies for their prevention. Blood. 2009;113:3406-3417.
2. Edens C, Haass KA, Cumming M, et al. Evaluation of the National Healthcare Safety Network Hemovigilance Module for transfusionrelated adverse reactions in the United States. Transfusion. 2019;59:524-533. 
3. McClelland DBL, Phillips P. Errors in blood transfusion in Britain: survey of hospital haematology departments. BMJ. 1994;308:1205-1206.
4. Chapman CE, Stainsby D, Jones H, et al. Ten years of hemovigilance reports of transfusion-related acute lung injury in the United Kingdom and the impact of preferential use of male donor plasma. Transfusion. 2009;49:440-452.  
5. SHOT Annual SHOT Report 2019 Summary. https://www.shotuk.org/wp-content/uploads/myimages/2019-SHOT-Summary.pdf. Accessed May 17, 2021.
6. Whitaker BI presented on behalf of FDA CBER Biologics Effectiveness and Safety Initiative  (BEST). OA4-SN5-36: Innovative hemovigilance methods: The FDA CBER Biologics Effectiveness and Safety (BEST) Initiative. Presented at: 2019 AABB Annual Meeting.