Blood Establishment Computer Software Devices Need Rigorous Safety & Effectiveness Reviews

May 31, 2016

Division of Dockets Management

Food and Drug Administration

5630 Fishers Lane, Room 1061 (HFA-305)

Rockville, Maryland 20852

Comments of the National Center for Health Research
on the proposed rule
“Classification of Blood Establishment Computer Software and Accessories”
[Docket No. FDA-2016-N-0406]


The National Center for Health Research (NCHR) opposes the proposed rule for Blood Establishment Computer Software (BECS) and Accessories, which would classify these devices as class II (special controls). Since there have been injuries and even a death associated with these devices, we strongly recommend they be reviewed through the more rigorous PMA (class III) approval process. PMA applications would provide physicians, researchers, and patients detailed information about the studies and tests submitted to approve the devices—information which is lacking on BECS devices that are currently cleared by the 510(k) process.

“BECS and BECS accessory devices are used in the manufacture of blood and blood components to assist in the prevention of disease in humans by identifying unsuitable blood donors.”[i]  The devices have been used to screen blood transfusions for HIV.

FDA cleared the first BECS as 510(k)s in 1996 as substantially equivalent to devices marketed prior to the 1976 medical device amendments.  However, there is no publicly available data to evaluate the safety and effectiveness of the predicate device. In fact, in a 1997 substantially equivalent letter from FDA to the American Blood Resources Association, the FDA does not even name the predicate device or identify it in any other manner.[ii]

FDA said that it cleared 220 BECS and BECS accessories under the 510(k) program. We examined FDA data for numerous BECS cleared by the 510(k) and none of them provided specific data about the safety and effectiveness of the devices because most of the 510(k) summary links were not available.

We partially agree with comments made by the Blood Product Advisory Committee (BPAC) at its December 3, 2014 meeting. We agree with BPAC’s statement that “BECS and BECS accessories presented a potential unreasonable risk of illness, injury, or death.”  However, we disagree with BPAC that “sufficient information exists to establish [class II] special controls for these devices.”1 If there is sufficient information, it is not available in the FDA’s 510(k) publicly accessible database.

In deciding on class II with special controls for BECS and BECS accessories, the BPAC considered an FDA review of scientific literature, recalls, adverse events reports, and “the history of safety and effectiveness of BECS and BECS accessories over many years of use.”  Safety history is not a solid basis to judge the safety of future devices that are constantly being changed (as a function of the iterative nature of medical devices).  For example, artificial hips had a good safety history until a change was made and metal-on-metal hips were marketed causing extreme injuries to patients.

Due to the voluntary nature of Manufacturer and User Facility Device Experience (MAUDE) reports, the true safety and effectiveness history of BECS is unknown.  Experts agree that most adverse events are never reported to MAUDE.  Even with undercounting, the FDA noted that from 1996 to 2014, it received 201 adverse events reports for BECS and BECS accessories (including one death and nine injuries).  In the 1990s, the FDA also found “that unsuitable blood and blood components had been released and distributed as a result of improperly designed software.”1

Recently, our Center examined FDA-reported recalls of medical software and we identified more than 600 software devices and 1.4 million units recalled for moderate or high risk patient safety issues. We are very concerned about the inadequacy of the 510(k) review of many medical software devices, which is why we recommend BECS and BECS accessories be categorized as class III devices.  Class III requires more rigorous evidence that BECS actually work and that evidence would be publicly available in the PMA database.

FDA currently lists four risks associated with BECS and BECS accessories: transfusion reaction or death due to “inadvertent release” of incompatible blood for transfusions; injury from transfusion of inaccurately labeled or stored blood components; injury or death from the release of blood from ineligible donors; and injuries to donors from inappropriate or excessive donation of blood.

FDA has proposed five special controls to mitigate the above risks: software performance requirements, validation testing, labeling about software limitations, traceability matrix, and performance testing. We are concerned that information about mitigating risks will never be shared with physicians, researchers, or patients because of the lack of information presented in 510(k) summaries (and in many cases, the lack of a publicly available summary altogether).  In 2014, our Center published a study in JAMA Internal Medicine, which found that despite the legal requirement that scientific evidence of substantial equivalence be publicly available for medical devices cleared by the FDA through the 501(k) process, such information is usually lacking.[iii]

For the above reasons, we oppose the proposed rule to classify BECs and BECS accessories as class II (special) control devices.  We strongly recommend the devices be classified as class III and require PMA approval.

National Center for Health Research

NCHR is dedicated to improving the health and safety of adults and children by using research to develop more effective treatments and polices. NCHR can be reached through Paul Brown at

[i] Federal Register (March 1, 2016).  Proposed Rule: Medical Devices; Hematology and Pathology Devices; Classification of Blood Establishment Computer Software and Accessories [Docket No. FDA-2016-N-0406].

[ii] Food and Drug Administration (March 20, 1997).  Letter to American Blood Resources Association (ABRS) stating FDA’s determination that ABRS’s device is substantially equivalent to devices marked prior to May 28, 1976 (BK960030 Letter).

[iii] Zuckerman D, (September 29, 2014). Lack of Publicly Available Scientific Evidence on the Safety and Effectiveness of Implanted Medical Devices,  JAMA Internal Medicine.