From regulation to recalls: Why implant informatics matter

Q

One of the primary purposes of the Food and Drug Administration’s (FDA) unique device identification (UDI) rule was to improve recall management, especially for implantable devices. But confusion over how implants are defined, classified, used and documented could result in hospitals being unable to effectively identify patients who may have been adversely impacted by a recalled device. That despite regulations from both the Office of the National Coordinator for Health IT (ONC) and the Centers for Medicare and Medicaid Services (CMS) that require capture, storage and sharing of UDI-related data for implants in electronic health records (EHRs) and other technology.

In this month’s column, we will review the various issues and challenges, many of which have come to light as the result of continued work by members of various Association for Healthcare Resource & Materials Management (AHRMM) Learning UDI Community (LUC) work groups.

Supply vs. Implant documentation

For most healthcare systems, the item master (on premise or cloud-based) serves as the primary source of product data for EHRs. Products that are flagged as implants trigger those responsible for clinical documentation to include not only the UDI-DI (device identifier) but also the UDI-PI (production identifier) such as lot, serial number, expiration data in the patient record. Supplies not flagged as implants do not have this additional information recorded, and sometimes are not documented at all.

Once a determination is made and implant status is noted in the item master and made available to the EHR, it can be difficult for hospitals to change the status. Some systems allow an override, but that can also create further variation in how implants and associated data are managed.

Identifying implants

There is not a specific implant flag in the FDA’s Global UDI Database (GUDID) that indicates if a product is an implant, but there is a field for a procode, which manufacturers assign to their products during the regulatory approval process according to an FDA classification system. Those codes determine which pathway a product goes through to receive market clearance. You can download the procodes from the FDA website and see additional attributes including implant flags. An API on the accessgudid.nlm.nih.gov website also enables you to download a list of products with implant flags based on the procodes, although there are some inaccuracies in the data.

Problems arise when an assigned procode incorrectly identifies whether a product is an implant or not. For example, former Shock Trauma Implant Coordinator Trent Pierce found a product that contained no drugs had been assigned a procode for a medication vs. an implant. Some LUC members have convinced the FDA to correct such errors, while others have been unsuccessful, suggesting inconsistencies in FDA processes to report and request corrections.

Classification counts

The Device Classification and High Risk Implant LUC work groups recommend use of the Global Medical Device Nomenclature (GMDN) as a better taxonomy for implant identification. A required field in the GUDID, the GMDN code is assigned by manufacturers based primarily on the approved use of a device but it can also reflect actual use; unlike procodes, GMDN codes can be changed by the manufacturer without regulatory implications.

A common, but mistaken practice in many hospitals is to use the United Nations Standard Products and Services Code (UNSPSC) to designate if a product is a supply or implant. While there are codes for various medical implants, e.g., surgical, cosmetic, orthopedic, UNSPSC was designed to support spend analysis in e-commerce and not designations of clinical use.

Variation in product use

One of the biggest challenges in determining whether a product is an implant has to do with variations in how products are used, e.g., if they stay in the body and for how long. The U.S. FDA definition of an implant is as follows:

“A device is regarded as an implantable device for the purpose of this part only if it is intended to remain implanted continuously for a period of 30 days or more, unless the Commissioner of Food and Drugs determines otherwise in order to protect human health.”

In practice, some devices that are intended (by the manufacturer) to remain implanted for 30 days or more are removed sooner. As an example, K-wires are used to stabilize bone fragments. While often removed after the bone has healed, they are intended to stay in the body for 30 days or longer. Other times they are used only temporarily during surgery. In such cases, the K-wire, while originally intended for longer implantation, may be documented as a supply and not recorded with UDI-PI or at all, in the patient record.

In other cases, implanted products are designed to be absorbed by the body, often faster than 30 days, but they may remain intact in the body longer than anticipated, sometimes causing complications.

Combination products create further confusion. For example, there are separate procodes for bowel staplers (classified as supplies) and the staples that remain in the body (and classified as implants). But what about a stapler that is preloaded with staples? As noted above, how products are flagged – implant or not – designates the level of downstream documentation and the ability to follow up with patients if there is an adverse event or recall.

Recommendations

Despite these challenges, the promise of using structured and standardized product identifiers to improve patient safety remains. With the highest risk recalls on the rise in recent years (more than a 64 percent increase from 2016 to 2017), overcoming these challenges should be a priority.

The Recall LUC work group is tackling some of these issues, and several LUC members offer these unofficial recommendations:

1.  If in doubt as to whether an implantable product will be in the body longer than 30 days, err on the side of caution and document it as an implant. In the event of a recall, it is easier to rule out patient exposure than to potentially miss identifying an impacted patient.
2. Help clinical staff understand the value of fully documenting implants to increase patient safety.
3. Promote the use of scanning to reduce the time clinicians spend on supply documentation. Scanning takes as little as eight seconds compared to up to nine minutes for manual documentation.
4. Create mechanisms for front line staff to report problems with implant designation, and request the FDA provide guidance on how best to resolve such issues.
5. Join the AHRMM Learning UDI Community (www.ahrmm.org/luc) to learn about best practices and offer your own input to accelerate UDI and value.