Companies that produce Class II devices know that special controls are both regulatory guideposts and a set of requirements that can be expensive to fulfill. In this blog, we’ll give our readers a look at how some of these controls stack up between similar products, but we’ll also take a look at the controls applied to a product that seems destined to serve as a bellwether for an emerging category of digital devices.
First, we’ll examine two novel digital devices with different indications for use but similar regulatory requirements. The first of these is the EndeavorRx digital therapy device for attention deficit/hyperactivity disorder, which was awarded a de novo in June 2020 to Akili Interactive Labs of Boston. The second is the June 2021 de novo for the Autism Spectrum Disorder (ASD) for diagnosis of autism spectrum disorder in pediatric patients, by Cognoa in Palo Alto, Calif.
Unfortunately, the FDA has not updated the regulation to reflect either of the products’ special controls as of the publication date of this blog. The Cognoa device’s special controls are Part 882.1491 of Title 21 of the Code of Federal Regulations while the EndeavorRx is at Part 882.5803. Since Part 882 lacks the regulations in question, we’ll offer the contents of both de novos to describe the FDA requirements. Both devices are subject to the requirements of Part 801.109, which deals with the exemptions from the usual directions for use for prescription devices.
For both products, device failure was a primary concern, which for the EndeavorRx is to be mitigated by labeling and software verification/validation. However, the FDA required clinical performance testing and software hazard analysis testing for the Cognoa ASD in addition to the mitigations set out for the EndeavorRx, because of a concomitant concern about incorrect analysis stemming from software failure.
The Cognoa has only one additional identified risk calling for mitigations, while the EndeavorRx has another five. This may be due to the fact that the EndeavorRx is a treatment device rather than a diagnostic device, but there is some overlap in the special controls. For both products, special controls include clinical performance testing, but the requirements begin to diverge after that.
There are only three groups of special controls listed in the de novo for the EndeavorRx and four for the Cognoa ASD, which may seem odd, given that the ASD is a diagnostic rather than a treatment device. The special controls for the EndeavorRx are:
- Clinical performance testing;
- A description of the software, including verification/validation data; and
- Product labeling.
The controls listed for the Cognoa system include both clinical performance and usability testing, while usability testing is not explicitly mentioned in the EndeavorRx de novo. It may be safe to assume that usability testing was part of the clinical study for the EndeavorRx, so it will be important to see whether the FDA provides more detail once the associated regulation appears at Part 882.
The labeling requirements for both de novos are substantial. For the EndeavorRx, these include:
- The use of images that demonstrate how to interact with the device;
- Minimal operating system requirements;
- A warning that the device is not intended as stand-alone therapy;
- A warning that the device is not a substitute for medication; and
- A summary of clinical performance testing.
For the Cognoa, the labeling requirements are somewhat more detailed, even if there are only three general labeling items. This is driven by the greater degree of specificity for the description of clinical testing to be included in labeling, which must include information on:
- Device output and clinical interpretation;
- Performance measures, such as sensitivity, specificity, and positive and negative predictive values;
- How cut-off values were determined for categorical classification of diagnoses; and
- Expected/observed adverse events and complications.
In both cases, the FDA has emphasized that the device is not intended as a stand-alone instrument for the intended use, but the special controls are at least somewhat determined by the intended use, even though both these devices are indicated for use outside of clinical settings.
CHIRP Heralds New Digital Device Category
One of the more novel products to come through FDA review programs in recent years is the Canary Health Implanted Reporting Processor (CHIRP), a digital device which is mated with the Persona knee replacement device manufactured by Zimmer. The need for this type of product was highlighted in a presentation made at the 2020 Current Concepts in Joint Replacement meeting. According to this summary, Daniel Berry of the Mayo Clinic, predicted that many orthopedic devices will have some sort of digital component over the next few years, and the Aug. 27, 2021, de novo for Canary Medical’s kinematic measurement device for knee replacements is a good example of this emerging trend.
The CHIRP device, along with the Canary tibial extension, is intended to provide clinicians with adjunctive kinematic data on the patient’s progress after total knee arthroplasty. Once again, the related regulation, Part 888.3600 has not yet appeared under the overarching regulation (Part 888), but the de novo lists several general controls, such as biocompatibility testing to control for infection and adverse tissue reaction. However, the list of special controls is considerable, which is expected for the digital component of what may be the first smart knee implant to reach the U.S. market.
Non-clinical testing requirements for this device included hermeticity, although it is not clear which bench and/or animal testing methods were used to demonstrate hermeticity. Preclinical testing was also used to ensure the accuracy, reliability and reproducibility of kinematic measurements. Overall, there are nine additional categories of special controls, and again, the most extensive of these is labeling, with eight specific requirements. Several of these are at least somewhat unique to this product code (QPP), such as:
- A statement regarding the limitations of the clinical significance of the kinematic data;
- Detailed instructions of the surgical technique needed to insert the device; and
- Shelf-life data.
Other special controls that may be more important to the QPP product code than to other digital devices are human factors testing for implant technique and information on device design and materials to ensure that the device’s geometric and material properties are appropriate for the intended use. Clearly the regulatory demands for this category of digital device are greater than those for other digital devices, but as is true for the EndeavorRx and the Cognoa de novos, the CHIRP is now available as a predicate for other devices under the 510(k) program, clearing a regulatory path for what will likely be a strong series of competitor digital products.