The Unreliable Nature of Self-reporting
Self-reporting is notoriously unreliable and in recent years has been further complicated by the increasing number of so-called professional patients who attempt to enter clinical trials for the remuneration offered, with little intention of being compliant. As the inadequacy of self-report, pill count and pharmacokinetic measurements became apparent, novel technologies were applied to the problem. One of the first and best known of these was the use of the Medication Event Monitoring System device (MEMS). This is typically a medication container equipped with a closure that records the date and time of opening and closing the container.
While certainly a step forward, there are issues with the technology. Caps can be lost, the containers can be considered clumsy and difficult to use for multi-day dosing, and it may be difficult to distinguish between a missed dose from one doubled at the next container opening (1). Patients with arthritis may have difficulty using the device, and the MEMS device does not report adherence in real time. Therefore, intervention to deal with non-adherence is not possible until the problem has been identified by a review of the cap data. On occasion, when the subject or patient was not aware of the significance of the opening of the container closure, repeated openings and closings have been noted shortly prior to a scheduled study visit. Nonetheless this was a significant step forward in assessing compliance.
Can Electronic Devices encourage greater compliance?
A number of other electronic devices have been developed to address compliance, including electronic pill boxes (2). With the availability of smart phones and similar technology, several systems have been developed to trigger alerts and remind patients to take medication at the appropriate time. However, these serve simply as reminders and do not address the issue of ensuring compliance. Recently, AICure has introduced a novel visual recognition platform (similar to voice recognition with audio) to assist in both detecting non-compliance and ensuring adherence. Using an app on the trial subject’s smartphone (or one provided by the sponsor), the study medication is identified using the device’s camera.
Subsequently, the subject is identified and the ingestion of the study medication is detected. No video review or recording is made, as the software, using artificial intelligence does the actual observation. The data are available to the study site in real time, and missed doses can be identified immediately. Additionally, the software can send prompts to the subject notifying them of the time for dosing or of a missed dose. This technology has proven very successful and the automatic computer vision processing software on the server verifies the identity of the subject, verifies the identity of the medication, and confirms ingestion. The ability to access adherence data in real time allows for pre-emptive intervention on the part of the investigational site. Adherence rates in excess of 90 percent have been reported. Indeed one stroke study reported 100 percent compliance as confirmed by blood levels – something almost unheard of in outpatient trials (3).
Technological advancements offer hope, but problems still remain
Other technologies have recently been introduced to assist in managing investigational product compliance during clinical trials. Proteus is an ingestible microchip that is incorporated into the pill or tablet and ingested. It transmits a signal upon ingestion that is received by a patch worn by the patient. While the technology works well, particularly in inpatient settings, it does have a disadvantage in that some patients find the patch irritating and discontinue use.
Additionally there is a prolonged setup time to operationalize the device due to the need to incorporate the sensor into the drug product, and each new use requires FDA approval at this point. Xhale also has a product now available, where a chemical additive is incorporated into the drug product. At a set time following ingestion, the subject exhales into a small gas chromatograph whereby the additive can be detected. Similarly to the Proteus sensor, additional time is required in formulating the drug product to incorporate the marker, and the patient must remember to exhale at a fixed period of time after the drug product is ingested.
As technology continues to advance, the issue of detecting non-compliance with study medication in clinical trials continue to be important and a focus of ongoing research and development.
Senior Scientific Director, Global Clinical Science
To read part one of Tackling Drug Non Compliance, click here.
1) Olivieri NF et al., Compliance Assessed by the Medication Event Monitoring System. Archives of Diseases in Childhood, 66:1399-1402, 1991
2) Hayes TL et al., An electronic pillbox for continuous monitoring of medication adherence. Proceedings of the 27th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 30Aug-3 Sept 2006, New York
3) Labovitz DL et al., Using artificial intelligence to reduce the risk of nonadherence in patients on anticoagulation therapy. Stroke 48(5) 1416-1419, 2017