Intelligent Engagement, through the use of virtual health assistants in conjunction with a variety of adherence technologies hold the promise to create an overall strategy for an intelligent adherence plecosystem.
Third in a Series: Understanding Health Behavior Theories
The first 2 articles in this series demonstrated an overall decrease in our nation's health since the war generation, the development of a plethora of health behavior theories, the conclusion of 2 major studies outlining our failings in trying to put health behavior theories into practice and finally my conclusion that an integrated technological approach is needed.
Numerous technologies are being marketed as the solution to our nation’s adherence problem. They can be broadly classified as:
The top 2—EHR and electronic prescribing—are for the most part, physician-focused attempts to contribute to adherence and improve overall quality of care. In particular, the EHR can trigger “target lists” of patients who need to be seen for a specific disease or condition. But in a recent Medical Economics article, author Ken Terry stated that the use of EHR to fill gaps in patient care, which includes control of a number of common chronic diseases, “remains confined to a small percentage of practices.”
Electronic prescribing loops in the retail pharmacist, who can then tie-in the retail pharmacy systems and processes to remind people to pick up their prescriptions and perhaps provide health behavior counseling. Although there is some evidence that these processes work, this approach cannot effectively address more than just the medication component of treating chronic disease. Another issue is that anyone even casually watching a pharmacy actually dispense the drug sees most patients “pass” on the counseling question built into the payment process in most pharmacies.
Telehealth (also termed telemedicine) is challenged with a clear definition, but basically consists of telecommunication, video conferencing, and remote monitoring (also termed telemonitoring). Much of the contemporary discussion on this subject revolves around the licensing requirements, payment issues, and the cobweb of state laws. But, there is much more. Much of telehealth involves nurse coaching, which the literature supports as effective but expensive. A recent review article entitled The eHealth Enhanced Chronic Care Model: A Theory Derivation Approach offers insight on numerous studies that demonstrated efficacy.
Telemonitoring has been around for more than 2 decades and grew rapidly after the widespread adoption of the Internet. Numerous articles have been published proving the overall utility of this technology. One British review of published studies found that telehealth had a significant positive effect on knowledge, social support, behavioral outcomes, and clinical outcomes.
Predictive analytics/modeling and adherence risk scoring are offered by a number of software vendors either as a service or as a software solution. These services rely on past medical and pharmacy claims and health risk assessments. Companies such as FICO and Optum use claims based as well as publically available data such as a physical address and other predictors, including retail purchase behavior, geo-credit profiles, and income/wealth indicators to create profiles on populations.
These predictive services assist health systems, large retail pharmacy chains, and managed care companies in predicting who might be less likely to adhere to medication or be nearing an acute event such as a hospital admission to again create a target list for outreach. But a person-to-person intervention is still needed in order to interact with the patient; a very resource-intense process.
Numerous organizations now offer IVR or text-based reminder services. The reminders are for daily meds as well as refills. The article referenced earlier found “modest health improvement using text messaging as a targeted intervention.”
In general, these services are one-way “push” messages. Two-way systems rely on a “word recognition” technology to attempt to understand any return message. Because of the limitations of these technologies, the ability of these systems to actually understand any return text is limited and most of these systems can only understand simple “yes” and “no” return text messages. For instance, if the text asked “Did you take your medication?” and the patient responded “Yes, and I think I am having a heart attack,” the system would understand the “Yes.” The, “I think I am having a heart attack,” would be captured but in most cases the true intent of the reply requires a human to review, interpret, and act upon; hopefully, earlier rather than later!
Cloud-based systems, for the most part, provide educational content and answers to questions using a FAQ format or search function. These systems also allow a patient to enter specific information into a database in a self-service manner; they rely on the patient to access the website, look up the information and type in the data. Some, such as one provided by Omada Health, is paired with a coach who assists with the training and goal setting.
There are literally tens of thousands of health apps available for smart phones. Many of these claim to be able to improve behavior and, ultimately, outcome. To date, most are little more than text-based reminders combined with a personal diary to record whether the medication was either taken or not taken, the date and dose. Many include a refill function. Apps are limited in their ability to analyze complex behaviors and responses. A search of PubMed will quickly demonstrate variety of digital apps to engage and improve health. Many have had positive results, but the number of subjects in the studies tends to be small with most studies suggesting further study.
The authors of a recent article in the Journal of Medical Internet Research reviewed 2254 articles and finally settled on a sample of 334 articles to analyze at the behavioral functionality of mobile apps in healthcare. In their research, they discovered that the most common behavioral health approaches used for the strategy of the app were: self-monitoring, cues to action; feedback; social support; social cognitive theory; and self determination theory. Other important findings were that acceptance was improved with “ease of use, limited time (needed) per use and… alerts to action." But, a primary conclusion of the authors was: “there is little evidence that apps are well received by users.”
The authors suggested a future study focused on combining multiple apps into a single intervention, something that is lacking in the current marketplace where consumers need one app for diet, one for exercise, another for diabetes management, and perhaps a 4th for hypertension; not a very desirable approach and one that lacks System 1 thinking!
Another category of technology taking aim at adherence is generally referred to as “smart packaging.” Starting with the legacy blister packs, electronics, wireless systems, and the Internet, smart packaging has led to a revolution in this area, which requires an entire article.
Next: Smart Packaging