… Through the work that I’ve been doing on gathering data from patients, it is becoming apparent that Patient Reported Outcomes (PRO) data is often of significantly greater clinical value than consumer health device data.
When most people think of digital health, they immediately think of consumer health devices. Fitbits. Blood pressure monitors. Heart rate monitors. Wi-fi enabled scales. All of the things that allow us to remotely monitor patient vital signs, either in the home or on the move. And as I’ve said before, the ability to capture this kind of data is undoubtedly a good thing, providing it doesn’t lead to further health information silos.
However, as I go about my work in the emerging digital health market, I meet a lot of organisations (particularly startups and new entrants to the health market, but also public health systems) who think that obtaining vast amounts of consumer health device data will somehow, magically, result in better health outcomes for patients.
It’s a classic case of what I call the ‘underpants gnomes strategy anti-pattern’…
In a classic episode of the ever-offensive South Park, the ‘underpants gnomes’ secretly set about stealing underpants in order to deliver a certain profit. The problem is, they’re not quite sure what the all important interim step, phase 2, will be. Similarly, the rush to gather vast amount of consumer health device data may seem to lead to good patient health outcomes, it’s just that not many people are giving thought to how…
And the more work that I’m doing on the ‘how’, the more I’m realising that there is really quite limited value in consumer health device data. As I said previously, such data has its place, which is largely in the remote management of pre-diagnosed patients, and in keeping them within defined boundaries in order to decide when to bring them in for an appointment. But there just isn’t great clinical value in most of the data that we gather. In fact, it’s worse than that. Self-measurement of vitals signs (such as blood pressure) can lead to undue worry for patients when values are outside recommended ranges, and in the worst case, self-diagnosis. Other measures, such as weight, can act as ‘negative reinforcers’, causing patients to lose heart or give up when they don’t experience progress in a measure that they are required to report.
Interestingly, through the work that I’ve been doing on gathering data from patients, it is becoming apparent that Patient Reported Outcomes (PRO) data is often of significantly greater clinical value than consumer health device data. Asking a patient how they are going (on a scale of 0 – 10) for measures like pain, nausea, fatigue, anxiety, depression, etc. turns out to be a better clinical indicator of a patient’s future health than any type consumer health data.
Although a patient’s PRO data is (by definition) subjective, it is normalised by the patient. Consequently, we are most interested in either high values, or noticeable changes in values. Such changes, over a sustained period, allow us to understand the changing risk profile of a patient and to respond appropriately. This facilitates clinical decision support systems that enable early intervention, patient risk stratification and the activation of a patient’s caregiver network (i.e. the ‘circle of care’).
All of this argument is to make that point that I think we’re putting too much effort into gathering consumer health device data, and not enough into gathering Patient Reported Outcomes data, which has far greater promise in terms of its clinical value. I know that consumer health devices are cool right now, but we mustn’t get distracted by the latest shiny thing. As we evaluate new technologies in the digital health toolbox then we must constantly be assessing how they can be used as part of intentional design processes to deliver the best possible outcomes for patients through innovative models of care.