In its Five-Year Forward View released in October 2014, the UK’s National Health Service (NHS) announced its intention to help people manage their own health. In keeping with this, the NHS has launched various self-care programs and has begun measure how effective these are through a patient activation measure (PAM).
By Clive Flashman, Global Head of Mobility for Healthcare and Life Sciences, CSC
As I discussed in my last blog, the journey patients take from engaged to empowered to activated is an important one, and PAM is an interesting resource in trying to discern their level of engagement. However, where similar tools often fail is in contextualizing the patient’s situation. By contextualization, I mean that the tool is relevant to where a patient is in his or her journey.
For example, the resources and information provided to a patient newly diagnosed with type 2 diabetes must be different from those directed to a patient who has been managing his or her diabetes for 10 years. The first person needs to understand a range of information, such as how the disease affects the body, what types of diets are needed, what should be avoid, etc. For the second person, who is potentially relatively expert at managing his or her health and wellness, the types of information might include new techniques for managing diabetes or new research on the condition.
That contextualization of information to patients and their particular set of circumstances can be software-enabled and relevant to the patient at any point in time. Let’s say, for argument’s sake, the patient has been told by the doctor that he or she is hypertensive and needs to lose weight. That’s all well and good when the patient is at home, but when out for a meal, the patient might not know what menu choices are healthier.
Now let’s say the person is going out for a Chinese meal. Typically, people who are trying to document their health and wellness might keep a food diary such as MyFitnessPal. The person might then log in and note what was eaten at the Chinese restaurant and calculate the calories and fat percentage. But it’s too late. Instead, what if that person had checked into the restaurant on Facebook and was then prompted, with the aid of an app, with the healthier menu choices?
There’s a fine line between an app that provides a useful way to pre-empt bad decisions and one that’s simply an annoyance. The way to manage that balance is through cognitive behavioral assessment: Does the individual respond better to prompts via a text message? Would the person respond better to a short video? And what timing makes most sense? Would the patient prefer a reminder about being on a certain diet and the need to be careful as soon as that person checks in on Facebook or an SMS 10 minutes later? In other words, what should be the format, timing and tone of those cognitive nudges that people need to help them maintain their own path to improved health and awareness?
The Age of Artificial Intelligence
The mass of data gathered by patients themselves, combined with the progression of artificial intelligence, is likely to be the enabler that helps patients manage their health better and become engaged, empowered and activated.
Now let’s return to the patient who goes out to eat. Let’s assume that individual has been keeping a food journal. If, every time that person goes out to eat, his or her blood glucose spikes and the person doesn’t feel well, the information recorded in the food journal could be replayed the next time that person plans to go out, as a reminder about how overindulgence feels.
In this situation, artificial intelligence is used to predict when a patient is more likely to want to see messages in a certain format or tone. It can be used to predict the way the patient will cognitively respond to the types of messages an application might want to send about how to manage health and wellness. And it can learn from the ways the user actually does respond to these cognitive nudges.
With authentication protocols such as Open ID, it becomes possible to access information across the various apps people might use to determine their state of mind and assess their receptiveness to certain types of message. Did that person post something positive on Twitter or Facebook? If so, the person might be more open at that point to talking about how to cope with health and wellness. Personal sentiment analysis is likely to be fundamental in assessing the behavioral context of app users.
Using such data to engage with patients is not without controversy, and certainly conversations need to be held about ethics, governance and privacy. But such apps have huge potential for helping clinicians gain a clear and honest perspective of their patients. It’s a step that might scare some people: Do they really want an app to know who and how they are? However, if patients and clinicians want the help a digital tool can provide, perhaps it’s a step more people will have to be willing to take.
As I said in my previous blog, it’s all down to whether it’s in the individual’s personal interest to embrace a tool that can help change behavior and improve his or her overall health.
In my next blog, I will explore possible concerns doctors may have about such apps and how these tools can be put to use by doctors to engage with their patients.
Clive Flashman will be at HIMSS17, February 19-23 in Orlando, CSC Booth 2773.