Eighteen luminaries (including Google co-founder Larry Page) recently developed a list of 14 Grand Engineering Challenges which they consider "essential for humanity to flourish." A couple of the challenges seem a bit...odd (Enhance Virtual Reality, Reverse Engineer the Brain), but one stands out from the rest (at least, for the Labs): Advance Health Informatics.
Labs researchers David Kil and Baiju Shah believe they have identified a significant opportunity in health informatics (i.e., use of information technology in health management). First, they think it should be possible to take demographic data (income, type of car, marital status, etc.) and claims data (like that already collected by your insurer) and correlate it with the likely future onset of various conditions (diabetes, high blood pressure, etc.). They've developed a tool that can make this type of prediction, actually--it's part of a suite called Integrated Health Management (IHM).
What would you do with these predictions once you had them? Well, that's the second part. The researchers contend that insurance companies need to become better at customer engagement and measuring outcomes. Today, the primary customer engagement channel is the nurse's call---a relatively expensive option used only in especially serious cases---that is, only when a patient is in danger of soon developing (or is in fact already hospitalized for) an expensive-to-treat condition. What David and Baiju are envisioning, by contrast, is a whole range of persuasive customer contact channels, some very inexpensive indeed. For example: phone calls, letters, e-mails, electronic "life coaches" (sort of Jiminy Crickets with sensors for heart rate and caloric expenditure, among other things), and SMS messages.
(By the way, in case you're (un)concerned: While it is illegal in the United States for insurers to use some types of demographic information to set premiums, they are allowed to use it for other purposes.)
Once you have cheap channels of engagement, it becomes possible to intervene in much less serious cases. For example, a "pre-diabetic" can benefit from nutritional and behavioral counseling---except that such counseling may not be covered by a benefit and would therefore be unavailable. This is because it may be literally cheaper for your insurer to pay for dialysis years from now than to provide counseling today. Insurers---like all businesses---operate on cost-benefit analysis, and today that analysis works against a large class of customers that could benefit from preventive intervention.
By contrast, sending out daily SMS messages that remind me to keep my calorie consumption down (plus a certainly futile exhortation to hula for 15 minutes) is fairly cheap---and may even keep me from developing diabetes. (Making these messages effective is part of why David and Baiju think insurers have to get a lot better at customer engagement.) Cost-benefit is still in the picture: The system calculates the ROI for SMS, e-mail, etc. to facilitate selection of the best-yet-cheapest channel for each customer/condition, taking into account the outcomes of previous interventions.
So that's it: better data for predicting onset of conditions and cheaper channels of engagement so that more people can get interventions and stay healthy---plus ROI calculations based on continuous measurement of past outcomes. Couldn't be simpler. (Much, anyway.)
One more point: IHM also works for optimizing consumer incentives, e.g., when you're paid to behave better (stop smoking, lose weight) or to do inconvenient things like take a health survey or wear some sort of monitor.
Accenture Technology Labs is actively looking for partners with whom to run a pilot of IHM. If you'd like to learn more, please get in touch with David Kil or Baiju Shah.