In our current business mythology of black swans, unicorns and software eating everything, everyone is either planning on or waiting for the “disruption” of healthcare. Spend time with entrepreneurs entering the field and you’ll hear a combination of exuberance and a wee bit of arrogance about the potential of the latest titanium wearable or cloud-based big data machine-learner or robotic doctor that’s going to blow up this dinosaur industry. It’s a mechanized menagerie out there! Before you launch a drone-based personalized medicine portal, here are a few considerations we make when we develop technology solutions for our clients, and we believe they apply broadly in product development for HIT.
Understand how you get paid. It’s not enough to get “Google smart” on the business: healthcare drives about twenty percent of the US economy, is growing at a rate that tends to outstrip inflation and is considered to have significant waste (at least thirty percent of healthcare services are thought duplicative or unnecessary), so there is definitely opportunity. The issue is understanding who specifically pays for the service and who gains value from potential savings or improved outcomes. Today, we have a reliance on third party intermediary payers who foot the bill, be they insurance companies or the taxpayer via CMS. With cost sharing on the increase, the patient pays more of the share of the bill directly. While this cost-shifting is an important factor in coming years, typically the vast majority of costs fall on an insurer. If you want to get paid for your idea, you’ll have to offer value to these stakeholders. The data in a Fitbit may be of interest to your doctor, but besides the challenge of integrating it accurately into a useful record, until there is clinically proven reason to have it, i.e. it improves the health of the patient or reduces the cost of care, there is no reason for the insurer to pay for it. With your technology, figure out who uses, who benefits and who pays – and align them.
Operate at scale. The good idea has to work across populations. The gold standard of scientific proof is the clinical trial, for good reason. While an n=1 can work for lifehacker bloggers, in order to implement an intervention that works with a diverse group of people in a variety of environments, it has to be tested thoughtfully across a good sample. You need to understand the concept of the Number Needed to Treat to appreciate how clinicians and payers will evaluate your approach. It’s our shared cognitive challenge – understanding how our individual experience may translate to the group’s experience – that requires well-structured pilots to assess the efficacy of every good idea. Just because the proposed idea works great for one person or small group doesn’t mean it works or should work across the spectrum; this is especially important as we think about the diversity of people even in a relatively small city or area. Income inequality, environmental factors and education are as impactful if not more to health than race and genetics. The power unleashed by software comes from using the tools to understand and manage complex adaptive systems.
The technology is (comparatively) easy. Culture is harder. If the idea delivers value across a diverse population, you still have to put it into operation. The patient/clinician relationship is still a complex mixture of humanity and science. While a robot or a video conference can do some things very well, it cannot provide the comfort and relationships that we all crave, especially when ill. Just because the algorithm does work at scale for the population, it may not be appropriate for the individual. Just as people dine in restaurants for more than efficient delivery of nutrients (slowing the adoption of some efficiency-enhancing technologies), we turn to our caregivers for counsel to address our whole selves and to benefit from their reasoned judgment best augmented but not supplanted by technology. And while the modern doctor or hospital visit can feel more like a factory than ever, it’s clearly not what the patient wants. We build technology solutions for and strongly recommend “disrupting” the experience of impersonal medicine.
It might sound curmudgeonly, but not a week goes by where we don’t read about the next Uber of this or that and find it hard to take it seriously. While we are truly living in a miraculous time, we have to keep in mind that the role of the provider is to be a partner to the patient and the technology that improves this relationship will be the winning technology.