“Hacking” Healthcare

8 09 2017
[Photo: Mayberry Health and Home]

I wish I had thought of this title before anyone else, but there is a book, scores of domains registered already, a lot of hand-wringing about evil hackers and I’m sure, a lot of events like this where people will try to hack healthcare to win contests. But this article isn’t really going to be about any of those things. This will be the first of a series of articles about how communities, thinkers and smart startups are taking back control of their healthcare.

Why does healthcare have to be hacked? In case it wasn’t painfully obvious, the state of healthcare in the US in 2017 is in basically the same shape that it has been for some time: too expensive, mediocre in terms of outcomes and nowhere near as safe as it should be. Compound that with epic gridlock in Washington and it becomes clear that these problems won’t be easily solved via policy. And yet, these constraints are forcing creativity. Think of it as the DIY movement in healthcare. In art, engineering and just about every endeavor, DIY is a concept that is near and dear to my heart and I want to celebrate it when it happens in healthcare.

So, let’s start with AI. So much hype.


Despite this hype, there is plenty to work with. While yes, AI is getting better at looking at x-rays or suggesting treatment options to physicians, the real value today comes from augmenting rather than replacing the caregiver. While teaching myself API.AI for a non-healthcare related project, I could see first hand the value of well designed tools to extend the caregiver relationship, leveraging the principles of CRM as well as the challenges inherent in training them. I like the chatbot, even though they still have a ways to go. Chatbots are hard to build well. If you’ve interacted with one, you likely figured it out pretty early on and hated it. But the technology is getting better and you will likely be engaging with many more bots  at your work or when you are applying for a job.

Some excellent use cases for bots include: improving adherence with provider recommendations (medication, diet, exercise) by gently nudging the patient via text messages to their phone or Facebook Messenger. Yes, HIPAA is hard. But it’s not insurmountable with a well crafted strategy. Also, most patients aren’t worried about the occasional reminder to eat better or work out via their phone after they leave the doctor’s office. And they work.


Another use case I love is augmenting a therapist for treating minor depression and other low-acuity behavioral health issues. Why aren’t we creating pathways for patients to get screened for depression in their doctor’s office and handed to an appropriate resource that doesn’t have to be the PCP? We know that depression is often associated with other chronic illness and that depression can prevent patients from adhering to medication and treatment plans, or just not showing at their doctor appointments. We also know that one-third of the adults that have major depressive episodes never see a professional. In addition to the costs of untreated illness, millions are needlessly suffering.

Here’s a hack worth doing: evidence-based chatbot tools overseen by a team of clinicians and technologists, regularly monitoring that the AI is providing quality support to patients and that it is aligned to their care plans and appropriately linked to technology.

If you read this far and you’re of a certain age and time, then maybe you’ll get this reference. At least it’s not  stock photography of people in lab coats in front of computers.

Value Based Payments 101 – Bundled Payments

17 05 2013


Source: http://www.flickr.com/photos/opensourceway/

Value Based Payment 101: Bundled Payments

In order to dive into value based payments, we need to first understand how health care is commonly paid for today and why many believe this needs to change to improve health care.  Today, most health care is paid for by private health insurance, public health insurance (Medicare/Medicaid) or by the individual, out of pocket as fee for service, essentially paying on a per procedure basis.  If you go to your doctor for a simple issue like a sore throat, there’s a line item for the office visit and exam, a line for the procedure to swab your throat, the lab fee for someone to look at the swabbing and if necessary, the prescription that needs to get filled to cure your sore throat.

This happens in a exponentially more complex fashion if you go to the doctor and she determines for example,  that you have an issue with your heart and refers you to a cardiologist, who then recommends a surgical procedure at your local hospital.  Every item is billed to someone on a line by line basis.  It’s this complexity that was the subject of the recent Time article. But to take this further, let’s say you have to be readmitted to the hospital a week after being discharged and you spend 5 extra days in the hospital, requiring treatment by new specialists as well as your cardiologist.  Who pays?  Today, the payer has to cover the costs of what was a potentially avoidable readmission.  To be fair, people are readmitted to hospitals regularly for reasons that are not anyone’s fault, but the causes of many readmissions are preventable and the costs are real.  For example, 20% of all Medicare patients are readmitted to the hospital within 30 days of discharge at a cost of $18 billion a year, not to mention the risks and impact of any hospital admission on the patient and their family.

So, how do you get this to change?  Well there are carrots and sticks.  First, the stick.  One example is that the Center for Medicare/Medicaid Services or CMS, has started a Readmissions Reduction program that was conceived as part of the health care reform legislation.  In this program, officials are reviewing the number of heart attack, heart failure and pneumonia patients who return to the hospital within 30 days of discharge. Hospitals with more readmissions than Medicare expected given their mix of patients are penalized by losing up to 1 percent of their regular payments, with penalties going up in coming years. Over 2,000 hospitals are currently expected to be impacted by reduced rates.  You can see the latest list of hospitals and their readmission rates here.  Readmission rates are coming down already, some say in direct response to the program.

Another approach to improving quality and efficiency is through “bundled payments” that are meant to encourage cooperation between all members of the medical team.  As per my example, for a hip replacement Medicare typically pays a fee to the anesthesiologist, the surgeon, the nurses, radiologists, the hospital, suppliers etc.  Under bundled payments, one sum is paid for the entire procedure based on historical data.  The goal of bundling payments is to create an incentive for all those groups to work together and find efficiencies including negotiating rates with suppliers that allow the surgery to cost less than the bundled amount and share the savings as a profit to be shared among the team members.  That’s the carrot.

The main problems with bundled payments are that some diseases like diabetes are hard to lump into episodes that lend themselves to bundled approaches.  However, many procedures and illnesses can be, like cardiac procedures and hip replacements.  So, while it may not be immediately feasible for all illnesses, it certainly can be for some.  Another issue is that there are not many providers that are willing to accept this level of risk.  This is changing as the successes that are being seen in gain more visibility.  Finally, a third reason for not doing bundled payments is the lack of software that accommodates this type of billing.  Virtually all health care billing software is designed to handle traditional fee for service approaches today, but that is a function of the fact that almost all billing is fee for service today.  The market will most certainly meet this need if and when it is necessary.

More Reading

Taking value even further when looking at new technologies and drugs.

Lessons from Medicare’s Demonstration Projects on Value-Based Payment (Congressional Budget Office)