Dying in the US is not cheap. For many people in the US, the final year of life is often very expensive. Patients enter the emergency room, are put under observation, and get hospitalized for several days or more. This whole sequence can occur multiple times in that final year, ending with a short stint in hospice care – here’s a quick reference on end-of-life spending for Medicare, courtesy of Kaiser Permanente. A far better approach is to steer our parents / grandparents towards a more comfortable end-of-life experience, by reducing unnecessary hospitalizations, whenever possible.
Why end-of-life planning matters. This is not a unique story, but I’ll share it regardless. A friend of mine (middling 60s) was diagnosed with an aggressive form of kidney cancer. He was presented with the option of surgery, followed by chemotherapy. The doctors gave him his chances for remission, and he decided to go down that road. It was a difficult recovery to say the least. A short time later they detected cancer in his remaining kidney. Now he was faced with another surgery (to remove his last kidney) and needing dialysis for the rest of his life. In case it’s not clear, routine dialysis is awful – it keeps you alive but leaves you exhausted. So he continued down this road. And less than a year later, his cancer came back! There he is again, being presented with more “advanced medical options”, but he finally said enough is enough. As of this writing, he’s living each day with terminal cancer, and the last I saw him, he was talkative and comfortable.
This is an unfortunate (and common) example of a fruitless journey through the US medical system. We want to avoid these cases, whenever possible, but it’s much easier said than done. Critical decisions on the course of care begin with a conversation about death – a clear and frank discussion between a medical team and the patient, focused on the care options, potential outcomes, and chances for success. And to support these decisions, medical teams need the best information available, including advanced analytics and rigorous predictions of life expectancy.
We need better prediction tools. I’ve discussed hospice initiatives with many health systems, almost always on the accountable care side, and I observed a couple themes. First, there is no consistent approach to hospice or advanced care planning across health systems – for many systems, it’s simply not a priority, despite the potential to improve quality of care and reduce healthcare spending. To be fair, health systems that are not fully invested in accountable care have zero interest in reducing healthcare spending. And second, the health systems that do focus on hospice often refer their patients too late – unnecessary hospitalizations have already occurred, costs are racking up, and the patient spends less than a week in hospice.
In addition to the financial burden placed on the system, patients and family members often get stuck with a portion of end-of-life medical bills that can ruin their financial standing. For more details, see this article on “financial toxicity” among cancer patients.
Ultimately, we need to harness advanced predictive analytics to get a better handle on when to start preparing patients for hospice care. Here’s a popular article that points to limitations for predicting mortality from claims data. To be fair, there’s a bit of controversy as to how much end-of-life spending accounts for total healthcare spending, but there’s little disagreement that end-of-life spending is significant.
There’s a great deal of public discussion around how to pay for healthcare, that healthcare should not cost so much, and the possibility of a single-payer system. But with respect to end-of-life spending, I don’t think these concerns matter all that much – for any redesigned healthcare system, we must identify wasteful spending and steer away from it.
Who’s poised to solve this problem? So here is my shortlist of companies and organizations that I think are well positioned to tackle this issue. I consider 3 elements for building this list: (1) access to sufficient data; (2) technical sophistication with predictive analytics; and (3) mission-oriented commitment to addressing our nation’s healthcare spending crisis. I will add to this list over time…
AHRQ – This government agency has access to large volumes of medical claims and recently issued predictive modeling challenges that are open to the public.
Haven Healthcare – This organization provides health care to a large pool of employees from Amazon, Berkshire Hathaway, and JPMorgan Chase. Enough said.
Google + Ascension partnership – Google requires no introduction; Ascension is a very large healthcare system (they bring the data to this relationship).
Happy Readings ~ James
