In JAMA, Joan Teno gives us a report card on end of life care that can function as a kind of Rorschach test for change-makers: there’s the good (deaths in the hospital have decreased from about 50% 30 years ago to 19%); and the not-so-good (deaths in the ICU increased to 29%). The editorial, by Zeke Emanuel, is a ‘glass-full’ interpretation–and yes, this is real progress. I’ve also heard a couple of palliative care leaders say, we’ve done it, now it’s just a matter of time. That’s where I beg to differ: my own opinion is that these data point to new challenges that our recent innovations, like inpatient PC teams, don’t address. For example, fewer in-hospital deaths mean that a major responsibility for this care has moved outside the hospital–where palliative care is just starting to make inroads–and how the community services fit together with the hospital services is something we haven’t started to look at as a big ecosystem–yet. [Teno pdf here; Emanuel here]

The other implication of the Teno study is the emergence of two big patterns. One pattern for end of life care is ‘leave the hospital to get community-based care’; the other pattern is 'try everything and die in the ICU’. What is important about these two patterns is that they require very different kinds of care. Most patients in the first pattern have a growing awareness accompanied by stepwise adaptation and acceptance; most patients in the second pattern pursue continued or escalating opportunities for disease-modifying intervention until the last moment, when life-extending technology is withdrawn for an immanent death. Of course, these two patterns are caricatures, but the implications could influence how we design resources. Patients in the first pattern might benefit from more upstream psychological intervention, and their families need to learn how to manage death at home; whereas the second pattern need only a bit of upstream work, and their families need to learn how to be surrogates who can make decisions about life-extending technology. Is it time for us to be looking for these sorts of patterns? Could our analyses, like the international comparison in JAMA last year, be more pattern-centric and less resource-centric?

Because this data discloses your health status. Remember the story about the woman who began to get mailers from Target featuring baby products–before she knew she was pregnant? It’s happening on a vast scale. In a new report from Propublica and NPR: “[Data brokers] are tracking your race, education level, TV habits, marital status, net worth. They’re collecting what you post on social media, whether you’re behind on your bills, what you order online. Then they feed this information into complicated computer algorithms that spit out predictions about how much your health care could cost them.” Yikes.

In the discussions about shifting systems and providers towards value, the elephant in the room is administrative burden. Austin Frakt summarizes the costs of administrative overhead nicely in the New York Times–although the data he cites is quite old. My personal observation is that administrative costs have continued to increase. Lance Downing takes a provocative look at an unintended side effect of billing complexity in the Annals of Internal Medicine: physician notes in the US are about 400% longer than a comparison using physician notes from Canada, UK, UAE, Denmark, and Singapore [full article here].

Of course, as clinicians document an increasing number of data points, often by checkbox, more administrative staff are needed to make sense of the data and report it. But the bigger issue for me is that as documentation length increases because of administrative needs, the content of the notes has decreased. We may have moved into a digital age for billing, but clinical care has actually regressed to an oral era: you have to talk to the patient’s physician to figure out what’s really going on because the progress note doesn’t tell you.

An instructive, if dry, review of strategies used across a variety of high-cost users written by a group from McMaster University provides a nice overview of interventions attempted to date. The thing that really stood out for me? “[An] RCT compared a housing first with a treatment then housing approach in more than 2,000 homeless individuals with severe mental illness in 5 Canadian cities.33 Every $10 invested in supportive housing resulted in average savings of $21.72 in healthcare use and community support services.” Maybe some of the $ we’re spending on administrative costs in the US would do more good directed to other places in the care ecosystem. [full article here]

Susannah Fox has long advocated for the importance of peer to peer networks–her latest blog points to Esther Perel, a psychotherapist whose podcasts have a big following. So it’s especially interesting that they’re converging on this peer thing–could it be the public activation lever that we’ve been looking for? The Boston Globe just featured our friends at the Conversation Project about their work with Chinese elders–not yet what Perel and Fox mean, but getting there… Fox is worth following–her last gig was Chief Technology Officer for HHS, and that was an outlier in her career.

This newsletter is made possible by the John A. Hartford Foundation. But the opinions, views and recommendations are mine alone. Thanks this week to Kate DeBartolo and Joan Teno.

I’m a little late this week because of jet lag–got back from Prague, where I was teaching a terrific group of PC physicians with Bob Arnold.