The hospital 25 years from now: a tower of ICU

At a celebration of a health care informatics lab’s first 25 years, Boston’s most experienced hospital leaders came in to speculate on what an American hospital would look like in 25 years. The experts agreed that more procedures would be doable on an outpatient basis. So our hospitals would essentially empty out? No! They’ll be filled with people who are incredibly sick and whose cases are extremely complex: “A Tower of ICU.”

David Nathan, who graduated from Harvard Medical School in 1955 and eventually became Physician-in-Chief of Children’s and then President of Dana Farber, pointed out that it would be difficult to train young people in this kind of environment where there are no simple cases. (He also shed light on the economics: “You cannot make money doing research. And teaching is hopeless.”) John Halamka, a doc-turned-CIO, quipped “Don’t teach the Krebs cycle; teach the revenue cycle.” Sandra Fenwick, the CEO of Children’s (a $2.3 billion/year enterprise), said that hospitals like hers would see “far more complex disease,” with the simpler problems being handled at home, by primary care providers, and community hospitals.

What about information technology? Electronic health records haven’t resulted in the savings, efficiencies, or improvements that were promised by the vendors and the Obama Administration. In the rare cases when a data exchange is accomplished from one hospital to another, the treating physician is “flooded with useless data”. There is no practical way, currently, to pull just the relevant material from another institution.

Yet computers will be critical to treatment, the speakers believed. “The doctor will Google you now,” was the joke circa 2000, but machine learning will soon transform this into “The Google will doctor you now.” Diagnostic procedures are producing more data than a human can inspect. “The average number of CT slices used to be 30,” one physician said. “Now it is 300. A radiologist cannot look at 300 slices in 10 minutes.” (It was noted that Vinod Khosla predicts that 80 percent of doctors will be obsolete; perhaps we should listen to him since he was smart enough to leave Kleiner Perkins before Ellen Pao could have sex with him.)

How about payments? Atul Butte envisioned a realtime link from Epic to the payor and every order will be screened instantaneously as currently happens with credit card transactions. The doctor will order an expensive test and the insurer will immediately come back with “no.”

(You might ask how good a job hospitals and doctors are doing today. A Harvard-trained pediatrician at the conference said “Only once I had kids did I realize that all of the advice I gave to parents during my pediatrics training was bad advice.”)

What about the disastrous patchwork of private insurance, government largesse, uninsured and undocumented migrants, and self-pay surviving another 25 years? The panelists thought that we would enter the Glorious Age of Single-Payer rather than continue as an international rogue outlier. Germany was cited as a success story for single-payer (Wikipedia says that Germany has a “universal multi-payer health care system”).

Systems-oriented doctors have always loved aviation. See The Checklist Manifesto, for example. The docs at this meeting enjoyed the phrase “Care Traffic Controller” for the physician of the future, coordinating all kinds of services to benefit a patient. None of them seemed to have reflected on the fact that the primary function of an Air Traffic Controller is to separate planes, not bring anyone together.