Scale in health care and Google

Sometimes I read a book that strikes so close to my passion that I can’t stop talking about it. I will recommend such a book to most everyone and often buy extra copies to give away to friends and colleagues (family members just tell me to go away). I found one of those books. It is In the Plex by Steven Levy. The book is a biography of Google, most of it centers on the founders and key managers who had to develop and take ownership of the growth of Google. There are a number of books that describe companies that have come from nothing to greatness, so that isn’t what makes this special. The key to this book is the notion of scale.

As I have talked with colleagues about this book I keep coming back to how the leaders of Google threw away the standards of their time to build at scale what they needed. For example, the standard in the 1990’s for servers was to buy the best you could afford to avoid having the system go down. Failure rates were kept to a minimum, as best as possible. This also meant that the servers were expensive. Google executives decided to look at it differently. They needed LOTS of servers. At a very low cost they bought parts that were rejected due to inferiority and build their own, for a fraction of the cost, adding redundancy to overcome the much higher failure rate. With this model they could scale up (or down) quickly and inexpensively to meet the demand of their product, search.

We have got to figure out how to do this in health care. yesterday I was on the phone with Matt Wojcik of GE Centricity to talk about interoperability and ICD-10. As I have many times lately, I mentioned In the Plex and we explored how an EMR search should be more like Google, building on data based on past use and a continuous reduction in time and errors. It also has to scale. As doctors engage with patients the system needs to do a better job at reducing clicks, drop down menus, and search so users can reduce non-patient care activities. At the same time doctors have to find ways to scale their own work.

For a doctor to see more patients, in the current way care is delivered, a clinic has to build more rooms, hire lots more staff, and lengthen the work day. That isn’t scale. We need to turn the work on its head to do things differently. Right now I am working on a group visit model, similar to what was done by Harvard Vanguard and Fairview Physicians (Noffsinger, E.B,, 2009). This allows scale through intense use of physician resources and increase in quality for the patient., or at least that is the idea. The data will reveal the facts.

So I recommend In the Plex. It is interesting and I learned a lot about Google and search. It is also provoking ideas about how we can scale health care so that it is meaningful to patients and meets the demand for access and quality.