In Healthcare Beyond Reform: Doing it Right For Half The Cost I lay out the five strategies that healthcare must adopt, and is adopting in various ways and places, to make healthcare better and cheaper at the same time.

Strategy Five is “Rebuild Every Process.” It’s about “lean manufacturing,” smart standardization, measurement, “big data,” evidence-based design, teaching the innovation, all the detailed, rigorous, hard attention to intelligent process re-design that healthcare is so obviously lacking — and that is absolutely necessary if healthcare is to improve its abysmal cost/benefit ratio.

Now in The New Yorker writer/surgeon Atul Gawande has done a brilliant turn on this theme, by diving into, of all things, the processes of a restaurant chain, comparing them to the duplicative, chaotic, mistake-prone processes of traditional healthcare, and finally to some examples of smart, rebuilt healthcare processes that drive down costs while killing fewer people.

Gawande shows how The Cheesecake Factory manages to deliver 308 dinner menu items and 124 beverage choices to exacting standards, on time, from fresh ingredients, with only 2.5% wastage, in a linen-napkin and silverware environment, at lower cost, then compares that with the disconnected, uncoordinated, messy environment that is most of US healthcare. He details several examples of how new drives toward standardization and control of processes in the operating room and the emergency department, for instance, are making a difference, lowering costs and improving not only outcomes but the patient experience, all at the same time.

There are still voices in medicine decrying standardization as “cookbook medicine” and insisting that every medical decision and action must be made on the spot, on the fly, on the doctor’s judgment and say-so alone. Yet there are right and wrong ways to do most of medicine. There is no clinical justification for not elevating the head of the bed for a patient with pneumonia on a ventilator; or for not fully covering a patient during a central line emplacement; or keeping a patient on 100% oxygen for longer than absolutely necessary. As one of the docs engineering the standardization of a group of EDs told Gawande: “Customization should be five per cent, not ninety-five per cent, of what we do.”

The contrast in the diffusion of innovation is quite stark. The Cheesecake Factory changes some of its menu every six months, and all of its cooks throughout the chain must learn the cooking and presentation of the new items. The chain has a dedicated system for teaching the new items to managing, and then teaching them how to teach them, so that they can propagate the new items throughout the system. Healthcare has no such system at all. When some new fact or technique is established through research (whether the proper use of negative and positive ventilation flows in patient rooms, or a superior wound closure technique, or better post-op therapy after knee replacement surgery) there is no mechanism to propagate that discovery throughout healthcare. Studies show that the average medical innovation takes 15 years to reach even half of patients in the system. That is simply unacceptable. People die and suffer because of such resistance to doing things a better way. We must rebuild every process in healthcare continually, striving for better care, better results, and lower cost.