NIH Logo
July 28, 2017
Data Is Critical for Better Care, Expanded Access, Says Gawande

Dr. Atul Gawande speaks in Masur Auditorium.
Dr. Atul Gawande speaks in Masur Auditorium.

There are 60,000 different ways the body can fail and medicine has developed 4,000 surgical procedures and 6,000 drugs to treat those failures, according to surgeon, New Yorker staff writer and public health researcher Dr. Atul Gawande.

“That volume of knowledge and skill has exceeded the knowledge of any one clinician to know how to manage it. We are all part of groups of people delivering health care,” he explained during a wide-ranging discussion with NIH director Dr. Francis Collins in Masur Auditorium recently.

Gawande began by relaying one of the best pieces of advice he received: say yes to everything before age 40. The idea is “you are experimenting in your life and paying attention to what really gives you energy.” What he loves is solving problems related to “how we deliver health care and access to health care” and “understanding how we create better outcomes of care.”

Over the past century, there have been many breakthroughs in science and medicine. In many cases, however, those breakthroughs aren’t applied in patient care, Gawande said.

Typically, the response when doctors fail to do something properly has been more training. If training doesn’t work, doctors face mandates, pay-for-performance programs and malpractice regulations, he said. These programs have only modest effects.

Gawande advocates a different approach—better delivery of the health care we know works. At Ariadne Labs, his health innovation center in Boston, he identifies faults in the health care system, develops, tests and then scales solutions. For example, he designed a 19-item surgical safety checklist to improve communication between members of surgical teams and patients. He then tested the checklist in 8 hospitals around the world. The average reduction in surgical complications was 36 percent and the average reduction in deaths was 47 percent.

“You cannot do science and learn without data,” he said. If, for instance, researchers can’t identify hospitals that have the best outcomes, they can’t share that information so other people can learn from it.

There are fields where information has been available for decades. In the U.S., data on infant and maternal mortality rates has been published since the 1920s. Back then, childbirth complications was the number one killer of women. Today, it’s a more rare cause of death for women. Gawande attributed the availability of data as one reason why the rate has fallen.

Collins mentioned that NIH is getting ready to launch the All of Us research program, a study that will gather data on 1 million Americans to learn more about how diseases happen and how to prevent illness. He asked what kind of information would be most useful to collect.

Gawande said researchers will be able to see “interconnections” by studying participants’ health, genetics, biological condition, sleep and eating habits, behavior over time and the quality of their health care system.

Gawande chats on stage with NIH director Dr. Francis Collins. the New Yorker author signs copies of his book.
At left, Gawande chats on stage with NIH director Dr. Francis Collins. At right, the New Yorker author signs copies of his book.


In response to a question about electronic health records, he said IT companies that sell EHR software tailor their pitch to chief financial officers instead of to those directly involved in patient care. The result, he explained, is a high-quality medical billing system, not a system that allows nurses, for example, to quickly enter information about a patient’s allergies.

Collins then asked Gawande about how to respond to the opioid crisis in the U.S. Six in 10 patients who receive a prescription for opioids after surgery still take the painkillers a year later. New data also suggests patients are prescribed more opioids than they need.

There are, however, steps officials can take right now. The state of New York, for instance, requires doctors to electronically prescribe painkillers. Gawande said this has made it easier to track prescriptions and detect forged ones.

“There’s an incredible role for NIH being able to define those patterns and then support the innovators making those kinds of systems come into place,” he said.

Gawande also took questions from online viewers. One asked about what can be done to improve end-of-life care.

“What I came to understand is that it really is a question about, okay, you want to fight. What do you want to fight for: your best possible day today or to sacrifice your day today for the sake of possible time later while we treat you?” he responded.

Patients “have goals for the quality of life and their purpose” and want medicine to help them reach those goals. The best way to find out what a patient wants is to ask. One study surveyed family members about experiences they had with a loved one who passed away recently. The study reported that 84 percent of respondents received excellent or very good care when a health care professional talked about end-of-life wishes ahead of time.

Gawande said, “We’re switching from what one of my colleagues called the century of the molecule to the century of the system,” whether it’s at the molecular level, the physiologic level or at the population level.

“The biggest insights and the biggest gains in health and understanding are going to come from” innovations at the health system level, he concluded.

back to top of page