What is the best way to make hospitals safe?
I'm a retired neurosurgeon who served for five years as a hospital chief of staff. I spent endless hours at meetings directed at improving patient safety. Eleven years ago patient safety was moved from the back to the front burners of hospital governance when the Institute of Medicine published To Err is Human on the Internet. It was a very influential book. It had two main messages. One was heard loud and clear by anyone who read a newspaper or kept informed via electronic media: there are too many deaths in American hospitals.
The second message was more muted but eventually may be the one with the greater impact: The blame and punish method of eliminating errors in a large complex institution such as a hospital is doomed to failure because it has been shown not to work in numerous other industries. Therefore we must use instead the systems approach, an engineering method, to keep patients safe.
The systems method has met with mixed success as stated by Drs. Robert M. Wachter and Peter J. Pronovost, M.D. in the New England Journal of Medicine. They focus on the need for balance between the systems approach and the need for individual accountability among doctors. The example they give the most emphasis to is hand washing between patient contacts. Only 20 to 70% of doctors are compliant, depending on the hospital, despite the well established fact that hand washing reduces hospital acquired infections of all kinds. Hand washing has been replaced by using antiseptic gels and foams and takes just a few seconds.
As a kid, I was impressed by how my pediatrician, Dr. Saul Starr, of Brooklyn, New York, would rigorously wash his hands before and after examining my brother and me during a house call. This had been a strong tradition among physicians initiated by Ignaz Semelweis, the Hungarian physician who advocated it to prevent infections related to childbirth being transmitted from one patient to another. Before Semmelweis, doctors would deliver babies with bare unwashed hands right after handling infected patients. Bacteria were yet to be discovered.
150 years later we are having problems getting compliance with this simple, effective and harmless procedure. Wachter and Pronovost also mention difficulties with sign outs between doctors from shift to shift, the use of “time outs” to get the doctor to confirm and announce the proper side of surgery, marking the site of the surgery in ink, and using a check list when inserting central venous catheters. I had problems as a medical staff leader with getting physicians to do sign outs and some surgeons to do time outs, including one who had already done wrong side surgery.
The authors have a table for restriction of hospital privileges based on the failure to do any of the above, including the suspension of patient care and/or operative privileges for one or two weeks. Here is where I part company with the doctors. It is not that I feel the punishment is inappropriate. It is just unenforceable. I was a chief of staff for a total of five years at two community hospitals. Community hospitals are where most of inpatient medical care is rendered in the United States. About half of the doctors are “independent contractors”, not salaried employees. This gives them an independence that the doctors, nurses and pharmacists employed by the hospital do not have. Those physicians can opt to admit or operate on patients at another hospital while suspended or resign from the hospital. The hospitals can not afford the loss of the cases they bring in.
Then there is the question of how to monitor the physicians. Any stranger or an identifiable inspector on the hospital floor would be noticed immediately. An attempt to use video monitoring would be so highly resented by a medical staff that none of them would be willing to discipline their peers. We are left with a stalemate. It occurred to me, only half seriously, that using an antiseptic dispenser on the outside of a hospital room door should trigger the unlocking of that door, a Rube Goldberg approach. I’m open to other suggestions.