Institute for the Study of
Healthcare Organizations & Transactions



Doctors Are Reminded: Wash Up!

So began a recent article in the New York Times (November 9, 1999) by Emily Yoffe. Physicians have long known that proper antiseptic procedures are critical to preventing the spread of infectious diseases. In 1843, Oliver Wendell Holmes (1809-1894), a prominent New England physician (later dean of Harvard Medical School) and father of the U.S. Supreme Court Justice, conducted a survey strongly suggesting that childbed (puerperal) fever was a contagious disease caused by an infection passed to pregnant women by their doctors, who frequently moved from patient to patient, and even from autopsy to patient, without washing their hands. He was derided by his colleagues. From their point of view, puerperal fever was caused by chance or God; no gentleman could have hands so dirty as to cause disease, and it was inconceivable that physicians could be responsible for the deaths of their own patients.

Later that same decade, Ignaz Philipp Semmelweiss (1818-1865), a physician working at the Vienna Medical School, bolstered the case for puerperal fever as a contagious disease. At the time, the Vienna Lying-In hospital had two maternity wards, one staffed by midwives, and the other by medical students supervised by staff physicians. The mortality rate among women attended by midwives was approximately 2-3%; however, the students’ ward had a rate of 10% or more. While hospital administrators blamed the high mortality rates on poverty, this could not explain the difference between the two wards. Instead, Semmelweiss believed that the students, who received much of their medical training in the autopsy room, were carrying infections from cadavers they dissected to the women in the ward. When students and physicians scrubbed their hands with chlorinated lime instead of washing with ordinary soap and water, mortality fell to the levels observed in the midwives’ ward. Semmelweis’s work is now recognized as a landmark in the history of medicine. But, as with Holmes, his conclusions were not accepted at the time, and physicians continued to ascribe puerperal fever to some constitutional predisposition on the part of the patients. After the death of Semmelweis, Joseph Lister published a series of studies of antisepsis, and Louis Pasteur identified the microbe associated with puerperal fever. With proof of the "germ" theory of disease, physicians acknowledged that germs could be passed from patient to patient, patient to doctor, and doctor to patient, and handwashing with antibacterial soaps was established as standard medical practice.

150 years after Holmes and Semmelweis, and more than a century after Lister and Pasteur, all physicians accept the germ theory of disease, and all acknowledge the importance of antisepsis (Lois Magner, A History of Medicine, 1992). The U.S. Centers for Disease Control and Prevention has concluded that handwashing is "the single most effective way to prevent the transmission of disease".

For a related press release, Click on 

Nevertheless, in 1981 Albert and Condie observed that handwashing rates in an intensive-care unit varied between 30-48% (New England Journal of Medicine, 304, 1465-1466, 1981). Apparently, handwashing doesn’t come any naturally to modern physicians than it did their 19th century forebears.

The problem persists. In 1996, Tibbals reported that only 12% of physicians in a pediatric intensive-care unit washed their hands after patient contact (Medical Journal of Australia, 164, 395, 1996). Even after an intensive program of education, monitoring, and feedback, handwashing rates rose only to 17%. When another sample of doctors were surveyed about their behavior, they reported that they washed their hands from 50-95% of the time; but when they were surreptitiously observed, their actual rate was as low as 9%. Pritchard and Raper, commenting on this study (Medical Journal of Australia, 164, 389-390, 1996), wrote that

It seems a terrible indictment of doctors that practices and protocols must be developed to take the place of something as simple… as hand washing. Perhaps an even bigger concern for current medical practice, and one which should lead us all to do some soul searching, is that careful and caring doctors can be extraordinarily self-delusional about their behavior.

More recently, Chang and associates traced an epidemic of yeast infection in a neonatal intensive care unit to the failure of some hospital staff members to wash their hands after playing with their pet dogs, who were carriers of the yeast (New England Journal of Medicine, 338(11), 706-711, 1998). Similarly, Moolenaar et al. suggested that an outbreak of a bloodstream infection in another neonatal ICU was due to bacteria carried under the long or artificial fingernails of some hospital nurses (Infection Control & Hospital Epidemiology, 21(2), 80-85, 2000). Both epidemics could have been prevented by appropriate handwashing (and, in the latter case, by requiring short, natural fingernails).

The U.S. Centers for Disease Control and Prevention has estimated that about 1.8 million patients will contract a "nosocomial" or hospital-transmitted infection every year. Of these, 20,000 patients will die directly from the infection, while the infection will contribute to the deaths of another 70,000 people. By contrast, fewer than 20,000 Americans died of AIDS in 1997, and fewer than 50,000 of breast cancer. Nosocomial infections add approximately $4.5 billion to annual health-care costs.

To some extent, probably, healthcare providers don’t wash their hands adequately for the same reason that children, food-service workers, and most other people don’t: it’s just too much trouble. While it may seem paradoxical that the recent epidemics of infection disease occurred in intensive-care units, that’s also where medical staff are the busiest. Part of the problem is perceptual: ordinary dirt is visible, and it can be removed with soap and water; but "germs" are invisible, and must be removed with chlorinated soaps and other antiseptics (a similar problem impairs compliance by patients with hypertension, a disease with no palpable symptoms). Part of the problem is cognitive: doctors and nurses think they wash their hands more frequently than they actually do. Handwashing is so "19th century", and 21st-century medical technology may give doctors a false sense of security. 

Providers now routinely wear gloves to eliminate skin-to-skin contact with patients, even during the most delicate procedures. But bacteria can adhere to the outsides of gloves, as well as to hands. A study by Thompson et al. indicates that while hospital staff usually wear gloves when they are required, they do not change their gloves as often as they should, so that the problem of patient-to-patient transmission persists (Infection Control & Hospital Epidemiology, 18(2), 97-103, 1997).

But part of the problem is also structural. Physicians and other healthcare providers must break the "code of silence" that makes it difficult for them to criticize their colleagues’ clinical practices. Senior staff must model handwashing for their subordinates and trainees. Somewhat surprisingly, there are no national medical standards for handwashing. In an era of cost-containment and managed care, hospital infection-control departments (which drain, but do not generate, revenue) may not be given enough staff to monitor staff members’ handwashing behavior and develop effective intervention programs. As well, patients may be discharged from the hospital before the infections take hold, making their sources more difficult to trace.

However important it may be in the abstract, handwashing is inconvenient and even aversive for staff. When one prominent medical school constructed a new pediatric hospital, the architects placed sinks only at the ends of each hallway, thus making it more difficult for doctors and nurses to wash between patients. Infection-control officers recommend 10-15 seconds of vigorous soaping, followed by a thorough water rinse, before and after any contact with body secretions, mucous membranes, or blood. But this could mean hundreds of handwashings per day. Such frequent and prolonged handwashing, coupled with drying with paper towels, dries and chafes the skin. Waterless, alcohol-based microbial agents, faster to use and much kinder to the skin, are available in Europe, but generally have not been adopted in the United States.

Improved compliance will require more than a change in attitudes and beliefs on the part of physicians, nurses, and other health-care workers. It will require a reorganization of the health-care environment to promote and maintain behavioral change. As a symbolic reflection of the kind of change required, a program at the University of Pennsylvania Medical School instructs new patients to ask every health-care worker they encounter whether they have washed their hands -- thus reversing the traditional roles of doctor and patient.

John F. Kihlstrom, PhD

Copyright © 2000 Institute for the Study of Healthcare Organizations & Transactions