Current Events Hospitals

Medical Errors kill between 50k and 100k annually

Most economists believe preferences are monotonic. This means that economic researchers believe the more of something you have (e.g.: money, burritos, cars, friends, etc.) the more well-off you become. This assumption likely holds if we view health as an argument in a person’s utility function; more health generally makes people better off. Putting ‘medical services’ into a person’s utility function does not work in a similar manner. Many medical services used in excess can actually do harm to an individual.

The monotonicity of utility in medical care would not be a problem if there were few injuries resulting from medical care; in fact, most people believe that physicians are highly-skilled individuals (which they are) and that the chance of an iatrogenic injury are extremely small. An article in the Milwaukee Journal Sentinel (“A dose of prevention“) contradicts this assumption by finding that “medical errors were killing between 44,000 and 98,000 people a year in U.S. hospitals, enough to rank among the top 10 causes of death in the U.S., in roughly the same league as diabetes and Alzheimer’s disease.”

Dr. Donald Berwick is a pediatrician and safety advocate. As CEO of the NGO Institute for Healthcare Improvement, Dr. Berwick is trying to reduce the amount of iatrogenic injuries with his 100,000 lives campaign. The campaign aims to eliminate unnecessary errors in the health care sector. Some of the IHI’s recommendations include:

  1. Preventing patients who receive medications and fluids through a central line from developing infections. The steps for stopping these infections include: proper hand washing, selecting the best site for the central line and cleaning the patient’s skin with an antiseptic called chlorhexidine.
  2. Taking steps known to reduce the risk of heart attacks, including giving patients aspirin and beta-blockers to prevent further damage to heart muscle.
  3. Avoiding drug errors by verifying the patient’s medication history and reviewing and updating medication lists, especially when patients move to different units or get released.
  4. Preventing patients on ventilators from getting pneumonia, through several steps, including raising their heads to between 30 and 45 degrees to prevent a buildup of fluids, and giving breaks in sedation that help determine the earliest point at which the ventilator can be removed.
  5. Dispatching rapid response teams to treat patients before a decline in condition becomes a full-blown crisis.
  6. Preventing surgical patients from developing infections through several steps, including timely use of antibiotics and appropriate hair removal.

Froedtert Hospital in Milwaukee instituted many of these changes and saw its mortality rate drop from 23.4 deaths/1000 discharges to 19.5 deaths/1000 discharges.

As an aside, measuring deaths per discharge is wise way to proxy for mortality rates from medical errors. If only pure mortality rates were measured, the hospital staff would have an incentive to send very sick patients to other hospitals to decrease the mortality rate at their hospital.


  1. Great post, Jason. Risks of medical interventions, be they tests or procedures, are often not discussed with patients, and clinicians often seem more concerned with the appearance of omission (i.e., not doing everything possible, perhaps out of fear of litigation) than with the problem of too much medicine.

    Thanks also for giving a plug to Don Berwick’s great work, and I’m looking forward to reading more of your stuff!

    Oh, and I agree that one can’t have too many burritos. 😉

Comments are closed.