'Best practices' . . . not always best

Dr. Jerome Groopman, whom I have mentioned before, has a fascinating article in the current New York Review of Books about the government's role in setting and requiring "best practices" in health care.

"Best practices" is in quotes because, as Dr. Groopman points out, such standards may not always be "best."

The government requires hospitals to adhere to its standards, and it reports on how well hospitals comply with them. Increasingly, Medicare, Medicaid, and non-government payers penalize hospitals with reduced or recouped payments for failing to comply with expert-defined best practices.

Yet, as Dr. Groopman writes: "[D]octors and other experts acting for the government and making use of research on comparative effectiveness. . . have repeatedly identified 'best practices,' only to have them shown to be ineffective or deleterious."

He ticks off a number of examples of "best practices" gone wrong, including:

- Medicare's tightly controlled blood sugar levels for intensive care patients "resulted in a higher likelihood of death."

- Medicare's recommendations for hip and knee replacements had "no effect" on complications.

- Federally-approved measures for congestive heart failure patients had "no major impact."

- Medicare's requirement for the speedy administration of antibiotics for pneumonia patients in emergency departments made things "worse."

Dr. Groopman even discusses a "best practice" that he helped develop that turned out to be in error. He also talks about common errors a researcher can make leading to a bogus "best practice."

This is not to say that research-based "best practices" are wrong. Adherence to such standards has improved hospital care, and I note Community General's success in such areas a low infection rates and low ventilator-acquired pneumonia.

It turns out, however, there is a difference between standards that are applicable in the same way for all patients (such as infection control practices) and standards as applied to an individual, given the complexity of the patient's illness (such a blood sugar management).

"[O]nce we depart from. . . mechanical procedures and impose a single 'best practice' on a complex malady," according to Dr. Groopman, "our treatment is too often inadequate."