Evidence based practice

AKA Life in the slow lane, part III

I filled my tank this morning and did the math. 33.27 mpg as compared to my more typical 29-30 mpg. I didn't even remember to keep it at 55 all week - just most of my highway miles. That works out to a 10% discount on gas, or based on the number of miles I drive, about $150/year if I can keep this up -- and if the week was truly typical of my driving pattern.

The benefit appears easy to document, it seems to outweigh the risks of changing my behavior. Keeping it up, that's the issue. I've been driving one way so long that changing my habits more or less permanently will be quite a challenge. That's because the risk of NOT changing my behavior isn't particularly devastating. Not even at $3/gallon. Then there's the time factor. Those 2 or 3 minutes I save on a short trip sometimes seem worth the expense in lower gas mileage.

There are questions about my "study": Was the study sample large enough? Was it consistent with the "control group" in terms of the type of miles driven? What sort of bias was introduced by the sampling technique? The studies we use to make decisions about evidence-based practice in healthcare can have similar concerns.

Sometimes the evidence for change in nursing or medical practice is impressive. Sometimes it isn't. Sometimes it is simply overwhelming. One example that comes to mind is the observation, 24 years ago, that led to the elimination of virtually all benzyl alcohol in products used in neonates.

Some very observant neonatologists noted that patients who received relatively larger amounts of flush solution containing benzyl alcohol as a preservative developed something they called "gasping syndrome". Those who received smaller amounts or none, did not. It was based on a VERY small patient population (16 infants), but the risk of behavior change (elimination of benzyl alcohol) was minimal, and the potential benefit in the elimination of benzyl alcohol toxicity was huge, so they reported to the FDA who quite promptly issued a warning.

As a NICU nurse who was practicing at the time this was published, I have no doubt at all of the validity of the observation. We stopped using benzyl alcohol in our NICU the day our head neonatologist got the word. It was very clear to us that we too stopped seeing benzyl alcohol toxicity in our patients. The babies who were smallest and sickest tended to have arterial lines which we used to draw arterial blood gases - and flushed with solution containing benzyl alcohol. The babies got progressively more and more acidotic and developed clotting problems. Once we eliminated benzyl alcohol, this pattern simply disappeared. I don't know how quickly other NICUs made the change, but evidence of the advantages of discontinuing benzyl alcohol accumulated very rapidly and preservative-free flush solutions became standard practice.

Most evidence-based practice changes require larger groups and double-blind studies. Some changes, like the elimination of benzyl alcohol, merely require careful observation. When the benefits of change overwhelmingly outweigh the risks, it is easier to decide to make a change and to stick to it. When the risks are greater or the benefits less clear, the evidence simply must be better.