I doubt many CIOs consider the physicians they work with enemies. But I venture to guess that some physicians do consider IT as the enemy.
Some of the animosity stems from statements like this, made by Jason Burke, of the SAS Institute, a business analytics company: "Evidence-based medicine, personal electronic health records [are causing] a transformative shift toward more information-based decision making related to patient care. …"
While most academic physicians will agree with this stance, many community doctors see evidence-based medicine (EBM), clinical practice guidelines, and decision-support systems as "cookbook medicine."
They simply don't believe EBM will have the transformative effect on patient care that Burke suggests. In their minds, medicine is as much art as science, and as such can't be distilled into a series of evidence-based guidelines and rules.
Many clinicians also question the philosophical assumptions upon which EBM is based. Understanding this skepticism is the first step toward getting buy-in from physicians who resist e-health records and clinical decision-support systems implementation.
One source of skepticism is that EBM-generated practice guidelines are usually based on the results of large clinical trials. One problem with these trials is their exclusion criteria. A trial evaluating a drug for hypertension, for example, often includes patients that have only hypertension. Such patient populations have to be free of any other chronic disorders that might skew the results.
Such exclusion criteria help investigators get pure data, but this doesn't mimic the real world, where doctors often treat patients suffering from a variety of these "co-morbidities."
Another source of skepticism is something called a Type 2 statistical error. Community physicians place a good deal of faith in their own clinical experience. When they see a patient respond to some treatment that doesn't have the blessing of the experts, many are inclined to believe their own eyes.
What they might be seeing in a clinical trial is a Type 2 error, which occurs when a study enrolls too few patients and the data analysis jumps to the conclusion that treatment X doesn't help disease Y. To detect relatively small but statistically significant effects of any treatment, one's sample size--namely the number of patients enrolled in the trial--has to be large enough. If not, you get false-negative results.
A 2004 review of the research shows that more than 300 studies have come to false-negative conclusions because their patient size was too small. Reason enough to question EBM.
It's unlikely the chasm between clinical experience and clinical experiment will be resolved anytime soon, but knowing it exists can help IT execs be more sensitive to the resistance they meet, and hopefully help them devise a more effective strategy to get buy-in despite these reservations. Coming up with such a strategy just might turn "enemies" into allies.