Last month, UCSF researchers reported on a study wherein investigators modified the default dispense number for various ambulatory prescription opioids. The goal was to determine if researchers could adjust the prescribing habits of physicians via minimal changes in the electronic health record (EHR). And the results were . . . they could indeed modify physician prescribing trends!
Prior to the study, when physicians ordered an opioid, they would automatically see a dispense number of 12 or 20; in other words, the order automatically queued up that the patient would get 12 or 20 pills at the pharmacy. During the research period, the default dispense number was changed to zero, 5, 10, or 15. Obviously, if the default was zero, the doctor would have to manually enter a number (Pro tip: you shouldn’t ask a patient to go to the drug store to get a prescription and then direct the pharmacist to dispense zero pills.) While physicians could always change the dispense instructions to any number of pills they desired, more than half the time, they didn’t do so. Take home message: humans that graduated from medical school aren’t that different from humans that didn’t.
The Oxford Dictionary assigns two definitions to the word inertia. From the physics perspective, inertia is “a property of matter by which it continues in its existing state of rest or uniform motion in a straight line, unless that state is changed by an external force.” I’m not trying to incorporate Newton’s laws of motion into my healthcare information technology (HIT) blog post, so let’s go with the alternative definition: “a tendency to do nothing or to remain unchanged.” I’m going to go out on a limb here and propose that most humans, in general, do indeed have a tendency to remain unchanged.
Marketers have used our love of inertia against us for a long time. If the box of cereal that I always buy is suddenly reduced by 20% in volume while the cost remains the same, do I revolt and stop buying the over-priced breakfast food? Nope, not typically. When I buy a shirt from a new website, there is often a message near the bottom of the screen asking me if I’d like to subscribe to their newsletter. Guess what: that box is typically pre-checked for me. The default, if I do nothing, is that I’m subscribed. I must take action to change that outcome. I believe that most of us are subscribed to a ton of newsletters that we’d rather not see.
The UCSF study lends credence to the idea that there is tremendous power in the EHR default. More often than not, doctors will not change values that automatically pop onto their screens. Of course, this only works if the defaults are within the user’s expected range. I bet that an emergency medicine physician would start to rapidly pay attention to default dispense values if 100 or 200 pills showed up as the initial value. Those numbers are way out of bounds for safe opioid dispensing. If the EHR doesn’t seem to be configured appropriately given prevailing medical practices, the default value will quickly become nothing more than noise that must always be questioned. This concept of inertia in the EHR seems to work only if the doctor doesn’t really notice it.
Those of us in the HIT trenches try to incorporate best practices (aka evidence-based medicine) by making it “easy” to do the “right” thing and just a little bit more difficult to do everything else. An order set for community-associated pneumonia (CAP) will have antibiotics that have been approved by infectious disease specialists and will likely be missing other antibiotics, whether they’re commonly prescribed or not. Can physicians order whatever antibiotics they want? Sure, but it’s not as easy as using those found in the order set for CAP. It’s that little bit of inertia that must be overcome that gives the default meds the edge much of the time.
We use defaults not only for ordering things like medications and blood tests, but often for documentation choices as well. Sometimes this causes a bit of a problem. For example, many physicians have pre-defined templates stored for certain kinds of physical exams. Imagine I’m a pediatrician (big stretch, I know) and I see a bunch of children with ear infections. Most of those patients have very similar exams: they have one or two red eardrums, a runny nose with clear rhinorrhea, a normal throat, and clear lungs. I’ll bet 80% of kids with ear infections fit this bill. The EHR might drop this template into my note once I’ve made the diagnosis of ear infection. This is very helpful, but . . . I have to remember to modify the documentation if this child is one of the 20% who perhaps has a red throat or some wheezing on the respiratory exam. Defaults giveth, but defaults can also taketh away.
It’s incumbent upon the clinical informatics team to use the power of inertia for good and not for evil. Make it easy to do the right, but not too easy. Of course, some might argue about what the right thing is, but that’s a topic for a different blog post.
Do you have examples of inertia and healthcare IT working to help clinicians and patients? If so, tweet at me at @CraigJoseph.