When Newton penned his laws of motion in the late 1600s, he was surely not thinking of surprise hospital inspections. (Hospitals date back to 350 BCE, but hospital inspectors did not come about until 1951 when the Joint Commission on Hospitals was born.) Nevertheless, his law can clearly be seen at work there: Medicare patients who had a hospital stay during an inspection week had a significantly lower risk of dying in the following 30 days, compared to similar patients who stayed during a non-inspection week. (JAMA Internal Medicine paper.)
The action of the inspection led to a reaction from the hospital staff – not surprising considering it is humans we are talking about here. Who doesn’t sit up straighter, let up on the gas, and then breathe a sigh of relief when a police car goes by? Similarly, hospital staffs start dotting i’s and crossing t’s when Joint Commission for Accreditation of Healthcare Organizations (JCAHO) inspectors show up unannounced.
It’s not clear exactly what the hospital improve upon when inspectors are watching, but it is clear that it is the presence of inspectors that is making the difference. Since the study included only surprise inspections and the impact was seen across all hospitals, it’s almost as good as a randomized controlled trial. Teaching hospitals did show slightly more impact from inspections than non-teaching hospitals.
Whenever we measure something – take an action – we will spur a reaction. The reaction may be positive, as in the case of better 30-day mortality for hospital patients. In other cases, the reaction perverts the purpose. Remember when the Veterans Administration clinics in Phoenix were caught gaming the system so that it appeared all patients were getting appointments within 14 days – which, coincidentally, was the goal set by VA national office? (I blogged on it back in 2014.) Here, the reaction was to manipulate the data – not to actually meet the goal.
Sadly, the inspected hospitals went back to their norm after the surprise JCAHO visit was over. This is akin to us putting our foot back on the gas, when the blue lights have pulled someone else over.
Measuring things, like any action, is rife with opportunity for unintended and perverse reactions, but also for unexpected gifts. The trick is to learn how to act such that the reaction is positive, at least in the net effect. And that is very tricky indeed!