Let me preface this by saying that we are all dumbasses, every one of us. And that overconfidence and misplaced self-regard are the usual source of our failings. Although it’s easy to rag on doctors and other healthcare workers for their foibles, they are no different than anyone else in this respect. It’s just that their mistakes matter more.
We’re all familiar with the Lake Wobegone effect, as exemplified by the classic study in which 89% of drivers rated themselves above average with respect to safety and skill.
I was reminded of this study when I came across this report on self-evaluation in hand hygiene compliance. It won’t surprise you that healthcare workers rated their own compliance with WHO hand hygiene standards significantly higher than that of their colleagues and supervisors–and much higher than that of practitioners in other specialties. In other words, this study exposed not only excessive self-regard among HCWs, but tribalism as well, a two-for-one deal.
Data replotted from Overconfidence in Infection Control Proficiency
The only surprise really in these results is that just 75% of respondents rated themselves positively. I suspect that the remainder don’t believe themselves deficient in hand hygiene but instead believe that the WHO standards are BS. If the survey question had been “are your hand hygiene practices adequate” or “do they endanger your patients?”, you can bet that responses would approach 100% positive and negative, respectively.
When I was at MicroPhage, developing a rapid test for S aureus bloodstream infections, we would call ID docs at hospitals around the country and ask them what fraction of their BSI patients received appropriate antibiotics. The answer, invariably, was “All of them”.
Anyone familiar with the appropriate antibiotic literature knows that this could not possibly be true. Although the definitions of appropriate therapy are subjective, these studies (e.g., here and here and here) typically report inappropriate therapy rates of 30-50%.
But rather than argue with docs, we would ask them to review their patient’s microbiology work ups and compare susceptibility reports with actual prescriptions.
Doctors rarely use these reports to guide therapies — micro work ups take 3 days to complete and BSI therapy needs to be initiated immediately. So they consult their hospital’s antibiogram and prescribe accordingly, confident that they are also prescribing appropriately. Susceptibility results are used for epidemiology and to develop the antibiogram. They are rarely used to guide or even retrospectively evaluate individual therapy.
When we would make a follow up call a few weeks later, most docs were shocked and surprised and a bit chagrined. They had been confident that they were doing right by their patients, but found that they were not. Many of them became advocates of our test (which not only ID’d S aureus, but distinguished MRSA from MSSA) and also became customers after we got FDA clearance.
ID docs are not unique in their overconfidence. When given two sets of patient histories- one requiring a modestly difficult diagnosis, one a very hard case – doctor’s diagnostic accuracies responded as expected, dropping from about 50% to 5%. But their confidence in their conclusions changed only a little, going from 70% to 65%.
And that’s the problem. There’s no shame in missing a hard diagnosis–that’s what defines it as hard. But not being able to distinguish hard cases from easy ones should be a concern. If you are always confident that you are right, then you never ask for more information, you never consult colleagues, you never make a Plan B. Or even check to see if you’ve made a mistake, like all those ID docs.
We tend to think of medical and scientific progress in terms of technology–new drugs, new instruments, new techniques. But the truth is, we have enough of those things already to make life better for the average human being. What we need is not better technology, but better minds.