More is Missed by Not Looking

Not always, but often enough to be embarrassing, the locum picks up something I missed. Often myxoedema – because I failed to notice over time the subtle changes in a patient I knew well.

The primary diagnosis

In hospital practice you are more exposed as a junior doctor because, especially in the Emergency Department (ED), you may be the “primary diagnoser”, i.e. the first doctor who has seen and assessed the patient. It’s a tight balance between being cautious and being decisive. While the longest delay in an ED is the time taken for a doctor to see the patient, another very common cause of delay is patients being kept for observation because someone is having difficulty in deciding whether to discharge or admit the patient.

Clinical indecision can arise from the fear: “I don’t know what I don’t know.” But that old adage, “More is missed by not looking than not knowing”, has been proven correct time and time again. Taking a proper history and performing an appropriately thorough examination should never be short-changed. Junior doctors routinely work under severe time pressure, which can lead to taking short cuts or jumping to conclusions too quickly.

Delayed diagnosis

We all know about red flag symptoms. Even so, perfectly competent and careful doctors can still be stung by pale red or pink ones. The eighteen-month-old baby with a temperature who was obviously teething and had nothing else significant to be found on thorough clinical examination could still be brought back in four hours with (by then) florid meningitis. That’s not negligence. Some conditions are simply not diagnosable at first presentation. If you do send home someone who turns out to have been developing meningitis or appendicitis, or whose very atypical mild chest pain turns out to be a brewing heart attack, that could well be an “unavoidably delayed diagnosis” rather than a “missed diagnosis”. Even so, I guarantee that you will over-investigate anyone you see later with those symptoms for the rest of your life!

There’s also the problem of “involuntary automaticity”¹ – the skilled action that people develop through repeatedly practising the same activity. This might be fine when driving a car, but dangerous when you’re trying to be alert to that one grain of wheat in a field of chaff. We rely on pattern recognition but must be alert to the one missing piece. Don’t try to explain away a pattern that doesn’t quite fit.

“Faulty clinical encounters — involving, for example, bad history-taking, physical exam, or failure to review the existing record — were to blame in about 80% of cases.”²

History and examination

Standardised thorough history taking and physical examination are essential. With many serious but rare conditions, the “signal to noise” ratio may be very low. One of the most common presentations in general or hospital paediatric practice is “a child with a temperature”, but the average GP might see only one child with bacterial meningitis in their practice during their professional lifetime. Thus, diagnosing meningitis requires the application of knowledge and a standardised approach to assessment, not previous experience. The fact that you’ve seen twenty children this week with upper respiratory tract infection does not exclude meningitis in the twenty first.

Yes, we may sort on the basis of “probabilities” but should also stay alert for obscure red flags. The World Health Organisation’s Surgical Safety Checklist is based on taking “time out” to run a final check before commencing a procedure. How about using “diagnostic time out”? You could get in the habit of taking a brief pause before settling on a diagnosis and commencing treatment to ask yourself, “Could I be wrong?” or “Could this be something more serious?” or, more viscerally, “What if I’m wrong?” and “What could happen to this patient?”

Diagnostic time out

• Could it be something else?
• Is there anything that doesn’t fit? (avoids confirmation bias)
• Could the patient have two diagnoses?
• Is this what the patient was worried about or thought was going on?

If you answer “yes” to any of these questions go through the history again, letting the patient tell their story without interruption.

Twelve tips to reduce the chance of errors in diagnosis³

1. Don’t jump to conclusions based on what’s common or what you’ve recently experienced, or because you don’t have time to think it through.
2. Avoid confirmation bias – keep an open mind. Don’t explain away symptoms or signs that don’t fit your presumptive diagnosis.
3. Use “diagnostic timeouts” – think again, with fresh eyes, before closing down the diagnostic process.
4. Think of the “worst case” – is there anything else serious that could be going on?
5. Be aware of your reaction to the patient as a person.
6. Be systematic in history taking and examination – don’t cut corners.
7. Don’t skip the physical examination because you think the x-ray or echo or blood test will make the diagnosis for you.
8. Ask yourself, “Why did this happen?” For example: “Yes, this is an acute attack of asthma which needs treatment, but what brought it on?”
9. Apply Bayesian theory. Never heard of it? If you want to stretch your brain, read the article at kevinboone.net/bayes.html or spend 10 minutes watching “Bayes’ Theorem – Explained Like You’re Five”. Warning: Could make your brain hurt!
10. Learn about “Occam’s Razor”. The common adage “If you hear hoof beats, think horses, not zebras” is usually – but not invariably – correct. But zebras exist. In medicine, “Occam’s Razor” translates as “diagnostic parsimony”. If the patient’s presentation can be explained by the existence of one rare condition or two co-existent common conditions, then you probably have one zebra, not two or three horses. Now, the contrary view is expressed in “Hickam’s Dictum”: “Patients can have as many diseases as they damn well please.” Hickham could be right – but at least test for a zebra, just in case.
11. If you’re learning, slow down. You’re under pressure to work quickly. You should feel greater pressure to work safely.
12. Admit your mistakes, learn from them – and then move on.

Dr Paul Nisselle AM (MDA National Member)
General Practitioner, VIC

References

1. Toft and Mascie-Taylor: Involuntary Automaticity: a Work-System Induced Risk to Safe Health Care. Health Services Management Research 2005;18:211-216.
2. Joe Elia: How to Miss a Diagnosis in Primary Care. NEJM Journal Watch 27 Feb 2013
Available at: jwatch.org/fw201302270000001/2013/02/27/how-miss-diagnosis-primary-care#sthash.p55FfBkd.dpuf.

3. Substantially modified by Paul Nisselle from: Robert L. Trowbridge: Twelve Tips for Teaching Avoidance of Diagnostic Errors. Medical Teacher 2008;30(5):496-500. Available at: informahealthcare.com/doi/abs/10.1080/01421590801965137?journalCode=mte.