When it's OK not to know

As GPs, our sense of success or failure rides overwhelmingly on whether not we are able to successfully diagnose or evaluate the conditions and cases that we see. On one level diagnosis is simply what we are paid to do, but if other GPs are anything like me, any pressure from the NHS or from patients to put our finger on the problem and avoid missing important diagnoses is minuscule compared to the pressure that we place on ourselves to perform in this area (as witnessed by the sense of achievement that a colleague and I felt on correctly deducing that a patient of ours was in the very early stages of HIV infection, although the patient, understandably, failed to share in our satisfaction).

The status of a good diagnostician is considerably higher in my profession than that of a skilled communicator or a judicious prescriber. However, a couple of recent encounters with very young patients have reminded me that there are some scenarios in which uncertainty can be perfectly acceptable, provided the not-knowing is handled appropriately.

One was a case of measles in a three-year-old girl, the first case I have ever seen in the GP setting. She had several of the textbook features of the disease, but I did not make the diagnosis - that happened about an hour later in the paediatric emergency assessment unit. My lack of diagnostic precision in this case had no adverse effect whatsoever, firstly because I knew that it could be measles and secondly because it was clear that this was the sickest child I had seen in quite some time; hence the management plan – straight to hospital, with the paediatric team duly notified by phone, was exactly what I’d have done if I’d known for sure.

The other case involved an even younger patient, presenting a few weeks prior to her date of birth. The midwife had suggested that she might be lying bottom-down instead of head-down, a breech presentation. On examination via her mother’s abdomen, I thought that I could feel her head in the correct position. However, I was simultaneously aware that it has been nearly a decade since I last worked in an obstetric team, palpating pregnant tummies on a daily basis, and that research published since then has showed that even practitioners who are ‘skilled up’ in this have an unacceptably low rate of accuracy in correctly identifying breech presentations by examination. I decided to test my opinion by asking a fellow GP to give an opinion on the baby’s position, without telling him my own impression. Lo and behold, my colleague came to the opposite conclusion, and the necessary action (although not the answer) became clear – a prompt ultrasound scan. Happily, the result of this was ‘head down’ indicating either that I had been correct or that the baby had changed position in the interim.

I suppose the reason why not-knowing can be acceptable is that we are not working in isolation. The GP’s assessment can be fortified with a formidable armoury of diagnostic tests and specialist colleagues’ opinions.

Hence the most important decision we routinely face is deciding which of the many uncertain scenarios presented to us actually requires the unleashing of said armoury.  The accompanying cost is that we may end up potentially medicalising everyday symptoms and provoking needless patient anxiety. In the majority of cases those symptoms are destined to resolve themselves before a definitive diagnosis is ever achieved.