The death of diagnosis

We have observed that junior doctors appear reluctant to attempt a diagnosis nowadays and tend to leave the diagnosis box in the clerking form blank, or record a descriptive term, eg ‘collapse ?cause’. This absence could lead to a delay in appropriate treatment or non-specific use of broad-spectrum antibiotics, resulting in iatrogenic complications and prolonged admission.¹,² A specific diagnosis would also avoid unnecessary investigations.

This service evaluation examined how frequently acute medical patients receive a diagnosis, as opposed to a clinical epithet, on admission or later during their hospital stay.

Over three days, the records of 100 consecutive patients from the Medical Admissions Unit (MAU) and four general medical wards at a UK provincial teaching hospital were analysed to see if, and at which point during their stay, a definite, provisional or differential diagnosis was recorded. Patient files were reviewed at three time points:

1. after junior review, within four hours of admission

2. after senior review, within 24 hours of admission

3. at the time of discharge from hospital, and on the discharge summary.

A diagnosis was defined by one fitting ICD-10 criteria. Cases were recorded as ‘no clear diagnosis’ if only a descriptive term was applied. The length of stay was examined according to the presence of a diagnosis.

These 100 patients had a median age of 71.5 years (range 19–96 years). 46 were male. Junior doctors saw all patients within four hours; 24% received a definite diagnosis, 24% had a provisional and 27% had a differential diagnosis. However, 25% received no level of diagnosis or a symptom such as ‘nausea’ recorded as their diagnosis (Table 1). They often had an unfocused management plan, such as ‘take bloods, chest X-ray and senior review’.

Table 1.

The prevalence and level of diagnosis from admission to discharge.

All patients received a more senior review, either from a registrar or consultant, within 24 hours of admission, so that 46 patients now had a clear diagnosis. The senior review radically changed six provisional diagnoses of junior doctors. Twenty patients still had no diagnosis; four were labelled with non-specific sepsis.

At the time of discharge, 69 records contained a discharge letter; 55 with a clear diagnosis, which also described appropriate treatment, while 14 had no clear diagnosis, usually containing only a description of their illness. However, 31% of patients left hospital without an EDAN (electronic discharge advice note), the system in use at the time, of whom two patients had died during their stay. The outcome of the other 29 patients without discharge letters was unknown.

Those patients who received a clear or provisional diagnosis made by junior doctors had a much shorter length of stay than those without one, seven versus 21 days (p=0.067).

Why do some junior or senior doctors not make a diagnosis? The history may be difficult to obtain, as 12 cases in this series presented with confusion and sometimes there are language barriers.

We speculate that time pressed junior doctors are reticent to make a definitive diagnosis because of inexperience and expect a senior colleague to make one.³,⁴ Alternatively, they may not have been taught that appropriate treatment results from applying the correct diagnosis. Certainly, appropriate and specific treatment cannot flow from labelling an acute patient with ‘collapse’, ‘sepsis’, or ‘chest pain’ of indeterminate origin. They commit themselves instead to ordering excessive investigations and may over-prescribe antibiotics. Most of the provisional diagnoses made by junior doctors in this series were correct and could have been used to start treatment immediately.⁵ After review by a senior doctor, half of patients had a clear diagnosis. More experienced senior doctors made more diagnoses, but juniors could benefit by attempting one too. Shorter length of stay seems associated with a clear diagnosis. We advocate an emphasis on the art of diagnosis in medical training.