Clinical Case 080: Head scratching diagnosis from the bush


This case comes from Dr Cathy Harris – a former Broome Doc, now living in Gunnedah NSW.  Cathy describes a great case in which the common traps of diagnostic momentum, bias and back-to-front reasoning come into play.  Here is the case as it played out:

I am handed the triage note of a 39 year old Aboriginal lady from a remote community.  The nurse tells me she has left flank pain and she has collected her urine and is off to find some haematuria for me.  The patient is an overweight Aboriginal lady, wearing a Coldplay t-shirt and looking very uncomfortable in one of the acute care beds.  My immediate thought is to agree with the nurse on the spot diagnosis of renal colic.

Initial History

Very poor historian especially regarding timelines.

–        1 week history of colicky left flank pain radiating to hip/groin

–        no precipitants/relieving factors

–        no macroscopic haematuria

–        no fevers

–        no urinary symptoms

–        no history of trauma


–        Afebrile, obs stable

–        CVS, Resp, GIT examination unremarkable

–        Left renal angle tenderness

–        Tenderness over left hip and groin


–        Poorly controlled  NIDDM

–        Medical notes show that in the previous month she was transferred from her community with suspected pyelonephritis.  Symptoms included fever and left flank pain.  She was admitted and diagnosed with and treated for LLL pneumonia (notes read later)


–        Oral hypoglycaemics (unknown) that she had left at home and therefore, not taken for the last week.

Initial management and investigations

–        Analgesia

–        Fluids

–        BSL= 18mmol/L

–        Initial Urine dipstick: NAD (found blood on the second urine dipstick, along with everything else…)

–        UEC, FBE, LFT: NAD

Second history and examination

Realising that I had less and less evidence for renal colic, I went back and asked a few more questions.  The patient insisted that her left flank, hip and groin were really sore.  I wondered if it could actually be a back or hip issue.  I took a lower limb neuro history which revealed a patch of numbness over her left lateral thigh.  I conducted a lower limb neuro exam and found what I thought was an L2 radiculopathy (I would later discover that I was a bit out with my dermatomes!).

The patient was able to mobilize but had a reduced range of movement of her hip.  The patient had spinal tenderness from approx T12-L2

Second attempt at management and Diagnosis

This was a Sunday, which heavily influenced management.  After discussion with a DMO and pain was well managed, it was decided to discharge her to the care of a cousin and get her back for a CT on Monday.  2 phone numbers and an address were taken before discharge.

Monday passed and the patient’s lack of attendance was disguised by a chaotic day.

Tuesday, phone calls were made but no response.

Wednesday the patient was finally retrieved by an ALO.  When she got to the ED it was clear that she had been self-medicating with alcohol and she was once again wearing the Coldplay t-shirt.  History and examination were unchanged.  She still had a ?L2 radiculopathy of unknown cause.  T10 –T12 + Lumbar CT were done (fortunately!) and revealed a likely T12 osteomeyelitis with T11/T12 facet joint septic arthritis (in a poorly controlled Type 2 diabetic).

Findings were discussed with Ortho at St Elsewhere and she was transferred for biopsy, PICC line and 6 weeks of IV antibiotics. 


–        Truckloads! Beware of diagnostic tunnel vision and be aware of diagnostic momentum, false confirmation and our tendency to make symptoms fit into boxes.

–        Take a history and examination prior to making a diagnosis

–        The potential for non-compliance in the context of a potentially serious diagnosis, should be an important consideration when developing a management plan.

–        A flick through the old notes is helpful (was the transfer for ?pyelonephritis actually the initial presentation of the this orthopaedic pathology?)

Thanks Cathy for sharing this case – tough case.  Really good examples of the errors we all face in our day-to-day practice.



  1. Great case! Having worked with this population a lot up in Darwin, if they’re a) worried enough to present b) complaining of bad pain, you can generally expect something bad to be happening. With the high incidence of poorly controlled diabetes my differential is always sepsis, sepsis, sepsis, (especially occult collections – even without fever), then occult trauma, rheumatic fever and its complications, then the usual differentials. Pays to keep an open mind, do thorough and serial examinations, and don’t stop til you have an answer.

  2. David Berger says

    After just four days of working in Broome ED this case has a ringing sense of familiarity to it. I’m also extremely impressed that you managed to take a history which elicited a small patch of numbness over the lateral thigh. I think I have a little way to go with my history-taking abilities.
    Presumably the initial bloods, especially white count and CRP, would also have pointed towards a septic cause.
    I love the lesson of this case to throw out all preconceptions when confronting a case and start off with ‘So what brought you here today?’, before moving on systematically from there. Thanks for posting it.

  3. Another great case. I suppose it reminds me to think about other systems (in this case neuro) and do a proper examination for them.

    Also given the ‘specific’ areas that pain presents in abdomens, for this has been another area of spot diagnosing in my head. Its hard to shake that voice going “come on, its appendicitis!”

Speak Your Mind