A 70 year old patient was recently admitted with fever of 'unknown' origin to a rural hospital.
The patient had developed the fever suddenly with associated shivering sensation and had presented to the same hospital. Other symptoms included some rhinorrhoea and a mild cough. The on-call physician took blood cultures and prescribed an antibiotic for an URT infection and the patient went home at her own request.
However, after several days the blood cultures grew Group G Streptococcus in both sets of culture bottles. The patient was contacted and admission was arranged for inpatient investigation and treatment.
The patient had a background history of suspected infective endocarditis 6 months previously with the same type of organism, which was treated accordingly. However, at that time the transthoracic and transesophageal echocardiograms were normal. Colonoscopy revealed several polyps but no source of bacterial translocation was identifiable.
The patient was treated for suspected I.E and made a good recovery at that time.
Otherwise, there was no other apparent medical history.
The patient was taking no regular medications.
Alcohol consumption was moderate with several cups of Japanese sake per day and the patient smoked several cigarettes per day for the last 30 years.
Of note, she had recently bought a puppy.
Physical Examination had failed to establish the focus of the fever.
The laboratory data revealed a raised white cell count, raised inflammatory markers, and mild elevations in the liver function tests only.
The chest radiograph was within normal limits.
Transthoracic echo revealed previously noted mild valvular insufficiencies but there was no worsening and no vegetation to be seen.
A CT scan was performed to look for an abscess but only some mild splenomegaly was identified. The CT was hence, quite unhelpful.
At this point, the suspected diagnosis was vegetation negative endocarditis. However, if one reviews the Modified Duke's Criteria, this patient does not meet the criteria for definite I.E. with only one major criteria (typical bacteria) and two minor criteria (Fever>38 degrees and underlying valvular heart disease).
On reviewing the patient, several additional questions were asked:
- The patient had previous tooth problems and the bottom left incisor was loose but no dental treatment had been performed recently
- The patient had experienced several flea bites recently from her dog
- There was no history of sinusitis or headaches
- The patient had been diagnosed with a type of eczema for which ointment had been prescribed but the patient could not remember the name of the therapy
- There was no foreign travel
- No history of consumption of raw or under cooked food stuffs
HEENT- eczematous changes of the upper eyelids and eyebrows consistent with seborrheic dermatitis. Scalp examination also revealed seborrheic dermatitis (fungal origin). No sinus pain. Loose front, lower incisor (when pushed with a chopstick at the bedside) but no acute gingivitis seen.
CVS: pulse 80 regular, BP 120/80mmHg, JVP not elevated. Heart sounds I + II. No III/IV sounds. Pansystolic murmur at the apex radiating to the left axilla - Levine II/VI. Murmur loudest on expiration. No carotid bruits. No peripheral stigmata of I.E.
RESP: RR 14/min, Sats 98% on room air, trachea central and no tug. Percussion resonant and no crackles or wheeze.
ABDO: Soft, non-tender, no masses, no clinical hepatosplenomegaly on palpation but dullness in Traub's space. Small, non-tender, lymph nodes in the groin areas.
Skin: Several healing flea bites on the abdomen. Redness and increased temperature of the left hand after failed insertion of a venous line on the admission day. As noted above, eczematous changes over the face. Her left leg revealed a well circumscribed area 3 x 3 cm with peripheral redness with central clearing consistent with fungal infection. Foot examination revealed extensive tinea pedis (fungal infection).
The extremities had not been mentioned during the original presentation of the history. It therefore came as a surprise to see a 30cm vertical anterior scar running down the right tibia with extensive erythema of the skin with soft tissue swelling. The patient admitted to a previous fracture of the tibia.
On noting the leg changes, further questioning revealed that the redness and swelling had got worse over the last 10 days but it was not painful and so the patient did not mention it especially as it was a chronic problem for which she thought her current fever was unconnected. The patient admitted that several months earlier there had been a flare-up which required antibiotic therapy. Examination of the leg revealed increased temperature of the lower limb and tapping the bone was non-painful. These changes were clearly acute on chronic but raised the possibilities of cellulitis +/- chronic osteomyelitis.
Group G streoptococcus is an uncommon cause of cellulitis but this patient had the risk factor of previous injury and extensive fungal skin infection which can allow entry of skin associated bacteria. Moreover, Group G strep is also an uncommon cause of osteomyelitis (please see Medline).
The fact that the same organism was identified 6 months previously when the diagnosis was 'endocarditis' and that several echocardiograms showed no vegetation, makes I.E. unlikely and recurrent cellulitis / chronic osteomyelitis of the right leg far more likely a cause.
The patient was commenced on benzylpenicillin for which the bacteria was sensitive and an MRI was considered to look for osteomyelitis.
Take Home Message
- Patients do not always tell you the history you want to hear to make a diagnosis. You have to ask the right questions! Then the history comes forth from the patient bit by bit.
- If you see a skin abnormality or a bony abnormality, ASK about it. 'What happened here?' In this case, the patient had chronic problems with the leg dating back 40 years and the patient did not think that it was relevant. But it is the physician's duty to recognise the problem to decide if it is relevant. Hence, using history and physical examination to establish the facts.
- If the history, physical, lab data and echocardiograms do fit the picture for I.E. then don't diagnose I.E. Use the Modified Duke's criteria, but if the criteria are not met, then I.E cannot be made as the diagnosis. Consider other causes of chronic infection e.g. sinusitis, osteomyelitis, abscess....
- Examine your patients from HANDS to HEAD to TOE and look actively for problems that might otherwise account for the fever and bacteraemia despite the 'lack of symptoms'. Do not just accept that because the patient 'did not say , that they therefore do not have'. ASK and you will KNOW.
- This is the importance of using the Review of Systems (ROS) questions to probe each system for problems. More focused questions can then be asked if clear abnormalities come to light. Without using the ROS you will miss important problems that could lead you to the diagnosis. ROS is like a verbal CT scan -- it is used to 'scan' each system with trigger questions that make the doctor consider various diagnoses if the patient says 'yes' to such questions. Indeed, it takes longer than a CT scan to perform but is more localising and without radiation. **Always remember that by subjecting your patient to a CT scan you increase their risk of malignancy and so CT should be avoided if at all possible**
- In this case, a good old fashioned history and physical examination made the diagnosis at the bedside rather than a barrage of blood tests, echocardiograms and a 'pan-man' CT. If the diagnosis does not seem right then it probably is not right, and as such, please go back to the patient and re-interview and re-examine to get more relevant information. The answer is sometimes staring you in the face!