Membership information 1800 932 916
Medicolegal advice 1800 936 077

TB or not TB?

01 August 2004

Carl was a healthy eight-year-old boy who presented to his GP, Dr P, with a rash on his legs. Dr P noted that the rash was ‘like insect bites’ and opted to ‘wait and see’.

A few days later, Carl’s parents brought him back to the surgery where he saw Dr R, who recognised the rash as erythema nodosum. Dr R ordered a chest x-ray, ESR and full blood count, prescribing a course of antibiotics.

Carl was reviewed a week later by Dr R. Carl’s tests revealed an elevated ESR of 50 mm/hr and right paratracheal lymphadenopathy with normal lung fields. Dr R saw Carl again shortly afterwards, noting that the rash had improved, and requesting a repeat ESR, which came back at 13mm/hr. Dr R took no further action.

Carl was seen by Dr R on a home visit 18 months later. He had been vomiting and complained of headache to his mother. Dr R was asked to visit again after another week, when he prescribed metoclopramide for the vomiting. Dr R kept no notes for either visit.

Dr P was asked to see Carl three times over the following week, noting, ‘Complaining of vomiting, poor appetite, general debility. O/E throat and chest normal, abdomen generalised tenderness, no rebound.’ At the third visit Carl was not significantly better and Dr P recorded his intention to refer Carl for a specialist opinion if he didn’t improve.

Two days after this, Carl was admitted to hospital in a semiconscious state after being seen by a deputising on-call GP. Carl was found to have tuberculous meningitis. He recovered but was left with significant long-term neurological disability.

His family instigated legal proceedings against Drs R and P, alleging negligence for not referring him for a paediatric opinion when he initially presented with erythema nodosum, and for their subsequent failure to diagnose tuberculous meningitis.

Expert opinion

We sought advice from a GP expert, who commented that the most important causes of erythema nodosum in a young child were streptococcal or TB infection, according to standard textbooks that he would expect a GP to consult, if unsure.

The expert felt that Dr R must have considered the possibility of TB, because she had ordered a chest x-ray. Whilst it was possible that streptococcal infection could have been the cause, especially as the ESR and the rash improved when antibiotics were given, the expert was surprised that Dr R hadn’t ordered a repeat chest x-ray or referred for advice, given the finding of paratracheal lymphadenopathy. The likelihood of TB was vastly increased by this finding, making streptococcal infection less likely.

With regard to Dr P’s management of the subsequent illness, the expert ‘believed that a general practitioner should be concerned that the child needed four visits [in two weeks] and I believe it would have been appropriate to have admitted the child.’

The expert was further concerned by the failure of Dr R to keep notes of her first two home visits. Further expert advice asserted that a tuberculin test would have been an appropriate route to the diagnosis, and that the insidious onset of symptoms prior to admission were classical of tuberculous meningitis.

Expert advice on causation held that, had the diagnosis been made at the outset, Carl could have been treated and made a full recovery. If earlier action had been taken when Carl first developed symptoms of meningitis, the likelihood was that he would have survived with significantly less impairment and disability.

Learning points

Tuberculosis remains a common killer worldwide. In the developed world, whilst it is less common now than historically, it still occurs and its incidence has been rising over the last decade. Certain ethnic groups, originally migrants from TB endemic areas, are more at risk, but the disease is no respecter of colour or creed, and its prevalence among these groups is considered by some as a marker of lower socioeconomic status rather than ethnic predisposition.

Doctors in the developed world are much less exposed to the varied presentations of TB than they once were. As the condition is largely treatable if diagnosed, it’s worth remembering as a possible differential diagnosis of disease in virtually all the functional/organ systems of the body.

A useful online tutorial on erythema nodosum and its many causes can be found at

In this case it appears that the doctors were falsely reassured by the improvement of the ESR after the course of antibiotics.The finding of paratracheal lymphadenopathy should have prompted further investigation.

Anti-streptolysin O (ASO) titre estimation is a useful way of confirming streptococcal infection, but it can be normal even where this is the cause. If one is in doubt of the cause of erythema nodosum in primary care, referral for advice is advisable.