Discussion

• 1. In this patient, VP shunt malfunction and infection must be considered since severe morbidity or mortality can occur if either goes undetected or untreated. Symptoms of acute shunt obstruction include headache, vomiting, papilloedema, cranial nerve VI palsy, change in personality and the ‘setting sun’ sign (lack of upward gaze) in infants. Shunt malfunction may be caused by obstruction, fracture, migration or kinking of the tubing, or damage to the shunt apparatus itself. Most shunt obstructions occur in the ventricular catheter by invasion of glioependymal tissue. The most common locations for distal tube fracture include the occipito-cervical junction, the base of the neck and the junction between the inferior border of the ribs and the abdominal wall.

The vast majority of shunt infections occurs within 6 months of shunt placement, typically owing to staphylococcus and streptococcus. Approximately 10% of infections occur later owing to primary abdominal infection or secondary bacteraemia. Shunt infections can present with few or no symptoms, but classically include fever, headache, altered mental status, redness along the surgical incision, tenderness over the reservoir or abdominal pain. Abdominal pain may be the presenting feature as the distal tip of the shunt empties into the peritoneal cavity which serves as a nidus for infection. Unfortunately, meningeal symptoms may not be reported because communication between the infected ventricles and the meninges may be absent.

Since this patient presents with symptoms consistent with both shunt malfunction and infection, we must consider both in the differential diagnosis, as well as simple meningitis and the benign viral syndrome. The fact that several of his family members are ill does not rule out shunt pathology.

• 2. The palpable portions of the shunt should be examined, looking for erythema, warmth, induration or fluctuance. Sometimes, focal breaks can be found where focal collections of cerebrospinal fluid (CSF) are palpated. In addition, the shunt reservoir should be compressed to assess for function. It should empty and refill easily and rapidly.

Given our patient's presentation, he should undergo a non-contrast head CT and a plain film shunt series. The head CT will assist in ruling out obstruction or pending herniation. An increase in ventricular size from previous head CT's suggests increased intracranial pressure, probably secondary to shunt malfunction. The plain film shunt series is important because it can show if the shunt has migrated, become kinked or has broken.

When there is concern about infection in patients with a VP shunt, the lumbar puncture (LP) is frequently negative even when the shunt is infected, thus some recommend a shunt tap rather than an LP to rule out meningitis or shunt infection.1 2 Those who recommend this approach argue that the CSF communicates directly with the shunt and tapping the shunt directly leads to one tap (shunt only) versus two (if an LP is done first and returns with a negative result, then a shunt tap would be required). On the other hand, some prefer to do a LP first in order to avoid the risk of infecting the shunt. Both approaches are reasonable and which approach is used will generally vary by practice setting. In a rural emergency department with limited or no access to neurosurgical services, the emergency doctor will often be required to perform the tap him/herself before considering a transfer of care.

To tap the shunt, the area over the reservoir should be wiped with a dry towel, cleaned with alcohol and allowed to dry. It should then be prepared with an iodine solution such as betadine or acceptable alternative such as chlorhexidine, allowed to dry again and finally, be tapped with a small butterfly needle to obtain the CSF for diagnostic purposes.

Some other important considerations include:

• A. If the patient has a source for the fever (eg, urinary tract infection), but symptoms consistent with shunt infection, should further evaluation be done? The answer is yes, as the VP shunt becomes a nidus for infection and the morbidity/mortality of shunt infections is high.

• B. If the shunt tap is negative, should an LP be done? This is harder to answer. While CSF communicates with the shunt and, therefore, suggests that a shunt tap would be adequate to rule out meningitis, there have been case reports where the shunt tap was negative but the LP positive. This is much less common than the reverse, but should be considered. This can occur in patients with two physiological CSF spaces owing, for example, to masses causing non-communicating hydrocephalus.

• 3. There are several reasonable approaches to managing this patient. If your suspicion is high for bacterial meningitis or shunt infection, then the patient should be admitted, even with negative studies. There have been case reports in the literature where early infections (confirmed by cultures and clinical course) presented with negative shunt and LP studies. On the other hand, if the patient looks well and you think the patient has a viral syndrome, it would be reasonable to discharge him with both primary care and neurosurgery follow-up within 24 h. In either case, the emergency doctor should coordinate care with both the primary doctor and the neurosurgeon.