Statistics from Altmetric.com
Editor,—A 12 year old boy was referred with a three day history of severe abdominal and chest pain radiating to the back. He had a vesiculo-haemorrhagic rash affecting his trunk and face for two days. He had been taking prednisolone (2 mg/kg/day) and diuretics since being diagnosed with rheumatic carditis four weeks previously. Two days before admission chickenpox was considered as a diagnosis in the referring hospital, but a vesicle scrape taken for electron microscopy was negative.
On admission ischaemic bowel was suspected, and a laparotomy and oesophagogastroduodenoscopy were performed. Apart from a haemorrhagic oesophagitis and a moderately engorged liver, no abnormalities were found. The next day his condition deteriorated with shock, hepatitis, and coagulopathy. A repeat vesicle scrape was again negative on electron microscopy, but immunofluorescence demonstrated varicella zoster virus (VZV); serum VZV IgG was not detected. Despite treatment with intravenous aciclovir and foscarnet along with full intensive care unit support, he developed multiorgan failure and died three days later.
Chickenpox specific, and possibly lifesaving,1 antiviral treatment was delayed by two days in this case as a result of negative electron microscopy of a vesicle scrape. This is not a sensitive technique, only detecting virus at a concentration of 106to 107/ml.2 In a review of paired samples sent to our laboratory over the past two years, nine of 33 samples positive for VZV by immunofluorescence gave false negative results by electron microscopy. Severe abdominal or back pain, frequently preceding the vesicular rash, appears to be a common feature of complicated varicella in immunocompromised patients,3 and aciclovir may be beneficial even at the time of visceral dissemination.1 4Treatment of these children should thus start without delay, and the results of examination of vesicular material, particularly where electron microscopy alone is available, must be interpreted with caution.