Early assessment of diaphragmatic dysfunction in children in the ITU: chest radiology and phrenic nerve stimulation

Abstract

Diaphragmatic dysfunction is a common postoperative complication of cardiac surgery in children, with important effects on respiratory morbidity. Its early diagnosis, followed by prompt surgical intervention, has been shown to reduce morbidity. However, the commonest method of diagnosis, based on hemi-diaphragmatic elevation on the chest radiograph, may be less accurate than direct techniques for assessing phrenic nerve function. We have compared electrophysiological and radiological diagnoses of diaphragmatic abnormality in 100 children (aged 3 days to 17.5 yrs) undergoing cardiac surgery, looking at respiratory morbidity as assessed by the duration of ventilation, the time spent on the cardiac intensive care unit (CICU), and the requirement for reintubation. Despite showing good reproducibility, radiological diagnosis was neither sensitive nor specific in identifying patients with electrophysiological phrenic nerve damage. Analysis of the measures of outcome supported the electrophysiological technique. Patients with electrophysiological evidence of damage had a longer duration of ventilation, spent longer on the CICU, and had a greater incidence of reintubation than either radiologically abnormal or "normal" patients. Chest X-rays are not a good method for diagnosing phrenic nerve damage in the early postoperative period in children. If early diagnosis is needed, then direct assessment of phrenic nerve function, such as the measurement of phrenic latency, may be a better technique.