We read with interest the recent report by Doyle and associates [1] on the necessity of routine chest radiography after thoracentesis. It was encouraging to see that operator suspicion of pneumothorax was useful in predicting the occurrence of pneumothorax. However, it is noteworthy that one of the nine patients who developed pneumothorax after thoracentesis did not have one of the characteristics (previous thoracic radiation, number of passes with the thoracentesis needle, aspiration of air, or operator suspicion of pneumothorax) noted by Doyle and colleagues to be associated with an increased risk ratio.

We are concerned by the authors' statement that "standard medical practice after thoracentesis is to obtain a posteroanterior chest roentgenogram" and their assertion that this practice opposes the 1988 American Thoracic Society guidelines [2]. These guidelines state that "a chest film should be performed after therapeutic thoracentesis in most instances." This does not seem to oppose standard practice.

We share with the authors the belief that most chest radiographs obtained after thoracentesis do not show a complication. We further agree that most unsuspected pneumothoraces do not result in serious clinical consequences. However, we disagree with the authors' statement "the individual clinician must decide what frequency of unsuspected pneumothorax is acceptable." We doubt that our current legal system would agree that even one episode of unsuspected pneumothorax resulting in excess morbidity or mortality without radiographic screening is acceptable.

We admire the authors' attempts to highlight correlations between procedures and pneumothorax, but we doubt that standard medical practice will change until standard legal practice changes.