Pulmonary diagnostic procedures in the critically ill.

When faced with a critically ill patient with new pulmonary infiltrates on chest roentgenograms, the physician must choose the appropriate diagnostic procedure on the basis of the expected yields versus the potential complications. The first steps in any patient should include discontinuation of any nonessential medications, careful evaluation of fluid status to exclude cardiogenic pulmonary edema, and a review of likely diagnoses based on the patient's underlying disease. Although not likely to be of immediate utility, obtaining cultures of blood and other body fluids or sites and serologic testing may provide helpful information when combined with other procedures. Bronchoscopy is a reasonable first step in patients with a slow progression of disease or in those in whom the pulmonary process is discovered early in its course. As these patients often present with several of the known risk factors for complications with bronchoscopy, the decision to perform this procedure should not be made lightly. Transbronchoscopic lung biopsy adds additional risk to bronchoscopy but also increases the diagnostic yield considerably over lavages, brushing, and bronchial washings. Open lung biopsy offers high diagnostic yields and relatively low rates of serious complications. Because of the invasive nature of the procedure, there is often reluctance to perform it. In patients with rapidly progressive disease or conditions that make the risk of bronchoscopy unacceptably high, such as severe hypoxemia, bleeding diathesis, or cardiac compromise, prompt diagnosis requires that the physician consider open lung biopsy as a first diagnostic procedure. The physician must also consider whether making a specific diagnosis will be of benefit to the patient. Potential benefits of a specific diagnosis include stopping unnecessary empirical (and potentially toxic) therapies, instituting correct and specific therapy, and thus decreasing morbidity and mortality. The impact of specific diagnosis on morbidity and survival is often difficult to demonstrate. Discouraging notes have been sounded by studies of the effect of bronchoscopic or surgical diagnosis on the ultimate outcome for patients. For bronchoscopy with transbronchoscopic lung biopsy, although the overall diagnostic rate was 60 per cent, no difference in survival was noted between patients in whom a diagnosis was made and those in whom the nature of the pulmonary process remained unknown. Similarly, in a series of patients who underwent open lung biopsy, although the results of biopsy led to a therapeutic change in 70 per cent of the patients, only 16.5 per cent of the patients benefited from this change.(ABSTRACT TRUNCATED AT 400 WORDS)