For several reasons, bronchoscopy is especially suited for diagnosing pulmonary infections in immunocompromised patients, but it is less suited for the diagnosis in immunocompetent patients:
(1) Immunocompromised patients are more prone to develop pulmonary infections with certain pathogens, such as P carinii, mycobacteria, HSV, СМУ and Aspergillus species (henceforth, these pathogens will be referred to as specific pathogens, requiring specific therapy, as opposed to empirical therapy). In our study, these organisms were found in 9/30 (30 percent) of the immunocompromised patients and 5/32 (16 percent) of the immunocompetent patients.
(2) The clinical picture in these infections is nonspecific and not conclusive enough to indicate specific organisms, and hence, guide specific therapy.
(3) With a standardized bronchoscopy program like ours, these agents can be found. New advancements in the field of molecular biology have further increased the possibilities to diagnose some of these pathogens. For example, after this study was completed, polymerase chain reactions (PCR) are in use at our hospital to diagnose pulmonary infections with CMV and M tuberculosis, and a PCR method to diagnose Mycobacterium avium is under development.
(4) Patients are seldom receiving specific treatment prior to the bronchoscopy. Among our patients, only one (patient 10, Table 3) was receiving such treatment before the bronchoscopic investigation.
(5) Presence of these pathogens is in most cases directly therapeutically conclusive. Major exceptions are fungi, where lung biopsy specimens generally are necessary to make a definitive diagnosis, and certain viruses. With specific and effective therapy against СМУ antiviral therapy in our hospital is used as empirical therapy, especially in patients who have undergone transplants. In this study, virus culture was still the method of choice. The final answer was often not at hand until after several weeks, and hence, specific therapy against CMV was not given. With the rapid diagnostic possibilities provided by the PCR technique, patients today can be given antiviral treatment within 48 h of the bronchoscopy if CMV is found in BAL fluid.
(6) Prior antibiotic therapy decreases the yield from bronchoscopy. This is demonstrated by the low yield among immunocompetent patients receiving empirical therapy and further underlines the fact that bronchoscopy is most rewarding in the search for specific pathogens in immunocompromised patients. Among the immunocompetent patients in our study, therapeutically guiding results were obtained in 4/14 (28.6 percent) if the patients were not receiving empirical antibiotic therapy, but only in 2/18 (11.1 percent) patients who had received antibiotics prior to the bronchoscopy. In conclusion, bronchoscopy is most cost-effective among immunocompromised patients.
In conclusion, our study, conducted over a long period of time and on a heterogeneous group of patients, indicates that diagnostic bronchoscopy is of great value for diagnosing pulmonary infections in immunocompromised patients. However, for the immunocompetent patients, the diagnostic yield is lower, especially if they have been receiving prior antibiotic therapy, and the indication for bronchoscopy must be established for each individual patient based on clinical importance, resources, and risk. Since the complication rate is not negligible, and bronchoscopy itself may be a risk factor for nosocomial pneumonia, the decision to perform bronchoscopy should always be preceded by careful consideration.