Decline in microbial studies for patients with pulmonary infections.

Received 1 March 2004; accepted 2 March 2004; electronically published 1 July 2004. Reprints or correspondence: John G. Bartlett, 1830 Monument St. #437, Baltimore, MD 21205 (jb@jhmi.edu). Clinical Infectious Diseases 2004; 39:170–2 2004 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2004/3902-0004$15.00 There appears to be a substantial decline in the quality of microbiology done in the context of pulmonary infections during the current era, compared with 3 decades ago, and the question posed to the medical community is whether this change is acceptable or reversible. The decline seems to be well documented. In the prepenicillin era, for example, Bullowa [1] reported his own experience with 14000 cases of pneumonia, from 80% of which S. pneumoniae was recovered. In the 1950s, most reports had a yield of pneumococci in patients hospitalized with community-acquired pneumonia of 40%– 70%. During the past 15 years, it has been a challenge to find any US-based study in which the yield of pneumococci was 118%. The experience at Johns Hopkins Hospital (Baltimore, MD) reflects these trends: the yield of S. pneumoniae from Gram staining and culture of sputum samples in cases of community-acquired pneumonia in 1970 was 60% [2]; in 1980, it was 40% [3]; and in 1991, it was 18% [4]. The conclusion from these reports is that either the pneumococcus is disappearing or microbiology is disappearing. Some would argue that other microbes have become far more prominent and thus fill the void, but this argument is weakened by the fact that few studies identify any likely pathogen in 50%–70% of cases. I queried Robert Fekety (University of Michigan) on his explanation for the 60% yield of S. pneumoniae in 1970, compared with the 18% yield in 1991 at the same hospital, and his response was not surprising: the difference was attributed to a dramatic change in emphasis on microbial detection. The previous era was characterized by house staff laboratories in every ward, and house staff and attending physicians often spent long periods performing a diagnostic examination of the expectorated sputum specimen, which had been obtained with great care and plated for prompt incubation in incubators on the ward. Therapeutic decisions were generally based on these results. During the past 30 years, there has been a notable decline in the quality of this exercise. Some of this decline can be ascribed to the Clinical Laboratory Improvement Amendments of 1988, which required that staff have credentials to interpret Gram stains of any specimens, thus essentially eliminating the house staff laboratory. Additional factors in the decline of microbiology were the outsourcing of specimens, which led to delays in processing and poor communication between the microbiologist and the physician. There were also the economic pressures to reduce cost: the chemistry laboratory became highly efficient because of technological advances, but the microbiology laboratory continued to depend on labor-intensive practices developed by Robert Koch in the 19th century. Some would argue that this transition simply reflects the current climate in medicine and has been accomplished with minimal loss. It can be argued correctly, for example, that microbial studies usually have a low yield even when there is an exhaustive search for every conceivable pathogen. There is also the lack of documented benefit in terms of reduced cost or better outcome. Perhaps most important is the apparent effectiveness of antibiotics that are selected empirically. One of our colleagues summarized his approach as follows: