Rethinking Pneumonia: A Paradigm Shift With Practical Utility

We associate the founders of germ theory, Louis Pasteur and Robert Koch, with a dusty, bygone age: gray beards, sepia tones, and antiquated techniques. However, in their own time, both Pasteur and Koch were what we would now call “early adopters,” embracing and advancing the leading edge of available technology in their efforts to identify microbes. Among his many innovations, Pasteur invented anaerobic cultivation and derived a variety of selective culture media. Koch pioneered the use of solid growth media, developed the techniques of pure culture to isolate individual species, and was the first to apply oil-immersion microscopy to respiratory specimens. [Rudolph Virchow, then the revered “Pope of Medicine,” was an old-guard skeptic. When presented with Koch’s oil-immersion discoveries, he scoffed that “What I can not see with my dry lens, I do not need to see” (1).] Though rightfully celebrated for their own scientific genius, both Pasteur and Koch were beneficiaries of recent advances in technology and both propelled their field forward with their own methodological innovations.Nearly a century and a half later, our diagnostic approach to pneumonia is in desperate need of such innovation. Although pneumonia remains a tremendous cause of morbidity, mortality, and healthcare expense, our diagnostic approach remains stuck in the 1880s. Among patients hospitalized for pneumonia in the United States, most have no pathogen identified (2, 3). In the absence of reliable, rapid diagnostic tools, most pneumonia cases are identified and treated based on clinical judgement alone, resulting in widespread and indiscriminate use of empiric antibiotics. Indeed, respiratory infections account for the majority of inappropriate antibiotic prescriptions in the United States (4, 5). Especially among critically ill patients, clinicians do not agree on which patients have pneumonia, with interobserver variability barely better than a coin flip (6). Pneumonia remains a 21st-century problem … [↵][1]1To whom correspondence should be addressed. Email: rodickso{at}med.umich.edu. [1]: #xref-corresp-1-1