Purpose of Review
Community-acquired pneumonia is a significant clinical and public health problem. Defining and predicting severe pneumonia is difficult but important.
Recent findings
Several new predictive models and more sophisticated approaches to describing pneumonia severity have been recently proposed, with subsequent validation in varied patient populations. Early data suggest that biomarkers may be useful in the future.
Summary
Definitions of pneumonia severity depend on the relevant clinical or public health question. A health-services reference definition seems most useful in most settings. The IDSA/ATS 2007 guidelines and SMART-COP are two recent promising methods for predicting severe pneumonia at the time of presentation. The traditional Pneumonia Severity Index and CURB-65 models are less useful. Accurate assessment of severity has important implications for triage, outcome, and defining populations for research applications. Novel biomarkers, while somewhat promising, do not yet have a validated role in pneumonia severity assessment.