Purpose of review: Patients with lower respiratory infections display a wide spectrum of disease severity. Management decisions regarding site of care, extent of investigations and level of treatment are mainly based on disease severity. Several severity assessment tools are still undergoing evaluation. This review highlights recent relevant studies.
Recent findings: Severity prediction rules such as the Pneumonia Severity Index cannot be relied upon as the sole means of identifying patients with lower respiratory infections who do not need hospital admission. Up to 40% of patients assigned to low-risk groups may require hospitalization. The most common medical reason for hospitalization in these circumstances is the presence of unstable comorbid illness. Social factors are equally important in the decision to admit. A new prediction rule based on the British Thoracic Society prediction rule has been proposed. This divides patients with community acquired pneumonia into three management groups based on risk of mortality. A prediction rule for use in patients with HIV-associated community acquired pneumonia has also been proposed. An IL-10 polymorphism has been identified as a prognostic factor in community acquired pneumonia. Audits examining the impact of adherence to severity-based guidelines on the outcome of community acquired pneumonia have revealed conflicting results. Differences in outcome may only be significant for patients with the most severe illness.
Summary: Severity of illness assessment is important in guiding management options at various stages in the clinical course of lower respiratory infections. However, no prediction rule should supercede clinical judgment.