Community acquired pneumonia: severity scores as an adjunct to clinical judgement

To the Editors:

We read with interest the article by Choudhury et al. [1] in a recent issue of the European Respiratory Journal. The authors investigated the reasons for hospital admission in pneumonia patients with a CURB-65 (confusion, urea >7 mmol·L⁻¹, respiratory frequency ≥30 breaths·min⁻¹, systolic blood pressure <90 mmHg or diastolic blood pressure ≤60 mmHg and age ≥65 yrs) score of 0–1, in order to identify the potential for improving outpatient management. However, the conclusion of the study is discordant with the objectives. The authors conclude that their study supports the recommendation from international guidelines that pneumonia severity scores should be used as an adjunct to clinical judgment when determining the need for hospitalisation. However, there are several issues regarding this study that should be mentioned. First, in order to avoid selection bias when recording the causes of admission, a prospective cohort study design should be applied and not a retrospective review by independent reviewers as mentioned by the authors. Secondly, in order to arrive at the authors conclusion that international guidelines should be used as an adjunct to clinical judgment when deciding on hospitalisation, the authors should have a priori strict criteria according to the severity of illness (CURB-65) as to which patients with community-acquired pneumonia need to be admitted and which do not. Therefore, strict admission criteria may facilitate the identification of variables that may be important for clinicians at the time of admission, in order to overrule CURB-65. Thirdly, we believe that clinical judgement is not equivalent, generalisable and infallible for every physician caring for patients with pneumonia. As suggested previously, certain physicians are more likely to appropriately manage patients with community-acquired pneumonia compared with others, and this may influence the results of the authors' observations [2, 3]. Finally, we agree with the authors that a prospective cohort design is desirable in order to appropriately address these questions.

• ©ERS 2012