Abstract
Background: CRB-65, a useful tool to predict community-acquired pneumonia (CAP) outcome in hospitals, has been recommended for use in primary care. CRB-65 scores are calculated by assigning one point for each of: the presence of Confusion, Respiratory rate ≥ 30 per minute, Blood pressure systolic < 90 mm Hg or diastolic ≤ 60 mm Hg, and age ≥ 65 years.
Aim: To assess the validity of CRB-65 to predict poor prognosis in adults presenting to primary care with lower respiratory tract infection (LRTI).
Methods: Clinicians prospectively recorded clinical features on a case registration form. Patients had a chest X-ray within 7 days after inclusion. A notes review was performed. We used a two-level logistic regression model (with patients nested within clinicians) to assess the association between a CRB-65 score ≥ 1 and mortality and a combination of re-consultation and hospital admission, and tested for interaction to assess differences in these outcome between patients with and those without CAP.
Results: None of the 3112 included patients died due to the LRTI. Complete data were available in 2627 (84%) patients. Of these, 866 (33%) had a CRB-65 score ≥ 1, 108 (4%) had CAP, and 488 (19%) re-consulted or were admitted to hospital. Both a CRB-65 score ≥ 1 and CAP were associated with more re-consultations or hospital admissions (odds ratio (95%CI): 1.32 (1.06–1.64) and 2.18 (1.13–4.21), respectively). The interaction term was not significant (1.17 (0.57-2.41)).
Conclusion: In patients presenting to primary care with LRTI, low mortality and low prevalence of CAP limit the usefulness of CRB-65. Nevertheless, a CRB-65 score ≥ 1 is associated with a significant increase in re-consultations and hospital admissions.
- © 2011 ERS