Community-acquired pneumonia distinguished from influenza infection based on clinical signs and symptoms during a novel (swine) influenza A/H1N1 pandemic.

AIMS: The numbers of patients with influenza-like illnesses increase during influenza outbreaks. A study was undertaken to distinguish community-acquired pneumonia (CAP) from influenza based on clinical signs and symptoms. METHODS: This retrospective study investigated patients with positive results in the rapid influenza antigen test and those diagnosed with CAP during an influenza A/H1N1 pandemic. Significant factors for predicting risk for CAP within 48 hrs from onset and at diagnosis were selected by multiple regression analysis. RESULTS: Within 48 hrs of onset and at diagnosis, age and coarse crackles significantly increased the risk of CAP whereas sick contact, sore throat, and rhinorrhoea significantly decreased the risk of CAP. Duration of illness, sputum, dyspnoea, chest pain, and coarse crackles also significantly increased the risk of CAP at diagnosis. CONCLUSIONS: CAP differed somewhat from influenza even within 48 hrs of onset and the differences became even more evident thereafter.

Significance of the progression of respiratory symptoms for predicting community-acquired pneumonia in general practice.

BACKGROUND AND OBJECTIVE: Early diagnosis improves outcomes in patients with community-acquired pneumonia (CAP). However, prediction of CAP based on symptoms and signs is difficult. The present study investigated the evaluation of progression of symptoms as a factor for predicting the occurrence of CAP in general practice. METHODS: Consecutive patients (n = 406) suspected of having CAP on routine clinical examination were studied retrospectively. Selection of patients with suspected CAP was based on progression of symptoms after 5 days, as well as published criteria. Diagnostic yields for the recommended criteria and our proposed criteria were then compared. Scoring systems for the prediction of CAP were designed, based on the results of multiple regression analysis. The diagnostic performance of these systems, including or excluding symptom progression, was compared using the areas under receiver operating characteristic curves. RESULTS: The sensitivity and specificity of the recommended criteria and our proposed criteria were 0.75 and 0.44, and 0.93 and 0.38, respectively. Sputum production, dyspnoea, fever > 38 degrees C, heart rate > 100 beats/min, decreased breath sounds, coarse crackles and progression of symptoms significantly increased the likelihood of CAP. Areas under receiver operating characteristic curves analysis showed that the diagnostic prediction of CAP was significantly improved when the scoring system included progression of symptoms. CONCLUSIONS: Progression of symptoms was a significant factor for predicting CAP and selecting patients who required CXR. Inclusion of progression of symptoms among the other recommended criteria, namely, dyspnoea, fever > 38 degrees C, heart rate > 100 beats/min and abnormal chest findings, improved prediction of the incidence of CAP in general practice.