Study designs of evaluations included in the review
Initially, case series with less than 10 observations and review articles without original data were excluded. Of the initial included articles, only primary prospective studies of the accuracy or precision of the clinical examination in CAP (studies rated as level I quality) were included in the main analyses. For accuracy studies, this required independent, blind comparisons of findings with a 'gold' standard of diagnosis or aetiology among a large number (greater than 50) of consecutive patients suspected of having CAP. For precision studies, this required at least two independent blinded raters of symptoms or signs in patients suspected of having CAP.
Specific interventions included in the review
The included studies had to use the clinical history and/or physical examination as the index test.
Reference standard test against which the new test was compared
No inclusion criteria relating to the reference standard were specified. The reference standard used in the four studies included in the main analysis was chest radiography; the diagnosis of pneumonia required the identification of a new infiltrate on a chest radiograph.
Participants included in the review
Studies of patients suspected of having CAP, who were aged over 16 years and who did not have known immunosuppression or nosocomial infections, were eligible for inclusion.
Outcomes assessed in the review
No inclusion criteria relating to the outcome measures were specified. The outcome measures reported in the review were the mean pair observer agreement rates and kappa values as measures of precision, and the positive and negative likelihood ratios (LRs) as measures of accuracy. Positive and negative predictive values were also discussed in the text.
How were decisions on the relevance of primary studies made?
Two of the authors applied the initial inclusion criteria.