Study designs of evaluations included in the review
Diagnostic studies were eligible for inclusion in the review. For studies that assessed clinical gestalt, the index test (pre-test probability) was to have been assessed before reference standard testing. It was unclear if this was also the case for studies in which clinical prediction rules were assessed. For studies in which a clinical prediction rule was derived, data were to have been collected systematically from consecutive patients and the studies had to have a minimum of 50 participants with confirmed PE.
Specific interventions included in the review
Studies estimating the pre-test probability of PE using either clinical gestalt or a clinical prediction rule were eligible for inclusion, provided the clinical assessment was made blind to the results of validated diagnostic tests. Clinical gestalt was defined as using information from the patient's history and findings from physical examination and routine investigations, without predetermined elements or a standardised score. A clinical prediction rule was defined as a mathematically derived formula that combined history, physical examination findings and routine laboratory results.
The included studies of clinical gestalt were of various combinations of access to patient history, physical examination, chest radiograph results, electrocardiogram, arterial blood gas analysis, patient symptoms, signs and risk factors, and the results of routine investigations. The clinical prediction rules were also diverse.
Reference standard test against which the new test was compared
The studies had to compare clinical gestalt or clinical prediction rules with a validated method of diagnosing PE to be eligible for inclusion. The reference standard tests included in the review were: pulmonary angiography; a combination of ventilation/perfusion lung scan, a D-dimer assay and a compression ultrasound; ventilation/perfusion lung scanning followed by angiography if the lung scan finding was nondiagnostic; and contrast spiral computed tomography of the chest with routine bilateral compression ultrasound of the legs, and lung scanning and pulmonary angiography if required.
Participants included in the review
Patients with suspected PE were eligible for inclusion in the review. Further characteristics of the included participants were not reported.
Outcomes assessed in the review
Studies that reported the pre-test probability of PE and the subsequent results of a validated diagnostic test were eligible for inclusion. The outcomes used in the review were percentage pre-test probability, actual disease prevalence and the likelihood ratios (LRs).
How were decisions on the relevance of primary studies made?
Three reviewers independently selected studies. Any disagreements were resolved by a senior reviewer. The decisions to include or exclude studies were based on the methods reported in the studies, before analysis.