Diagnostic performance of FDG PET/CT in critically ill patients with suspected infection: A systematic review and meta-analysis.

BACKGROUND/PURPOSE: Nuclear imaging, including gallium scintigraphy and fluorodeoxyglucose (FDG) positron emission tomography (PET), has been widely used to identify focus of infection in fever of unknown origin. However, little is known about its role in critically ill patients, who are usually with multiple inflammatory foci and unable to tolerate long image acquisition time. This systematic review aimed to evaluate the diagnostic performance of FDG PET for suspected infection in critically ill patients.

METHODS: PubMed and Embase were searched up to July 24th, 2019 to identify studies evaluating the diagnostic performance of FDG PET for finding infection focus in critically ill patients following the PRISMA guidelines. The bivariate mixed-effects model was used to pool the measure for diagnostic performance. Publication bias was evaluated by Deeks' method.

RESULTS: A total of 4 studies with 87 patients were included. All the four studies evaluated FDG PET. Majority of the patients were either mechanically ventilated (76%) or shocked requiring vasopressors (61%). Test and transportation related adverse events were rare (2%). The summary sensitivity and specificity were 0.94 (95% CI, 0.79-0.99) and 0.66 (95% CI, 0.45-0.83), respectively. The AUC for summary ROC curve was 0.83.

CONCLUSIONS: FDG PET was a very sensitive tool with acceptable specificity for detecting the origin of infection in critically ill patients. However, current available studies have limitation in evaluating safety issue. Further research should investigate both benefit and risk of doing this test for this group of vulnerable patients.