Comparative diagnostic accuracy of respective nuclear imaging for suspected fracture-related infection: a systematic review and Bayesian network meta-analysis.

Abstract

The aim of this study was to compare the accuracy of available nuclear imaging modalities in the diagnosis of suspected fracture-related infection (FRI). We conducted a comprehensive literature search of PubMed, EMBASE and the Cochrane Library to retrieve diagnostic accuracy studies in which FRI was investigated using different nuclear imaging modalities. The pooled sensitivity, specificity, likelihood ratios and diagnostic odds ratios were constructed using the bivariate meta-analysis framework, while the superior index was pooled using Bayesian network meta-analysis. 22 eligible studies (1,565 patients) were included in the quantitative analysis. A broad overlapping confidence interval (CI) of pooled sensitivity was observed among bone scintigraphy (0.94; 95% CI 0.85-0.98), 18F-FDG PET and PET/CT (0.91; 95% CI 0.85-0.94) and leukocyte scintigraphy (0.86; 95% CI 0.53-0.97). Bone scintigraphy (0.34; 95% CI 0.08-0.75) seemed to be less specific than all the other modalities, while leukocyte scintigraphy (0.96, 95% CI 0.92-0.98) was notably more specific than 18F-FDG PET and PET/CT (0.78; 95% CI 0.69-0.85). Based on the superiority index, 18F-FDG PET/CT (3.78; 95% CI 0.14-11.00), 18F-FDG PET (2.98; 95% CI 0.14-9.00) and leukocyte scintigraphy (1.51; 95% CI 0.11-7.00) all achieved high accuracy in detecting FRI. Bone scintigraphy is a highly sensitive nuclear imaging technique but lacks the specificity needed to unequivocally differentiate among various conditions suspected to be FRI. Leukocyte scintigraphy, 18F-FDG PET/CT and PET all present good satisfactory accuracy for the diagnosis of FRI, but their costs should be further reduced to promote their wide application.