Forty-four diagnostic accuracy studies were included in the review (2,867 patients with 2,896 nodules): 10 DCE CT; six DCE MRI; 22 FDG PET; and seven SPECT. Just over half the included studies were prospective. All studies reported an appropriate reference test. Most indicated complete verification by the reference test. Less than half reported blinded interpretation of test results. Few reported adequate description of the index test. Only three were described as having adequate reporting of results.
Positive likelihood ratios for detecting malignancy in an solitary pulmonary nodule for the diagnostic tests were: DCE CT 3.91 (95% CI: 2.42 to 5.40, I2=82%); DCE MRI 4.57 (95% CI: 3.03 to 6.10, I2=0%); FDG PET 5.44 (95% CI: 3.56 to 7.32, I2=56%); and SPECT 5.16 (95% CI: 4.03 to 6.30, I2=75%). These indicated small to moderate changes in likelihood of disease.
Negative likelihood ratios were: DCE CT 0.10 (95% CI: 0.03 to 0.16, I2=0%); DCE MRI 0.08 (95% CI: 0.03 to 0.12, I2=0%); FDG PET 0.06 (95% CI: 0.02 to 0.09, I2=0%); SPECT 0.06 (95% CI: 0.04 to 0.08, I2=98%). These indicated moderate to large changes in likelihood.
Differences in performance for all tests were negligible.