A total of 15 studies (n=4378) were included in this review comparing screening strategies for cancer detection. Thirteen studies were cohort designs (4 retrospective, 9 prospective), one study prospectively compared exposed and non-exposed cohorts and the final study was an RCT, although the data had been extracted and used as additional cohorts.
CT of abdomen/pelvis: analysis of five studies found that this technique significantly increased the proportion of undiagnosed cancer detected following extensive screening to 66.1% (95%CI: 59.0, 73.2) compared to limited screening which detected only 47.6% of cases (95% CI: 40.0, 55.1) in all patients with VTE. In the subgroup of patients with unprovoked VTE there was a significant increase in the proportion of detected cancer, from 49.4% (95% CI: 40.2, 58.5) in the limited screening groups, to 69.7% (95% CI: 61.1, 77.8) in the extensive screening groups.
US of abdomen/pelvis: analysis of eight studies did not find a significant difference in the proportion of undiagnosed cases of cancer detected between screening types.
Four studies compared the rate of cancer detection in early-stage, previously undiagnosed cancer between limited and extensive screening programs. Pooled data across diagnostic tests suggested that extensive screening increased the number of cases detected from 13.5% (95% CI: 8.8, 21.0) under limited screening to 22.0% ((95% CI: 15.6, 30.1).
One study reported rate of cancer-related mortality by screening strategy: an absolute difference of 1.9% cancer-related mortality (95% CI: -5.5, 10.9) in favour of extensive screening was detected.
US of the abdomen/pelvis and screening for carcinoembryonic antigen (CEA) and prostate-specific antigen (PSA) tumour factors did not significantly increase the frequency of cancer detection over and above limited screening alone.
No studies reported screening complications.