| Purpose To investigate the so far hidden additional prognostic value of data provided by new diagnostic imaging techniques by estimating the prognostic value of ancillary (unrequested) findings for the occurrence of several relevant diseases (figure 1). Research question Does ancillary imaging data have additional predictive value for the occurrence of relevant morbidity? Imaging plays a key role in diagnosis. Over 9 million investigations are performed annually in the Netherlands. Advances in imaging technology over the last decade have provided the opportunity to abstract prognostic information from multislice Computed Tomography (msCT) scans obtained for diagnostic reasons. This is often referred to as ancillary (unrequested) information. Currently, these imaging characteristics are largely ignored because their relevance is not known nor quantified. In this project, we will investigate the extent to which unrequested imaging findings are of prognostic relevance for the occurence of several relevant diseases, other than those clinically indicated. Using a case-cohort approach, we will follow up 20.000 patients with an indication for chest msCT. The msCT of all patients who experience a cardiovascular event, hip or vertebral fracture or lung disease (cases) and of a random sample of the baseline cohort (controls) will be retrieved and assessed with regard to information about a) known pathology, b) abnormalities of unknown relevance and c)other imaging findings. These imaging characteristics will be used together with clinical characteristics and routine classification of CT to construct prognostic models to predict future morbidity and mortality (cardiovascular, hip and vertebral fracture and admission for lung cancer or chronic obstructive pulmonary disease). The derived predictors will then be prospectively validated in patients not included in the derivation cohort. The novelty of the proposal lies in the recognition that imaging data obtained to serve a particular diagnostic purpose in daily clinical practice not only contain information related to the original indication, but also ancillary information of potential clinical benefit. It will identify which imaging information is medically relevant to patients and their physicians and which information is normal and should not lead to therapeutic intervention. The approach and findings may serve as a role model for other diagnostic techniques. |
