This Article Abstract Full Text Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ng, C. Articles by Palmer, C. Search for Related Content PubMed PubMed Citation Articles by Ng, C. Articles by Palmer, C.

Analysis of diagnostic confidence and diagnostic accuracy: a unified framework

¹ Department of Radiology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA, ² Centre for Applied Medical Statistics, University of Cambridge, Cambridge CB2 2SR, UK

View larger version (15K): [in this window] [in a new window]   Figure 1. Routes schematically showing changes in diagnoses on admission, following CT and at final diagnosis. The upper diagram schematically represents scenarios where the post-CT diagnosis is of high confidence and is correct. The middle diagram schematically represents scenarios where the post-CT diagnosis is of high confidence and is incorrect, and the final diagnosis is more severe than the post-CT diagnosis. The lower diagram schematically represents scenarios where the post-CT diagnosis is of high confidence and is incorrect, and the final diagnosis is less severe than the post-CT diagnosis. Left-hand labels: Route labels. For consistency the same nomenclature is used in Figure 2. x (y, z), Confidence score assigned to route travelled by patient from admission to final diagnosis: x=best estimate (y=pessimistic, z=optimistic) score.

View larger version (34K): [in this window] [in a new window]   Figure 2. Flow chart demonstrating the 36 possible combinations of confidence levels, diagnostic changes and ultimate diagnostic accuracy. Route Label: the italiczed labels (such as A1 and A3) are the same route labels as those in Figure 1. *Route score: this score was derived from a weighted average of the best estimate, pessimistic, and optimistic score (see Materials and methods section). For clarity, the underlying component scores have been presented, denoted as x (y, z) = best estimate (pessimistic, optimistic) scores. **Last row of flow chart = number of patients in each route in the case study.

View larger version (29K): [in this window] [in a new window]   Figure 3. Conventional analyses of confidence without and with Omary correction. Note: the analyses are independent of the correctness, or otherwise, of the post-test diagnosis.

View larger version (43K): [in this window] [in a new window]   Figure 4. Flow of patients through the study.*Represents patients in routes F4, F6 in Figure 2. **Represents patients in routes C4, C5, E2, E3, E4, E6 in Figure 2. ***Number of post-CT diagnoses correct at follow up. Represents patients in routes A5, A6, B5, B6 in Figure 2 and is the number of admitting diagnoses correct at follow up.

View larger version (15K): [in this window] [in a new window]   Figure 5. Frequency distributions, and superimposed normal curves, of within-person changes in diagnostic confidence on common [–4, 4] scale, for n = 62 study subjects, according to: (a) score-based weighted average method; (b) conventional analysis, uncorrected for post-CT change in diagnosis; and (c) conventional analysis, with Omary correction for post-CT change in diagnosis.