Sensitivity and significance of disease outcome measures in inflammatory bowel disease

In inflammatory bowel disease (IBD), proxy measures of clinical outcome are often collected into summary indices of qualitative self-rated disease markers, clinical observations, and quantitative biochemical analyses. In Crohn's disease (CD), a frequently used index is the Crohn's disease activity index (DCAI). This index consists of six qualitative variables and two quantitative variables. The aim of this presentation is to illustrate the use of this index to calculate its range, to estimate errors in the index, its sensitivity, and the number of significant steps in the index. The measure of sensitivity of the summary index was analyzed for the signal-to-noise ratio (SNR), the reference change value (RCV) and the confidence interval (CI). If identical errors were assumed in patient self-rated health and clinically judged disease manifestations, such as tumours and fistulas, the majority of the variance of the index was caused by the self-rated experience of health, the number of days over which the individual variable was rated, and the prognostic multiplier of each variable.The range of the index has no upper limit, but can be estimated to 403 units, of which patient self-rating of well-being account for up to one-third of the summary index maximal score range. The median signal noise measure of index sensitivity was 18 SDs. The two disease classification limits of 150 units for moderate disease and 450 for severe disease on average cover an interval of limit ±41.5 units vs. ±60.5 units. In judgments on change in clinical outcome the RCV interval of steps of 121 units are valid. Conclusion: Both variance and range of the CDAI summary score are primarily decided by the self-rated experience of well-being. Variables on disease signs have a minor impact on the index. Rating of the two important outcome parameters: Self-experienced health and medical outcome would favourably be given in two individual scores.