In addition to the summaries of the proportions with improvement and worsening, the proportion of patients who improved one step (eg, improvement from A to B or B to C) or three steps (eg, improvement from A to D) in a given organ domain from A or B at baseline to week 52 was evaluated.
The proportion of improvement of the overall model fit relative to the independence model was good (NFI = 0.912), indicating the proposed model was 91.2 % better than the independence model.
Proportion of improvement in palliation score was evaluated by Fisher's exact test.
Conversely, when reporting results at the specific level, both literatures use a higher proportion of improvement rather than explicit enhancement language.
In each dataset, we computed the proportion of improvement in terms of AUC brought by iASeq over the best single-dataset based ranking method (i.e., dAUC = AU C iAseq − AU C bestsingle AU C bestsingle ).
Figure 6h shows the distribution of dAUC (i.e., the proportion of improvement of AUC by iASeq over the best single-dataset based ranking method in each dataset) in all 40 datasets.
With 34 participants per group, we will have 97% power to find a significant difference between at least two of the 6 groups and 80% power to find a significant difference between Group 6 and Group 2, the treatment groups assumed to differ most in proportion of improvement.
To incorporate deteriorations prevented, we calculated the difference in the proportion of improvements minus deteriorations in the intervention group and improvements minus deteriorations in the control group.
Greatest proportion of improvements were noted in 2009 2010 (87-88% compared to 74-77% in other years), but in multivariate models including age, baseline severity and time period, no statistically significant associations were found with improvement or recovery and time period.
The total sample size for this study (adding phases II and III) is computed as the minimal number of patients per group to detect differences from 15%, even in the case of maximal requirements that would be having proportions of improvement in HRQoL close to 50%.
