Table 3

Intraclass correlation coefficient of CGI ratings and their 95% confidence intervals.

CGIs1

95% C.I.

CGIs2

95% C.I.

CGIs1–2

95% C.I.

CGIi

95% C.I.


Group 1

0.64

[0.38, 0.77]

0.70

[0.53, 0.81]

0.64

[0.49, 0.77]

0.65

[0.39, 0.79]

Group 2

0.65

[0.40, 0.79]

0.80

[0.70, 0.87]

0.74

[0.54, 0.82]

0.74

[0.56, 0.84]

Group 3

0.86

[0.72, 0.93]

0.74

[0.57, 0.85]

0.69

[0.41, 0.84]

0.71

[0.53, 0.81]

Group 2'

0.88

[0.73, 0.94]

0.94

[0.90, 0.96]

0.91

[0.79, 0.95]

0.92

[0.83, 0.95]


CGIs measures severity at T1 (CGIs1) at T2 (CGIs2) and CGIi improvement between T2 and T1. CGIs1–2 measures the difference between the CGI severity scores measured at T2 and T1. Group 1 corresponds to 3 raters using the traditional CGI response format. Group 2 corresponds to 4 raters using the improved response format. Group 3 corresponds to a pair of ratings derived from a restricted Delphi procedure (each rating is obtained from 2 clinicians) and using the improved response format. Group 2' is Group 2, but the intraclass correlation coefficient computed here estimates the reliability of the average of 4 ratings derived from independent raters.

Kadouri et al. BMC Psychiatry 2007 7:7   doi:10.1186/1471-244X-7-7

Open Data