Page 140 - 2018_10-Haematologica-web

H.M. Schoemans et al.
Table 5. Detailed results of participants for graft-versus-host disease (GvHD) vignettes compared to the Expert Gold Standard – GvHD Severity Scoring.
Expert Gold Standard Grade Severity Scoring II
GradeI°
GradeII° 0 (0%)
Results from the "App" group given by 37 participants - n (%)
Grade Grade Grade Mild Moderate Severe Missing Other Total III IV
1 0 0 2 1 0 0 0 37
Grade III ° 00
(0%) (0%)
Grade IV ° 013
(0%)
(0%) (0%) (5%) (0%) (0%) (10%) 0 0 0 0 0 37
33 (89%)
(0%) (3%) (8%)
Mild °° 0000
(0%)
(0%) (0%) (0%) (0%) (10%) 22 1 1 1 74
(0%) (0%) (0%) (0%) Moderate °° 00000
(30%)
(1%) (1%) (1%) (20%) 11 1 1 74
(0%) (0%) (0%) (0%) (0%)
Severe °° 0 0 0 0 2 18
(15%)
(1%) (1%) (20%) 0 0 74
tion that evaluates the patient at a single point in time. Expanding on our promising accuracy results and user-feed- back, future plans include the development of a module to perform longitudinal patient evaluations (with an integrat- ed disease response evaluation according to international criteria18,19) and a module to capture patient-reported GvHD evaluation based on the Lee symptom scale.20 These added functionalities will dramatically increase the clinical useful- ness of the tool in following patients over time.
However, a challenging issue with eHealth tools is how to approach their constant and rapid change over time. This evolution is driven by evolving clinical practices, user feed- back, and updates in computer programs and/or operating
systems. The results reported in this study, for instance, have been obtained with a version of the eGVHD app which has already become obsolete, as a new version (using additional skip-logic features) has been developed to address the valid criticism expressed about the time-con- suming aspect of its use. The constant evolution of the vir- tual world is a challenge in the current context of European regulation (EU Directive 93/42/EEC MEDDEV 2. 4/1 Rev. 9 June 2010), which requires eHealth applications to be for- mally validated by a tedious quality assurance process at every new adaptation of the tool. This is not practically fea- sible in real life, and is probably, more often than not, unnecessary. Health regulation agencies will need to adjust
(3%)
(0%) (0%) (5%) (3%) (0%) (0%) (0%) (10%) 0 0 0 0 0 0 0 37
37 (100%)
(0%)
(0%) (0%) (0%) (0%) (0%) (0%) (10%) 0 0 0 2 0 0 37
35 (95%)
33 (89%)
49 (66%)
61 (82%)
(0%) (0%) (20%) (10%) (10%) (9%) (14%) (27%) (18%) (0%) (0%) (100%)
Results from the "No App" group given by 40 participants - n (%)
Grade II Grade III Grade IV Mild Moderate Severe Missing Other Total
4 0 0 4 1 0 0 2 40
(0%) (0%) (0%) (0%) (3%) (24%)
Total 33 39 38 33 53 102 68 2 2 370
(9%)
Expert Gold Standard
Severity Scoring Grade I
GradeI°
Grade II ° 3 (8%)
(10%)
54 (73%)
29 (73%)
(0%) (0%) (10%) (3%) (0%) (0%) (5%) (10%) 4 1 3 13 4 1 0 40 (10%) (3%) (8%) (33%) (10%) (3%) (0%) (10%)
9 0 1 1 0 2 40 (23%) (0%) (3%) (3%) (0%) (5%) (10%)
GradeIV° 1 8 19
(3%) (20%) (48%) (3%)
Mild°° 13 12 0 0
(16%) (15%) (0%) (0%) Moderate°° 5 8 4 0 5
GradeIII° 0 0 (0%) (0%)
(6%) (10%) (5%) (0%) (6%) Severe°° 1 9 9 1 8 23
(19%)
(0%) (4%) (20%) 0 2 80 (0%) (3%) (20%)
11 (28%)
27 (68%)
(1%) (11%) (11%) (1%) (10%) (29%)
Total 52 52 63 18 53 99 49 2 12 400
7 (28%)
(13%) (13%) (16%) (4.5%) (13%) (25%) (12) (0%) (3%) (100%)
n: number; "Missing" corresponds to a lack of answer; "Other" corresponds to any answer not matching the proposed choices. The number of ° symbols used indicates the num- ber of clinical vignettes involved.The highlighted diagonal corresponds to a perfect agreement between participants and expert results.
1 2 0 0 2 40 (5%) (0%) (0%) (5%) (10%)
32 (40%)
19 2 1 1 80
(24%) (1%) (1%) (1%) (20%) 15 0 3 80
40 (50%)
27 (34%)
1706
haematologica | 2018; 103(10)