Anzeige
Erschienen in:
01.06.2014 | Original Article
Optimizing study design for interobserver reliability: IUGA–ICS classification of complications of prostheses and graft insertion
Erschienen in: International Urogynecology Journal | Ausgabe 6/2014
Einloggen, um Zugang zu erhaltenAbstract
Introduction and hypothesis
Results of interobserver reliability studies for the International Urogynecological Association–International Continence Society (IUGA-ICS) Complication Classification coding can be greatly influenced by study design factors such as participant instruction, motivation, and test-question clarity. We attempted to optimize these factors.
Methods
After a 15-min instructional lecture with eight clinical case examples (including images) and with classification/coding charts available, those clinicians attending an IUGA Surgical Complications workshop were presented with eight similar-style test cases over 10 min and asked to code them using the Category, Time and Site classification. Answers were compared to predetermined correct codes obtained by five instigators of the IUGA-ICS prostheses and grafts complications classification. Prelecture and postquiz participant confidence levels using a five-step Likert scale were assessed.
Results
Complete sets of answers to the questions (24 codings) were provided by 34 respondents, only three of whom reported prior use of the charts. Average score [n (%)] out of eight, as well as median score (range) for each coding category were: (i) Category: 7.3 (91 %); 7 (4–8); (ii) Time: 7.8 (98 %); 7 (6–8); (iii) Site: 7.2 (90 %); 7 (5–8). Overall, the equivalent calculations (out of 24) were 22.3 (93 %) and 22 (18–24). Mean prelecture confidence was 1.37 (out of 5), rising to 3.85 postquiz. Urogynecologists had the highest correlation with correct coding, followed closely by fellows and general gynecologists.
Conclusions
Optimizing training and study design can lead to excellent results for interobserver reliability of the IUGA-ICS Complication Classification coding, with increased participant confidence in complication-coding ability.