Abstract
Background
Lamb et al. developed the metric for the observation of decision-making tool (MTB-MODe) to evaluate the quality of urologic multidisciplinary cancer conferences (MCCs) in the United Kingdom. We used generalizability theory to assess the reliability of a modified version of MTB-MODe in a North American context. Specifically, we wished to determine if the tool could distinguish between high- and low-quality MCC decision-making.
Methods
Two assessors independently evaluated two MCCs (MCC1, MCC2) using the modified MTB-MODe. Generalizability theory was used to assess overall tool reliability and to identify sources most likely to contribute to variance in reliability scores. A total of 60 cases were evaluated.
Results
The overall reliability scores of MCC1 and MCC2 were 0.72 and 0.74, respectively. Inter-rater reliability scores were reasonable (>0.55) and raters did not contribute significantly to variance in reliability scores. Internal consistency of the individual MTB-MODe items was low, demonstrating that items were not highly correlated.
Conclusions
The MTB-MODe reliably assessed the quality of individual MCC cases. Raters did not contribute significantly to reliability scores, suggesting that the tool can be successfully implemented using a single rater. Low internal consistency of the MTB-MODe items demonstrates that the tool can be used to provide feedback on individual tool items. Such data can be used by stakeholders to help improve MCC quality.
Keywords
Get full access to this article
View all access options for this article.