Kappa statistic to measure agreement beyond chance in free-response assessments

The usual kappa statistic requires that all observations be enumerated. However, in free-response assessments, only positive (or abnormal) findings are notified, but negative (or normal) findings are not. This situation occurs frequently in imaging or other diagnostic studies. We propose here a kappa statistic that is suitable for free-response assessments.

We derived the equivalent of Cohen's kappa statistic for two raters under the assumption that the number of possible findings for any given patient is very large, as well as a formula for sampling variance that is applicable to independent observations (for clustered observations, a bootstrap procedure is proposed). The proposed statistic was applied to a real-life dataset, and compared with the common practice of collapsing observations within a finite number of regions of interest.

The free-response kappa is computed from the total numbers of discordant (b and c) and concordant positive (d) observations made in all patients, as 2d/(b + c + 2d). In 84 full-body magnetic resonance imaging procedures in children that were evaluated by 2 independent raters, the free-response kappa statistic was 0.820. Aggregation of results within regions of interest resulted in overestimation of agreement beyond chance.

The free-response kappa provides an estimate of agreement beyond chance in situations where only positive findings are reported by raters.