Abstract
This study examines the dependability of naive nursing assessments of cough effort intensity (CEI) in unstable, but realistic, clinical situations. Sixty-two medical and surgical patients were recruited while on mechanical ventilation and studied on 3 consecutive days. Each day, three nurses measured two CEI ratings 3 minutes apart. Using a generalizability theory approach, the researchers estimated multiple sources of error in the measurement of CEI. Patient and day of measurement represented systematic variance, planned to increase variability in CEI. Time of assessment and nurse rater represented random sources of error. The design accounted for all sources of measurement error in the model. Variability in CEI was accounted for by individual patient differences and raters nested within patient day. The dependability coefficient was .39. Thus, the investigators concluded that naive nurse assessments of CEI are undependable in clinical situations.