Simplified Therapeutic Intervention Scoring System

Abstract

To validate a simplified version of the Therapeutic Intervention Scoring System, the TISS-28, and to determine the association of TISS-28 with the time spent on scored and nonscored nursing activities. Prospective, multicenter study. Twenty-two adult medical, surgical, and general Dutch intensive care units (ICUs). A total of 903 patients consecutively admitted to the ICUs. TISS-28 was constructed from a random sample of 10,000 records of TISS-76 items. The respective weights were calculated using multivariable regression analysis through the origin; TISS-76 scores were used as predicted values. Cross validation was performed in another random sample of 10,000 records and the scores of TISS-76 were compared with those scores obtained with TISS-28 (r equals .96, r sup 2 equals .93). The analysis of validation included 1,820 valid pairs of TISS-76 and TISS-28 records. The mean value of TISS-28 (28.8 plus minus 11.1) was higher (p less than .00) than that value of TISS-76 (24.2 plus minus 10.2). TISS-28 explained 86% of the variation in TISS-76 (r equals .93, r2 equals .86). The simplified TISS-28 explains 86% of the variation in TISS-76 and can therefore replace the original version in the clinical practice in the ICU.