Validation of the simplified therapeutic intervention scoring system on an independent database

Objective: To evaluate the performance of the Simplified Therapeutic Intervention Scoring System on an independent database and determine its relation with the Therapeutic Intervention Scoring System in the quantification of nursing workload in intensive care. Design: Analysis of the database of a multicenter prospective Portuguese study. Setting: 19 intensive care units (ICUs) in Portugal. Patients: Data on 1094 patients consecutively admitted to the ICUs were collected during a period of 3 months. Methods: Collection of the data necessary for the calculation of the Therapeutic Intervention Scoring System (TISS-76) and the Simplified Therapeutic Intervention Scoring System (TISS-28) during the first 24 h in the ICU. Basic demographic statistics and all the variables necessary for the computation of the Simplified Acute Physiology Score II were also collected. Vital status at discharge from the hospital was registered. Regression techniques, Pearson's correlation and paired sample t-test were used. Results are presented as mean ± standard deviation except when stated otherwise. Reliability was evaluated by the use of intraclass correlation coefficients in a 5 % random sample. Measurements and results: After exclusion of all the patients with missing data, 1080 patients were analysed. The overall mean TISS-28 (29.82 ± 10.64) was significantly lower than the mean TISS-76 (31.14 ± 11.95). Both systems showed very significant differences between ICUs (p < 0.001). The correlation between the two was good, with TISS-28 explaining 72 % of the variation of TISS-76 (r = 0.85, r ² = 0.72). The relation between the two systems was TISS-28 = 6.22 + 0.85 TISS-76. In this cohort, reliability of data collection was very high, with intraclass correlation coefficients greater than 0.90 for both systems. Conclusions: TISS-28 was validated on this independent population. The results indicate that TISS-28 can replace TISS-76 for the measurement of the nursing workload in Portuguese ICUs Received: 29 October 1996 Accepted: 27 February 1997