Automated computerized intensive care unit severity of illness measure in the Department of Veterans Affairs: Preliminary results

Objective: To evaluate the feasibility of an automated intensive care unit (ICU) risk adjustment tool (acronym: SISVistA) developed by selecting a sub set of predictor variables from the Acute Physiology and Chronic Health Eva luation (APACHE) III available in the existing computerized database of the Department of Veterans Affairs (VA) healthcare system and modifying the AP ACHE diagnostic and comorbidity approach. Design: Retrospective cohort study. Setting: Six ICUs in three Ohio Veterans Affairs hospitals. Patient Selection: The first ICU admission of all patients from February 19 96 through July 1997. Outcome Measure: Mortality at hospital discharge. Methods: The predictor variables, including age, comorbidity, diagnosis, ad mission source (direct or transfer), and laboratory results (from the +/- 2 4-hr period surrounding admission), were extracted from computerized VA dat abases, and APACHE III weights were applied using customized software. The weights of ail laboratory variables were added and treated as a single vari able in the model. A logistic regression model was fitted to predict the ou tcome and the model was validated using a bootstrapping technique (1,000 re petitions). Main Results: The analysis included all 4,651 eligible cases (442 deaths). The cohort was predominantly male (97.5%) and elderly (63.6 +/- 12.0 yrs). In multivariate analysis, significant predictors of hospital mortality incl uded age (odds ratio [OR], 1.06; 95% confidence interval [GI], 1.04-1.09), comorbidity (OR, 1.11; 95% CI, 1.08-1.15), total laboratory score (OR, 1.07 ; 95% GI, 1.06-1.08), direct ICU admission (OR, 0.39; 95% CI, 0.31-0.49), a nd several broad ICU diagnostic categories. The SISVistA model had excellen t discrimination and calibration (C statistic = 0.86, goodness-of-fit stati stics; P > .20). The area under the receiver operating characteristic curve of the validated model was 0.86. Conclusions: Using common data elements often found in hospital computer sy stems, SISVistA predicts hospital mortality among patients in Ohio VA ICUs. This preliminary study supports the development of an automated ICU risk p rediction system on a more diverse population.