Model for predicting morbidity and mortality and need for ICU postoperative surgical patients with cancer

Abstract

Several scores for predicting morbidity and mortality in patients undergoing surgery, who will possibly use ICU beds, are being tested. Currently, the indication of intensive postoperative support for elective surgeries is established in a subjective way and may even endanger the performance of the hospital since it could spawn the suspension of operations due to lack of vacancies in the ICU. The most used score was proposed by the American Society of Anesthesiologists, ASA, as a risk predictor, however this classification is frequently submitted to different interpretations. It is also considered the use of POSSUM, which takes into account twelve physiological variables and six surgical variables. We could mention the Karnofsky scale as well, which classify patients according to the degree of functional disability caused by a neoplastic disease. The objective of this study was to correlate these elements in order to find a better predictor for ICU. Secondly, we intended to verify whether POSSUM physiological variables could be used as predictors to ICU reservation in surgeries. We reviewed 186 patients medical records undergoing oncologic surgeries. As a result, we can infer that the physiological variables that compose the POSSUM are more accurate as predictors than all the other analysed elements that could also predict bed reservation in ICU. Moreover, we may notice that the POSSUM variables age, serum hemoglobin and systolic blood pressure would be sufficient to predict properly the vacancy allocation in the ICU.