Korean Journal of Nephrology 1996;15(2):156-163.

중환자실 급성신부전 환자에서 APACHE III System의 예후적 유용성

장재원 , 김경조 , 이무송 , 임채만 , 김형호 , 이종수 , 김순배 , 박정식

Abstract

Despite advances in medical knowledge and technology, the mortality of patient with acute renal failure(ARF) is 50%, and reach up to 80% in patients in ICU. But there is no proper method to predict prognosis of them. On the day of admission, APACHE III score can predict prognosis of critically ill patients and daily update of APACHE III score confirm of uncertainty of final outcome. Therefore, this study was performed to investigate prognostic utility of APACHE III system in patient with ARF. Consecutive 146 patients were admitted to ICU for renal replacement therapy(RRT) under the diagnosis of ARF. They were estimated in the following items', age, sex, type and duration of RRT(renal replacement therapy), duration of mechanical venti- lator, the number of organ failure, history of sepsis, cardiopulmonary arrest, and upper gastrointestinal bleeding, presence of HBs antigen, anti-HBs anti- body, and underlying diseases. APACHE III total score was measured on the day of admission, at the initiation of RRT, 3rd and 7th day post-RRT. Among 146 patients with ARF(M, 104; F 42), 89 patients died. Overall motality of this group was 61%. Mean age of survivors was 45 (±17yrs S.D.) as compared with that of the nonsurvivors 52( ±18yrs S.D.). At the initiation of RRT, all of those with a APACHE III total score below 50 survived, whereas all with a total score over 109 died. By univariatae analysis on 100 cases with APACHE III total score between 50 and 109 at the initiation of RRT, there were statistically significant difference between survivors and nonsurvivors in the presence of sepsis, cardiopulmonary arrest, underlying disease, the numbers of failured organ, APACHE III score at the initiation of RRT(SCO-O), APACHE III score on the 3rd day post-RRT(SCO-3), and the change between the two scores. By multivariate logistic regression analysis, the following were independently significant factors for the outcome: SCO-O, FD SCO30(=[SCO-3 SCO-0] x 100 / SCO-O, p<0.001) the presence of sepsis and severity of underlying disease in APACHE III system. Applying these 4 independent variables, we formulated an equation which could predict the probability of death Probability of death = exp(Z) / [1 + exp(Z)], where Z = -16.5 + 0.1 x SCO-0 + 1.12 x FaSCO30 + 2.08 x sepsis + 1.78 X underlying disease [underlying disease .' 1; chronic health score =0; 2, 1 chronic health score 14; 3, chronic health score > =15, Sepsis :' 0, none : 1, yes]. Cut-off point of 0.6 obtained by this equation showed sensitivity 7296 and specificity 96%. APACHE III prognostic system could be used to predict outcome of patients with ARF requiring RRT.