Zimmerman JE, Kramer AA, McNair DS, et al. Acute Physiology and Chronic Health Evaluation (APACHE) IV: Hospital mortality assessment for today’s critically ill patients

George Washington University, Washington, Washington, D.C., United States
Critical Care Medicine (Impact Factor: 6.15). 06/2006; 34(5):1297-310. DOI: 10.1097/01.CCM.0000215112.84523.F0
Source: PubMed

ABSTRACT To improve the accuracy of the Acute Physiology and Chronic Health Evaluation (APACHE) method for predicting hospital mortality among critically ill adults and to evaluate changes in the accuracy of earlier APACHE models.
: Observational cohort study.
A total of 104 intensive care units (ICUs) in 45 U.S. hospitals.
A total of 131,618 consecutive ICU admissions during 2002 and 2003, of which 110,558 met inclusion criteria and had complete data.
We developed APACHE IV using ICU day 1 information and a multivariate logistic regression procedure to estimate the probability of hospital death for randomly selected patients who comprised 60% of the database. Predictor variables were similar to those in APACHE III, but new variables were added and different statistical modeling used. We assessed the accuracy of APACHE IV predictions by comparing observed and predicted hospital mortality for the excluded patients (validation set). We tested discrimination and used multiple tests of calibration in aggregate and for patient subgroups. APACHE IV had good discrimination (area under the receiver operating characteristic curve = 0.88) and calibration (Hosmer-Lemeshow C statistic = 16.9, p = .08). For 90% of 116 ICU admission diagnoses, the ratio of observed to predicted mortality was not significantly different from 1.0. We also used the validation data set to compare the accuracy of APACHE IV predictions to those using APACHE III versions developed 7 and 14 yrs previously. There was little change in discrimination, but aggregate mortality was systematically overestimated as model age increased. When examined across disease, predictive accuracy was maintained for some diagnoses but for others seemed to reflect changes in practice or therapy.
APACHE IV predictions of hospital mortality have good discrimination and calibration and should be useful for benchmarking performance in U.S. ICUs. The accuracy of predictive models is dynamic and should be periodically retested. When accuracy deteriorates they should be revised and updated.

  • Source
    • "ODERN clinical data acquisition systems are capable of 38 continuously monitoring and storing measurements of 39 patient vital signs, such as heart rate (HR) and blood pressure 40 (BP), over multiple days of hospitalization [1]. Despite this 41 continuous feed of data, commonly used acuity scores, such as 42 APACHE and SAPS [2]–[5], are based on snap-shot values of 43 these vital signs, typically the worst values during a 24-h period. 44 However, physiologic systems generate complex dynamics in 45 their output signals that reflect the state of the underlying control 46 systems [6]–[8]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Cardiovascular variables such as heart rate (HR) and blood pressure (BP) are regulated by an underlying control system, and therefore the time series of these vital signs exhibit rich dynamical patterns of interaction in response to external perturbations (e.g., drug administration) as well as pathological states (e.g., onset of sepsis and hypotension). A question of interest is whether "similar" dynamical patterns can be identified across a heterogeneous patient cohort, and be used for prognosis of patients' health and progress. In this work, we used a switching vector autoregressive (SVAR) framework to systematically learn and identify a collection of vital sign time series dynamics, which are possibly recurrent within the same patient and may be shared across the entire cohort. We show that these dynamical behaviors can be used to characterize the physiological "state" of a patient. We validate our technique using simulated time series of the cardiovascular system, and human recordings of HR and BP time series from an orthostatic stress study with known postural states. Using the HR and BP dynamics of an intensive care unit (ICU) cohort of over 450 patients from the MIMIC II database, we demonstrate that the discovered cardiovascular dynamics are significantly associated with hospital mortality (dynamic modes 3 and 9, p = 0:001, p = 0:006 from logistic regression after adjusting for the APACHE scores). Combining the dynamics of BP time series and SAPS-I or APACHE-III provided a more accurate assessment of patient survival/mortality in the hospital than using SAPS-I and APACHE-III alone (p = 0:005 and p = 0:045). Our results suggest that the discovered dynamics of vital sign time series may contain additional prognostic value beyond that of the baseline acuity measures, and can potentially be used as an independent predictor of outcomes in the ICU.
    IEEE Journal of Biomedical and Health Informatics 06/2014; 19(3). DOI:10.1109/JBHI.2014.2330827 · 1.98 Impact Factor
  • Source
    • "pulse rate, mean blood pressure, temperature, respiratory rate, etc.) retrieved from the first 24 h vital signs and the age variable are two of the main items that have to be collected for predicting hospital mortality among patients admitted to ICUs [35]. In [35], Zimmerman et al. analyzed the unique relative contribution of each risk factor in APACHE to hospital mortality prediction. They found that APS is the major contribution (65.6%) and age provided the 3rd priority (9.4%) in hospital mortality prediction. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a computational-intelligence-based model to predict the survival rate of critically ill patients who were admitted to an intensive care unit (ICU). The prediction input variables were based on the first 24h admission physiological data of ICU patients to forecast whether the final outcome was survival or not. The prediction model was based on a particle swarm optimization (PSO)-based Fuzzy Hyper-Rectangular Composite Neural Network (PFHRCNN) that integrates three computational intelligence tools including hyper-rectangular composite neural networks, fuzzy systems and PSO. It could help doctors to make appropriate treatment decisions without excessive laboratory tests. The performance of the proposed prediction model was evaluated on the data set collected from 300 ICU patients in the Cathy General Hospital in 2012. There were 10 input variables in total for the prediction model. Nine of these variables (e.g. systolic arterial blood pressures, systolic non-invasive blood pressures, respiratory rate, heart rate, and body temperature) were routinely available for 24h in ICU and the last variable is patient's age. The proposed model could achieve a 96% and 86% accuracy rate for the training data and testing data, respectively.
    Computers in Biology and Medicine 01/2014; 47C:13-19. DOI:10.1016/j.compbiomed.2013.12.012 · 1.90 Impact Factor
  • Source
    • "K.S. Lee et al. / Journal of Critical Care 29 (2014) 185.e9–185.e12 systems retain age as a significant component [14] [16]. From our regression analysis, the age factor showed less impact on prognosis than other factors. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The Sequential Organ Failure Assessment (SOFA) score, originally developed to assess organ failure status, is widely used as a prognostic indicator in intensive care unit patients. Additional prognostic factors, such as age and comorbidities, may complement the predictive performance of the SOFA. In total, 1049 consecutive patients were enrolled prospectively. SOFA and other admission-based intensive care unit scores were recorded during the first 24 hours. A complemented SOFA (cSOFA) score model was constructed by adding age and comorbidity scores to the original SOFA score, based on logistic regression analysis. The predictive performance was evaluated with regard to hospital mortality by receiver operating characteristics analysis. The Hosmer-Lemeshow goodness-of-fit test was used to assess calibration of the model, and leave-one-out cross-validation was performed. The cSOFA score (maximum 30 points) was calculated as the SOFA score (24 points) + age score (2 points) + comorbidity score (4 points). The cSOFA score model showed satisfactory calibration and cross-validation performance. The AUC (95% CI) of the cSOFA score (0.812 [0.787-0.835]) was higher than the SOFA score (0.743 [0.715-0.769], P < .0001). The performance of the SOFA score to predict hospital mortality can be improved by considering age and comorbidity factors.
    Journal of critical care 10/2013; DOI:10.1016/j.jcrc.2013.10.006 · 2.19 Impact Factor
Show more