Factors associated with physicians' predictions of long-term mortality of critically ill patients
CCCF ePoster library. Ferreyro B. Oct 2, 2017; 198224; 42
Dr. Bruno Ferreyro
Dr. Bruno Ferreyro
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Abstract
Rate & Comment (0)
#42



Factors associated with physicians’ predictions of long-term mortality of critically ill patients

Ferreyro, Bruno L1.; Detsky, Michael E.1 



1. Interdepartmental Division of Critical Care Medicine, Department of Medicine, Mount Sinai Hospital/University Health Network, University of Toronto, Ontario, Canada.


Introduction
Physician’s estimate of the prognosis of patients’ is essential to shared decision making in the intensive care unit (ICU). The variables influencing these judgments are not well understood.

Objectives
The aim of this study is to determine which individual physician and patient characteristics influence physician´s predictions of patient’s mortality at 6 months. Additionally, we explored which of these factors are associated with the confidence of the predictions.

Methods
Secondary analysis of a prospective cohort study that evaluated how well ICU physicians predicted patients’ long-term outcomes. Recruitment occurred from October 2013 to May 2014 in 5 ICUs in the University of Pennsylvania. Patients were included if they required mechanical ventilation > 48 hours, vasopressors >24 hours, or both. ICU attending physicians were asked to predict whether patients were expected to live or die in 6-months, and to report their confidence in these predictions. Using univariate and multivariable logistic regression, we assessed the association between baseline physician and patient characteristics with study outcomes: 6-month mortality prediction (yes/no) and confidence of the prediction (yes/no). We graphically explored the linear association between continuous predictors and outcomes. We calculated c statistics for each logistic regression model and goodness of fit was measured using Hosmer Lemeshow.

Results
Of 340 eligible patients, 303 were enrolled. Forty-seven physicians contributed at least one prediction per patient. Patient’s median age was 62 years old (IQR 53-71) and 57% (N=157) were male. Physicians median age was 41 (IQR 38-53) and 80% (N=37) were male. Mortality at 6 months was predicted for 33% (N=99), of which 43% (N=130) died. The final multivariable model (Table 2) explaining physicians’ predictions of mortality included patient’s age (OR 1.03; 95%CI 1.01-1.05), the presence of malignancy (OR 2.70; 1.54-4.74); medical (vs. surgical patients) (OR 5.87; 95%CI 2.70-12.7) and patient’s APACHE score (OR 1.02; 95%CI 1.01-1.02). This model deemed good discrimination (C statistic=0.77; 95% CI 0.72-0.83) and calibration (difference between expected and observed p value=0.87). The same variables were also associated with patient’s actual mortality at 6 months (Table 3). Physician’s age (OR 0.93; 95% CI 0.90-0.96), physician’s male gender (OR 0.53; 95% CI 0.32-0.88) and years since graduation (OR 0.94; 95% CI 0.91-0.97) were associated lower confidence of the prediction (Table 3). A higher APACHE score was also associated with lower confidence (OR 0.99; 95% CI 0.98-0.99).

Conclusion
Physician and patient factors are associated with predictions of mortality and patient’s actual mortality at 6 months. These include patient age, presence of malignancy, being a medical patient and severity of illness. Physician age, being male, years of experience and severity of illness were inversely associated with confidence in these predictions. This information should be considered when physicians reflect on how they make predictions for critically ill patients.

    This eLearning portal is powered by:
    This eLearning portal is powered by MULTIEPORTAL
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.


Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.


Save Settings