Katharine E. Hartnack, Psy.D., is a 2018 graduate of the Psy.D. Program in Clinical Psychology at Antioch University, New England

Dissertation Committee:

  • Theodore Ellenhorn, PhD, Committee Chair
  • Vince Pignatiello, PsyD, Committee Member
  • Patrick Purdon, PhD, Committee Member


delirium, mortality, ICU, critically ill patients and delirium

Document Type


Publication Date



Delirium is a neurocognitive disorder defined as an acute disturbance in attention, awareness, and cognition with a fluctuating course not better explained by a preexisting condition (American Psychiatric Association, 2013). It is prevalent in up to 70% of hospital patients and 82% of patients in the intensive care unit (ICU; Ely, Speroff, Gordon, & Bernard, 2004; Kavanagh & Gottfried, 2007; Mcnicoll, Pisani, Ely, Gifford, & Inouye, 2005). The impact of delirium on mortality is inconsistent in the literature. Many studies have concluded that delirium prevalence is associated with increased risk of mortality (Cole, 2004; Kavanagh & Gottfried, 2007; Moskowitz et al., 2017; Pandharipande et al., 2013; Pauley et al., 2015), while others suggest delirium does not increase risk of patients dying (Levkoff et al., 1992; Wolters et al., 2014). The objective of this retrospective observational study was to determine if delirium is an independent predictor of mortality and develop a new model predicting three-month mortality of critically ill patients. Of the 165 patients followed in this study, 42 (25.5%) were deceased at three months and 123 (74.5%) survived. The most accurate model of predicting three-month mortality had an area under the curve of 0.89 (CI: 0.81 to 0.94), which included delirium burden defined as the fraction of the number of days patients were positive for individual features of delirium during their hospital stay. The main finding of the present study is the development of a new model that accurately predicts three-month mortality of critically ill patients. This study provides further evidence that delirium is an independent predictor of mortality and new evidence that delirium fraction improves the accuracy of a predictive models of mortality. We also identified individual features of delirium that are more predictive of mortality than others. Future research is needed to develop prevention measures and treatment interventions for delirium in the ICU and on hospital floors to reduce risk of patient mortality.


Katharine E. Hartnack

ORCID Scholar ID# 0000-0002-4318-8459