Clinical Predictors for Unsafe Direct Discharge Home Patients From Intensive Care Units Academic Article uri icon

  •  
  • Overview
  •  
  • Research
  •  
  • Identity
  •  
  • Additional Document Info
  •  
  • View All
  •  

abstract

  • Purpose: To describe factors (demographics and clinical characteristics) that predict patients who are at an increased risk of adverse events or unplanned return visits to a health-care facility following discharge direct to home (DDH) from intensive care units (ICUs). Methods: Prospective cohort study of all adult patients who survived their stay in our medical–surgical–trauma ICU between February 2016 and 2017 and were discharged directly home. Patients were followed for 8 weeks postdischarge. Univariable and multivariable logistic regression analyses were performed to identify factors associated with adverse events or unplanned return visits to a health-care facility following DDH from ICU. Results: A total of 129 DDH patients were enrolled and completed the 8-week follow-up. We identified 39 unplanned return visits (URVs). There was 0% mortality at 8 weeks postdischarge. Eight potential predictors of hospital URVs ( P < .2) were identified in the univariable analysis: prior substance abuse (odds ratio [OR] of URV of 2.50 [95% confidence interval: 1.08-5.80], hepatitis (OR: 6.92 [1.68-28.48]), sepsis (OR: 11.03 [1.19-102.29]), admission nine equivalents of nursing manpower score (NEMS) <24 (OR: 2.28 [1.03-5.04], no fixed address (OR: 22.9 [1.2-437.3]), ICU length of stay (LOS) <2 days (OR: 2.95 [1.28-6.78]), home discharge within London, Ontario (OR: 2.44 [1.00-5.92]), and left against medical advice (AMA; OR: 6.06 [2.04-17.98]). Conclusions: Our study identified 8 covariates that were potential predictors of URV: prior substance abuse, hepatitis, sepsis, admission NEMS <24, no fixed address, ICU LOS <2 days, home discharge within London, Ontario, and left AMA. The practice of direct discharges home from the ICU would benefit from adequately powered multicenter study in order to construct a clinical prediction model (that would require further testing and validation).

authors

  • Lau, Vincent Issac
  • Priestap, Fran
  • Lam, Joyce
  • Basmaji, John
  • Ball, Ian M

publication date

  • October 2020

has subject area