Do Specialized Units Improve Outcomes? Gordon D. Rubenfeld, MD MSc Professor of Medicine, University of Toronto Chief, Program in Trauma, Emergency, and Critical Care Sunnybrook Health Sciences Centre
One model MICU is the overflow no bedspacing in subspecialty ICUs
Another model Very rarely off service patient in Burn ICU managed by CCM Otherwise a single ICU team manages patients in all critical care beds but CCU
Many organizational factors might affect ICU outcome Training of intensivist Board certified, surgeon/internist/anesthesia/respirologist Dose of intensivist 24/7, time to ICU, responsibilities outside of ICU Training and dose of nurses, respiratory therapists, physiotherapy, pharmacy, nutrition Ratios Presence of multi-disciplinary rounds Non-physician providers Nurse practitioners, PA Team culture Tele-ICU
The real question is how to integrate subspecialty expertise into the ICU Potential value of subspecialty model Experience/expertise of subspecialty intensivists Experience/expertise of non-physician clinicians Relationships with primary specialty Control over beds
The Question How best to incorporate subspecialty expertise into critical care? The Problem Limited (but some!) data, strong opinions Money, control, and ego at stake The Data
ICU patients are getting older, sicker, and have multiple (not single) organ problems Crit Care Med 2010; 38:375 381
Increasing sepsis, renal failure, respiratory failure, and pneumonia
Reduced PAC and IABP Increased CVC and mechanical ventilation
The authors conclude With a growing body of evidence extolling the benefits of critical care-trained providers in traditional intensive care unit settings, and with influential advocate groups promoting such strategies, necessary qualifications for CCU clinicians must be addressed. Should these units be directed by specially trained cardiologists, by intensivists, or through a collaborative approach? Which strategy translates into better patient outcomes?
Subspecialty boarding is bad or Don t let a Neuro ICU look after your acute coronary syndrome or a Surgical ICU look after your ischemic stroke 5.00 * * * * General Ideal Boarding 3.75 2.50 1.25 0 ACS Stroke ICH Pneumonia Abdominal Surgery CT Surgery No mortality benefits of putting patients in to their subspecialty ICUs Potential harm of creating subspecialty ICUs and using them generally Am J Respir Crit Care Med Vol 179. pp 676 683, 2009
Conflicting data Project Impact admission to a NeuroICU associated with better outcomes after ICH Crit Care Med 2001; 29:635 640
Conflicting data Project Impact admission to a NeuroICU associated with better outcomes after ICH Note: study of 2 Neuro ICUs that differed substantially from other ICUs in cohort in many ways? selection bias for which ICH patients admitted to the NeuroICU ICU based, not population based, cohort
Effect of closed ICU in trauma care Does the training of the director matter? Nathens AB, et al.: The Impact of an Intensivist-Model ICU on Trauma-Related Mortality. Ann Surg 2006; 244(4): 545-554.
Effect of closed ICU in trauma care Does the training of the director matter? Nathens AB, et al.: The Impact of an Intensivist-Model ICU on Trauma-Related Mortality. Ann Surg 2006; 244(4): 545-554.
Intensivist staffing in patients with ALI reduces mortality but pulmonary consultation does not Treggiari MM, Am J Respir Crit Care Med. 2007;176(7):685-90.
Intensivist staffing in patients with ALI reduces mortality but pulmonary consultation does not Pulmonary consult in open ICUs Treggiari MM, Am J Respir Crit Care Med. 2007;176(7):685-90.
Are intensivists safe? Ann Intern Med. 2008;148:801-809. Studied the effect of having an intensivist as your doctor (who), not the effect of having an intensivist led ICU (how) Unique question in the literature
Majority of patients and ICUs in this study were cared for in non-intensivist led ICUs Effect of an intensivist on care of an individual patient Patients triaged in these hospitals to be cared for by an intensivist or not Potential for selection bias huge
Only 1 comparison in this study might reflect intensivist staffing Care in ICU with >95% CCM versus no care by CCM associated with 47% increase in odds of death Might be a comparison of intensivist staffing versus not But
What else is being compared Apples and Oranges in the Levy study = Nearly all intensivist managed 90.0 = No one intensivist managed 67.5 45.0 22.5 0 Urban Academic CC Fellowship
Novel models for incorporating subspecialty expertise in ICU Retired neurosurgeons as rounding neurosurgeon in shared model NICU Clinical experience and buy-in from primary team Physical presence on rounds Neurosurgical teaching for intensivists Neurosurgery 67:234-236, 2010
Conclusions ICU models and hospital deployment of subspecialty ICUs are very complex Difficult to tease apart variables in existing studies Intensivist staffing (closed, open, consultative) Intensivist training Geographic location of ICU Subspecialist role Consultative model and elective admission to intensivist model not effective and may be harmful
Conclusions Admission of critically ill patients to subspecialty ICUs does not confer benefit over general ICU possible exception of neuro ICUs Admission of critically ill patients to the wrong specialty ICU appears to be harmful Hospitals with a subspecialty ICU model need to consider the potential for benefit (outside of the neuro ICU) and for harm (bed flexibility)
Email gordon.rubenfeld@sunnybrook.ca for slides