The number of intensive care unit (ICU) beds in the United Stateshascontinuedto increaseover the last3decades, ashave ICU utilization rates and costs,1,2 and this despite the lack of any federal, regional, or critical care society mandates to justify these increases. Some experts believe that the increase in the number of ICU beds has led to inappropriate use of these beds by patients who are either too healthy or too sick to benefit from intensive care.3,4 This may in part explain the stable national ICU occupancy rate of approximately68%between 1985and2010 and suggests that ICU utilization has simply risen tomeet the increased number of beds.1,2 Furthermore, it hasbeenover adecadeandahalf sincenational societyguidelines for ICUand intermediate careunit admission, discharge, and triagewere published.5,6 In the intervening years, changes have occurred in the demographic and other characteristics of ICUpatientpopulations; the ICUworkforce has evolved; the legislative landscape has altered; and newethical and end-of-life factors have been introduced that affect the appropriate utilization of ICU beds. Ideally, US hospitals and ICUs would follow some consistent and clinically appropriate approaches to the triage of sick patients. At a minimum, we would expect that triage would be similar for day andnight,weekdayandweekendwithin the same hospital or hospital network, andwewould expect it to be independent of ICU and hospital occupancy rates. The reality, however, is far more complex and confusing: ICU triage decisions are based on many factors, some quantifiable and others intangible. These factors can be classified into 2 categories:hospital (institutional)basedandICUrelated.Hospitalbased factors include size (small,medium, or large), teaching or nonteaching status, nurse to patient ratios, rules and regulations of the limits ofward-based care, attitudes toward risk, practice styles of physicians andnurses, availability of nurses andsupport staff (eg,hospitalists, advancepractitioners, rapid response teams), privileging of clinicians for ICU admission, and the presence of step-down units (ie, intermediate or progressive). The ICU-based parameters include ICU to hospital bed ratios, model of care (ie, open, closed, or collaborative), types (ie, multiple-specialty ICUs or single, large ICU), intensivist staffing (ie, high or low intensity, full time or part time), presence of resident trainees and/or advance care practitioners, and coverage (ie, in-house or telemedicine). In this issueofJAMAInternalMedicine,ChangandShapiro7 retrospectively analyze ICU utilization for 4 medical conditions (diabetic ketoacidosis, pulmonary embolism,upper gastrointestinalbleeding,andcongestiveheart failure) in94acutecare nonfederal hospitals in Washington state and Maryland between2010and2012. These common illnessesmaybe classified as “in-between” conditions if they are not presenting at extreme levels of severity. The authors found great variability in the ICU utilization between high and low ICU utilizers. Higher ICUutilizationoccurred in smallerhospitals and teaching hospitals. Of note, the vast majority of hospitals had concordant ICU utilization (high or low) for all 4 conditions, suggesting consistent ICU vs ward triage patterns. The ICU and hospital occupancy did not affect ICU utilization for each of these conditions. Similarly, risk-adjustedhospitalmortality (at 30 days) did not differ between the high and low ICU utilizers;however, onceapatientwasadmitted to the ICU, therewas an increase in the number of invasive procedures (eg, central venous catheters) and higher hospital costs compared with patients cared for on the wards. Although this study showed greater use of standardized care pathways (represented by invasive procedures and increased costs) in the higher ICUutilization centers than in the lower, outcomes were similar. Many factors influence hospiAuthor Audio Interview at jamainternalmedicine.com
Read full abstract