Pilot study of nurse directed patient flow to a surgical intensive care unit on ED length of stay

Introduction: There are over 2 million admissions to intensive care units (ICU) from emergency departments (ED) in the United States every year. Prolonged waits for ICU admission from the ED have been associated with higher rates of complications and higher mortality, not to mention increased patient and provider dissatisfaction. The reasons behind these increased rates are multi-factorial. The added resources needed to care for critically ill patients in the ED have a compounding effect on the care of other patients within the department as well. Similarly, the causes of increasing ED length of stay (LOS) for critically ill patients has many roots and has been anatomized into input, throughput, and output factors. Improving any of the factors, potentially, can help decompress the system and improve patient flow out of the department. This has the downstream effect of decreasing patient ED LOS and possible adverse outcomes. A working group within our hospital identified the nursing transition of care from ED to ICU as an area of opportunity for potentially decreasing ED LOS for patients admitted to the ICUs. The team identified time delays in coordinating availability of ED and ICU nurses for hand-off and the need for ED nurses to accompany their patient to the ICU (making them unavailable for their other patients) as potential targets. We hypothesized that an ICU nurse driven protocol to pull patients from the ED would result in decreased ED LOS. Methods: As part of a Trauma Performance Improvement Process/Registry at our busy, academic, urban, Regional Trauma Center, the ED LOS data and injury severity score (ISS) is captured for patients admitted directly to the Trauma/Surgical ICU from the ED. A pilot program was launched whereby SICU nurses, on receipt of a bed request from the ED, would proactively present to the ED and receive face-to face report from the ED nurse. The SICU nurse would then accompany the patient from the ED to the SICU. Comparative analysis of ED LOS and ISS was then performed on a quarterly basis pre- (period 1) and post- (period 2) intervention. Results: Evaluation during period 1 revealed 104 patients with an average ISS of 20.8 admitted directly from ED to ICU with average ED LOS 7.2 hours (95% CI, 6.7-7.7). During period 2, 123 patients were admitted with an average ISS of 20.2. Average ED LOS was 6.4 hours (95% CI, 6.1-6.8). During the initial trial period, there was an 18% increase in admissions without significant change in acuity (based on ISS). Despite a trend in decreased ED LOS between Period 1 and Period 2, the change is not statistically significant (p value >0.1). Conclusions: This pilot study attempted to demonstrate a reduction in ED LOS by utilizing a proactive, ICU nurse driven, pull through of patients from the ED to the Surgical ICU. A trend towards reduction in ED LOS between the control and study period was seen despite significant increased patient volume. This reduction in ED LOS was not statistically significant. The potential reasons contributing to this include (a) a lack of power in the study, and/or (b) a lack of any or enough reduction in the ED LOS between the standard and intervention groups. There was no noticed increase in ED LOS associated with the higher volume of patients admitted to the Trauma/Surgical ICU during the study periods. As ED and hospital overcrowding increase due to the increasing acuity and volumes of patients, new techniques need to be developed to decrease the negative impact on patient care and outcomes.