In the hospital setting, traditional practice in the event of a cardiopulmonary arrest or “code” consists of alerting a team of health care professionals to administer lifesaving treatments.  Survival after a code has historically been poor (Brindley et al, 2002, Schein et al, 1990). As a result, there has been a world-wide initiative to address this problem.  Codes occurring on general care units are a particular problem for a number of reasons.  The general care nurse may be unable to recognize early patient deterioration due to a lack of experience or due to a challenging patient care assignment.  This nurse may also be unable to articulate the problems that he/she is witnessing.  Additionally, this nurse lacks critical-care expertise necessary to provide appropriate care.
Because of the low survival-to discharge rate post-arrest, researchers have focused on predictors of adverse events in an effort to intervene before a patient deteriorates to the point of arrest.  Multiple studies have cited warning signs predictive of cardiopulmonary arrest which may include: acute change in heart rate and rhythm, blood pressure, oxygenation, and mentation (Ashcraft, 2004, Bellomo et al, 2004, Clark, 2001, Considine & Botti, 2004, Franklin & Mathew, 1994, Peck, 2004, Peden-McAlpine & Clark, 2002, Schein et al, 1990, & Simchen, 2004).  Other studies discuss the issue of failure to recognize signs of clinical instability (Peden-McAlpine & Clark, 2002, McQuillan et al, 1998). The phenomenon of “failure to rescue” is often cited and described as a failure on the part of the nursing and/or medical staff to report clinical signs to the appropriate party, to adequately assess the unstable patient, and/or to appropriately treat the patient (Franklin & Mathew, 1994).
In response to these practice problems, there has been worldwide research into best practices to improve patient outcomes.  Australia and Great Britain have studied the effect of implementation of the Medical Emergency Team (MET) while in the United States (US), the Rapid Response Team (RRT) or the MET model has been researched.  Bellomo (2004) studied the difference in outcomes after the implementation of the MET for surgical patients in Australia and concluded that there was a statistically significant reduction in mortality, length of stay, and adverse events.  There was also an increase in awareness of consequences of physiologic instability.  Galhotra et al (2006) found that nurses surveyed from two hospitals in the US felt that the MET improved patient care.  These team models are collectively known as Rapid Response Systems (RRS) by the core international researchers who have pioneered the concept and process.  The report of the first International Conference on Medical Emergency Teams was published in 2006 and identified the characteristics of effective RRS.  Four components were identified:
·        An afferent process for event detection and response triggering,
·        An efferent process for deployment of the RRT/MET,
·        An administrative process which supports and organizes resources, and
·        An evaluation process for continuous quality improvement.
A literature search was performed using key words including failure to rescue, adverse events, medical emergency teams, rapid response teams, and cardiac arrest.  Evidence reviewed included studies addressing clinical antecedents to adverse events, the failure to rescue phenomenon, and the effect of the RRT/MET intervention.
A committee was formed to explore the development of a version of the RRT concept guided by the Institute for Healthcare Improvement’s (IHI) Getting Started Kit (IHI.org, 2005). A nurse-led team was selected as the model which best fit our organizational culture.  The team named the Early Nursing Intervention Team (ENIT) was defined as the team of the Medical Intensive Care Unit (MICU) charge nurse and the general care nurse caring for the unstable patient.  The ENIT procedure was planned and a research proposal was approved by the Institutional Review Board.  Once the ENIT concept was well-defined by the planning committee, key stakeholders were briefed including nursing and medical staff. 
The ENIT strategy was implemented after education of all involved staff and a transition period to become accustomed to the process.  Budgeted staffing was increased to free the charge nurse of a patient assignment around the clock.  Similar to the first component of the RRS listed above, our afferent process involves twice daily rounding on general care units.  ENIT is also available by pager.  The general care staff was given criteria for activating ENIT stressing that they could call if they were simply worried about a patient.  A method for organizing thoughts and efficiently reporting concerns was introduced during the education process.   The ENIT committee continues to meet to provide feedback and ensure a consistent, effective process.
A chart review of 100 patients transferring to MICU from general care units was completed retrospectively – before the ENIT intervention – and prospectively for the first 100 patients during the study phase.  Data analysis included codes outside MICU, time to transfer to MICU, survival to discharge, and length of stay.
Codes outside MICU have decreased to just 25 in the first year of the ENIT program compared to 45 per number discharged in the same time frame one year prior to implementation (p=0.017).  The time to transfer the unstable patient to MICU decreased from a mean time of 242.12 minutes before ENIT to 88.65 minutes during the study phase (p = 0.000). Mortality was 55.6 percent in the baseline phase compared to 44.4 percent in the study phase (n=81) but did not change significantly.  Length of stay did not change significantly from a mean of 17.22 days in the baseline period compared to 26.71 days in the study phase.
The key benefits have been the improvement in early transfer to MICU and the reduction in codes on general care units.  Twice daily rounding has led to side benefits of increased collegiality and recruitment of experienced general care nurses to our MICU. Also of note, the MICU staff vacancy rate has dropped dramatically from a high of 36 percent in early 2004 to 13 percent in early 2006.