More data is available now to healthcare providers than ever before. Real time laboratory and physiological vital signs are a key element in the ongoing assessment and evaluation of the hospitalized patient. Nursing assessment and judgment has a significant role in promoting health as well as preventing adverse events in the hospitalized patient. An adverse event is defined as “an unintentional injury or complication resulting in disability, death or prolonged hospital stay that is a result of healthcare management rather than the patient underlying disease” (Considine, 2004). Assessment includes the integration of multiple data sources into a cohesive picture of the patient. This skill can be an elusive skill and is often dependent upon the education and experience of the healthcare provider as well as hospital staffing patterns of the moment (Aiken et al., 2003; Stanton 2004,). An assessment that indicates patient deterioration is often documented, but subsequent action to prevent further decompensation does not happen.
Ashcraft, 2004 states that once the pre-arrest variables have been identified, the patient may live (rescue) or die (failure-to-rescue) depending on actions / inaction of the involved staff. Unintended inaction can lead to serious adverse events or a cardiopulmonary resuscitation event. It is not the dramatic changes such as myocardial infarction or respiratory arrest that are missed; it is the subtle changes that occur over time and are overlooked by front line clinicians. Theses changes present a less dramatic picture but have just as lethal an outcome if not recognized and treated with evidence-based clinical intervention.
The clinical question was can a structured approach based on a numerical rating system that implements evidence-based interventions preemptively prevent adverse events?
Failure-to-rescue is the inability to save a hospitalized patient's life after the development of a complication that occurs after the second day of hospitalization (Aiken et al., 2002; Clarke & Aiken, 2003). Failure to rescue is considered a nurse-sensitive indicator. Much of the center of hospital quality improvement focuses on this indicator with the intention of decreasing the episodes of failure-to-rescue. (Thompson et al. 2008). This indicator has a direct relationship to the quality of nursing care as well as the mortality rate in any healthcare facility (Schmid et al. 2007). Failure-to-rescue is considered an adverse event that can often lead to cardiac arrest. Hourihan et al. (1995) indicated that patients who progressed from failure to rescue to a cardiac arrest had an 87% mortality rate at hospital discharge when for other emergencies, such as sepsis or respiratory arrest, the rate was only 27%. Further analysis would lead to the conclusion that the ability to recognize and prevent this particular adverse event would influence mortality rates. Additionally, nursing must be prepared to initiate actions that can negate or prevent failure-to-rescue events.
Early recognition of complications and implementation of evidence-based management of that complication ultimately can improve the quality of care by rescuing these at risk patients (Friese & Aiken, 2008). There is a continuing need for ongoing investigation into the causes of failure to rescue events to develop empirically sound interventions. (Schmid, 2007).
Since 1995, the literature reports and outlines the development and implementation of Rapid Response Teams (RRT) as a way to identify and intervene earlier in patient deterioration. The Institute for Healthcare Improvement (IHI) identified such teams as a way to improve death in patients outside of critical care units. (Institute for Healthcare Improvement, 2005). Additionally, in 2007, the Joint Commission on Accreditation mandated implementation of RRT as a national patient safety goal, becoming a standard in 2010. (http://www.jointcommission.org/standards_information/standards.aspx).
Upon review of the Rapid Response Team (RRT) data at a suburban 208 bed acute care facility it was felt that opportunity existed to intervene earlier in patient deterioration, prior to summoning the RRT or calling a Code Blue (cardiopulmonary arrest event). The question of what would prompt the nursing staff to recognize and react in a more proactive fashion to prevent adverse events was asked. The follow up question was how can systems and process already in place be used to prompt nursing action?
In 2001, the Modified Early Warning Scoring tool (MEWS) and escalation pathway was implemented in two healthcare facilities in Great Britain as a way to improve recognition of adverse events (when a patient's condition is deteriorating). (Higgins, Maries-Tillot, Quinton,& Richmond 2008). Issues identified in the implementation of the process included: difficulty in using the tool, failure to adhere to the escalation pathway and poor compliance by the staff. Additionally, a number of staff were unable to accurately calculate the MEWS score with consistent accuracy. With additional training and monitoring, the system was implemented throughout the hospital trust in 2007 and became the standard of care for the system.
After completion of a review of the literature, the decision was made to implement an early warning scoring system with escalation pathway as a pilot study on a 29 bed PCU (progressive care unit) located within a 208 bed acute care hospital. The term “PAR” (patients at risk) was first coined by Goldhill (1995). It was felt that this term better described and was more easily understood than MEWS, the term used in the Higgins et al study.
The established modified early warning physiological parameters used by Higgins, Maries-Tillot, Quinton, & Richmond (2008) suited the selected patient population, with exceptions as noted (asterisk on chart). For the PCU patient population, the “0” score for respiratory rate per minute of 16–20 was closer to the norm rather than 09–14 respirations per minute as was used in the MEWS tool. Additionally, a stand alone score of “4” was added for any acute neurological change. This was not part of the original MEWS scoring system. To utilize the tool, and an aggregate score is calculated from established baseline physiological parameters consisting of blood pressure, heart rate respiratory rate, temperature, central nervous system assessment, urine output, and oxygen saturations. (table 1). A threshold number, in this case “4” then triggers the implementation of an escalation pathway (table 2)
Utilizing physiological data already being collected on a routine basis by the nursing staff, a clinical prediction rule (PAR score) rates the level of patient stability. Since the facility used an electronic health record, the system could automatically generate the PAR score for the bedside primary nurse and could automatically print a unit report every 6 hours for the charge nurse to review. A specific score (4) mandates an evaluation of the patient and implementation of an escalation of patient care pathway. (See table 1).
A 29 bed PCU was selected since this unit has the highest number of patient transfers from the Intensive Care Unit (ICU) and higher patient acuity within the facility This unit averaged 5–6 adverse / failure-to-rescue events (RRT or Code Blues) per month. The rationale for selection of this unit was that the higher the patient acuity the more at risk is the patient population. Additionally, these at risk patients have increased opportunity for undetected deterioration. All patients admitted to this unit were included in the study.
The proposed pilot study was presented to the UPC (unit practice council) of the PCU. Unit practice councils are composed of members of the department that review practice issues and relevant changes that affect the quality of care of that department. The UPC embraced the project and the nurses from the council were role models for implementation within the unit. Education was required of all staff, both licensed and unlicensed, and the education was done through on-line learning modules. In person education was continues on a one-to-one basis as needed for staff that did not understand the process.
The PAR score pilot project was implemented on August 24, 2009 in the PCU. The goal of the project was: “Zero Code Blues in PCU”. The PAR score was automatically generated for the primary nurse after each set of vital signs were completed. The PAR report was automatically printed for the charge nurse every six hours. Rounding by the clinical educators during implementation allowed for discussion regarding the project. For example, if the PAR score was above “4” and if an intervention was in place to treat the source of the score elevation/ patient deterioration, then it was felt that the process was working. Additionally, some specific patient populations were noted to always have elevated scores such as dialysis patients and as long as this was noted as the cause of the elevation (in other words, no urine output) then is was an acceptable score for that patient.
The PAR pilot study utilized a quasi-experimental design. The dependent variable was defined as the number of adverse/failure-to-rescue events per 1000 in-patient days. The independent variable was defined as the PAR assessment score/tool. All patients admitted to the PCU were included in the study. The PAR scoring tool was not implemented on any other unit within the facility. The only identified factor influencing the recognized number of adverse/failure-to-rescue events was the PAR scoring tool. IRB (internal review board) approval was obtained prior to initiation of the pilot study.
Evaluation took place during the 6 week pilot study. Data (RRT, Code Blue and mortality) from prior to implementation and data collected during implementation were used to conduct a series of binomial exact tests. These tests were computed on the difference between the mean number of events that occurred before and after the PAR scoring system was implemented. The results indicated that a statistically significant reduction in Code Blues, RRT and Mortality had occurred after implementation. In fact, during the pilot study there were no RRTs or Code Blues. The dependent variable, adverse event (RRT, Code Blues, and mortality) was calculated per 1000 patient days.
The mean number of adverse events per 1000 patient days significantly dropped in the year that the PAR scoring system has been in use for those patients who met the inclusion criteria.(figure 1). This data reflects that the intervention and its impact have been able to be sustained. Since there was considerable impact on the number of adverse events on PCU, the PAR scoring system was then expanded next to a 58 bed medical-oncology unit with similar results, (figure 2), again reflects that the impact can be sustained. Total facility PAR scoring was implementation by September, 2010.
The statistical analysis reveals utilization of a clinical prediction rule (PAR score) and the that use of a specific course of action (Escalation Pathway) can impact adverse/failure-to-rescue events in all types of patient care units. This process can often stop or slow patient deterioration is a marked advancement in the care of hospitalized patients.
This approach was successfully implemented on a 29 bed progressive care unit, initially, and then throughout the entire 208 bed facility. The electronic medical record that automatically generated the score successfully supported the implementation of the PAR Score. This factor successfully negated one of the downfalls identified by Higgins et al – that nurses have difficulty calculating an accurate score. Utilization of other resources and processes already in place, such as summoning the RRT as part of the escalation pathway, that did not increase the work load of the nurse allowed the nurse to work smarter, not harder. The clinical quality indicator of adverse events / failure-to-rescue should be monitored within healthcare facilities. Evaluation of practices associated with these events should occur. Processes that change the practice that surrounds these events should be evaluated, and if found relevant, should be used to improve patient outcomes.
Buist, M., Moore, G., Bernard, S., Waxman, B., Anderson, J., & Nguyen, T. (2002). Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: Preliminary study. British Journal of Medicine, 324,
Chan, P.S., Jain, R., Nallmothu, B.K., Berg, R.A., & Sasson, C. (2010). Rapid response
teams: A systematic review and meta-analysis. Archives of Internal Medicine. 170
Clark, S.P. (2004). Failure to rescue: Lessons from missed opportunities in care. Nursing inquiry, 11
Clarke, S.P. & Aiken, L.H. (2003). Failure to rescue. [Electronic version]. American journal of nursing. 103
Considine, J. Botti, M. (2004) Who, when where? Identification of patients at risk of an in-hospital adverse event: Implications for nursing practice. International journal of nursing practice, 10,
Friese, C., & Aiken, L. (2008). Failure to rescue in the surgical oncology population: Implications for nursing and quality improvement. Oncology Nursing Forum, 35
Goldhill, D.R., White, S.A., & Sumner, A. (1999). Physiological values and procedures in the 24h before ICU admission from the ward. [Electronic version]. Anesthesia. 54
Goldhill, D.R., Worthington, L., Mulcahy, A., Tarling, M. & Sumner, A. (1999). The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia, 54
Higgins, Y. et al (2008). Promoting patient safety using an early warning scoring system. Nursing Standard, 22
Hourihan, F., Bishop, G., Hillman, K.M., Duffurn, K. & Lee, A. (1995). The medical emergency team: a new strategy to indentify and intervene in high risk patients. Clinical Intensive Care 6
McArther-Rouse, F. (2001). Critical care outreach services and early warning scoring systems: A review of the literature. [Electronic version]. Journal of advance nursing, 36
Simpson, K.R. (2005), Failure to rescue: Implications for evaluating quality of care during labor and birth. Journal of perinatal-neonatal nursing, 19
Schmid, A., Hoffman, L., Happ, M.B., Wolf, G., & DeVita, H. (2007), Failure to rescue: A literature review. Journal of nursing administration, 37
March 2001. Stanton, M.W. (2004). Hospital nurse staffing and quality of care. Agency for healthcare research and quality.
Thompson, C., Dalgleish, L., Bucknall, T., Estabrooks, C., Hutchinson, A.M., Fraser, K, deVos, R., Binnekade, J.,& Saunders, J. (2008). The effects of time pressure and experience on nurses' risk assessment decisions: a signal detection. Nursing research, 57