Automated Clinical Inference and Rapid Response Teams Improve Patient Safety


November / December 2006

Feature Article

Automated Clinical Inference and Rapid Response Teams Improve Patient Safety

In 1999, the Institute of Medicine (IOM) released To Err Is Human: Building a Safer Health System. The report alleged between 44,000 and 98,000 people die unnecessarily in hospitals in the United States each year due to medical errors. To people who work in the field of patient safety, the conclusions were not new — they were based on decade-old data from the Harvard Medical Practice Study (Brennan, et al., 2004) — but the report struck a chord in the media and public.

To Err Is Human helped energize subsequent investigation, regulation, legislation, and other initiatives intended to reduce error and improve patient outcomes (AHRQ, 2005; Berwick et al., 2006; Bates & Gawande, 2000). Although the report’s 5-year goal of reducing medical errors by 50% is not believed to have been met, it succeeded in significantly raising the national profile of the safety of health systems. It furthered the now widely held view that the root of most medical error is not individual incompetence or carelessness, but the complexity of the medical system itself as it has developed over time (Leape & Berwick, 2005). A corollary is that initiatives to improve patient safety will be more effective if they focus on building safety into the system rather than blaming individuals for mistakes.

New technology has been integral to several proposed systemic changes. The potential impact of these systems has been discussed elsewhere (Bates & Gawande, 2003). This article focuses on the potential of automated clinical inference in patient monitoring systems to improve rapid response on the medical/surgical wards of acute care hospitals.

Patient Safety as a National Priority
A broad array of stakeholders have responded to To Err Is Human in the 6 years since its publication, but measurable progress has been modest (Leape & Berwick, 2005). No nationwide system tracks every facet of patient safety, so we must piece together the larger picture from individual studies and reports, and the composite shows measured gains at best. The Agency for Healthcare Research and Quality (AHRQ) publishes annually the National Healthcare Quality Report, which includes some patient safety indicators. According to the AHRQ’s 2005 report, the “overall quality” of U.S. healthcare improved 2.8% between 2002 and 2003. Specifically in terms of patient safety, the AHRQ reports rates of some preventable adverse events declined during the period (e.g., post-operative pneumonia, hospital-acquired bloodstream infections in ICU patients), but others continued to rise (e.g., ventilator-associated pneumonia, post-operative sepsis, infections due to medical care). If approximately 90,000 people die each year due to hospital-acquired infections (CDC, 2000), and if an estimated 240,000 patients died in hospitals between 2001 and 2003 due to safety incidents (HealthGrades, 2005), improvement to date has been at the margins.

Patient Safety Initiatives
The modest pace of measurable improvement does not mean little has been accomplished. The IOM report and its subsequent public attention have helped energize several public and private initiatives to advance patient safety.

The United States federal government continues to take a leading role. In 2001 Congress began appropriating funds earmarked for patient safety research. The AHRQ established a Center for Quality Improvement and Safety to facilitate the development of quality and safety standards, and the Veteran’s Health Administration (VHA) has established several patient-safety research programs (Leape & Berwick, 2005).

Non-governmental organizations have also played key roles. The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) has adopted an aggressive set of patient safety goals and is requiring hospitals to implement safe practices. The National Patient Safety Foundation has worked to raise awareness, identify best practices, and disseminate information to improve patient safety, and the Accreditation Council on Graduate Medical Education has promulgated residency training work-hour limitations intended to reduce fatigue and resulting error (Bates & Gawande, 2003).

The Institute for Healthcare Improvement’s 100,000 Lives Campaign is arguably one of the most ambitious private patient safety efforts. The campaign is motivated by the idea that a small number of proven interventions, implemented on a wide scale, could prevent 100,000 deaths over the first 18 months of the project, and every year thereafter. In the first year, more than 3,000 hospitals signed on to the campaign, agreeing to make the following six changes:


  • Deliver Reliable, Evidence-Based Care for Acute Myocardial Infarction (AMI) to prevent deaths from heart attack.
  • Prevent Adverse Drug Events (ADEs) by implementing medication reconciliation.
  • Prevent Central Line Infections by implementing a series of interdependent, scientifically grounded steps.
  • Prevent Surgical Site Infections by reliably delivering the correct perioperative care.
  • Prevent Ventilator-Associated Pneumonia by implementing a series of interdependent, scientifically grounded steps.
  • Deploy Rapid Response Teams at the first sign of patient decline.


In short, the IHI changes were designed to optimize care for AMI patients, reduce the number of hospital-acquired infections, and avoid unnecessary morbidity and mortality due to preventable cardiopulmonary arrest (Berwick, et al., 2006).

Rapid Response to Early Warning Signs
The Rapid Response Team (RRT, also known as the Medical Emergency Team) is a tool recommended by the IHI to reduce preventable cardiopulmonary arrest. RRTs were first reported in Australia by Ken Hillman and his colleagues at the University of New South Wales (2001, 2002, 2005). Their purpose is to provide immediate care to patients on the medical/surgical ward who show signs of physiological instability or clinical deterioration. RRTs do not replace Code Teams; they provide intervention to prevent, rather than treat, cardiopulmonary arrest. After the AHRQ began using “failure to rescue” as a quality measure and IHI started promoting the use of RRTs as part of its 100,000 Lives Campaign, RRT implementation has grown rapidly in the US, from less than 50 hospitals in 2003 to roughly 1,400 in 2005 (Robeznieks, 2005).

RRTs are primarily motivated by two observations. First, the overall outcome for patients who experience in-hospital cardiopulmonary arrest is poor and has changed little since the introduction of cardiopulmonary resuscitation (CPR) more than 40 years ago (Naeem & Montenegro, 2005). Second, signs and symptoms of patient instability and deterioration are often present well before cardiopulmonary events. A number of studies have evaluated the efficacy of CPR after in-hospital arrest. One of the largest was a meta-analysis of 98 published reports of in-hospital CPR outcomes between 1966 and 1990 (Schneider, et al., 1993). Of nearly 20,000 patients who received CPR, only 15% survived to hospital discharge. More recent outcomes studies have had similar results. The type and severity of arrest is an important predictor of survival. Patients with an initial rhythm of ventricular fibrillation or ventricular tachycardia generally have higher rates of survival to discharge than patients with asystole or pulseless electrical activity. Other factors associated with survival include rapid restoration of spontaneous circulation and whether the event was witnessed (Gwinnutt, et al., 2000; Cohn, et al., 2004; Brindley, et al., 2002).

Several studies have shown that predictive physiological changes are observable in patients on medical/surgical wards well before they experience a cardiopulmonary event (Naeem & Montenegro, 2005; Fieselmann et al., 1993; Goldhill et al., 1999). Buist, et al. (1999) found that 76% of medical/surgical patients with cardiac arrest or unplanned ICU admission displayed signs of clinical instability more than 1 hour prior to the critical event (mean 6.5 h). In one-third of these events, the signs had been present for more than 24 hours. This delay has a significant effect on survival. In a recent study, Young, et al. (2003) and Kaboli and Rosenthal (2003) found a delay of as little as 4 hours in the transfer of deteriorating patients from the medical/surgical ward to the ICU increased the risk of mortality by a factor of four.

RRTs usually include a critical-care nurse, a respiratory therapist, and physicians or other support personnel as required. In the general model, RRTs can be called by any hospital staff member involved in the care of patients, including nurses, physicians, and respiratory therapists, among others. Calls come primarily from floor nurses, when a patient’s physiological parameters fall outside predefined criteria. These criteria vary by institution, but usually include acute and abnormal changes in vital signs, oxygenation, or level of consciousness. In addition, the model emphasizes the ability of staff to call the RRT when they are concerned about the patient (intentionally left undefined).

Despite the motivating premise and significant adoption of RRTs, clinical evidence for their efficacy is mixed. The earliest RRT studies were descriptive in nature (Daly, et al., 1998; Lee, et al., 1995). Some later studies, mostly small, non-randomized, and using historical controls, found that RRT implementation was associated with a reduction in unplanned ICU admission (Bristow, et al., 2000), cardiac arrest, and death. Others found no significant differences before or after RRT implementation (Kenward, et al., 2004; Salamonson, et al., 2001).

The first published study to provide evidence that RRTs could reduce all-cause hospital mortality was performed by Bellomo, et al. (2003) at the University of Melbourne. The study had a before-and-after interventional design with a 4-month observational period before implementation of the RRT, a 1-year preparatory period to educate staff on the use of the RRT, a 2-month run-in period, and a 4-month intervention period with the RRT operating at full capacity. In the study, the investigators observed a 56% relative risk reduction in cardiac arrest deaths and a 26% reduction in total inpatient deaths (p=0.004) after implementation of the RRT.

A study with similar design performed in a 700-bed general hospital in England had fewer promising results. During the 12-month intervention period, the RRT was called 136 times for 130 patients, but no significant differences in unexpected cardiac arrest or overall hospital deaths were observed for the year before or the year after implementation (Kenward, et al., 2004).

These early studies had limitations in design. For example, it is not possible to determine from these reports how many patients had physiological abnormalities warranting intervention but for whom the RRT was not called. These data could help determine whether the absence of an observable effect was due to lack of awareness or underutilization of the RRT rather than a lack of efficacy, although institutional awareness and utilization and the interventions of the team itself are considerations in the overall assessment of efficacy.

Hillman et al. (2005) performed a large and rigorous study in 23 Australian hospitals. The study used a cluster-randomized design to compare functioning as usual (in 11 hospitals) to implementation of a RRT (in 12 hospitals). The primary outcome measure was the composite of cardiac arrest, unexpected death, or unplanned ICU admission during the 6-month period after RRT activation. At the end of the study period, no significant differences were observed between groups in the primary outcome measure. A significant reduction in the rate of cardiac arrests and unexpected deaths were observed from baseline to the end of the study period in both groups of hospitals, but in these secondary measures as well, no differences were observed in relationship to whether the hospitals had implemented an RRT.

The lack of significant effect in the Hillman study was disappointing. In the report, the authors discussed several possible reasons for this outcome, including the possibility that the RRT system is an ineffective intervention and the possibility that the study period was too short to see an effect. An interesting finding from the perspective of identifying ways to improve patient outcomes was that for patients without a documented do-not-resuscitate order, a record of blood pressure, heart rate, and respiratory rate in the 15-minute period before an event occurred was absent in 3,657 (62%) cases, incomplete in 1,122 (19%) cases, and complete in only 1,120 (19%) cases. Hillman concluded that these findings suggest, “the need for improved intensive monitoring of patients in general wards; frequent and rigorous documentation of patients’ condition; and increased attention to education to ensure a timely response by appropriately trained clinicians.”

These conclusions were echoed by Daryl Jones of Bellomo’s research group at the University of Melbourne (2005a). The Melbourne group has consistently reported some of the best results with RRTs. In addition to a reduction in all-cause mortality, their hospital saw a sustained and progressive reduction in cardiac arrests over the 4-year period after RRT implementation (2005b). In a study to determine the circadian pattern of RRT activation at the hospital and to relate it to nursing and medical activities, Jones found that peak levels of RRT calls occurred around the time of scheduled caregiver visits. Jones concluded that, “It is likely that a substantial proportion of these patients would have been ill for some time before the call was made, and were only identified during routine observations or at the time of nursing handover”(Jones, et al., 2005a).

Increasing Surveillance in Medical/Surgical Wards
Increasing surveillance on the medical/surgical ward has the potential to reduce the incidence of adverse events. Perhaps the best evidence comes from studies that have examined the relationship between nurse staffing levels and patient outcomes. The largest was jointly funded by AHRQ, the Health Resources and Services Administration, the Centers for Medicare & Medicaid Services, and the National Institute of Nursing Research (Needleman, et al., 2001). It examined the records of 5 million medical patients and 1.1 million surgical patients treated at 799 hospitals during 1993. The study found that higher nursing staff levels (RNs, LPNs, aides) were associated with:


  • Lower rates of urinary tract infections, pneumonia, shock, upper gastrointestinal bleeding, and long hospital stays in medical patients.
  • Lower rates of urinary tract infections and failure to rescue in surgical patients.
  • A 2% to 25% reduction in adverse outcomes, depending on the outcome.


A second, smaller study of Pennsylvania hospitals (funded by AHRQ and the National Science Foundation) found a lower incidence of nearly all adverse outcomes in hospitals with higher levels of licensed nurses (Unruh, 2003). For example, in the study a 10% higher proportion of licensed nurses were estimated to decrease lung collapse by 1.5%, pressure ulcers by 2%, and falls by 3%.

The relationship between nurse staffing levels and patient mortality is less clear, but some studies have found that 30-day mortality and failure to rescue are higher when nurse staffing levels are lower (Stanton & Rutherford, 2004; Aiken, et al., 2002b; Aiken, et al., 1999).

Facilitating surveillance on the medical/surgical floor by increasing nurse staffing levels offers promise for improving patient outcomes, but current and predicted nursing shortages make this approach problematic. According to a 2002 report by the workforce commission of the American Hospital Association (HRSA), the nursing shortage “reflects fundamental changes in population demographics, career expectations, work attitudes, and worker dissatisfaction” that indicate the shortages are not temporary or cyclical in nature. In addition, the Health Resources and Services Administration predicts that hospital nursing vacancies will reach 800,000 by 2020 as the number of nurses is expected to grow by 6%, while demand for nursing care is expected to grow by 40% over the period.

As importantly, the burden on nurses continues to increase as well. In one survey of 13,471 nurses, 83% reported an increase in the number of patients assigned to them during the previous year (Aiken, et al., 2002a). Moreover, these patients tended to be sicker and require higher levels of care than in previous years as new medical technologies and efforts to control healthcare costs reduced average length of stay and diverted many healthier patients to the outpatient setting (Stanton & Rutherford, 2004).

The Potential of Automated Monitoring
Novel automated systems may help nurses better monitor patients on medical/surgical wards. They can continuously track vital signs as well as bed position (especially for patients at risk of falling). Improved vigilance through automated monitoring has the potential to more rapidly identify patients who are physiologically unstable, in clinical decline, or exiting the bed without help. And given the underlying rationale for RRTs, more rapid identification and assistance for these patients may reduce preventable adverse events including cardiopulmonary arrest.

Physiological parameters that include blood pressure, temperature, heart rate and rhythm, O2 saturation, and respiration rate are all routinely measured in the operating room and the ICU. The potential for significant improvement in vigilance lies in bringing a basal level of vital sign monitoring out of the ICU and onto the medical/surgical floor.

An effective monitoring system for the medical/surgical floor would include several characteristics:


  • High Accuracy (sensitivity and specificity). The medical/surgical environment is unique in the hospital, with staffing levels and workflows that permit very low tolerance for false alarms that take caregivers away from other important duties.
  • Be accepted by patients. It should not have wires, cuffs, or other connections that are uncomfortable or restrict patient movement.
  • Be well-integrated into the existing workflows on the floor.
  • Not interfere with normal patient care activities.
  • Provide economic value either by demonstrably reducing the incidence of adverse events or by improving flow through the ICU and medical/surgical wards. New technology represents a significant investment of limited hospital resources and must demonstrate its value before it becomes widely adopted.


A primary challenge in implementing an automated medical/surgical monitoring system is false alarms. Auditory warnings are used throughout the hospital, but even in the operating room and ICU, many are perceived as a hindrance by medical staff because of the high incidence of false alarms (Block, et al., 1999; Chambrin, 2001; Edworthy & Hellier, 2005).

Most alarms on monitoring equipment are generated when a parameter (e.g., respiration rate) crosses a limit. In practice, patients can experience a wide variation in a parameter without significant physiological effect. Several studies have examined the rate of false alarms in patient monitoring in the critical care setting, and in general they’ve found that the vast majority of alarms have no clinical significance (Chambrin, 2001; Lawless, 1994; Tsien & Fackler, 1997). Research by Bliss and Dunn (2000) and Bliss et al. (1995) has shown that if an alarm is perceived to be 90% reliable, then people will respond slightly more than 90% of the time. If the alarm is 10% reliable, people will respond only 10% of the time. It’s no wonder, then, that hospital staff ignore alarms in the operating room and ICU. On the medical/surgical floor, with its lower staffing levels, false alarms are likely to be even less tolerable.

The next generation of monitoring devices will be distinguished from those in the past by the power of the algorithms they use to intelligently identify clinically important events. Several approaches are possible. For example, one would be to monitor more than one physiological parameter (e.g., heart rate, respiratory rate, etc.) at a time and trigger an alarm only when both near their threshold values. Another would be to average a sensor’s readings over a carefully defined window of time to filter out transient changes unlikely to have clinical significance.

Patient safety remains a critical national need. Despite the increasing interest in patient safety and many new patient safety initiatives, too many people still die unnecessarily in hospitals. RRTs are a rapidly growing approach to prevent unnecessary morbidity and mortality, and increased monitoring on medical/surgical floors may help them better realize their potential. Automated systems that monitor patient vital signs may extend the value of RRTs by providing early warning of patient decline. For medical/surgical monitoring devices to become widely accepted they must provide value to the hospital and find the right balance between sensitivity and specificity so that they the increased burden on nursing staff is justified by the benefits they achieve.

Joshua Jacobs serves as an associate professor of medicine at the University of Hawaii. He is Board certified by the American Board of Family Medicine and has direct responsibilities for clinical research, medical advisory board activities, and clinical aspects of regulatory affairs and product development as director of medical affairs at Hoana Medical, Inc. He is a principal investigator for various research activities with the United States National Institutes of Health.ÝJacobs has been an invited speaker at numerous national and international meetings and workshops with a focus on quality improvement and medical education, and medical informatics. He may be contacted at


Agency for Healthcare Research and Quality (AHRQ). (2005, December). 2005 National healthcare quality report. Rockville, MD: U.S. Department of Health and Human Services. AHRQ Publication No. 06-0018.

Aiken, L. H., Clarke, S. P., Sloane, D. M. (2002a). International Hospital Outcomes Research Consortium. Hospital staffing, organization, and quality of care: cross-national findings. International Journal of Quality Health Care, 14(1), 5-13.

Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., & Silber, J. H. (2002b). Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. Journal of the American Medical Association, 288(16), 1987-1993.

Aiken, L. H., Sloane, D. M., Lake, E. T., Sochalski, J., & Weber, A. L. (1999). Organization and outcomes of inpatient AIDS care. LDI Issue Brief, 1, 1-4.

Bates, D. W., & Gawande, A. A. (2003). Improving safety with information technology. New England Journal Medicine, 348(25), 2526-2534.

Bates, D. W., & Gawande, A.A. (2000). Error in medicine: what have we learned? Annals of Internal Medicine, 132(9), 763-767.

Bellomo, R., Goldsmith, D., Uchino, S., Buckmaster, J., Hart, G. K., Opdam, H., Silvester, W., & Doolan, L., & Gutteridge, G. (2003). A prospective before-and-after trial of a medical emergency team. Medical Journal of Australia, 179(6), 283-287.

Berwick, D. M., Calkins, D. R., McCannon, C. J., & Hackbarth, A. D. (2006). The 100,000 Lives Campaign: Setting a goal and a deadline for improving health care quality. Journal of the American Medical Association, 295(3), 324-327.

Bliss, J. P., & Dunn, M. C. (2000). Behavioural implications of alarm mistrust as a function of task workload. Ergonomics, 43(9), 1283-1300.

Bliss, J. P., Gilson, R. D., & Deaton, J. E. (1995). Human probability matching behaviour in response to alarms of varying reliability. Ergonomics, 11, 2300-2312.

Block, F. E. Jr., Nuutinen, L., & Ballast, B. (1999). Optimization of alarms: a study on alarm limits, alarm sounds, and false alarms, intended to reduce annoyance. Journal of Clinical Monitoring and Computing, 15(2), 75-83.

Brennan, T. A., Leape, L. L., Laird, N. M., Hebert, L., Localio, A.R., Lawthers, A. G., Newhouse, J. P., Weiler, P. C., & Hiatt, H. H. (2004). Harvard Medical Practice Study I. Incidence of adverse events and negligence in hospitalized patients: Results of the Harvard Medical Practice Study I. 1991. Quality & Safety in Health Care, 13(2), 145-151.

Brindley, P. G., Markland, D.M., Mayers, I., & Kutsogiannis, D. J. (2002). Predictors of survival following in-hospital adult cardiopulmonary resuscitation. Canadian Medical Association Journal, 167(4), 343-348.

Bristow, P. J., Hillman, K. M., Chey, T., Daffurn, K., Jacques, T. C., Norman, S. L., Bishop, G. F., & Simmons, E. G. (2000). Rates of in-hospital arrests, deaths and intensive care admissions: The effect of a medical emergency team. Medical Journal of Australia, 173(5), 236-240.

Buist, M. D., Jarmolowski, E., Burton, P. R., Bernard, S. A., Waxman, B. P., & Anderson, J. (1999). Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Medical Journal of Australia, 171(1), 22-25.

Centers for Disease Control and Prevention (CDC). (2000). Monitoring hospital-acquired infections to promote patient safety — United States, 1990-1999. Morbidity & Mortality Weekly Report, 49(8):149-153.

Chambrin, M. C. (2001). Alarms in the intensive care unit: how can the number of false alarms be reduced? Critical Care, 5(4), 184-188.

Cohn, A. C., Wilson, W. M., Yan, B., Joshi, S. B., Heily, M., Morley, P., Maruff, P., Grigg, L. E., & Ajani, A. E. (2004). Analysis of clinical outcomes following in-hospital adult cardiac arrest. Internal Medicine Journal. 34(7), 398-402.

Daly, F. F., Sidney, K. L., & Fatovich, D. M. (1998). The Medical Emergency Team (MET): a model for the district general hospital. Australia & New Zeland Journal of Medicine, 28(6), 795-798.

Edworthy, J., & Hellier, E. (2005). Fewer but better auditory alarms will improve patient safety. Quality & Safety in Health Care, 14(3), 212-215.

Fieselmann, J. F., Hendryx, M. S., Helms, C. M., & Wakefield, D. S. (1993). Respiratory rate predicts cardiopulmonary arrest for internal medicine inpatients. Journal of General Internal Medicine, 8(7), 354-360.

Goldhill, D. R., White, S. A., & Sumner, A. (1999). Physiological values and procedures in the 24 h before ICU admission from the ward. Anaesthesia, 54(6), 529-534.

Gwinnutt, C. L., Columb, M., & Harris, R. (2000). Outcome after cardiac arrest in adults in UK hospitals: Effect of the 1997 guidelines. Resuscitation, 47(2), 125-135.

Health Resources and Services Administration (HRSA), Bureau of Health Professions, National Center for Health Workforce Analysis. (n.d.). Projected supply, demand, and shortages of registered nurses: 2000-2020. HRSA Web site: http://bhpr.

HealthGrades. (2005). HealthGrades quality study: Second annual report on patient safety in American hospitals. Golden, CO: HealthGrades, Inc.

Hillman, K. M., Bristow, P. J., Chey, T., Daffurn, K., Jacques, T., Norman, S. L., Bishop, G. F., & Simmons, G. (2001). Antecedents to hospital deaths. Internal Medicine Journal, 31(6), 343-348.

Hillman, K. M., Bristow, P. J., Chey, T., Daffurn, K., Jacques, T., Norman, S. L., Bishop, G. F., Simmons, G. (2002). Duration of life-threatening antecedents prior to intensive care admission. Intensive Care Medicine, 28(11), 1629-1634.

Hillman, K., Chen, J., Cretikos, M., Bellomo, R., Brown, D., Doig, G., Finfer, S., & Flabouris, A. (2005). MERIT study investigators. Introduction of the medical emergency team (MET) system: A cluster-randomised controlled trial. Lancet, 365(9477), 2091-2097.

Institute of Medicine (IOM). (2000). To err is human: Building a safer health system. L. T. Kohn, J. Corrigan, M. S. Donaldson (Eds). Washington, DC: National Academies Press.

Jones, D., Bates, S., Warrillow, S., Opdam, H., Goldsmith, D., Gutteridge, G., & Bellomo, R. (2005a). Circadian pattern of activation of the medical emergency team in a teaching hospital. Critical Care, 9(4), R303-306.

Jones, D., Bellomo, R., Bates, S., Warrillow, S., Goldsmith, D., Hart, G., Opdam, H., & Gutteridge, G. (2005b). Long term effect of a medical emergency team on cardiac arrests in a teaching hospital. Critical Care 9(6), R808-815.

Kaboli, P. J., & Rosenthal, G. E. (2003). Delays in transfer to the ICU: a preventable adverse advent? Journal of General Internal Medicine, 18(2), 155-156.

Kenward, G., Castle, N., Hodgetts, T., & Shaikh, L. (2004). Evaluation of a medical emergency team one year after implementation. Resuscitation, 61(3), 257-263.

Lawless, S. T. (1994). Crying wolf: False alarms in a pediatric intensive care unit. Critical Care Medicine, 22(6), 981-985.

Leape, L. L., & Berwick, D. M. (2005). Five years after To Err Is Human: what have we learned? Journal of the American Medical Association, 293(19), 2384-2390.

Lee, A., Bishop, G., Hillman, K. M., & Daffurn, K. (1995). The medical emergency team. Anaesthesia Intensive Care, 23(2), 183-186.

Naeem, N., & Montenegro, H. (2005). Beyond the intensive care unit: A review of interventions aimed at anticipating and preventing in-hospital cardiopulmonary arrest. Resuscitation, 67(1), 13-23.

Needleman, J., Buerhaus, P., Mattke, S., et al. (2001). Nurse-staffing levels and patient outcomes in hospitals. Final report for Health Resources and Services Administration. Contract No. 230-99-0021. Boston, MA: Harvard School of Public Health.

Robeznieks, A. (2005). Quick assistance. Surge of popularity puts rapid-response teams in 1,400 U.S. hospitals. Modern Healthcare, 35(36), 33.

Salamonson, Y., Kariyawasam, A., van Heere, & B, O’Connor, C. (2001). The evolutionary process of Medical Emergency Team (MET) implementation: Reduction in unanticipated ICU transfers. Resuscitation, 49(2), 135-41.

Schneider, A. P., Nelson, D. J., & Brown, D. D. (1993). In-hospital cardiopulmonary resuscitation: a 30-year review. Journal of the American Board of Family Practitioners, 6(2), 91-101.

Stanton, M. W., & Rutherford, M. K. (2004). Hospital nurse staffing and quality of care. Rockville (MD): Agency for Healthcare Research and Quality, Research in Action, Issue AHRQ Pub. No. 04-0029.

Tsien, C. L., & Fackler, J. C. (1997). Poor prognosis for existing monitors in the intensive care unit. Critical Care Medicine, 25(4), 614-619.

Unruh, L. (2003). Licensed nurse staffing and adverse events in hospitals. Medical Care, 41(1), 142-52.

Young, M. P., Gooder, V. J., McBride, K., James, B., & Fisher, E. S. (2003). Inpatient transfers to the intensive care unit: delays are associated with increased mortality and morbidity. Journal of General Internal Medicine, 18(2), 77-83.