Rapid response systems

INTRODUCTION — Rapid response systems identify deteriorating hospitalized patients prospectively and seek to alter their clinical trajectory through increasing the clinical resources directed to them [1]. As hospitalized patients may exhibit warning signs prior to deterioration [2,3], rapid response systems have the potential to prevent adverse clinical outcomes, including cardiac arrest and death.

Rapid response systems are being utilized increasingly throughout the world, particularly in developed countries [4,5]. In 2005, the Institute for Healthcare Improvement made implementation of rapid response systems a key part of the 100,000 Lives Campaign to improve the quality of care in hospitals and reduce mortality rates [6]. However, the evidence supporting rapid response systems is limited, and it remains controversial whether these systems of care are effective [7-9].

In this topic, we discuss the rationale for rapid response systems and current evidence for their use in adults. Rapid response systems in the pediatric population and specific treatments in the management of critically ill patients are reviewed separately. (See "Overview of sudden cardiac arrest and sudden cardiac death" and "Advanced cardiac life support (ACLS) in adults" and "Pediatric advanced life support (PALS)", section on 'Rapid response teams'.)

DEFINITIONS — Rapid response systems are programs that are designed to improve the safety of hospitalized patients whose condition is deteriorating quickly [10]. They are based on prospective identification of high-risk patients, early notification of a team of responders who have been preselected and trained, rapid intervention by the response team, and ongoing evaluation of the system’s performance.

Several terms are used to refer to rapid response systems. These terms include critical care outreach, medical emergency teams, medical response teams, and rapid response teams. There are subtle differences between these terms, but all maintain two key features: an afferent limb (ie, how the team is activated) and an efferent limb (ie, the response of the team). Medical emergency teams typically refer to physician-led teams that have the ability to manage complex airway issues, establish central access, and initiate intensive care unit (ICU) level care at the bedside [10]. Rapid response teams are generally nurse-led teams [10]. Critical care outreach is slightly different from the other terms in that critical care outreach also focuses on educating non-critical care staff and improving transfers between ICUs and the general hospital wards.

There are no studies to date that compare different models with one another. Each model has different advantages and disadvantages. As an example, physician-led groups may have more clinical resources immediately available to them but may be more expensive to establish, while nurse-led groups might be more immediately accessible to the bedside nurse who activates the rapid response system but require additional consultants for complicated procedures such as intubation (table 1) [11].

KEY COMPONENTS — An international consensus group was convened by members of patient safety, critical care medicine, and hospital management organizations to clarify designs of rapid response systems and make recommendations for their implementation [10]. Key features of rapid response systems include:

●The criteria for activation are mostly objective but often include at least one subjective criterion. Objective criteria usually include changes in vital signs or acute neurologic change. The subjective criterion usually involves language such as “Staff member is worried about the patient” [12]. The subjective criterion provides an opportunity for clinical judgment to trigger an escalation of care in the absence of vital sign or neurologic change. Inclusion of a subjective criterion has been shown to identify patients who would otherwise be missed by the objective criteria [13]. Examples of activation criteria are included in the table (table 2). Hospitals that implement rapid response systems provide frequent reminders to all staff about the activation criteria, including placing posters around the hospital and stickers on identification badges.

●The triggering mechanism for activation is available 24 hours a day and leads to a response within 15 minutes. The response includes direct patient examination and assessment by team members. There should be no negative consequences for activation. Typically these systems are reactive (responding to calls), but some designs include active (prospective) surveillance of triggers for patient deterioration.

●The members of the rapid response team are chosen depending upon the severity of illness of the patients and the specific goals and resources of the institution. The members of the rapid response teams and traditional “code” or cardiac arrest teams may be similar, often including at least one nurse and one physician. However, in rapid response teams, the intensive care unit (ICU) personnel tend to be involved earlier in the patient’s course than with traditional cardiac arrest teams. The rapid response system should include team members with:

•Ability to diagnosis the clinical problem

•Ability to initiate therapy

•Authority to transfer the patient to higher levels of care

●The rapid response system should also have a feedback loop that collects data and improves the system to maximize patient safety. At the patient level, patient data are reviewed to improve clinical strategies. At the administrative level, resources are coordinated to ensure adequate staffing, equipment, and training for the entire system.

GOALS OF INTERVENTION — The major goal of rapid response systems is to react promptly and appropriately to deteriorating clinical status, such as respiratory failure, cardiac failure, altered consciousness, hypotension, arrhythmias, pulmonary edema, and sepsis. Rapid response systems aim to achieve this by matching increasing patient acuity with increased health care resources and expertise at a point when disease trajectory can still be modified. Failure to recognize and respond to these conditions is associated with adverse events and is labeled “failure to rescue.”

Other specific goals of care of a given system vary substantially from one institution to another. The concept of a rapid response system and its goals are summarized in a figure (figure 1).

Identify patients at risk — Rapid response teams differ from traditional cardiac arrest teams in that rapid response teams assess patients in whom respiratory, cardiac, or neurologic deterioration is developing (eg, vital sign changes or altered mental status), whereas traditional cardiac arrest teams generally respond to later events, including respiratory or cardiac arrest. Rapid response systems are put in place to intervene and prevent deterioration that might lead to cardiac arrest, as the survival of inpatients with cardiac arrest is low; several studies using different methodologies indicate that the risk of death is greater than 80 percent after in-hospital cardiopulmonary resuscitation (CPR) for cardiac arrest [14,15].

In multicenter, prospective cohorts, clinical antecedents, such as a change in vital signs, are present in the majority of patients (60 to 70 percent) before in-hospital cardiac arrest occurs [2,3]. In one large prospective study, the presence of any one of the following was associated with a 6.8-fold increase in mortality (95% CI 2.7-17.1) [16]:

●Bradycardia <30 beats/minute

●Systolic blood pressure <90 mmHg

●Oxygen saturation <90 percent

●Respiratory rate <6 breaths/minute

●A decrease in Glasgow Coma Scale by 2 points

●Onset of coma

The optimal thresholds for activation criteria in different patient populations are not well-established. In addition, it is not known how often hospitalized patients have abnormal vital signs but do not go on to have an adverse event. Some hospitals have implemented automated activation systems based on the electronic health record, but results have been mixed. Some studies found that, compared with using paper-based systems, risk identification using algorithms in electronic health records was more accurate for predicting a composite outcome of in-hospital cardiac arrest, intensive care unit (ICU) transfer, and death within 24 hours of observation [17], and that alerts to the rapid response team from an electronic health record were associated with lower ICU admission rates than were call-activated responses, whether used alone or in combination with electronic alerts [18]. However, other studies do not show electronic systems to be clearly superior to manually based activation systems [19].

However, patients for whom rapid response systems are activated have substantial risks for death in the short term. As an example, in one study, 18.6 percent of patients seen by the rapid response team died during the hospitalization [20]. Another study found that patients who met activation criteria had a 7.9 percent 30-day mortality rate and 15.8 percent died by 180 days [21].

Communicate change in patient status — Rapid response systems mandate and accelerate communication among team members caring for the patient. The literature suggests that a sizable percentage of cardiac arrests occur due in part due to poor communication. In one study, four independent reviewers assessed pre-arrest care of 118 inpatient cardiac arrests in a large public hospital [22]. The panel determined that 68 percent of cardiac arrests were avoidable. Of the avoidable cases, 35 percent had a delay in nursing staff communicating with the physician, and 29 percent had a delay in the physician responding to the nurse’s request.

Rapid response systems improve communication at both the nursing and physician levels by systematic activation of a team of providers and ensuring no negative feedback for appropriate activation [1].

Provide effective and timely interventions — The quality of care of critically ill patients prior to ICU transfer may be suboptimal and associated with increased hospital mortality [23]. Common causes of suboptimal care include lack of knowledge or experience by the treating clinician, failure to appreciate urgency, and failure to seek advice. As an example, in 45 percent of cases of preventable cardiac arrest in one hospital, a physician in training was the most senior physician to review clinical signs of deterioration during the 24 hours prior to cardiac arrest [22].

Although physicians in training are frequently involved in the care of patients by a rapid response team, specialty-trained physicians (eg, critical care specialists, cardiologists) may be systematically involved sooner than with traditional models of care, while also encouraging less-specialized providers to seek help when needed. Ideally, members of a rapid response team are also regularly trained and audited and may undergo simulation training, which improves performance both as a team and as individuals.

Avoid delays in ICU transfer — Rapid response systems facilitate early transfer of patients to the ICU when needed, with less likelihood of being “refused” admission by the ICU service.

Delays in ICU transfer are common, occurring in 39 percent of patients in one prospective cohort study [23]. Observational studies have found that mortality rates are higher when patient transfer to higher levels of care is delayed [24,25]. In one study, there was a 36 percent mortality rate in patients “refused” by the ICU compared with 14 percent of those admitted, a difference that persisted after adjustment for severity of illness [24]. In another study, patients with vital sign abnormalities greater than four hours prior to ICU transfer had an increased mortality risk compared with those who were transferred more quickly (relative risk [RR] 3.5, 95% CI 1.4-9.5) [25]. Similarly, delay in activation of the rapid response system for decompensating patients is independently associated with increased mortality [26].

In addition, some critically ill patients are either inappropriately admitted to the general ward, rather than to the ICU, from the emergency department or decompensate rapidly shortly after admission. Patients transferred to the ICU within 24 hours of ward admission from the emergency department had an increased risk of 30-day mortality compared with similar patients directly admitted to the ICU (RR 2.5, 95% CI 1.2-5.2) [27]. Rapid response systems can expedite ICU admission of critically ill patients from the emergency department and of those who are inappropriately sent to the general ward [28].

ROLES OF THE RAPID RESPONSE TEAM — Rapid response team interventions depend on the specific condition of the patient, similar to care provided by traditional cardiac arrest teams, but earlier in the development of physiologic instability [10]. Common interventions in a specific clinical situation may include administration of oxygen, intravenous fluids, bronchodilators, diuretics, and respiratory therapy (eg, suction, chest percussion with postural drainage), among others. Patients who are more acutely ill may need more aggressive treatment such as cardiopulmonary resuscitation (CPR), vasopressors, and intubation with mechanical ventilation.

Many patients for whom the rapid response team is activated may remain on the general ward with more frequent monitoring. Others will require prompt transfer to the intensive care unit (ICU). The rate of ICU transfer has varied widely in studies. As an example, one study found that 15 percent of patients were transferred to the ICU after rapid response team activation [12] while, in another study, 45 percent of patients with rapid response team activations were transferred to the ICU [20].

Some rapid response systems (eg, critical care outreach) also facilitate communication between providers staffing the general wards and the ICUs during handoffs [29]. Rapid response systems may also facilitate discussion and implementation of end-of-life care [30-32].

For other patients, particularly those patients who remain on the medical ward, activation of the rapid response system may serve to realign nursing staffing. In many cases, nurse managers or other senior, experienced nurses serve as team members to both provide guidance to more junior nursing staff and to facilitate changes in staffing levels. The rapid response system serves as a marker of patients who require additional nursing staffing and clinician vigilance.

Given the heterogeneity of rapid response models in the literature, there are no consistent models or recommendations for how much responsibility the rapid response team should assume in the care of the patient. In some models, the rapid response team assumes primary control over the patient’s care, while in other hospitals, the team functions as a consult service. Some teams mandate discussion with the attending physician responsible for the patient’s care. It is unclear which model minimizes patient harm.

EFFICACY — A fundamental goal of rapid response systems is to identify patients, particularly those not in an intensive care unit (ICU), who might benefit from a higher level of care and implement that care to decrease morbidity and mortality. The limited data currently available suggest that rapid response systems may achieve this goal, but are not definitive. Interpretation of studies is limited by differences in the interventions used, populations studied, and outcomes that were measured.

Systematic reviews and meta-analyses of mostly observational studies found the evidence suggested that use of rapid response systems was associated with a reduction in mortality in-hospital [33,34] and with a reduction in cardiorespiratory arrests in non-ICU patients [33,34]; reduced cardiac arrests were also seen in a large time-series study [35]. Additionally, studies found that a delay in activation of rapid response systems may be associated with increased in-hospital and 30-day mortality [36-38]. However, these data are not definitive given the low quality of the available evidence and the wide variation in rapid response systems, particularly with regard to team structure, participants, and capabilities.

Only two randomized trials have evaluated rapid response systems, both with unblinded design, and conflicting findings on overall mortality:

●The MERIT investigators enrolled 125,132 patients from 23 Australian hospitals in a cluster-randomized trial of medical emergency team implementation [39]. Hospitals were randomly assigned to medical emergency teams or usual care. There was no detectable difference in unexpected mortality, cardiac arrests, or unplanned ICU admissions.

●In a study of 2903 hospitalized patients in the United Kingdom, 16 hospital wards, in pairs matched for patient acuity and usual conditions treated, were randomly assigned in four-week slots to a critical care outreach team or to usual care [29]. Implementation of the critical care outreach team reduced in-hospital mortality (odds ratio [OR] 0.52, 95% CI 0.32-0.85).

Although some studies have reported that rapid response systems can decrease other clinical outcomes, such as postoperative complications, respiratory failure, stroke, severe sepsis, acute renal failure, and unplanned ICU transfer [12,39,40], further study is required.

POTENTIAL DISADVANTAGES — Rapid response systems may have potential disadvantages. It is possible that the reliance on rapid response systems may induce a decreased sense of responsibility on the part of the hospital ward team. Hospital ward teams may be less likely to provide optimal care knowing that there is a backup system of rapid response systems to identify and treat patients if they worsen. The presence of rapid response teams may also inadvertently lead to confusion as to who has primary responsibility for the patient [1].

In addition, implementation of rapid response systems can be expensive and time-consuming. It is unclear whether these systems divert resources away from more effective interventions. One analysis done at a pediatric hospital suggested that the costs of a rapid response system could be offset by a decrease in the number of clinical deterioration events [41].

Because it is not clear what the optimal activation criteria are to prevent unnecessary evaluation and intensive care unit (ICU) transfer, rapid response systems may lead to excessive assessment of patients who have abnormal vital signs but are not at risk for serious adverse events. Activation criteria in several sites have focused on changes in vital signs rather than fixed values in an effort to identify deteriorating patients more accurately [12]. However, it is our experience is that it is difficult to implement calculations at the bedside unless it is embedded in an electronic medical record.

Finally, some authors have suggested that implementation of a rapid response system is, in essence, creating a system of rescue for patients who have been mis-triaged to lower levels of care. Alternative strategies for managing patient flow and optimizing limited critical care resources may prove more cost-effective than an entirely new clinical system [8].

SUMMARY AND RECOMMENDATIONS

●Rapid response systems are programs that are designed to improve the safety of hospitalized patients whose condition is deteriorating quickly. (See 'Definitions' above.)

●The key components of a rapid response system include (see 'Key components' above):

•Objective and subjective activation criteria (table 2)

•A triggering mechanism available 24 hours a day, leading to a first response within 15 minutes of activation

•Members of the rapid response team who have the ability to diagnose the clinical problem, initiate therapy promptly, and transfer the patient to higher levels of care if needed

•A feedback loop to ensure safety improvement at both the patient and administrative levels

●Rapid response systems have several goals that are focused on improving the quality of patient care as soon as clinical deterioration is detected, up until the time that patients are stabilized or transferred to the intensive care unit (ICU). (See 'Goals of intervention' above.)

●Systematic reviews of cohort and randomized trials have suggested that implementation of rapid response systems reduce the rate of cardiopulmonary arrest and may reduce mortality in hospitalized patients. (See 'Efficacy' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Michael D Howard, MD, MPH, who contributed to an earlier version of this topic review.