Avoiding failure-to-rescue: rapid response systems (Subbe, Welch)

Failure to rescue of patients with quantifiable signs of deterioration is a preventable complication of hospital treatment. Rapid response systems represent a comprehensive approach to reduce failure to rescue and consist of an afferent limb and an efferent limb. The afferent limb aims to monitor for relevant signs of deterioration and usually includes a standardised method of evaluation of vital signs to identify changes from a stable status. Escalation of deteriorating patients can be through any member of staff. In a patient centred service patients and relatives should be able to raise concerns too. The team that responds to the deterioration is labelled a rapid response team, medical emergency team or critical care outreach team. Systematic reviews and meta-analyses of published studies show a reduction in cardiac arrests and mortality in hospitals deploying rapid response systems.


The configuration and funding of health care at local, regional and national levels vary across the world, but the epidemiological challenges of ageing populations, an increasing burden of chronic disease and rising costs are common themes.

Modern hospitals manage larger numbers of increasingly complex cases. For example, ‘Finished Admission Episodes’ have grown by 21% in 10 years in England, with patients aged 70 to 74 making up the single largest group when broken down into 5-year age bands (with the exception of patients aged zero to 4, including babies born in hospital).491 Such patients commonly have several medical conditions. At the same time, treatment regimens are ever more complicated, and many patients take a variety of often-interacting and sometimes immunosuppressive medications. Surgical procedures are becoming more sophisticated and are performed on patients with significant comorbidities. Beyond the burden of disease, ageing is related to increasing frailty with a decreasing ability to adapt to physiological and psychological insults such as acute respiratory or heart failure, acute kidney injury, sepsis, clotting disorders and delirium.

Not surprisingly, ward staff in hospitals worldwide struggle to manage the acuity and dependency of patients either at-risk or actually experiencing deterioration. Insufficient staff numbers, an inadequate skill-mix and a higher proportion of temporary staff are associated with gaps in care and increased mortality.492 Moreover, acute deterioration in one patient increases the risk of a critical illness event in neighbouring patients, illustrating that wards tend to lack the resilience to ensure patient safety when unexpected additional demands are brought to bear.493

Typically, 5 to 10% of hospital patients have periods of significant instability. Deterioration is usually revealed by abnormal vital signs and/or laboratory/diagnostic results. Altered physiology may be transient and resolve with little or no treatment, but deterioration to the level of critical illness, organ failure and death can occur precipitously; one multinational study found that one in 10 ward patients referred for a rapid response died within 24 h.494 The absence of a reliable response to warning signs of impending harm was initially described in the surgical literature as ‘failure to rescue’ but the term is now used in a broader sense including missed opportunities to recognise or act on signs of deterioration in the general hospital population.

There is considerable evidence that deteriorating ward patients often receive suboptimal care. Inadequate monitoring is a major problem, with signs of abnormal mental and respiratory function frequently overlooked. This can result in cardiopulmonary arrest and (often belated) transfer to an ICU. Delayed treatment of deterioration is associated with worse outcomes even when ICU admission does occur, and patients and families are left with physical and psychological complications including increased dependency and posttraumatic stress disorder. Furthermore, missed deterioration is a growing cause of complaints and litigation as patients are better informed and less tolerant of suboptimal care.

Based on the observation that deterioration is common, often predictable and sometimes preventable, hospitals in Australia, the United Kingdom, the United States of America, Scandinavia and the Netherlands developed systems of surveillance and escalation to reduce the number of avoidable deterioration events in the 1990s and 2000s. In Australia, Ken Hillman established a medical emergency team triggered by abnormalities in any one vital sign, in the UK critical care outreach teams alerted by a whole range of indicators were formed, while in the USA Michael DeVita’s rapid response team was activated by predefined abnormalities.495–497 The commonalities were that personnel with experience in ICU and training in the management of critical illness responded to altered physiology and ward staff concern about patients outside ICU. The insight that a whole system is needed for timely identification and management of potential or actual deterioration has led to the term “rapid response system” being used to describe the essential components required.

The current chapter describes the structure and processes of a rapid response system and the current understanding of its potential impact on clinical outcomes, particularly with regard to peri-operative patients.


The first international consensus conference on rapid response systems in 2010 summarised the key elements of a functioning system (Fig. 9) that detects early deterioration and provides timely escalation and treatment using the terms ‘afferent limb’ for a mechanism of deterioration detection and ‘efferent limb’ for the mechanism of response.498

Fig 9: The structure of a rapid response system, adapted from the findings of the first Consensus Conference of medical emergency teams. (499)


Afferent limb: Patients at risk of significant deterioration are generally identified by physiological ‘Track and Trigger’ systems. Vital signs including respiratory rate, oxygen saturation, pulse rate, BP, level of consciousness, temperature and other indicators such as urine output and reported pain are ‘tracked’, with ‘trigger’ of an escalation if threshold values are reached. track and trigger systems may use threshold values of single physiological parameters, summary scores (generally known as ‘Early Warning Scores’, Fig. 10), or complex composites including laboratory data and other markers of acute and chronic illness. Escalation to the efferent limb might include activation by patients or family members.500–502

Fig 10: National Early Warning Score, AVPU, alert, voice, pain, unresponsive; BP, blood pressure; bpm, beats per minute; O2, oxygen. (502)


Efferent limb: Triggers can lead to escalation to a patient’s own medical teams or to specialist personnel with critical care skills. The latter can be critical care nurses or doctors or mixed teams including medical residents and respiratory therapists. In the United Kingdom, dedicated critical care outreach nurses usually provide the response. In Australia and the United States, it is more typically a medical emergency team or rapid response team that leaves the ICU to review patients on general wards. Rapid response teams can expedite transfers into higher care areas as well as facilitating invasive and noninvasive ventilation, insertion of central venous catheters and administration of inotropic and vasoactive drugs on general wards. The quality of the co-ordination between the patient’s primary, ward-based team and the response team is important;503 the location and characteristics of the patient, the nature of the episode of deterioration and the skill mix of the clinical teams are all factors. At an organisational level, the underlying concept of care is ‘right time, right people, right place’.

The afferent and efferent limbs of the rapid response system are supported by administrative functions and data collection for audit, quality assurance and improvement. A recent systematic review found that ‘clear leadership and continuous quality improvement provide the foundation for the continuing collaboration to manage deteriorating patients’.499 At an organisational level, regular evaluation of the safety culture should underpin other metrics.504 Governance of a hospital-wide system to provide safe care requires interprofessional training tailored to the needs of the different areas to assure whole system performance as well as structures to ensure widespread learning from both best practices and serious adverse events.

Process: The risk of deterioration is usually derived from considerations of physiological instability and pre-existing morbidity such as conditions that might suppress the immune response to infection. In addition, staff concern (nurse worry) and patients’ self-reports are increasingly understood to enable a more complete understanding of the patient’s condition and trajectory.505

At microlevel, management of the individual patient at-risk of deterioration can be structured with a ‘record, recognise, report, respond and repeat’ framework (Fig. 11).506 The regular, reliable recording of vital signs is essential in detecting patients in impending or actual crisis;507 noting that recognition of deterioration is a complex problem and many patients deteriorate ‘in plain sight’. This has led to the development of rule-based systems to encourage awareness of the need for escalation once threshold values have been reached. The UK National Early Warning Score (NEWS) is an example of a widely validated scoring system with an associated escalation protocol (Fig. 10).502 The ambition is to have a ‘common language of deterioration’ with an expected response that is understood across both primary and secondary care.

Fig 11: Process of activation and response of a rapid response system. Chain of survival. (506)


Healthcare technicians or nurses usually undertake recording of vital signs. The translation of the underlying pathology that leads to deterioration into the ordering of confirmatory tests and prompt treatments generally requires the involvement of other healthcare professionals such as advanced nurse practitioners or doctors. The process of reporting to more senior staff is one of the weak points in the translation of data about deterioration into active management. Automated electronic systems for vital sign capture might support more reliable escalation.508 Structured communication tools such as ‘Situation–Background–Assessment–Recommendation’ are recommended to increase the amount of actionable information relayed to the responding team.509

Rapid response to deterioration requires the immediate availability of suitably skilled staff. Cognitive aids can help these personnel deliver the required actions in intrinsically stressful situations in a more timely and complete manner.510 Nonetheless, acute illness is by definition unstable and variable. Many patients will improve for a period only to deteriorate again, meaning that surveillance of those at-risk does not end with the initial escalation event. In addition, many deteriorating patients may not have a reversible condition but may instead be suffering from terminal illness. De-escalation and referral to more palliative orientated services is not infrequently the most appropriate action.494,511 Equally, unstable patients who do have potentially reversible conditions are likely to need safe transfer to a higher level of care such as ICU or the operating theatre.

The expert consensus is that patients who require organ support or other critical care type interventions should receive such treatments in a timely fashion,512 certainly within 6 h of documented deterioration; and that all patients who trigger on whichever criteria is locally agreed should have a documented plan of the goals of care within 24 h of triggering.504 The plan should specify if, how and when a patient’s treatment should be escalated or de-escalated.

Outcomes: The impact of a functioning rapid response system can be described along the dimensions of the quadruple aim: around clinical outcomes, patient and staff satisfaction and health-economic metrics. In 2018, the International Society for rapid response systems convened a consensus conference to agree on a set of metrics that can be applied to hospitals worldwide.504 Given that the majority of cardiac arrests are known to be preceded by documented deterioration, the number of cardiac arrests is consistently found to be one of the outcomes that decreases in hospitals with a functioning rapid response system.513 It would be expected that a reduction in the number of cardiac arrests occurring 30 min or more after abnormal vital signs are recorded would be seen.

Meta-analyses and systematic reviews of the peer reviewed literature on rapid response systems suggest a measurable impact on mortality at a hospital level.514,515 Nonetheless, the only two randomised controlled trials of rapid response systems – from the UK and Australia – showed mixed results with reduced mortality in the step wedge design UK trial but no improvement in the cluster randomised controlled trial from Australia.497,516 The reasons for the difference in results are unclear, but may be linked to a difference in day-to-day engagement with ward teams. While there is a clear link between delayed initiation of life-saving treatment and mortality, this is difficult to monitor outside a research setting. Many patients who die in hospital are expected to die and the attributable mortality of acute complications is often difficult to quantify.517

Transfer to ICU is sometimes used as a surrogate indicator of timely escalation but the relationship between early detection of deterioration and ICU admission is somewhat tenuous; with early detection leading appropriately to avoidance of admission in some cases and to more admissions in other patient groups or healthcare systems. It is also often difficult to define where rapid response systems incur or save costs; prompt intervention might lead to shorter courses of acute illness in and outside ICU, but the set-up and maintenance of a system with monitoring equipment, acquisition of vital signs and the staff required to respond to escalation in a timely manner will incur costs too. Other effects of rapid response systems are documented in qualitative studies; for example, in the suggestion that the support of a rapid response system improves ward staff satisfaction and retention – as well as patient satisfaction – but this requires further investigation.518

Rapid response and peri-operative care

In the context of peri-operative care, rapid response systems have additional safety functions, especially in the critical care outreach format where a follow-up service for patients discharged from the operating theatre recovery area or ICU after major surgery is available. Critical care outreach systems allow potentially more flexible use of bed-capacity with the ability to support the ‘flex-up’ beds on general wards so as to provide care for relatively unstable patients for a limited period after surgery. This model also incorporates pro-active rounds with the ward staff with regards to patients who are causing concern but can be effectively managed with pre-emptive measures.

After major surgery or a prolonged stay in ICU, many patients struggle with complications such as critical illness neuromyopathy, anxiety, depression and posttraumatic stress disorder. The rapid response critical care outreach team can assist with the surveillance of patients’ physical and mental health after recovery from critical illness, guide goal setting and aid continuity between critical and general care, even if the most effective interventions to reduce the burden of disease after critical care are less clear.519

Discussion and conclusion

At the third international consensus conference on rapid response systems, we reviewed metrics for evaluating the function of an individual hospital’s rapid response system with the intention to provide a universally applicable model of quality assurance, independent of any particular healthcare system or size of hospital. It is arguable whether or not some of the metrics should be made public and how this would affect the engagement of clinical teams and hospital administrators. Due to the complexity of hospitals that within much larger systems that support patients, it may not be possible to compare metrics directly between hospitals – especially those from different regions or countries – but this should not preclude an open debate about outcomes informed by transparent data. Regardless of the goals of care for individual patients, timely delivery of the right care for deteriorating ward patients will reduce the impact of acute illness on patient morbidity and, importantly, suffering.

Rapid response systems do not operate in an organisational vacuum but can be used as the safety-thermometer of an organisation or indeed as the engine of patient safety across all general wards. Metrics of organisational culture including the ability of staff to correct care decisions of senior colleagues and to speak out and escalate care outside a primary care team might be the measures that give managerial teams the confidence that patients are safe. In a similar vein, standards of staff training and assurance of competence to detect and care for vulnerable patients are key to overall hospital safety and can be delivered and monitored by a good quality rapid response system. The financial impact of a deficit in the safety of a hospital could include many other indicators such as poor staff retention, costs of litigation and the broader allocation of a value to patient and staff satisfaction.

Hence, catastrophic deterioration of patients in hospital might be predictable and preventable with the whole-system approach to safety of a rapid response system.

Please visit the following link to see this article in full on the EJA website