March 29, 2016


Prompt recognition of patients with acute clinical deterioration enables appropriate medical and surgical treatments by staff with critical care skills. It is well known that delays in transfer the patients from the ward to the Intensive Care Unit (ICU) and delays in responding to the clinical signs of acute neurological, respiratory and haemodynamic impairment represent key determinants of clinical outcome. In order to improve the safety of potentially rapid deteriorating patients out of the ICU setting, many countries have introduced into the hospital practice the Rapid Response Teams (RRTs) with the aim to early identify and provide a rapid response in presence of predefined warning clinical signs and symptoms. Despite many data support the usefulness of such interventions to reduce the incidence of in-hospital cardiac arrest, to expedite ICU transfers, and to better manage the end-of-life-care, there is not consensus yet on their potential benefits in terms of decreased hospital mortality.

In light of that, the authors of this paper conducted a retrospective analysis on the implementation of an intensivist-led RRT in one hospital, compared with three other institutions belonging to the same regional healthcare centre. Three temporal cohorts were identified (the pre-RRT period [18 months], the implementation period [6 months] and the RRT period [6 months]) and unexpected mortality rate was chosen as primary endpoint. The RRT included an ICU resident, an ICU fellow/attending and, if requested, and ICU nurse. The team was provided with a fully equipped crash cart and the communication with the ward was based on the SBAR method (Situation, Background, Assessment, Recommendation). Clinical warning signs included circulatory, respiratory and consciousness parameters: the recognition of a single criterion allowed patient caregivers to directly contact the RRT using a dedicated phone number. These alarm criteria were spread trough posters displaying, simulation-based training courses, practical education sessions and local hospital newspaper.

During the study period, 564 RRT interventions were performed: median time of arrival was 5 min, main activation criteria was hypoxaemia, most common initial diagnosis was sepsis. In the interventional hospital, the unexpected mortality rate significantly decreased from 21.9 to 17.4 for 1000 discharges between the pre-RRT and post-RRT periods (p=0.002), without observing any significant variation in the other three hospital which did not adopt the RRT. Interestingly, with an average of about 18,000 admissions/year and one RRT activation per day, 1.5 lives were expected to be saved every week. Similarly mean SOFA score for ICU unplanned admissions significantly decreased from 7 to 5 (p<0.01). Finally, both overall mortality rate and cardiac arrest percentage decreased, in absence of any change in the rate of deaths with ‘do not resuscitate’ orders.

However, given the above intriguing findings, the retrospective design of the study, the inclusion of all admissions (non only the unplanned ones) in the control hospitals and the use of a single alarm criterion rather a composite scoring system, should be taken into account as main limitations of the study.

Take Home Message

RRTs may represent a useful tool to improve prompt recognition of acute deteriorating patients and decrease unexpected hospital mortality. Further prospective investigations may better clarify the ‘real life’ clinical impact of  their potential implementation on a larger scale.

This article review was submitted by Gennaro De Pascale on behalf of the NEXT Committee.


Jung B et al. Rapid response team and hospital mortality in hospitalised patients. Intensive Care Medicine; Seven-Day Profile Publication, Volume 42, Issue 4 / April, 2016; Pages 494 – 504

Comment on this news