In search of the "Holy Grail": will we ever prove the efficacy of Rapid Response Systems (RRS)?

of analysis.22 This is because the intervention is not clean. A RRS is It is hard to think of an intervention in the last two decades hat has been implemented in such a widespread fashion, and so ffectively, but with so little evidence as Rapid Response Systems.1 o one from an administrative, clinical or research perspective can uestion the underlying assumptions that underpin RRS utilizaion. A plethora of research articles in the 1990s and 2000s have escribed the risk factors for unexpected patient clinical deterioation, that eventually results in either in-hospital cardiac arrest IHCA), unplanned Intensive Care Unit (ICU) admission from the eneral ward or unexpected hospital mortality.2–10 The executive ummary of this basic epidemiological work is that such patients o not on the whole suddenly deteriorate, arrest, die, or if they re lucky get an ICU admission with an associated high mortality ate; rather they deteriorate slowly over hours and even days, with he clinical deterioration beautifully documented in the observaion chart and the written medical record. The obvious solution s to intervene in some way during this patient deterioration, ith appropriate, and timely, diagnosis, treatment and if necesary, resuscitation.6,11,12 Alternatively an earlier decision about the ppropriateness of resuscitation can be made with the patient and alliative options considered. Based on this logic, the case for the RRS is overwhelming. Howver, demonstrating this in practice has been more difficult. In the arly to mid-2000s a number of single centre historically controlled tudies, demonstrated quite significant benefit with the introducion of RRS with up to a 50% reduction IHCA.13–16 However, there ave been only two prospective randomised studies that have rawn data from more than a single hospital.17,18 There is only the luster-randomised work of the MERIT group that has extended his type of analysis to more than three sites.17 Neither of these tudies could demonstrate superiority of the RRS with respect to he traditional endpoints of IHCA arrest, unplanned ICU admission rom the ward, or unexpected hospital mortality as defined as being death that occurred with no “do not resuscitate” order in place. he MERIT study, despite being ambitious in its scope, was underowered and suffered from intervention contamination in both the ontrol and intervention hospitals. However, not deterred, by the ack of superiority of this study, the MERIT investigators performed everal post-hoc analysis of the data to demonstrate that the incience of and timing of the emergency call to a RRS were what really attered, not the randomised allocation to the RRS intervention or ontrol system.19,20