Reduced frequency of cardiopulmonary arrests by rapid response teams Redução de paradas cardiorrespiratórias por times de resposta rápida

Conflict of interest: none. ABSTRACT Objective: To evaluate the impact of the implementation of a rapid response team on the rate of cardiorespiratory arrests in mortality associated with cardiorespiratory arrests and on in-hospital mortality in a high complexity general hospital. Methods: A retrospective analysis of cardiorespiratory arrests and in-hospital mortality events before and after implementation of a rapid response team. The period analyzed covered 19 months before intervention by the team (August 2005 to February 2007) and 19 months after the intervention (March 2007 to September 2008). Results: During the pre-intervention period, 3.54 events of cardiorespiratory arrest/1,000 discharges and 16.27 deaths/1,000 discharges were noted. After the intervention, there was a reduction in the number of cardiorespiratory arrests and in the rate of in-hospital mortality; respectively, 1.69 events of cardiorespiratory arrest/1,000 discharges (p<0.001) and 14.34 deaths/1,000 discharges (p=0.029). Conclusion: The implementation of the rapid response team may have caused a significant reduction in the number of cardiorespiratory arrests. It was estimated that during the period from March 2007 to September 2008, the intervention probably saved 67 lives.


INTRODUCTION
Success in treating cardiorespiratory arrest (CRA) depends on immediate resuscitation measures performed by trained professionals and with adequate resources (1) .The creation of specialized teams for emergency care came about with the purpose of quick and effective rescue of CRA victims.In these situations, a team of professionals (Code Blue) is activated and moves immediately to the site of the event (2,3) .
Based on this concept, in 2004 the Institute for Healthcare Improvement (IHI) recommended the implementation of rapid response teams in hospitals as part of a strategy to avoid CRAs, and consequently, to reduce in-hospital mortality.The recommendation, followed by hundreds of North-American hospitals, was part of the successful 100.000Lives Campaign: Setting a Goal and a Deadline for Improving Health Care Quality, carried out between December 2004 and June 2006 (2) .
The function of the rapid response teams, triggered by signs and symptoms of sudden and unexpected clinical worsening, is to improve the safety of hospitalized patients, providing early intervention and preventing the occurrence of CRAs (2)(3)(4) .In December 2006, a new campaign began, called The Five Million Lives Campaign, with new goals (11) .
In August 2005, Hospital Israelita Albert Einstein (HIAE) first implemented Code Blue for the treatment of adult victims of CRA.Made up by a team of two physicians (a cardiologist and a physician from intensive care or on call in the emergency room), a nurse and a physical therapist, Code Blue is set off by a single key of the telephone by any member of the treatment team (12) .
In February 2007, 19 months after the introduction of Code Blue, the rapid response team was introduced at HIAE.Called Code Yellow, this rapid response team is triggered by the nursing team via telephone when it identifies worsening in vigilance and cardiac, neurological and respiratory monitoring parameters, or when any member of the team is seriously concerned with a patient's general status (13) .Code Yellow care is provided by a physician from the Adult Intensive Care Unit (AICU), which has up to 5 minutes to give initial treatment.
Although consistent, the concept of prevention, which is the basis for implementation of rapid response teams in hospitals, is still a debatable issue (9,14) .Many studies have failed to demonstrate a reduction in in-hospital mortality, which is its main purpose (15) .

OBJECTIVE
The objective of this study was to evaluate the impact of implementation of Code Yellow on the incidence of CRAs, on in-hospital mortality, and on CRA-related mortality.

METHODS
To evaluate the impact of the implementation of Code Yellow on the total number of CRAs, in-hospital mortality, and CRA-related mortality, a retrospective study was performed using patient medical records.
HIAE is a private general hospital focused on treating highly complex diseases, which during the period of analysis had 477 beds.During this time, 309153 patients were seen in the emergency unit (monthly average of 8136) and 13593 hospital discharges were given (monthly average of 2989).
For Code Blue, the record of treatment data was made on a specific form based on the Utstein method, in accordance with guidelines of the American Heart Association (AHA) (16) and the Brazilian Society of Cardiology (SBC) (17) .For Code Yellow, treatment data recording was also made on a specific form, as per models suggested by the IHI (2) (Appendix 1).
All the cases of CRA in adult patients seen at HIAE were analyzed, encompassing 19 months before and 19 months after implementation of Code Yellow.These periods were called the pre-and post-intervention periods, respectively.The initial date of analysis coincides with the date of implantation of Code Blue, which standardized treatment and collection of data relative to adult patients victims of CRA.The treatment records of the patients admitted in CRA status to the Primary Care Unit were excluded, since these patients were already in a critical situation and would not benefit from Code Yellow.
For patients that presented with CRA while in the AICU, seriousness was assessed by means of the prognostic Acute Physiology and Chronic Health Disease Classification System II (APACHE II), verified for each patient admitted to the unit (18) .If the patient had more than one APACHE score, the one considered was that closest to the CRA event.
The outcomes evaluated were cases of CRA per 1,000 hospital discharges, CRA-related deaths per 1,000 discharges, lethality of CRA cases, all-cause mortality per 1,000 discharges, and percentage of CRA-related deaths relative to the total number of all-cause deaths.CRA-related deaths were considered as those that occurred within 24 hours of the event.Gonçales PD, Polessi JA, Bass LM, Santos GP, Yokota PK, Laselva CR, Fernandes Junior C, Cendoroglo Neto M, Estanislao M, Teich V, Sardenberg C To evaluate outcomes, the following rates were constructed monthly: -rate of CRAs per 1,000 discharges = (total number of CRAs/number of discharges excluding those from the day clinic, pediatrics and neonatology)*1,000; -rate of CRA-related deaths per 1,000 discharges = (number of deaths due to CRA within 24 hours of the event/number of discharges excluding those from the day clinic, pediatrics and neonatology)*1,000; -lethality of CRA cases = (number of deaths due to CRA within 24 hours after the event/total number of CRAs); -rate of all-cause deaths per 1,000 discharges = (total number of deaths excluding pediatrics and neonatology/number of discharges excluding those from the day clinic, pediatrics and neonatology)*1,000; -percentage of CRA-related deaths relative to the total number of all-cause deaths (number of deaths due to CRA within 24 hours after the event/total number of deaths excluding from pediatrics and neonatology).

Statistical analysis
To compare characteristics of the patients and of the preand post-intervention outcomes, exploratory analyses were performed by means of descriptive measurements.
A possible statistical significance of the difference between these parameters between the pre-and postintervention periods was assessed using Student's t test for non-paired samples, when normal distributions were found; otherwise, Mann-Whitney's non-parametric test was used.
To evaluate the behavior of rates over time, univariate regression models were adjusted, with Poisson distribution, stratified for the pre-and post-intervention periods, having as response variable the respective rates, and as explanatory variable the reference month of measurement.Based on the adjustments of the models, the trend coefficients and respective p values were obtained.All the decisions were based on a 5% significance level (19,20) .Statistical analyses were performed using IBM ® SPSS ® Statistics software.

RESULTS
During the period analyzed, 82,829 hospital discharges were recorded, 40,033 of which in the 19 months prior to implementation of Code Yellow and 42,796 during the subsequent 19 months.During this time, 214 CRAs were documented (44 outside the AICU or Stepdown Unit and 170 CRAs treated in this units), as well as 127 CRA-related deaths.
The profile of the CRA victims included in the analysis is described on table 1.
As to the characteristics of the patients that suffered CRA, no statistically significant difference was observed in the mean age pre-and post-intervention (p=0.230).The mean value of the APACHE II score was 20, during the pre-intervention period, and 26 during post-intervention (p<0.001),indicating greater seriousness and worse prognosis for CRA patients in the post-intervention period.
The rates of CRA per 1,000 discharges, CRA-related deaths per 1,000 discharges, lethality of CRA cases, rate of all-cause mortality per 1,000 discharges, and the proportion of CRA-related deaths relative to the total number of deaths, segmented for the pre-versus postintervention periods, are described on table 2.
The progression graph of the monthly rates of CRA per 1,000 discharges is shown in figure 1.The dotted line represents the mean of monthly rates during the period pre-versus post-intervention.
The implementation of the rapid response team reduced by 52% the rate of CRA per 1,000 discharges (p<0.001).Before intervention, the CRA rate showed no significant trend towards a decline, with a trend Reduced frequency of cardiopulmonary arrests einstein.2012;10(4):442-8 coefficient approximately equal to 0 (p=0.851).After intervention, this trend to decline became significant, with a trend coefficient of -0.126 (p=0.0117).
Figure 2 shows the monthly progression of the CRA-related death rate per 1,000 hospital discharges.
As to the number of CRA-related deaths, there was a 66% reduction rate with the implementation of the  Figure 3 shows the monthly progression of the allcause mortality rate per 1,000 discharges.
As to all-cause mortality, a 12% reduction was observed after the rapid response team was implemented (p=0.029).

DISCUSSION
After implementation of the rapid response team (Code Yellow), a 52% drop was noted in the number of CRAs at HIAE.This decline was observed in all units of the hospital, including Intensive Therapy and Stepdown Unit.Furthermore, a significant reduction in the number of deaths associated with this event was seen, demonstrating that the rapid response teams not only reduce the incidence of CRAs, but also save lives.
The incidence of CRAs in the hospital environment is rarely published by hospitals.It is estimated that for every 1,000 admissions, there are 1 to 5 events.Even though it is an uncommon complication during hospitalization, the rates of morbidity and mortality associated with the event are enormous, varying between 80 and 85% (21) .
Even with modern cardiopulmonary resuscitation techniques and intensive training, little has changed as to the outcomes most often associated with this event: death, prolonged hospitalization, and neurological sequelae.Studies have shown that during hospitalization in general hospitals patients present with signs and symptoms of physiological deterioration a few hours before suffering a CRA, suggesting that strategies for early identification and treatment are able to improve the outcome for these patients (22)(23)(24) .Based on this hypothesis, the implementation of the medical emergency teams (MET), also called rapid response teams (RRT), was the intervention adopted by various hospitals with the objective of reducing the incidence of CRAs, unplanned admissions to ICUs, and primarily, reducing in-hospital mortality.More than taking on care of the patient, the rapid response teams work primarily as an immediate second opinion in potentially serious cases (7,9,25,26) .In this way, the presence of professionals adequately prepared for these situations is a critical factor for success.
The controversy regarding the subject comes from the lack of multicentric randomized studies (14) .A metaanalysis conducted by Chan et al. in 2010 concluded that although they have great appeal, there is no robust evidence showing that rapid response teams are effective in reducing in-hospital mortality (15,27) .

Primary findings
This is the first study to analyze the impact of the introduction of a rapid response team in the country, and one of the longest follow-up periods described in literature.In the 19 months prior to the implementation of Code Yellow, the rate of CRA-related deaths (that occurred within 24 hours after the event) was 2.33 per 1,000 discharges.Making a projection, if the CRArelated deaths remained on the same level, 100 CRArelated deaths would be expected during the postintervention period (2.33 deaths/1,000 discharges x 42,796 discharges).As 33 deaths were reported during that period, in the 19 months after its implementation Code Yellow may have saved 67 lives.
Another important finding concerns lethality.After intervention, the CRA became a less lethal event.When the seriousness and prognosis were analyzed by means of APACHE II of the patients hospitalized in the ICU who suffered a CRA, it was noted that, even though they were more serious and had a worse prognosis, the number of CRA-related deaths in these patients was lower.

Limitations
The study data do not allow one to affirm a cause and effect relation between implementation of the rapid response team and the reduction of in-hospital mortality, since other strategies focused on patient safety were implemented at HIAE during this same period.Additionally, the benefit of implementation of the rapid response team may be associated not only with early intervention, but also with the capacity acquired by the team in rapid recognition of signs and symptoms of clinical deterioration (9) .Thus, it is likely that the continued training of the team, advances in resuscitation techniques, and greater availability of resources also contributed toward the outcome observed.

CONCLUSION
Although there may be a strong probability of reduction in the incidence of CRA, of in-hospital mortality, and of associated mortality by the implementation of a rapid response team, the limitations of the study do not allow this assertion.New studies are necessary to solve this controversy.
The pre-intervention period corresponded to the interval between implementation of Code Blue, in August 2005, until the month prior to implementation of Code Yellow, in February 2007 (19 months of follow-up).The post-intervention period began in March 2007 and continued until 19 months after the implementation of Code Yellow, in September 2008.

Table 1 .
Characteristics of cardiorespiratory arrest victims before and after implementing the Code Yellow (intervention) * Mann-Whitney non-parametric test; ** Student's t test.CRA: cardiorespiratory arrest; APACHE II: Acute Physiology and Chronic Health Disease Classification System II.