Clinical pathway intervention compliance and effectiveness when used in the treatment of patients with severe sepsis and septic shock at an Intensive Care Unit in Spain

The purpose of this quasi-experimental study was to assess levels of compliance with the intervention bundles contained in a clinical pathway used in the treatment of patients with severe sepsis and septic shock, and to analyze the pathway's impact on survival and duration of hospital stays. We used data on 125 patients in an Intensive Care Unit, divided into a control group (N=84) and an intervention group (N=41). Levels of compliance increased from 13.1% to 29.3% in 5 resuscitation bundle interventions and from 14.3% to 22% in 3 monitoring bundle interventions. In-hospital mortality at 28 days decreased by 11.2% and the duration of hospital stay was reduced by 5 days. Although compliance was low, the intervention enhanced adherence to the instructions given in the clinical pathway and we observed a decline in mortality at 28 days and shorter hospital stays.


Introduction
Severe sepsis represents a great challenge for health care.Considerable resources have been invested in research, as well as in the development of new treatments.Until the publication of the Guidelines for the Management of Severe Sepsis and Septic Shock by the International Sepsis Forum (1) in 2001, however, many aspects of care delivery to septic patients were not based on scientific evidence.
The publication of these guidelines served as the base to start the international campaign "Surviving Sepsis" in 2004 (2) , an international program designed for awareness-raising, planning and clinical guideline dissemination purposes.In 2005, after a program review, the so-called "Severe sepsis bundles" (3) are published, a range of activities and treatments designed for teams who attend to these patients to follow the times, sequences and objectives of individual components, with the final aim of reducing mortality due to severe sepsis by 25%.There is great heterogeneity, however, in levels of compliance with these recommendations (4)(5) .
In this sense, clinical pathways, protocols seem to be proper instruments to plan and coordinate the sequence of medical, nursing and administrative procedures needed to achieve maximum efficiency in Laguna Pérez A, Chilet-Rosell E, Delgado Lacosta M, Alvarez-Dardet C, Uris Selles J, Muñoz-Mendoza CL.
the care process (6)(7) .The great quality improvement opportunities these instruments offer by standardizing care and establishing a multidisciplinary cooperation climate justified their rapid dissemination in the USA (8) and explain the fact that, in most hospitals, they are used for the most frequent processes; that contrasts with the lack of implementation in Spain (9) .As a result, we are interested in studying the management of the severe sepsis and septic shock process by comparing two models, the traditional and another based on the use of protocols and a clinical pathway developed based on the "Surviving Sepsis", with a view to assessing the degree of compliance with the intervention bundles proposed in the clinical protocol and analyze their impact on survival rates and on the duration of hospital stays.

Sample, design and study context
In this quasi-experimental study, developed at a general teaching hospital (HGU) in Spain, the application (ACCP/SCCM) (10) .
Patients who had previously been treated for severe sepsis during more than 24 hours, younger than

Patient management
A consensus existed on the definition of the diagnosis moment or "baseline" as the point when the criteria were met to start the specific intervention:   2).

Data collection and analysis
Clinical and demographic data for all patients were extracted from the Clinical Histories through two data collection forms, designed for this purpose in accordance with the Universidad de Loma Linda model (11) and the ARIAM model v1 to register severe sepsis and septic shock (12) .
The patients' characteristics were subject to descriptive analysis and differences between the control and intervention groups were compared (Student's T, Mann-Whitney's U and Chi-Square).
The standardized mortality rate was calculated as the ratio between observed mortality at 28 days and expected mortality (standardized by age and calculated in the ARIAM sepsis record based on the APACHE II score).
To determine the impact of the intervention and six

Clinical-epidemiological characteristics
The incidence of severe sepsis/septic shock at the In Table 1, it can be observed that patients in both groups were similar in terms of baseline epidemiological data, primary origin of the infection, severity at the moment of its presentation and expected mortality.In case of refractory hypotension, ‡ Median glucose levels in 24 hours, § Adequate management of activated C protein, *The definition of the indicator was included between parentheses When considering both groups, adherence to antibiotics administration in the first 120 min.was achieved in 40.9 % of the patients.The intervention did not change compliance rates (40.7% vs. 41.5%,p=0.9).When the cut-off point for antibiotics administration is limited to the first 60 minutes after the diagnosis (31.1% of patients), mortality rates among these patients (26.3%) are lower when compared with patients who receive antibiotics after more than three (50%) or more than six hours (70 %), evidencing how delayed antibiotics administration decreases survival rates (Figure 1).

Discussion
The study showed that the incidence rate of severe sepsis in the study context was in line with descriptions in Spanish epidemiological studies, with high lethality levels, prolonged hospital stays and high prevalence rates of multiple organ failure at the time of the diagnosis.Severe sepsis entails high lethality rates in the first 48 hours, which suggests delays in the diagnosis, reanimation and establishment of antibiotics therapy.Results from earlier studies range from a mild trend towards improved survival (13) to a significant drop in mortality rates (14)(15)(16) .These studies differ from the present in that most were accomplished at Emergency Services and include many patients with severe sepsis, who usually demand less aggressive reanimation and show lower mortality rates than patients in septic shock.Rev. Latino-Am.Enfermagem 2012 July-Aug.;20(4):635-43.
In our series, the protocol was predominantly developed at the ICU, were patients are in more severe conditions and at a high risk of developing nosocomial infections, which contribute to aggravate their condition (17) .

Concerning adherence to different interventions, all
results improved after the education program, similar to other research results (4,18) and lower than in other investigations (14,19) , although the latter considered less interventions.
The intervention did not change adherence to antibiotics administration within the first 120 minutes.This is probably due to the fact that baseline figures are relatively high.Our data show, however, how mortality rises when antibiotics administration is delayed.A study showed that, for each hour of delay in antibiotics administration, mortality increases by 7.6% (20) .When comparing the intervention group with the control group, Recently, the results of data accumulated between 2005 and 2008 have been published, considering 15,022 patients from 165 centers who actively participated in the monitoring of the "surviving sepsis" campaign (19) .
The conclusions were that adherence to compliance with the pathways increased over time, so that, at the end of two years of campaign, adherence to the six-hour reanimation bundle" rose linearly from 10.9% to 31.3%, while adherence with the "24-hour monitoring protocol" rose from 18.4 % to 36.1%.Non-adjusted mortality dropped from 37% to 30.8% after two years.
As observed, although increased adherence tends to be statistically significant in large series, rates continue relatively low.
Compliance with individual components is much higher, on the other hand.Although the defense of the "bundles" is based on the fact that they are more effective if administered as a whole, some experts appoint that it may not be strictly necessary to complete from the regression model one by one, the rest of the "bundle" maintained its protective effect, with statistical significance (18) .
Regarding the benefits of applying the "24-hour bundle", disagreements exist in the interpretation of trials.Various researchers have reported decreased death risks among compliers with the "bundle" in comparison with non-compliers (15,18) .Our results did not show a drop in mortality risks in any of the interventions.The results of a recently published meta-analysis showed higher survival rates with the six-hour than with the 24-hour bundle (22) .

Conclusion
The intervention was associated with greater adherence to the clinical pathway, although compliance levels remained low.After the intervention, a trend towards improved survival rates was observed, with lower 28 th -day mortality rates and shorter hospital stays.
The main study limitations are related to the sample size and selection form.Our results are significant though, and contribute to knowledge on sepsis management in clinical practice, showing that standardization, organization and local consensus are key quality components, suggesting that the clinical pathway can be a principle to put in practice the planning and continuous improvement of a highly complex care process.
of a clinical pathway is analyzed by comparing a control group with an intervention group.The control group (N=84) included all adult patients, over 18 years of age, consecutively hospitalized at the Intensive Care Unit (ICU) between June 2008 and July 2009, coming from the Emergency Service, medical-surgical hospitalization wards or the emergency surgical services, who were diagnosed with severe sepsis or septic shock, according to the ICU discharge report.The intervention group (N=41) comprised all adult patients, over 18 years of age, who were hospitalized for treatment at the ICU with a confirmed diagnosis or considerable suspicion of severe sepsis or septic shock, coming from hospitalization wards, Urgency Services and/or emergency surgical service, between October 2009 and March 2010.The diagnosis or suspicion of severe sepsis or septic shock were based on the criteria of the American College of Chest Physicians/Society of Critical Care Medicine Based on the variables related to the different treatment interventions, care and time intervals, quality criteria were constructed to reflect changes in daily practices and variability modifications (interventions made and interventions made within the preset time).The choice of the criteria was expected to comply with the following conditions: 1) Being based on a generally accepted clinical practice and supported by evidence.2) Being adapted to the context where the clinical pathway is developed.3) The achievement of each intervention can be determined by a "yes" or "no" in the clinical history review.Compliance indicators were constructed for the interventions in the six-hour reanimation bundle and the 24-hour monitoring bundle individually and as a whole.Also, a result indicator was constructed, measuring the effectiveness of compliance with the different interventions on 28-day mortality.The indicators were expected to comply with validity (identifying care quality) and specificity conditions (detecting only those cases in which the problem actually exists, no false positives) (Table ICU of the HGU corresponded to 96 cases/year, which corresponds to a calculated accumulated incidence rate of 85±27 cases of severe sepsis/septic shock per 100,00 inhabitants/year, a mean first-day SOFA score of 7.6±3.0and APACHE II: 25.8±8.1, and an expected mortality of 54.8±23.4.At the moment of the diagnosis, the mean number of organs presenting dysfunctions was 3±1 and failure of at least two organs was found in 21.9%; the most frequent diagnosis was pneumonia (32.8%).

Figure 1 -
Figure 1 -Antibiotics administration delay and mortality

Figure 2 -( 4 -
Figure 2 -Representation of death probability in patients who received intervention in the correct period (in relation to patients who did not or beyond the times established in the clinical pathway) the decreased mortality seems to be due to the lower number of patients who receive antibiotics treatment after six hours, as well as to the drop in mortality rates among patients who receive this medication in the first hour.Thus, considering the low compliance levels with Early Goal-Directed Treatment (EGDT), improved antibiotics administration could better explain the lower mortality rate detected in our series than the rest of the interventions.The findings are consistent with another study(21) , and the recommendations of the "surviving sepsis" campaign that emphasize appropriate antibiotics administration is one of the interventions in the six-hour reanimation protocol that should be done, if possible, in the first hour after the diagnosis.
all interventions to gain benefits, although survival rates do improve significantly with the number of interventions received.In another study, achieving more than four interventions from the six-hour reanimation bundle (no matter which) was associated with a lower death risk.The highest probability was obtained with six or more interventions but, when they were removed

Table 2 -
Level of compliance with the interventions in the six-hour and 24-hour intervention bundles proposed in

Table 2 -
(continuation) Percentage of patients who needed mechanical ventilation and have a mean inspiratory plateau pressure <30 cm H 2 O during the first 24 hours after presenting severe sepsis or septic shock [the median of all measures is obtained]) MBP: Mean Blood Pressure; SBP: Systolic Blood Pressure; CVP: Central Venous Pressure; ScvO 2 : Central Venous Oxygen Saturation; ICU: Intensive Care Unit †