Characteristics and Outcomes for Hospitalized Patients With Recurrent Clinical Deterioration and Repeat Medical Emergency Team Activation* Henry T. Stelfox, MD, PhD, FRCPC1; Sean M. Bagshaw, MD, MSc, FRCPC2; Song Gao, MSc3
Objective: To describe the occurrence of recurrent clinical deterioration and repeat medical emergency team activation and assess its effect on processes and outcomes of care. Design: Retrospective cohort study. Setting: Two community hospitals and two tertiary care hospitals, Alberta, Canada. Patients: Consecutive hospitalized adult patients with sudden clinical deterioration and medical emergency team activation without admission to ICU. Intervention: None. Measurement and Main Results: We compared ICU admission rates (admissions > 2 hr following index medical emergency team), hospital length of stay, and hospital mortality for a cohort of 3,200 patients with and without recurrent clinical deterioration following medical emergency team activation adjusting for patient, provider, and hospital characteristics.The cohort consisted of 3,200 patients. Ten percent of patients (n = 337) experienced *See also p. 1725. 1 Departments of Critical Care Medicine, Medicine, and Community Health Sciences, Institute for Public Health, University of Calgary and Alberta Health Services, Calgary, AB, Canada. 2 Division of Critical Care Medicine, Faculty of Medicine and Dentistry, University of Alberta and Alberta Health Services, Edmonton, AB, Canada. 3 Alberta Health Services, Edmonton, AB, Canada. Dr. Stelfox and Mr. Gao had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. This study was performed at University of Calgary. Supported, in part, by operating grant (H1N-104065), Canadian Institutes of Health Research. Dr. Stelfox is supported by a New Investigator Award from the Canadian Institutes of Health Research (CIHR) and a Population Health Investigator Award from Alberta Innovates. He received support for article research from CIHR. His institution received grant support from the CIHR. Dr. Bagshaw is supported by a Canada Research Chair in Critical Care Nephrology and a Clinical Investigator Award from Alberta Innovates. Funding sources had no role in the design, conduct, or reporting of this study, and we are unaware of any conflicts of interest. Mr. Gao has disclosed that he do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0000000000000315
Critical Care Medicine
recurrent clinical d eterioration and repeat medical emergency team activation during their hospital stay. Patients more likely to experience recurrent clinical deterioration and repeat medical emergency team activation included those with chronic liver disease (odds ratio, 1.75; 95% CI, 1.14–2.69) or who received airway suctioning (odds ratio, 1.66; 95% CI, 1.23–2.25), noninvasive mechanical ventilation (odds ratio, 1.67; 95% CI, 0.94–2.94), or central IV catheter insertion (odds ratio, 1.81; 95% CI, 1.02–3.21) during the index medical emergency team activation. Patients with recurrent clinical deterioration were more likely than patients without recurrent clinical deterioration to be subsequently admitted to ICU (43% vs 13%; odds ratio, 6.11; 95% CI, 4.67–8.00; p < 0.01), to have longer lengths of hospital stay (median, 31 d vs 13 d; p < 0.01), and to die during their hospital stay (34% vs 23%; odds ratio, 1.98; 95% CI, 1.47–2.67; p < 0.01). Conclusions: Recurrent clinical deterioration and repeat medical emergency team activation are common and associated with increased risk of subsequent ICU admission, increased length of hospital stay, and increased hospital mortality. It may be possible to identify patients at risk of recurrent clinical deterioration following medical emergency team activation and target interventions to improve patient care. (Crit Care Med 2014; 42:1601–1609) Key Words: cohort studies; critical care; hospitalization; hospital rapid response team; length of stay; mortality
R
apid response systems (RRS) have been developed to improve the safety of hospital care. They are designed to identify patients with sudden clinical deterioration and activate a medical emergency team (MET) to provide rapid evaluation and treatment. The underlying principle is that early intervention in patients with sudden clinical deterioration may improve patient outcomes (1–3). Although the effectiveness of RRS is controversial (4–8), they have been promoted by prominent healthcare organizations (9, 10) and broadly implemented. As a result, each day hospitalized patients around the world with sudden clinical deterioration are identified by RRS and receive a “rapid second opinion” and brief intervention from METs (1). Most of these patients are not admitted to an ICU and remain under the care of their physician on a medical or surgical hospital unit. www.ccmjournal.org
1601
Stelfox et al
There is a relatively small body of literature describing the epidemiology of recurrent clinical deterioration following an initial MET activation (11, 12). Although these studies are an important contribution, they were conducted in single tertiary care hospitals and reported conflicting outcomes. For these reasons, we identified a multicenter cohort of hospitalized patients with sudden clinical deterioration and MET activation managed without admission to ICU, but whose goals of care designation allowed for ICU-level care (i.e., cohort of patients that received a rapid second opinion and continued with full medical care under their attending physician). Our objective was to describe the occurrence and determinants of recurrent clinical deterioration and repeat MET reactivation (i.e., patients with multiple MET a ctivations) and its effect on outcome.
MATERIALS AND METHODS Study Questions We asked four specific study questions regarding the epidemiology of patients with recurrent clinical deterioration and repeat MET activation. 1. What is the incidence of recurrent clinical deterioration and repeat MET activation? 2. Do patients with recurrent clinical deterioration and repeat MET activation receive similar healthcare resources as patients with a single MET activation? 3. Do patients with recurrent clinical deterioration and repeat MET activation have similar outcomes of care as patients with a single MET activation? 4. What patient, provider, and hospital characteristics are associated with recurrent clinical deterioration and repeat MET activation? Study Cohort We identified consecutive adult patients (excluding cardiac surgery and coronary care units as some of these units are not serviced by METs) admitted to four hospitals (two tertiary academic, two community) in Alberta, Canada, with sudden clinical deteriorations identified by the RRS and triggering a MET activation from January 1, 2007, to December 31, 2009. Patients were eligible for inclusion if they were not admitted to an ICU within 2 hours of the MET activation (local policy is to decide on patient disposition within 30 minutes of MET arrival) (13) and their goals of care designation (i.e., resuscitative care vs medical care vs comfort care) upon completion of the MET activation allowed for ICU-level care (i.e., resuscitative care). Cardiac arrest activations (separate activation) and planned MET follow-up visits (e.g., scheduled MET follow-up of patients discharged from ICU) were not included. The cohort was purposefully defined to comprise patients who had experienced a sudden clinical deterioration triggering an initial MET activation and were managed on a hospital ward and left under the care of their admitting physician with goals of care designations that allowed for ICU-level care. 1602
www.ccmjournal.org
The four study hospitals were managed by Alberta Health Services and included Foothills Medical Centre (tertiary care hospital with 1,088 hospital beds, ≈ 42,000 annual admissions, and 25 medical-surgical ICU beds), University of Alberta Hospital (tertiary care hospital with 650 beds, ≈ 30,000 annual admissions, and 30 medical-surgical ICU beds), Rockyview General Hospital (community hospital with 616 beds, ≈ 33,000 annual admissions, and 10 medical-surgical ICU beds), and Peter Lougheed Centre (community hospital with 577 beds, ≈ 30,000 annual admissions, and 18 medical-surgical ICU beds). Each hospital’s MET comprised an ICU physician (attending, fellow, resident, or physician extender), nurse, and respiratory therapist, and it provided services 24 hours per day 7 days per week. The Alberta Health Services RRS employs criteria based on respiratory status, heart rate, blood pressure, mental status changes, or provider “worried” to trigger MET activation (13). Physicians make management decisions for patients with MET activation (including patient disposition) on a case-by-case basis without a guideline, protocol, or decision-making support. Sources of Data We used data from Alberta Health Services clinical (14) and administrative databases (15, 16) that have previously been successfully used for program evaluation and research (13, 17). The MET databases capture reason for assessment, vital signs, diagnostic and therapeutic interventions, and patient disposition, with data acquired at the time of patient assessment. The ICU databases are electronic patient information systems that capture demographic, clinical, and outcome data for all patients admitted to ICU. Alberta Health Services administrative databases capture data on all hospitalized patients, including vital status at discharge, dates of admission and discharge, up to 25 International Classification of Diseases, 10th Edition (ICD-10), diagnostic codes, and up to 20 Canadian Classification of Health Interventions procedure codes. Patient, Physician, and Hospital Factors We identified patient, provider, and institutional factors that may impact processes and outcomes of care for patients with clinical deterioration and MET activation. Patient factors included demographic variables, comorbid conditions, and reason for MET activation. Presence of comorbid conditions was derived using the Deyo classification of Charlson comorbidities and validated ICD-10, coding algorithms (summarized as a single comorbidity score for multivariable analyses) (18). Physician factors included whether an ICU attending physician was present at the MET activation (13). Hospital factors included ICU occupancy (percentage of ICU beds occupied) (13), the day of the week (weekday [Monday 08:00 to Friday 17:00] vs weekend [Friday 17:01 to Monday 07:59]), and time of day (daytime [08:00 to 17:00] vs night time [17:01 to 07:59]) of the MET activation (19). Process and Outcome Measures The primary outcome was admission to ICU during the remainder of hospitalization, defined as more than 2 hours July 2014 • Volume 42 • Number 7
Feature Articles
following initial MET activation, but prior to hospital discharge (13). We examined three secondary outcome measures: 1) use of healthcare resources, evaluated by investigations and interventions performed during MET activation, 2) ICU and hospital length of stays, and 3) in-hospital mortality. Statistical Analysis The strategy for the primary analysis was to answer each of the four study questions. The unit of analysis was MET activation. We adjusted for patient, physician, and hospital covariates (all baseline variables measured at the time of the initial MET) when analyzing outcome measures. We selected a generalized estimating equations model as it is an extension of standard logistic regression which adjusts for correlation among observations (i.e., patients with more than one hospitalization with a MET activation during the study period and patients clustered within hospitals) (20, 21) and provides “population-average” effect (average response for observations sharing the same covariates) that has familiar interpretation and is more useful when estimating effects at a population level (22, 23). Statistical analyses were performed with the SAS system (SAS version 9.2; SAS Institute, Cary, NC), and a two-sided p value of less than 0.05 was considered significant. The Health Research Ethics Boards at the University of Calgary and University of Alberta approved this study and waived the need for informed consent from patients and physicians.
RESULTS
the initial activation to the second activation was 5 days (IQR, 2–17). Patient characteristics by number of MET activations were similar with a few notable exceptions (Table 1). Patients with recurrent clinical deterioration and repeat MET activation had their index MET activation later in their hospital stay (median hospital day, 8; IQR, 3–20) than patients with a single activation (median day, 5; IQR, 2–14). ICU attending physicians (compared with other ICU-based physicians) were more likely to attend the index MET activation for patients with multiple activations. Process of Care We compared processes of care during the index MET activation according to the number of activations patients experienced during their hospital stay (Table 2). Processes of care were similar for the study groups. Patients’ with recurrent clinical deterioration and repeat MET activation were more likely to have arterial blood gases (p = 0.04) and their airways suctioned (p = 0.07) during the index MET activation. Patients with a single MET activation were more likely to receive an electrocardiogram (p = 0.08). Among patients with recurrent clinical deterioration and repeat MET activation, clinical presentation and processes of care during subsequent activations were similar to the index activation with a few notable exceptions (Table 3). Compared with the index MET activation, subsequent activations were more likely to be for a respiratory reason (p = 0.01) and patients were more likely to have their airway suctioned (p = 0.04), to be intubated and mechanically ventilated (p
Objective: To describe the occurrence of recurrent clinical deterioration and repeat medical emergency team activation and assess its effect on processes and outcomes of care. Design: Retrospective cohort study. Setting: Two community hospitals and two tertiary care hospitals, Alberta, Canada. Patients: Consecutive hospitalized adult patients with sudden clinical deterioration and medical emergency team activation without admission to ICU. Intervention: None. Measurement and Main Results: We compared ICU admission rates (admissions > 2 hr following index medical emergency team), hospital length of stay, and hospital mortality for a cohort of 3,200 patients with and without recurrent clinical deterioration following medical emergency team activation adjusting for patient, provider, and hospital characteristics.The cohort consisted of 3,200 patients. Ten percent of patients (n = 337) experienced *See also p. 1725. 1 Departments of Critical Care Medicine, Medicine, and Community Health Sciences, Institute for Public Health, University of Calgary and Alberta Health Services, Calgary, AB, Canada. 2 Division of Critical Care Medicine, Faculty of Medicine and Dentistry, University of Alberta and Alberta Health Services, Edmonton, AB, Canada. 3 Alberta Health Services, Edmonton, AB, Canada. Dr. Stelfox and Mr. Gao had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. This study was performed at University of Calgary. Supported, in part, by operating grant (H1N-104065), Canadian Institutes of Health Research. Dr. Stelfox is supported by a New Investigator Award from the Canadian Institutes of Health Research (CIHR) and a Population Health Investigator Award from Alberta Innovates. He received support for article research from CIHR. His institution received grant support from the CIHR. Dr. Bagshaw is supported by a Canada Research Chair in Critical Care Nephrology and a Clinical Investigator Award from Alberta Innovates. Funding sources had no role in the design, conduct, or reporting of this study, and we are unaware of any conflicts of interest. Mr. Gao has disclosed that he do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0000000000000315
Critical Care Medicine
recurrent clinical d eterioration and repeat medical emergency team activation during their hospital stay. Patients more likely to experience recurrent clinical deterioration and repeat medical emergency team activation included those with chronic liver disease (odds ratio, 1.75; 95% CI, 1.14–2.69) or who received airway suctioning (odds ratio, 1.66; 95% CI, 1.23–2.25), noninvasive mechanical ventilation (odds ratio, 1.67; 95% CI, 0.94–2.94), or central IV catheter insertion (odds ratio, 1.81; 95% CI, 1.02–3.21) during the index medical emergency team activation. Patients with recurrent clinical deterioration were more likely than patients without recurrent clinical deterioration to be subsequently admitted to ICU (43% vs 13%; odds ratio, 6.11; 95% CI, 4.67–8.00; p < 0.01), to have longer lengths of hospital stay (median, 31 d vs 13 d; p < 0.01), and to die during their hospital stay (34% vs 23%; odds ratio, 1.98; 95% CI, 1.47–2.67; p < 0.01). Conclusions: Recurrent clinical deterioration and repeat medical emergency team activation are common and associated with increased risk of subsequent ICU admission, increased length of hospital stay, and increased hospital mortality. It may be possible to identify patients at risk of recurrent clinical deterioration following medical emergency team activation and target interventions to improve patient care. (Crit Care Med 2014; 42:1601–1609) Key Words: cohort studies; critical care; hospitalization; hospital rapid response team; length of stay; mortality
R
apid response systems (RRS) have been developed to improve the safety of hospital care. They are designed to identify patients with sudden clinical deterioration and activate a medical emergency team (MET) to provide rapid evaluation and treatment. The underlying principle is that early intervention in patients with sudden clinical deterioration may improve patient outcomes (1–3). Although the effectiveness of RRS is controversial (4–8), they have been promoted by prominent healthcare organizations (9, 10) and broadly implemented. As a result, each day hospitalized patients around the world with sudden clinical deterioration are identified by RRS and receive a “rapid second opinion” and brief intervention from METs (1). Most of these patients are not admitted to an ICU and remain under the care of their physician on a medical or surgical hospital unit. www.ccmjournal.org
1601
Stelfox et al
There is a relatively small body of literature describing the epidemiology of recurrent clinical deterioration following an initial MET activation (11, 12). Although these studies are an important contribution, they were conducted in single tertiary care hospitals and reported conflicting outcomes. For these reasons, we identified a multicenter cohort of hospitalized patients with sudden clinical deterioration and MET activation managed without admission to ICU, but whose goals of care designation allowed for ICU-level care (i.e., cohort of patients that received a rapid second opinion and continued with full medical care under their attending physician). Our objective was to describe the occurrence and determinants of recurrent clinical deterioration and repeat MET reactivation (i.e., patients with multiple MET a ctivations) and its effect on outcome.
MATERIALS AND METHODS Study Questions We asked four specific study questions regarding the epidemiology of patients with recurrent clinical deterioration and repeat MET activation. 1. What is the incidence of recurrent clinical deterioration and repeat MET activation? 2. Do patients with recurrent clinical deterioration and repeat MET activation receive similar healthcare resources as patients with a single MET activation? 3. Do patients with recurrent clinical deterioration and repeat MET activation have similar outcomes of care as patients with a single MET activation? 4. What patient, provider, and hospital characteristics are associated with recurrent clinical deterioration and repeat MET activation? Study Cohort We identified consecutive adult patients (excluding cardiac surgery and coronary care units as some of these units are not serviced by METs) admitted to four hospitals (two tertiary academic, two community) in Alberta, Canada, with sudden clinical deteriorations identified by the RRS and triggering a MET activation from January 1, 2007, to December 31, 2009. Patients were eligible for inclusion if they were not admitted to an ICU within 2 hours of the MET activation (local policy is to decide on patient disposition within 30 minutes of MET arrival) (13) and their goals of care designation (i.e., resuscitative care vs medical care vs comfort care) upon completion of the MET activation allowed for ICU-level care (i.e., resuscitative care). Cardiac arrest activations (separate activation) and planned MET follow-up visits (e.g., scheduled MET follow-up of patients discharged from ICU) were not included. The cohort was purposefully defined to comprise patients who had experienced a sudden clinical deterioration triggering an initial MET activation and were managed on a hospital ward and left under the care of their admitting physician with goals of care designations that allowed for ICU-level care. 1602
www.ccmjournal.org
The four study hospitals were managed by Alberta Health Services and included Foothills Medical Centre (tertiary care hospital with 1,088 hospital beds, ≈ 42,000 annual admissions, and 25 medical-surgical ICU beds), University of Alberta Hospital (tertiary care hospital with 650 beds, ≈ 30,000 annual admissions, and 30 medical-surgical ICU beds), Rockyview General Hospital (community hospital with 616 beds, ≈ 33,000 annual admissions, and 10 medical-surgical ICU beds), and Peter Lougheed Centre (community hospital with 577 beds, ≈ 30,000 annual admissions, and 18 medical-surgical ICU beds). Each hospital’s MET comprised an ICU physician (attending, fellow, resident, or physician extender), nurse, and respiratory therapist, and it provided services 24 hours per day 7 days per week. The Alberta Health Services RRS employs criteria based on respiratory status, heart rate, blood pressure, mental status changes, or provider “worried” to trigger MET activation (13). Physicians make management decisions for patients with MET activation (including patient disposition) on a case-by-case basis without a guideline, protocol, or decision-making support. Sources of Data We used data from Alberta Health Services clinical (14) and administrative databases (15, 16) that have previously been successfully used for program evaluation and research (13, 17). The MET databases capture reason for assessment, vital signs, diagnostic and therapeutic interventions, and patient disposition, with data acquired at the time of patient assessment. The ICU databases are electronic patient information systems that capture demographic, clinical, and outcome data for all patients admitted to ICU. Alberta Health Services administrative databases capture data on all hospitalized patients, including vital status at discharge, dates of admission and discharge, up to 25 International Classification of Diseases, 10th Edition (ICD-10), diagnostic codes, and up to 20 Canadian Classification of Health Interventions procedure codes. Patient, Physician, and Hospital Factors We identified patient, provider, and institutional factors that may impact processes and outcomes of care for patients with clinical deterioration and MET activation. Patient factors included demographic variables, comorbid conditions, and reason for MET activation. Presence of comorbid conditions was derived using the Deyo classification of Charlson comorbidities and validated ICD-10, coding algorithms (summarized as a single comorbidity score for multivariable analyses) (18). Physician factors included whether an ICU attending physician was present at the MET activation (13). Hospital factors included ICU occupancy (percentage of ICU beds occupied) (13), the day of the week (weekday [Monday 08:00 to Friday 17:00] vs weekend [Friday 17:01 to Monday 07:59]), and time of day (daytime [08:00 to 17:00] vs night time [17:01 to 07:59]) of the MET activation (19). Process and Outcome Measures The primary outcome was admission to ICU during the remainder of hospitalization, defined as more than 2 hours July 2014 • Volume 42 • Number 7
Feature Articles
following initial MET activation, but prior to hospital discharge (13). We examined three secondary outcome measures: 1) use of healthcare resources, evaluated by investigations and interventions performed during MET activation, 2) ICU and hospital length of stays, and 3) in-hospital mortality. Statistical Analysis The strategy for the primary analysis was to answer each of the four study questions. The unit of analysis was MET activation. We adjusted for patient, physician, and hospital covariates (all baseline variables measured at the time of the initial MET) when analyzing outcome measures. We selected a generalized estimating equations model as it is an extension of standard logistic regression which adjusts for correlation among observations (i.e., patients with more than one hospitalization with a MET activation during the study period and patients clustered within hospitals) (20, 21) and provides “population-average” effect (average response for observations sharing the same covariates) that has familiar interpretation and is more useful when estimating effects at a population level (22, 23). Statistical analyses were performed with the SAS system (SAS version 9.2; SAS Institute, Cary, NC), and a two-sided p value of less than 0.05 was considered significant. The Health Research Ethics Boards at the University of Calgary and University of Alberta approved this study and waived the need for informed consent from patients and physicians.
RESULTS
the initial activation to the second activation was 5 days (IQR, 2–17). Patient characteristics by number of MET activations were similar with a few notable exceptions (Table 1). Patients with recurrent clinical deterioration and repeat MET activation had their index MET activation later in their hospital stay (median hospital day, 8; IQR, 3–20) than patients with a single activation (median day, 5; IQR, 2–14). ICU attending physicians (compared with other ICU-based physicians) were more likely to attend the index MET activation for patients with multiple activations. Process of Care We compared processes of care during the index MET activation according to the number of activations patients experienced during their hospital stay (Table 2). Processes of care were similar for the study groups. Patients’ with recurrent clinical deterioration and repeat MET activation were more likely to have arterial blood gases (p = 0.04) and their airways suctioned (p = 0.07) during the index MET activation. Patients with a single MET activation were more likely to receive an electrocardiogram (p = 0.08). Among patients with recurrent clinical deterioration and repeat MET activation, clinical presentation and processes of care during subsequent activations were similar to the index activation with a few notable exceptions (Table 3). Compared with the index MET activation, subsequent activations were more likely to be for a respiratory reason (p = 0.01) and patients were more likely to have their airway suctioned (p = 0.04), to be intubated and mechanically ventilated (p
