- Email: [email protected]
Impact of delayed admission to intensive care units on patients with acute respiratory failure
American Journal of Emergency Medicine xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contributions
Impact of delayed admission to intensive care units on patients with acute respiratory failure☆,☆☆,★,★★,☆☆☆ Chih-Chia Hsieh, MD a, Ching-Chi Lee, MD, MSc b,c, Hsiang-Chin Hsu, MD a, Hsin-I Shih, MD, MPH a,d, Chien-Hsin Lu, MD a, Chih-Hao Lin, MD a,⁎ a
Department of Emergency Medicine, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan, Taiwan Department of Internal Medicine, Madou Sin-Lau Hospital, Tainan, Taiwan Graduate Institute of Medical Sciences, College of Health Sciences, Chang Jung Christian University, Tainan, Taiwan d Department of Public Health, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan, Taiwan b c
a r t i c l e
i n f o
Article history: Received 27 May 2016 Received in revised form 31 August 2016 Accepted 29 September 2016 Available online xxxx
a b s t r a c t Background/Purpose: To determine the impact of delayed admission to the intensive care unit (ICU) on the clinical outcomes of patients with acute respiratory failure (ARF) in the emergency department (ED). Methods: This retrospective cohort study included non-traumatic adult patients with ARF and mechanical ventilation support in the ED of a tertiary university hospital in Taiwan from January 1, 2013, to August 31, 2013. Clinical data were extracted from chart records. The primary and secondary outcome measures were a prolonged hospital stay (N30 days) and the in-hospital crude mortality within 90 days, respectively. Results: For 267 eligible patients (age range 21.0-98.0 years, mean 70.5 ± 15.1 years; male 184, 68.9%), multivariate analysis was used to determine the significant adverse effects of an ED stay N1.0 hour on in-hospital crude mortality (odds ratio 2.19, P b .05), which was thus defined as delayed ICU admission. In-hospital mortality significantly differed between patients with delayed ICU admission and those without delayed admission, as revealed by the Kaplan-Meier survival curves (P b .05). Moreover, a linear-by-linear correlation was observed between the length of ICU waiting time in the ED and the lengths of total hospital stay (r = 0.152, P b .05), ICU stay (r = 0.148, P b .05), and ventilator support (r = 0.222, P b .05). Conclusions: For patients with ARF who required mechanical ventilation support and intensive care, a delayed ICU admission more than 1.0 hour is a strong determinant of mortality and is associated with a longer ICU stay and a longer need for ventilation. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Despite considerable advancements in respiratory support techniques that have improved the survival of patients with acute respiratory failure (ARF) over the years [1-3], ARF in critical patients is still
☆ Disclosures of conflict of interest: The authors disclose no conflict of interest. ☆☆ Ethical Adherence: The study was in accordance with the ethical standards and was approved by the institutional review board in the hospital. ★ Funding: This research was supported by National Cheng Kung University Hospital (No. NCKUH-10504011). ★★ Writing Assistance: None. ☆☆☆ Author Contributions: CC Hsieh and CH Lin conceived and supervised the study. CC Hsieh, HC Hsu, HI Shih, and CH Lu were involved in acquisition of data. CC Hsieh and CC Lee interpreted the data and performed statistical analysis. CC Hsieh and CH Lin drafted the manuscript, and all authors contributed substantially to its revision. HC Hsu, HI Shih, and CH Lu offered administrative and technical supports. CH Lin is the corresponding author who takes responsibility for the paper as a whole. ⁎ Corresponding author at: Department of Emergency Medicine, National Cheng Kung University Hospital, 70403, No.138, Shengli Rd., North District, Tainan City, Taiwan (R.O.C.). Tel.: +886 6 2353535x2237, +886932989778; fax: +886 6 2359562. E-mail address: [email protected] (C.-H. Lin).
associated with mortality rates of 40%–65% [1,4,5]. ARF that necessitates mechanical ventilator support was reported to be crucial in the development of intensive care medicine as a specialty [6]. ARF remains a common reason for admission to the intensive care unit (ICU) [7]. When managing a patient who needs ICU admission in the emergency department (ED), but there are no available ICU beds, the clinical physician often faces a dilemma of whether to transfer the patient. Transferring the patient for admission to another hospital means risking possible complications during transportation, while keeping the patient in the same ED means risking potential deterioration during the uncertain length of waiting for an ICU bed. This issue is especially important in an era in which ED overcrowding is becoming a common scene [8]. ED overcrowding interferes with the delivery of effective and timely care [9] and is associated with an increased length of ED stay and waiting time for admission [10]. The phenomena of ED overcrowding could spread from one hospital to other hospitals in a region, which increases the difficulty of transferring critical patients [11,12]. Although adverse effects of prolonged ICU waiting time were reported [13-17], the association between delayed ICU admission and the outcomes of patients with ARF was rarely validated. Regarding
http://dx.doi.org/10.1016/j.ajem.2016.09.066 0735-6757/© 2016 Elsevier Inc. All rights reserved.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
2
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
whether to transfer a critical patient, the definition of an optimal waiting time for an ICU bed to avoid adverse effects is crucial for physicians. Thus, we conducted this study to define the optimal time for ICU admission and to determine the adverse effects of delayed ICU admission in ED patients with ARF who require mechanical ventilation support. 2. Methods
organ dysfunction or hypoperfusion: metabolic acidosis, arterial hypoxemia (PaO2b 75 mmHg or PaO2/FiO2b250), oliguria (b0.03 L/h for 3 h or 0.7 L/24 h), coagulopathy (increase in prothrombin time or a drop in platelet count by 50% or to b100 × 107/L), and encephalopathy (defined as a Glasgow Coma Scale score b14) [20]. Septic shock was defined as the presence of systemic inflammatory response syndrome and a systolic blood pressure ≤ 90 mmHg after a crystalloid fluid challenge of 20 to 30 mL/kg of body weight over a 30-minute period or a blood lactate concentration ≥4 mmol/L [21].
2.1. Study design, setting and population This retrospective, observational cohort study was conducted in the ED of a tertiary university hospital in Taiwan with a capacity of 1045 general ward beds and 148 ICU beds. The institutional review board of the study hospital approved this study. The ICU in the study hospital was a closed system managed by fixed physician staff. The ED adopted a 5-level triage system (i.e., resuscitation, emergency, urgent, less urgent, and not urgent). The ED patients who experienced ARF and required ventilator support were to be admitted to the ICU as soon as possible if beds were available. Patients on mechanical ventilation remained in the ED only when ICU beds were unavailable. Subsequently, an ICU bed was booked after connecting the patient to a ventilator. 2.2. Data collection and processing All of the ED visits between January 1, 2013 and August 31, 2013 were screened according to the International Classification of Diseases, 9th Revision, Clinical Modification codes (ARF, 518.81) using a computer database. The data for eligible ED patients were retrieved from medical records by using a predetermined form that included their medical information, namely demographic data, vital signs in the ED, consciousness levels, use of mechanical ventilation, use of inotropic agents, comorbidities, laboratory findings, hospitalization period, the cause of ARF, and patient outcomes. Patients who met the following criteria were excluded: a lack of information on hospitalization or transfer before the ED visit, an unsuccessful return of spontaneous circulation for out-of-hospital and in-hospital cardiac arrest, and a lack of complete information after ED arrival (eg, patients transferred out or left against medical advice). Furthermore, to provide a clear definition of delayed ICU admission, patients who died while waiting during the ED stay were also excluded. Medical records of the eligible patients were reviewed for the aforementioned clinical information by two of the authors, who collectively resolved any discrepancy observed in the medical records. All of the patients were followed for at least 90 days after ED arrival. The primary outcome was a prolonged length of hospital stay (N30 days), and the secondary outcome was in-hospital crude mortality within 90 days. 2.3. Definitions After the ICU bed booking, the ICU waiting time was measured as the number of hours from ventilator connection in the ED until ICU admission. The length of hospital stay was the period from ED presentation to discharge from the hospital; the length of ICU stay was the period between ICU admission and hospital discharge or transfer to the general ward. ARF was defined on the basis of a PaO2/fraction of inspired oxygen (FIO2) ratiob 200 mmHg and the requirement of mechanical respiratory support, including all methods of artificial ventilation with or without an artificial airway [6,7]. As previously reported [18], the severity of acute illness was measured using the modified rapid emergency medicine score at ED arrival, and a score ≥ 8.0 indicated critical illness. Comorbidities were defined as previously reported [19]. Severe sepsis was defined as the coexistence of sepsis and at least one of the following signs or symptoms of acute
2.4. Statistical analysis Statistical analyses were performed using the Statistical Package for the Social Sciences for Windows, Version 15.0 (SPSS, Chicago, IL). Continuous variables were expressed as the means ± SDs and compared using the Student t test. Categorical variables, expressed as numbers and percentages, were compared using the chi-square or Fisher exact test. The Pearson correlation coefficient was used to measure the strength of the linear-by-linear relationship between 2 continuous variables. The primary and secondary outcomes were first analyzed using a univariate regression model, and all of the variables with a P b .1 in the univariate analysis were then incorporated into a stepwise, backward logistic regression model. The time demarcating either a delayed or not delayed admission was proposed to be the time when the effect of ICU waiting on in-hospital mortality began to emerge. To determine this time demarcation, the overall data were divided into subsets on the basis of different lengths of ICU waiting time. To compare the adverse effects of delayed ICU admission on survival rate, Kaplan–Meier plots along with a log-rank test were used. P b .05 was considered statistically significant.
2.5. Ethics The study was in accordance with the ethical standards and was approved by the institutional review board in the hospital.
3. Results 3.1. Demographics and clinical characteristics The total ED visits during the study period was 57 272 (age range 0.5-113.0 years, mean ± SD 46.0 ± 26.3 years; male 29 272, 51.1%) and the percentages of triage acuities were 2.2% resuscitation, 8.8% emergency, 52.1% urgent, 36.2% less urgent, and 0.8% not urgent. A total of 267 patients (age range 21.0-98.0 years, mean 70.5 ± 15.1 years; male 184, 68.9%) were enrolled in the study, as shown in Fig. 1. The overall mean (±SD) of the modified rapid emergency medicine score was 8.0 (±3.2). The median (interquartile range) lengths of ICU waiting time and ED stay were 5.0 (1.0-17) hours and 5.1 (1.5-17.0) hours, respectively. The leading diagnostic category of ARF was pneumonia (140 patients, 52.4%); other major categories included acute stroke (29, 10.9%), acute coronary syndrome (17, 6.4%), massive gastrointestinal bleeding (16, 6.0%), arrhythmia (10, 3.7%), and intraabdominal infection (9, 3.4%). The most common comorbidity was hypertension (145 patients, 54.3%), followed by diabetes mellitus (97, 36.3%), malignancy (76, 28.5%), chronic kidney disease (72, 27%), old stroke (48, 18.0%), chronic obstructive pulmonary disease (42, 15.7%), coronary artery disease (45, 16.9%), and liver cirrhosis (24, 9%). The median (interquartile range) lengths of total hospital stay and ICU stay were 15.0 (9.0-27.0) and 8.0 (8.0-16.0) days, respectively. The 30-day and in-hospital crude mortality rates were 21.0% (n = 56) and 26.6% (n = 71), respectively.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
3
Non-traumatic patients with acute respiratory failure who required mechanical ventilation supports in the ED n = 295
Exclusion transfered to other hospitals: 4 discharged against medical advice: 1 failed to have sustained ROSC > 2 hours after resuscitation: 6 expired during the ED stay: 17
Patients who needed ICU admission were enrolled in the study n = 267
Patients who were hospitalized more than 30 days n = 57
Patients who were discharged within 30 hospital days n = 210
Patients who expired during hospitalization n = 16
Patients who expired during hospitalization n = 55
ED = emergency department; ROSC = return of spontaneous circulation; ICU = intensive care unit. Fig. 1. Patient selection flowchart.
3.2. Clinical predictors for prolonged length of hospital stay A univariate analysis was used to compare patients with a prolonged length of hospital stay (N30 days) and those with a length of hospital stay ≤30 days. The variables for comparison included clinical characteristics, demographics, the major cause of ARF, the severity of illness, initial syndrome in the ED, major comorbidities, and the length of waiting time to ICU admission, as shown in Table 1. Only one variable, the underlying malignancy, was significantly associated with the prolonged length of hospital stay. Notably, various lengths of waiting time to ICU admission were not associated with a prolonged length of hospital stay. In the multivariate analysis (Table 1), the following 2 factors were independently associated with a prolonged length of hospital stay: comorbidities with malignancy and ARF caused by acute stroke. 3.3. Risk factors of in-hospital mortality In the univariate analysis, the following variables were compared between patients who died and those who recovered: patient demographics, the major cause of ARF, the severity of illness, initial syndrome in the ED, major comorbidities, and the length of waiting time to ICU admission (Table 2). Patients who had a critical illness (modified rapid emergency medicine score ≥8) in the ED, a length of waiting time to ICU admission N1 hour, and comorbid liver cirrhosis were frequently
associated with in-hospital mortality. In the multivariate regression (Table 2), the following variables were independently associated with patients who died during the hospital stay: critical illness (modified rapid emergency medicine score ≥ 8.0) in the ED, a length of waiting time to ICU admission N1.0 hour, and comorbid diabetes mellitus and malignancy. Thus, the multivariate analysis showed that the length of waiting time to ICU admission N1.0 hour was defined as delayed ICU admission. Further analysis of the survival curves showed a significant difference in the in-hospital crude mortality rate between patients who had a waiting time to ICU admission ≤1.0 hour and those who experienced delayed ICU admission (Fig. 2).
3.4. Correlation of ICU waiting time and length of hospital stay For all of the eligible patients, a significant linear-by-linear correlation was observed between the ICU waiting time in the ED and the lengths of total hospital stay, ICU stay, and ventilator support (Table 3). Among the various correlation coefficients between the indicated periods and the ICU waiting time, the length of ventilator support was the highest (r = 0.222, P b .05). Moreover, a strongly positive correlation was observed between the length of ventilator support and the lengths of total hospital stay (r = 0.744, P b .05) and ICU stay (r = 0.941, P b .05).
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
4
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
Table 1 Association of a prolonged hospital stay (N30 days) with clinical demographics, initial syndromes, severity, comorbidities, causes of acute respiratory failure, and the length of ED stay
Variable
Old age,≧65 years Gender, male Nursing-home residents Severity-of-illness marker in the ED Initial syndrome Severe sepsis Septic shock Modified rapid emergency medicine score ≧8 Major cause of acute respiratory failure Sepsis-related, n = 178 Pneumonia Intra-abdominal infection Urinary tract infection Non-sepsis-related, n = 89 Acute stroke Acute coronary syndrome Gastrointestinal bleeding Arrhythmia Length of waiting for ICU admission N1 hour N2 hours N4 hours Major comorbidities Hypertension Diabetes mellitus Malignancy Chronic kidney disease
Patients with prolonged hospital stay n (%)
Univariate analysis
Multivariate analysis
Yes (N = 57)
No (N = 210)
Odds ratio (95% CI)
p value
Odds ratio (95%CI)
p value
38 (66.7) 40 (70.2) 5 (8.8)
149 (71.0) 144 (68.6) 30 (14.3)
0.82 (0.44-1.53) 1.08 (0.57-2.04) 0.58 (0.21-1.56)
0.53 0.82 0.27
-
-
43 (75.4) 10 (17.5) 26 (45.6)
135 (64.3) 34 (16.2) 116 (55.2)
1.71 (0.88-3.32) 1.10 (0.51-2.39) 0.68 (0.38-1.22)
0.11 0.81 0.20
-
-
35 (61.4) 4 (7.0) 1 (1.8)
105 (50.0) 5 (2.4) 7 (3.3)
1.59 (0.88-2.89) 3.09 (0.90-11.93) 0.52 (0.06-4.30)
0.13 0.09 1.00
3.71 (0.92-14.88) -
0.06 -
10 (17.5) 1 (1.8) 3 (5.3) 0 (0)
19 (9.0) 16 (7.6) 13 (6.2) 10 (4.8)
2.14 (0.93-4.90) 0.22 (0.03-1.67) 0.84 (0.23-3.06) -
0.07 0.13 0.79 0.09
2.53 (1.07-5.98) NS
0.04 NS
46 (80.7) 39 (68.4) 33 (57.9)
154 (73.3) 130 (61.9) 106 (50.5)
1.52 (0.74-3.14) 1.33 (0.71-2.49) 1.34 (0.75-2.43)
0.26 0.37 0.32
-
-
30 (52.6) 22 (38.6) 24 (42.1) 19 (33.3)
115 (54.8) 75 (35.7) 52 (24.8) 53 (25.2)
0.92 (0.51-1.65) 1.13 (0.62-2.07) 2.21 (1.20-4.08) 1.48 (0.79-2.79)
0.78 0.69 0.01 0.22
2.63 (1.39-4.98) -
0.003 -
ED = emergency department; 95% CI = 95% confidence interval; NS = no significance after processing of the stepwise, backward multivariate regression.
4. Discussion Our study results indicated that delayed ICU admission, defined as N1.0 hour, was a strong predictor of in-hospital
crude mortality for patients with ARF and ventilator support in the ED. Our findings are consistent with previous studies that have examined the adverse effects of ICU waiting time on patient outcomes [13-16].
Table 2 Association of in-hospital mortality with clinical demographics, initial syndromes, severity, comorbidities, causes of acute respiratory failure, and the length of ED stay Patient number, n (%)
Univariate analysis
Multivariate analysis
Variable Old age,≧65 years Gender, male Nursing-home residents Severity-of-illness marker in the ED Initial syndrome Severe sepsis Septic shock Modified rapid emergency medicine score ≧8 Major cause of acute respiratory failure Sepsis-related, n = 178 Pneumonia Intra-abdominal infection Urinary tract infection Non-sepsis-related, n = 89 Acute stroke Acute coronary syndrome Gastrointestinal bleeding Arrhythmi Length of waiting for ICU admission N1 hour N2 hours N4 hours Major comorbidities Hypertension Diabetes mellitus Malignancy Chronic kidney disease Old stroke Liver cirrhosis
Death (N = 71)
Survivor (N = 196)
Odds ratio (95% CI)
p value
Odds ratio (95%CI)
p value
50 (70.4) 49 (69.0) 5 (7.0)
137 (69.9) 135 (68.9) 30 (15.3)
1.03 (0.57-1.86) 1.01 (0.56-1.81) 0.42 (0.16-1.13)
0.93 0.98 0.08
0.28 (0.10-0.80)
0.08
44 (62.0) 14 (19.7) 50 (70.4)
134 (68.4) 30 (15.3) 92 (46.9)
0.75 (0.43-1.33) 1.36 (0.67-2.74) 2.69 (1.50-4.82)
0.33 0.39 0.001
3.16 (1.70-5.87)
b0.001
32 (45.1) 5 (7.0) 1 (1.4)
108 (55.1) 4 (2.0) 7 (3.6)
0.67 (0.39-1.15) 3.64 (0.95-13.95) 0.39 (0.05-3.19)
0.15 0.06 0.36
3.62 (0.83-15.84) -
0.09 -
12 (16.9) 7 (9.9) 5 (7.0) 4 (5.6)
17 (8.7) 10 (5.1) 11 (5.6) 6 (3.1)
2.14 (0.97-4.74) 2.03 (0.74-5.57) 1.27 (0.43-3.80) 1.89 (0.52-6.91)
0.06 0.17 0.77 0.46
NS -
NS -
60 (84.5) 51 (71.8) 33 (46.5)
140 (71.4) 118 (60.2) 95 (48.5)
2.18 (1.07-4.45) 1.69 (0.99-1.31) 1.08 (0.63-1.87)
0.03 0.08 0.77
2.19 (1.04-4.64) -
0.04 -
41 (57.7) 32 (45.1) 26 (36.6) 22 (31.0) 11 (15.5) 11 (15.5)
104 (53.1) 65 (33.2) 50 (25.5) 50 (25.5) 37 (18.9) 13 (6.6)
1.21 (0.70-2.09) 1.65 (0.95-2.88) 1.69 (0.95-3.01) 1.31 (0.72-2.38) 0.79 (0.38-1.64) 2.58 (1.10-6.06)
0.50 0.07 0.08 0.37 0.53 0.03
1.94 (1.06-3.55) NS 2.87 (1.13-7.28)
0.03 NS 0.03
ED = emergency department; ICU = intensive care unit; 95% CI = 95% confidence interval; NS = no significance after processing of the stepwise, backward multivariate regression.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
Fig. 2. Kaplan-Meier survival curves of clinical outcomes in patients with delayed ICU admission (waiting time to ICU admission N1.0 hour) compared with those in patients without delayed admission.
We found that a cutoff point of 1.0 hour of ED waiting should define delayed ICU admission. Despite the adverse effects that were extensively reported in previous studies, few studies have measured the length of the ICU waiting time as a continuous variable for statistics, and thus, the cutoff time was not recognized [15,16]. Nevertheless, a recent report has attempted to answer the aforementioned question [17]. Hung et al. defined an ICU waiting time of N4 hours as a delay for adult patients with ARF in the ED [17]. We believe that the differences in the patient population (including the severity of medical illness and comorbidities, and the cause of ARF), the capacity of the ICUs and EDs, and the number of physicians and nursing staff contributed to this discrepancy in the definition of the ICU waiting period. Our study revealed a positive correlation of the ICU waiting time in the ED with the lengths of hospital stay, ICU stay, and ventilator support. Although several reports have also not detected differences in the length of ICU or hospital stay between patients with delayed ICU admission and those with immediate ICU admission [15,22], our findings were resonant with those in southern Taiwan that focused on patients with ARF [17]. From our perspective, this discrepancy may be due to the differences in the capacity and occupancy of the ICU beds and general ward beds, the health insurance system, patient characteristics, and the cause of ARF. ED overcrowding threatens patient safety because it might increase morbidity and mortality in several heterogeneous patients, particularly critical patients [13,14,23]. Critical patients require both intensive and longitudinal care. The extended stay of critical patients intensifies ED overcrowding. Avoiding delayed ICU admission is a crucial strategy to achieve desirable outcomes in critical patients in the ED. If the waiting time to ICU admission is potentially longer than 1.0 hour, then the ED physician should proactively justify a strategy to transfer the patient or not.
5
Though some researchers considered the socioeconomic status of patients was one of possible factors affecting the waiting time of ICU admission [24,25], we did not include this factor in our analysis. The public health insurance system covered almost 100% of population in Taiwan [26,27]. While the expense of treatment and hospitalization of critically ill patients are generally covered by the insurance system, the patient's socioeconomic status usually does not confound the ICU waiting time in the clinical practice in Taiwan. However, this may arouse the concern of overusing intensive care resources [28]. It is not uncommon that terminally ill patients were admitted to the ICU, which might actually decimate the resources that were needed by those medically salvageable patients. This phenomenon may worth further investigation. The nurse-to-patient ratio in the ED is generally less than that in the ICU. Nurse staffing issues are an ongoing concern, one that influences the safety of both the patient and the nurse. Ensuring adequate nurse staffing levels has been shown to reduce medical and medication errors and thus, to decrease patient complications and mortality [29,30,31]. ED overcrowding or shortage of nurses also affect ICU admission of the critical care patient from ED [32,33]. An optimal staffing model, using innovative and collaborative strategies, requires an approach that recognizes unique patient care settings to improve the quality of patient outcomes [34,35]. Health care leaders should develop strategies to timely improve nurse staffing in the ED when a patient awaits ICU admission more than 1 hour. 5. Limitations The study design had several limitations. First, to appropriately measure the time spent in the ED and to provide a clear definition of delayed ICU admission, patients who died during the ED stay and those who were transferred or who left against medical advice were excluded, which may have led to selection bias. However, only 7.5% (22/295) of the patients with ARF were excluded from our population, thus having minor effects on the present findings. Second, our study was restricted by its retrospective study design; some data could not be accessed because of information attrition. Several parameters that are not recorded in the chart records, such as ED capacity, the extent of ED crowding, the number and quality of physicians, the resources of the nursing staff, and the ICU capacity and occupancy, cannot be adjusted in the statistical model. These factors may have caused a change in our results that were obtained using the multivariate regression model. Therefore, an additional study is warranted to investigate the etiology of delayed ICU admission and whether an administrative process shortens the ICU waiting time in the ED. Finally, because of discrepancies in patient epidemiological data and hospital capacity, our definition of delayed ICU admission may not be generalizable to other hospitals. Prospective multicenter studies with adjustments for potential confounding factors are warranted. 6. Conclusions Regarding the original design and purpose of an ED setting, critical patients may experience adverse outcomes if they are not promptly admitted to the ICU. Delayed ICU admission may prolong the length of hospital stay and ventilator support; notably, this delay may result in a poor prognosis. For patients with ARF who require ventilator support, a waiting time ≤ 1.0 hour for ICU admission is recommended in the
Table 3 Correlation coefficients (p value) of the ICU waiting time in the ED and lengths of total hospital stay, ICU stay, and ventilator support
ED stay Length of total hospital stay Length of ICU stay Ventilation day
ICU waiting time (hours)
Length of total hospital stay (days)
Length of ICU stay (days)
Ventilation day (days)
1.000 (−) 0.152 (0.013) 0.148 (0.016) 0.222 (b0.001)
1.000 (−) 0.783 (b0.001) 0.744 (b0.001)
1.000 (−) 0.941 (b0.001)
1.000 (−)
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
6
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
present study. A prospective multicenter study of a large population should be conducted in the future to validate this cutoff time and thereby reduce mortality rates. Acknowledgement This research was supported by National Cheng Kung University Hospital (No. NCKUH-10504011). All of the authors declare that they have no conflict of interest. References [1] Milberg JA, Davis DR, Steinberg KP, Hudson LD. Improved survival of patients with acute respiratory distress syndrome (ARDS): 1983-1993. JAMA 1995;273:306–9. [2] Abel SJ, Finney SJ, Brett SJ, Keogh BF, Morgan CJ, Evans TW. Reduced mortality in association with the acute respiratory distress syndrome (ARDS). Thorax 1998;53: 292–4. [3] Jardin F, Fellahi JL, Beauchet A, Vieillard-Baron A, Loubieres Y, Page B. Improved prognosis of acute respiratory distress syndrome 15 years on. Intensive Care Med 1999;25:936–41. [4] Brun-Buisson C, Minelli C, Bertolini G, Brazzi L, Pimentel J, Lewandowski K, et al. Epidemiology and outcome of acute lung injury in European intensive care units. Results from the ALIVE study. Intensive Care Med 2004;30:51–61. [5] Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000; 342:1334–49. [6] Vincent JL, Thijs L, Cerny V. Critical care in Europe. Crit Care Clin 1997;13:245–54. [7] Vincent JL, Akca S, De Mendonca A, Haji-Michael P, Sprung C, Moreno R, et al. The epidemiology of acute respiratory failure in critically ill patients. Chest 2002;121:1602–9. [8] McCabe JB. Emergency department overcrowding: a national crisis. Acad Med 2001; 76:672–4. [9] Rathlev NK, Chessare J, Olshaker J, Obendorfer D, Mehta SD, Rothenhaus T, et al. Time series analysis of variables associated with daily mean emergency department length of stay. Ann Emerg Med 2007;49:265–71. [10] Asaro PV, Lewis LM, Boxerman SB. The impact of input and output factors on emergency department throughput. Acad Emerg Med 2007;14:235–42. [11] Lin CH, Kao CY, Huang CY. Managing emergency department overcrowding via ambulance diversion: a discrete event simulation model. J Formos Med Assoc 2015; 114:64–71. [12] Kao CY, Yang JC, Lin CH. The impact of ambulance and patient diversion on crowdedness of multiple emergency departments in a region. PLoS One 2015;10(12), e0144227. http://dx.doi.org/10.1371/journal.pone.0144227. 2015 eCollection. [13] Cowan RM, Trzeciak S. Clinical review: emergency department overcrowding and the potential impact on the critically ill. Crit Care 2005;9:291–5. [14] Chalfin DB, Trzeciak S, Likourezos A, Baumann BM, Dellinger RP, and DELAY-ED study group. Impact of delayed transfer of critically ill patients from the emergency department to the intensive care unit. Crit Care Med 2007;35:1477–83. [15] Cardoso LT, Grion CM, Matsuo T, Anami EH, Kauss IA, Seko L, et al. Impact of delayed admission to intensive care units on mortality of critically ill patients: a cohort study. Crit Care 2011;15:R28. http://dx.doi.org/10.1186/cc9975. [16] Servia L, Badia M, Baeza I, Montserrat N, Justes M, Cabre X, et al. Time spent in the emergency department and mortality rates in severely injured patients admitted to the intensive care unit: an observational study. J Crit Care 2012;27:58–65.
[17] Hung SC, Kung CT, Hung CW, Liu BM, Liu JW, Chew G, et al. Determining delayed admission to intensive care unit for mechanically ventilated patients in the emergency department. Crit Care 2014;18:485. http://dx.doi.org/10.1186/s13054-014-0485-1. [18] Howell MD, Donnino MW, Talmor D, Clardy P, Ngo L, Shapiro NI. Performance of severity of illness scoring systems in emergency department patients with infection. Acad Emerg Med 2007;14:709–14. [19] Schellevis FG, van der Velden J, van de Lisdonk E, van Eijk JT, van Weel C. Comorbidity of chronic diseases in general practice. J Clin Epidemiol 1993;6:469–73. [20] Brun-Buisson C, Doyon F, Carlet J, Dellamonica P, Gouin F, Lepoutre A, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults. A multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis. JAMA 1995;274:968–74. [21] Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2008. Intensive Care Med 2008;34:17–60. [22] Engoren M. The effect of prompt physician visits on intensive care unit mortality and cost. Crit Care Med 2005;33:727–32. [23] Sprivulis PC, Da Silva JA, Jacobs IG, Frazer AR, Jelinek GA. The association between hospital overcrowding and mortality among patients admitted via western Australian emergency departments. Med J Aust 2006;184:208–12. [24] Khan BA, Shakeel N, Siddiqui EU, Kazi G, Khan IQ, Khursheed M, et al. Impact of delay in admission on the outcome of critically ill patients presenting to the emergency department of a tertiary care hospital from low income country. J Pak Med Assoc 2016;66:509–16. [25] Harris S, Singer M, Rowan K, Sanderson C. Delay to admission to critical care and mortality among deteriorating ward patients in UK hospitals: a multicentre, prospective, observational cohort study. Lancet 2015;385(Suppl. 1):S40. [26] Cheng TM. Taiwan's new national health insurance program: genesis and experience so far. Health Aff (Millwood) 2003;22:61–76. [27] Chen LS, Wang YR. A conceptual framework for Taiwan's hospital clinical performance indicators. J Formos Med Assoc 2015;114:381–3. [28] Milea D, Azmi S, Reginald P, Verpillat P, Francois C. A review of accessibility of administrative healthcare databases in the Asia-Pacific region. J Mark Access Health Policy 2015;20:3. http://dx.doi.org/10.3402/jmahp.v3.28076 [eCollection 2015]. [29] McHugh MD, Rochman MF, Sloane DM, Berg RA, Mancini ME, Nadkarni VM, et al. Better nurse staffing and nurse work environments associated with increased survival of In-hospital cardiac arrest patients. Med Care 2016;54:74–80. [30] West E, Barron DN, Harrison D, Rafferty AM, Rowan K, Sanderson C. Nurse staffing, medical staffing and mortality in intensive care: an observational study. Int J Nurs Stud 2014;51:781–9. [31] Checkley W, Martin GS, Brown SM, Chang SY, Dabbagh O, Fremont RD, et al. Structure, process, and annual ICU mortality across 69 centers: United States critical illness and injury trials group critical illness outcomes study. Crit Care Med 2014;42: 344–56. [32] Clark K, Normile LB. Delays in implementing admission orders for critical care patients associated with length of stay in emergency departments in six mid-Atlantic states. J Emerg Nurs 2002;28:489–95. [33] Park SH, Blegen MA, Spetz J, Chapman SA, De Groot H. Patient turnover and the relationship between nurse staffing and patient outcomes. Res Nurs Health 2012; 35:277–88. [34] Carson SS, Stocking C, Podsadecki T, Christenson J, Pohlman A, MacRae S, et al. Effects of organizational change in the medical intensive care unit of a teaching hospital: a comparison of ‘open’ and ‘closed’ formats. JAMA 1996;276:322–8. [35] Blegen MA, Goode CJ, Spetz J, Vaughn T, Park SH. Nurse staffing effects on patient outcomes: safety-net and non-safety-net hospitals. Med Care 2011;49:406–14.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contributions
Impact of delayed admission to intensive care units on patients with acute respiratory failure☆,☆☆,★,★★,☆☆☆ Chih-Chia Hsieh, MD a, Ching-Chi Lee, MD, MSc b,c, Hsiang-Chin Hsu, MD a, Hsin-I Shih, MD, MPH a,d, Chien-Hsin Lu, MD a, Chih-Hao Lin, MD a,⁎ a
Department of Emergency Medicine, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan, Taiwan Department of Internal Medicine, Madou Sin-Lau Hospital, Tainan, Taiwan Graduate Institute of Medical Sciences, College of Health Sciences, Chang Jung Christian University, Tainan, Taiwan d Department of Public Health, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan, Taiwan b c
a r t i c l e
i n f o
Article history: Received 27 May 2016 Received in revised form 31 August 2016 Accepted 29 September 2016 Available online xxxx
a b s t r a c t Background/Purpose: To determine the impact of delayed admission to the intensive care unit (ICU) on the clinical outcomes of patients with acute respiratory failure (ARF) in the emergency department (ED). Methods: This retrospective cohort study included non-traumatic adult patients with ARF and mechanical ventilation support in the ED of a tertiary university hospital in Taiwan from January 1, 2013, to August 31, 2013. Clinical data were extracted from chart records. The primary and secondary outcome measures were a prolonged hospital stay (N30 days) and the in-hospital crude mortality within 90 days, respectively. Results: For 267 eligible patients (age range 21.0-98.0 years, mean 70.5 ± 15.1 years; male 184, 68.9%), multivariate analysis was used to determine the significant adverse effects of an ED stay N1.0 hour on in-hospital crude mortality (odds ratio 2.19, P b .05), which was thus defined as delayed ICU admission. In-hospital mortality significantly differed between patients with delayed ICU admission and those without delayed admission, as revealed by the Kaplan-Meier survival curves (P b .05). Moreover, a linear-by-linear correlation was observed between the length of ICU waiting time in the ED and the lengths of total hospital stay (r = 0.152, P b .05), ICU stay (r = 0.148, P b .05), and ventilator support (r = 0.222, P b .05). Conclusions: For patients with ARF who required mechanical ventilation support and intensive care, a delayed ICU admission more than 1.0 hour is a strong determinant of mortality and is associated with a longer ICU stay and a longer need for ventilation. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Despite considerable advancements in respiratory support techniques that have improved the survival of patients with acute respiratory failure (ARF) over the years [1-3], ARF in critical patients is still
☆ Disclosures of conflict of interest: The authors disclose no conflict of interest. ☆☆ Ethical Adherence: The study was in accordance with the ethical standards and was approved by the institutional review board in the hospital. ★ Funding: This research was supported by National Cheng Kung University Hospital (No. NCKUH-10504011). ★★ Writing Assistance: None. ☆☆☆ Author Contributions: CC Hsieh and CH Lin conceived and supervised the study. CC Hsieh, HC Hsu, HI Shih, and CH Lu were involved in acquisition of data. CC Hsieh and CC Lee interpreted the data and performed statistical analysis. CC Hsieh and CH Lin drafted the manuscript, and all authors contributed substantially to its revision. HC Hsu, HI Shih, and CH Lu offered administrative and technical supports. CH Lin is the corresponding author who takes responsibility for the paper as a whole. ⁎ Corresponding author at: Department of Emergency Medicine, National Cheng Kung University Hospital, 70403, No.138, Shengli Rd., North District, Tainan City, Taiwan (R.O.C.). Tel.: +886 6 2353535x2237, +886932989778; fax: +886 6 2359562. E-mail address: [email protected] (C.-H. Lin).
associated with mortality rates of 40%–65% [1,4,5]. ARF that necessitates mechanical ventilator support was reported to be crucial in the development of intensive care medicine as a specialty [6]. ARF remains a common reason for admission to the intensive care unit (ICU) [7]. When managing a patient who needs ICU admission in the emergency department (ED), but there are no available ICU beds, the clinical physician often faces a dilemma of whether to transfer the patient. Transferring the patient for admission to another hospital means risking possible complications during transportation, while keeping the patient in the same ED means risking potential deterioration during the uncertain length of waiting for an ICU bed. This issue is especially important in an era in which ED overcrowding is becoming a common scene [8]. ED overcrowding interferes with the delivery of effective and timely care [9] and is associated with an increased length of ED stay and waiting time for admission [10]. The phenomena of ED overcrowding could spread from one hospital to other hospitals in a region, which increases the difficulty of transferring critical patients [11,12]. Although adverse effects of prolonged ICU waiting time were reported [13-17], the association between delayed ICU admission and the outcomes of patients with ARF was rarely validated. Regarding
http://dx.doi.org/10.1016/j.ajem.2016.09.066 0735-6757/© 2016 Elsevier Inc. All rights reserved.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
2
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
whether to transfer a critical patient, the definition of an optimal waiting time for an ICU bed to avoid adverse effects is crucial for physicians. Thus, we conducted this study to define the optimal time for ICU admission and to determine the adverse effects of delayed ICU admission in ED patients with ARF who require mechanical ventilation support. 2. Methods
organ dysfunction or hypoperfusion: metabolic acidosis, arterial hypoxemia (PaO2b 75 mmHg or PaO2/FiO2b250), oliguria (b0.03 L/h for 3 h or 0.7 L/24 h), coagulopathy (increase in prothrombin time or a drop in platelet count by 50% or to b100 × 107/L), and encephalopathy (defined as a Glasgow Coma Scale score b14) [20]. Septic shock was defined as the presence of systemic inflammatory response syndrome and a systolic blood pressure ≤ 90 mmHg after a crystalloid fluid challenge of 20 to 30 mL/kg of body weight over a 30-minute period or a blood lactate concentration ≥4 mmol/L [21].
2.1. Study design, setting and population This retrospective, observational cohort study was conducted in the ED of a tertiary university hospital in Taiwan with a capacity of 1045 general ward beds and 148 ICU beds. The institutional review board of the study hospital approved this study. The ICU in the study hospital was a closed system managed by fixed physician staff. The ED adopted a 5-level triage system (i.e., resuscitation, emergency, urgent, less urgent, and not urgent). The ED patients who experienced ARF and required ventilator support were to be admitted to the ICU as soon as possible if beds were available. Patients on mechanical ventilation remained in the ED only when ICU beds were unavailable. Subsequently, an ICU bed was booked after connecting the patient to a ventilator. 2.2. Data collection and processing All of the ED visits between January 1, 2013 and August 31, 2013 were screened according to the International Classification of Diseases, 9th Revision, Clinical Modification codes (ARF, 518.81) using a computer database. The data for eligible ED patients were retrieved from medical records by using a predetermined form that included their medical information, namely demographic data, vital signs in the ED, consciousness levels, use of mechanical ventilation, use of inotropic agents, comorbidities, laboratory findings, hospitalization period, the cause of ARF, and patient outcomes. Patients who met the following criteria were excluded: a lack of information on hospitalization or transfer before the ED visit, an unsuccessful return of spontaneous circulation for out-of-hospital and in-hospital cardiac arrest, and a lack of complete information after ED arrival (eg, patients transferred out or left against medical advice). Furthermore, to provide a clear definition of delayed ICU admission, patients who died while waiting during the ED stay were also excluded. Medical records of the eligible patients were reviewed for the aforementioned clinical information by two of the authors, who collectively resolved any discrepancy observed in the medical records. All of the patients were followed for at least 90 days after ED arrival. The primary outcome was a prolonged length of hospital stay (N30 days), and the secondary outcome was in-hospital crude mortality within 90 days. 2.3. Definitions After the ICU bed booking, the ICU waiting time was measured as the number of hours from ventilator connection in the ED until ICU admission. The length of hospital stay was the period from ED presentation to discharge from the hospital; the length of ICU stay was the period between ICU admission and hospital discharge or transfer to the general ward. ARF was defined on the basis of a PaO2/fraction of inspired oxygen (FIO2) ratiob 200 mmHg and the requirement of mechanical respiratory support, including all methods of artificial ventilation with or without an artificial airway [6,7]. As previously reported [18], the severity of acute illness was measured using the modified rapid emergency medicine score at ED arrival, and a score ≥ 8.0 indicated critical illness. Comorbidities were defined as previously reported [19]. Severe sepsis was defined as the coexistence of sepsis and at least one of the following signs or symptoms of acute
2.4. Statistical analysis Statistical analyses were performed using the Statistical Package for the Social Sciences for Windows, Version 15.0 (SPSS, Chicago, IL). Continuous variables were expressed as the means ± SDs and compared using the Student t test. Categorical variables, expressed as numbers and percentages, were compared using the chi-square or Fisher exact test. The Pearson correlation coefficient was used to measure the strength of the linear-by-linear relationship between 2 continuous variables. The primary and secondary outcomes were first analyzed using a univariate regression model, and all of the variables with a P b .1 in the univariate analysis were then incorporated into a stepwise, backward logistic regression model. The time demarcating either a delayed or not delayed admission was proposed to be the time when the effect of ICU waiting on in-hospital mortality began to emerge. To determine this time demarcation, the overall data were divided into subsets on the basis of different lengths of ICU waiting time. To compare the adverse effects of delayed ICU admission on survival rate, Kaplan–Meier plots along with a log-rank test were used. P b .05 was considered statistically significant.
2.5. Ethics The study was in accordance with the ethical standards and was approved by the institutional review board in the hospital.
3. Results 3.1. Demographics and clinical characteristics The total ED visits during the study period was 57 272 (age range 0.5-113.0 years, mean ± SD 46.0 ± 26.3 years; male 29 272, 51.1%) and the percentages of triage acuities were 2.2% resuscitation, 8.8% emergency, 52.1% urgent, 36.2% less urgent, and 0.8% not urgent. A total of 267 patients (age range 21.0-98.0 years, mean 70.5 ± 15.1 years; male 184, 68.9%) were enrolled in the study, as shown in Fig. 1. The overall mean (±SD) of the modified rapid emergency medicine score was 8.0 (±3.2). The median (interquartile range) lengths of ICU waiting time and ED stay were 5.0 (1.0-17) hours and 5.1 (1.5-17.0) hours, respectively. The leading diagnostic category of ARF was pneumonia (140 patients, 52.4%); other major categories included acute stroke (29, 10.9%), acute coronary syndrome (17, 6.4%), massive gastrointestinal bleeding (16, 6.0%), arrhythmia (10, 3.7%), and intraabdominal infection (9, 3.4%). The most common comorbidity was hypertension (145 patients, 54.3%), followed by diabetes mellitus (97, 36.3%), malignancy (76, 28.5%), chronic kidney disease (72, 27%), old stroke (48, 18.0%), chronic obstructive pulmonary disease (42, 15.7%), coronary artery disease (45, 16.9%), and liver cirrhosis (24, 9%). The median (interquartile range) lengths of total hospital stay and ICU stay were 15.0 (9.0-27.0) and 8.0 (8.0-16.0) days, respectively. The 30-day and in-hospital crude mortality rates were 21.0% (n = 56) and 26.6% (n = 71), respectively.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
3
Non-traumatic patients with acute respiratory failure who required mechanical ventilation supports in the ED n = 295
Exclusion transfered to other hospitals: 4 discharged against medical advice: 1 failed to have sustained ROSC > 2 hours after resuscitation: 6 expired during the ED stay: 17
Patients who needed ICU admission were enrolled in the study n = 267
Patients who were hospitalized more than 30 days n = 57
Patients who were discharged within 30 hospital days n = 210
Patients who expired during hospitalization n = 16
Patients who expired during hospitalization n = 55
ED = emergency department; ROSC = return of spontaneous circulation; ICU = intensive care unit. Fig. 1. Patient selection flowchart.
3.2. Clinical predictors for prolonged length of hospital stay A univariate analysis was used to compare patients with a prolonged length of hospital stay (N30 days) and those with a length of hospital stay ≤30 days. The variables for comparison included clinical characteristics, demographics, the major cause of ARF, the severity of illness, initial syndrome in the ED, major comorbidities, and the length of waiting time to ICU admission, as shown in Table 1. Only one variable, the underlying malignancy, was significantly associated with the prolonged length of hospital stay. Notably, various lengths of waiting time to ICU admission were not associated with a prolonged length of hospital stay. In the multivariate analysis (Table 1), the following 2 factors were independently associated with a prolonged length of hospital stay: comorbidities with malignancy and ARF caused by acute stroke. 3.3. Risk factors of in-hospital mortality In the univariate analysis, the following variables were compared between patients who died and those who recovered: patient demographics, the major cause of ARF, the severity of illness, initial syndrome in the ED, major comorbidities, and the length of waiting time to ICU admission (Table 2). Patients who had a critical illness (modified rapid emergency medicine score ≥8) in the ED, a length of waiting time to ICU admission N1 hour, and comorbid liver cirrhosis were frequently
associated with in-hospital mortality. In the multivariate regression (Table 2), the following variables were independently associated with patients who died during the hospital stay: critical illness (modified rapid emergency medicine score ≥ 8.0) in the ED, a length of waiting time to ICU admission N1.0 hour, and comorbid diabetes mellitus and malignancy. Thus, the multivariate analysis showed that the length of waiting time to ICU admission N1.0 hour was defined as delayed ICU admission. Further analysis of the survival curves showed a significant difference in the in-hospital crude mortality rate between patients who had a waiting time to ICU admission ≤1.0 hour and those who experienced delayed ICU admission (Fig. 2).
3.4. Correlation of ICU waiting time and length of hospital stay For all of the eligible patients, a significant linear-by-linear correlation was observed between the ICU waiting time in the ED and the lengths of total hospital stay, ICU stay, and ventilator support (Table 3). Among the various correlation coefficients between the indicated periods and the ICU waiting time, the length of ventilator support was the highest (r = 0.222, P b .05). Moreover, a strongly positive correlation was observed between the length of ventilator support and the lengths of total hospital stay (r = 0.744, P b .05) and ICU stay (r = 0.941, P b .05).
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
4
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
Table 1 Association of a prolonged hospital stay (N30 days) with clinical demographics, initial syndromes, severity, comorbidities, causes of acute respiratory failure, and the length of ED stay
Variable
Old age,≧65 years Gender, male Nursing-home residents Severity-of-illness marker in the ED Initial syndrome Severe sepsis Septic shock Modified rapid emergency medicine score ≧8 Major cause of acute respiratory failure Sepsis-related, n = 178 Pneumonia Intra-abdominal infection Urinary tract infection Non-sepsis-related, n = 89 Acute stroke Acute coronary syndrome Gastrointestinal bleeding Arrhythmia Length of waiting for ICU admission N1 hour N2 hours N4 hours Major comorbidities Hypertension Diabetes mellitus Malignancy Chronic kidney disease
Patients with prolonged hospital stay n (%)
Univariate analysis
Multivariate analysis
Yes (N = 57)
No (N = 210)
Odds ratio (95% CI)
p value
Odds ratio (95%CI)
p value
38 (66.7) 40 (70.2) 5 (8.8)
149 (71.0) 144 (68.6) 30 (14.3)
0.82 (0.44-1.53) 1.08 (0.57-2.04) 0.58 (0.21-1.56)
0.53 0.82 0.27
-
-
43 (75.4) 10 (17.5) 26 (45.6)
135 (64.3) 34 (16.2) 116 (55.2)
1.71 (0.88-3.32) 1.10 (0.51-2.39) 0.68 (0.38-1.22)
0.11 0.81 0.20
-
-
35 (61.4) 4 (7.0) 1 (1.8)
105 (50.0) 5 (2.4) 7 (3.3)
1.59 (0.88-2.89) 3.09 (0.90-11.93) 0.52 (0.06-4.30)
0.13 0.09 1.00
3.71 (0.92-14.88) -
0.06 -
10 (17.5) 1 (1.8) 3 (5.3) 0 (0)
19 (9.0) 16 (7.6) 13 (6.2) 10 (4.8)
2.14 (0.93-4.90) 0.22 (0.03-1.67) 0.84 (0.23-3.06) -
0.07 0.13 0.79 0.09
2.53 (1.07-5.98) NS
0.04 NS
46 (80.7) 39 (68.4) 33 (57.9)
154 (73.3) 130 (61.9) 106 (50.5)
1.52 (0.74-3.14) 1.33 (0.71-2.49) 1.34 (0.75-2.43)
0.26 0.37 0.32
-
-
30 (52.6) 22 (38.6) 24 (42.1) 19 (33.3)
115 (54.8) 75 (35.7) 52 (24.8) 53 (25.2)
0.92 (0.51-1.65) 1.13 (0.62-2.07) 2.21 (1.20-4.08) 1.48 (0.79-2.79)
0.78 0.69 0.01 0.22
2.63 (1.39-4.98) -
0.003 -
ED = emergency department; 95% CI = 95% confidence interval; NS = no significance after processing of the stepwise, backward multivariate regression.
4. Discussion Our study results indicated that delayed ICU admission, defined as N1.0 hour, was a strong predictor of in-hospital
crude mortality for patients with ARF and ventilator support in the ED. Our findings are consistent with previous studies that have examined the adverse effects of ICU waiting time on patient outcomes [13-16].
Table 2 Association of in-hospital mortality with clinical demographics, initial syndromes, severity, comorbidities, causes of acute respiratory failure, and the length of ED stay Patient number, n (%)
Univariate analysis
Multivariate analysis
Variable Old age,≧65 years Gender, male Nursing-home residents Severity-of-illness marker in the ED Initial syndrome Severe sepsis Septic shock Modified rapid emergency medicine score ≧8 Major cause of acute respiratory failure Sepsis-related, n = 178 Pneumonia Intra-abdominal infection Urinary tract infection Non-sepsis-related, n = 89 Acute stroke Acute coronary syndrome Gastrointestinal bleeding Arrhythmi Length of waiting for ICU admission N1 hour N2 hours N4 hours Major comorbidities Hypertension Diabetes mellitus Malignancy Chronic kidney disease Old stroke Liver cirrhosis
Death (N = 71)
Survivor (N = 196)
Odds ratio (95% CI)
p value
Odds ratio (95%CI)
p value
50 (70.4) 49 (69.0) 5 (7.0)
137 (69.9) 135 (68.9) 30 (15.3)
1.03 (0.57-1.86) 1.01 (0.56-1.81) 0.42 (0.16-1.13)
0.93 0.98 0.08
0.28 (0.10-0.80)
0.08
44 (62.0) 14 (19.7) 50 (70.4)
134 (68.4) 30 (15.3) 92 (46.9)
0.75 (0.43-1.33) 1.36 (0.67-2.74) 2.69 (1.50-4.82)
0.33 0.39 0.001
3.16 (1.70-5.87)
b0.001
32 (45.1) 5 (7.0) 1 (1.4)
108 (55.1) 4 (2.0) 7 (3.6)
0.67 (0.39-1.15) 3.64 (0.95-13.95) 0.39 (0.05-3.19)
0.15 0.06 0.36
3.62 (0.83-15.84) -
0.09 -
12 (16.9) 7 (9.9) 5 (7.0) 4 (5.6)
17 (8.7) 10 (5.1) 11 (5.6) 6 (3.1)
2.14 (0.97-4.74) 2.03 (0.74-5.57) 1.27 (0.43-3.80) 1.89 (0.52-6.91)
0.06 0.17 0.77 0.46
NS -
NS -
60 (84.5) 51 (71.8) 33 (46.5)
140 (71.4) 118 (60.2) 95 (48.5)
2.18 (1.07-4.45) 1.69 (0.99-1.31) 1.08 (0.63-1.87)
0.03 0.08 0.77
2.19 (1.04-4.64) -
0.04 -
41 (57.7) 32 (45.1) 26 (36.6) 22 (31.0) 11 (15.5) 11 (15.5)
104 (53.1) 65 (33.2) 50 (25.5) 50 (25.5) 37 (18.9) 13 (6.6)
1.21 (0.70-2.09) 1.65 (0.95-2.88) 1.69 (0.95-3.01) 1.31 (0.72-2.38) 0.79 (0.38-1.64) 2.58 (1.10-6.06)
0.50 0.07 0.08 0.37 0.53 0.03
1.94 (1.06-3.55) NS 2.87 (1.13-7.28)
0.03 NS 0.03
ED = emergency department; ICU = intensive care unit; 95% CI = 95% confidence interval; NS = no significance after processing of the stepwise, backward multivariate regression.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
Fig. 2. Kaplan-Meier survival curves of clinical outcomes in patients with delayed ICU admission (waiting time to ICU admission N1.0 hour) compared with those in patients without delayed admission.
We found that a cutoff point of 1.0 hour of ED waiting should define delayed ICU admission. Despite the adverse effects that were extensively reported in previous studies, few studies have measured the length of the ICU waiting time as a continuous variable for statistics, and thus, the cutoff time was not recognized [15,16]. Nevertheless, a recent report has attempted to answer the aforementioned question [17]. Hung et al. defined an ICU waiting time of N4 hours as a delay for adult patients with ARF in the ED [17]. We believe that the differences in the patient population (including the severity of medical illness and comorbidities, and the cause of ARF), the capacity of the ICUs and EDs, and the number of physicians and nursing staff contributed to this discrepancy in the definition of the ICU waiting period. Our study revealed a positive correlation of the ICU waiting time in the ED with the lengths of hospital stay, ICU stay, and ventilator support. Although several reports have also not detected differences in the length of ICU or hospital stay between patients with delayed ICU admission and those with immediate ICU admission [15,22], our findings were resonant with those in southern Taiwan that focused on patients with ARF [17]. From our perspective, this discrepancy may be due to the differences in the capacity and occupancy of the ICU beds and general ward beds, the health insurance system, patient characteristics, and the cause of ARF. ED overcrowding threatens patient safety because it might increase morbidity and mortality in several heterogeneous patients, particularly critical patients [13,14,23]. Critical patients require both intensive and longitudinal care. The extended stay of critical patients intensifies ED overcrowding. Avoiding delayed ICU admission is a crucial strategy to achieve desirable outcomes in critical patients in the ED. If the waiting time to ICU admission is potentially longer than 1.0 hour, then the ED physician should proactively justify a strategy to transfer the patient or not.
5
Though some researchers considered the socioeconomic status of patients was one of possible factors affecting the waiting time of ICU admission [24,25], we did not include this factor in our analysis. The public health insurance system covered almost 100% of population in Taiwan [26,27]. While the expense of treatment and hospitalization of critically ill patients are generally covered by the insurance system, the patient's socioeconomic status usually does not confound the ICU waiting time in the clinical practice in Taiwan. However, this may arouse the concern of overusing intensive care resources [28]. It is not uncommon that terminally ill patients were admitted to the ICU, which might actually decimate the resources that were needed by those medically salvageable patients. This phenomenon may worth further investigation. The nurse-to-patient ratio in the ED is generally less than that in the ICU. Nurse staffing issues are an ongoing concern, one that influences the safety of both the patient and the nurse. Ensuring adequate nurse staffing levels has been shown to reduce medical and medication errors and thus, to decrease patient complications and mortality [29,30,31]. ED overcrowding or shortage of nurses also affect ICU admission of the critical care patient from ED [32,33]. An optimal staffing model, using innovative and collaborative strategies, requires an approach that recognizes unique patient care settings to improve the quality of patient outcomes [34,35]. Health care leaders should develop strategies to timely improve nurse staffing in the ED when a patient awaits ICU admission more than 1 hour. 5. Limitations The study design had several limitations. First, to appropriately measure the time spent in the ED and to provide a clear definition of delayed ICU admission, patients who died during the ED stay and those who were transferred or who left against medical advice were excluded, which may have led to selection bias. However, only 7.5% (22/295) of the patients with ARF were excluded from our population, thus having minor effects on the present findings. Second, our study was restricted by its retrospective study design; some data could not be accessed because of information attrition. Several parameters that are not recorded in the chart records, such as ED capacity, the extent of ED crowding, the number and quality of physicians, the resources of the nursing staff, and the ICU capacity and occupancy, cannot be adjusted in the statistical model. These factors may have caused a change in our results that were obtained using the multivariate regression model. Therefore, an additional study is warranted to investigate the etiology of delayed ICU admission and whether an administrative process shortens the ICU waiting time in the ED. Finally, because of discrepancies in patient epidemiological data and hospital capacity, our definition of delayed ICU admission may not be generalizable to other hospitals. Prospective multicenter studies with adjustments for potential confounding factors are warranted. 6. Conclusions Regarding the original design and purpose of an ED setting, critical patients may experience adverse outcomes if they are not promptly admitted to the ICU. Delayed ICU admission may prolong the length of hospital stay and ventilator support; notably, this delay may result in a poor prognosis. For patients with ARF who require ventilator support, a waiting time ≤ 1.0 hour for ICU admission is recommended in the
Table 3 Correlation coefficients (p value) of the ICU waiting time in the ED and lengths of total hospital stay, ICU stay, and ventilator support
ED stay Length of total hospital stay Length of ICU stay Ventilation day
ICU waiting time (hours)
Length of total hospital stay (days)
Length of ICU stay (days)
Ventilation day (days)
1.000 (−) 0.152 (0.013) 0.148 (0.016) 0.222 (b0.001)
1.000 (−) 0.783 (b0.001) 0.744 (b0.001)
1.000 (−) 0.941 (b0.001)
1.000 (−)
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066
6
C-C. Hsieh et al. / American Journal of Emergency Medicine xxx (2016) xxx–xxx
present study. A prospective multicenter study of a large population should be conducted in the future to validate this cutoff time and thereby reduce mortality rates. Acknowledgement This research was supported by National Cheng Kung University Hospital (No. NCKUH-10504011). All of the authors declare that they have no conflict of interest. References [1] Milberg JA, Davis DR, Steinberg KP, Hudson LD. Improved survival of patients with acute respiratory distress syndrome (ARDS): 1983-1993. JAMA 1995;273:306–9. [2] Abel SJ, Finney SJ, Brett SJ, Keogh BF, Morgan CJ, Evans TW. Reduced mortality in association with the acute respiratory distress syndrome (ARDS). Thorax 1998;53: 292–4. [3] Jardin F, Fellahi JL, Beauchet A, Vieillard-Baron A, Loubieres Y, Page B. Improved prognosis of acute respiratory distress syndrome 15 years on. Intensive Care Med 1999;25:936–41. [4] Brun-Buisson C, Minelli C, Bertolini G, Brazzi L, Pimentel J, Lewandowski K, et al. Epidemiology and outcome of acute lung injury in European intensive care units. Results from the ALIVE study. Intensive Care Med 2004;30:51–61. [5] Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000; 342:1334–49. [6] Vincent JL, Thijs L, Cerny V. Critical care in Europe. Crit Care Clin 1997;13:245–54. [7] Vincent JL, Akca S, De Mendonca A, Haji-Michael P, Sprung C, Moreno R, et al. The epidemiology of acute respiratory failure in critically ill patients. Chest 2002;121:1602–9. [8] McCabe JB. Emergency department overcrowding: a national crisis. Acad Med 2001; 76:672–4. [9] Rathlev NK, Chessare J, Olshaker J, Obendorfer D, Mehta SD, Rothenhaus T, et al. Time series analysis of variables associated with daily mean emergency department length of stay. Ann Emerg Med 2007;49:265–71. [10] Asaro PV, Lewis LM, Boxerman SB. The impact of input and output factors on emergency department throughput. Acad Emerg Med 2007;14:235–42. [11] Lin CH, Kao CY, Huang CY. Managing emergency department overcrowding via ambulance diversion: a discrete event simulation model. J Formos Med Assoc 2015; 114:64–71. [12] Kao CY, Yang JC, Lin CH. The impact of ambulance and patient diversion on crowdedness of multiple emergency departments in a region. PLoS One 2015;10(12), e0144227. http://dx.doi.org/10.1371/journal.pone.0144227. 2015 eCollection. [13] Cowan RM, Trzeciak S. Clinical review: emergency department overcrowding and the potential impact on the critically ill. Crit Care 2005;9:291–5. [14] Chalfin DB, Trzeciak S, Likourezos A, Baumann BM, Dellinger RP, and DELAY-ED study group. Impact of delayed transfer of critically ill patients from the emergency department to the intensive care unit. Crit Care Med 2007;35:1477–83. [15] Cardoso LT, Grion CM, Matsuo T, Anami EH, Kauss IA, Seko L, et al. Impact of delayed admission to intensive care units on mortality of critically ill patients: a cohort study. Crit Care 2011;15:R28. http://dx.doi.org/10.1186/cc9975. [16] Servia L, Badia M, Baeza I, Montserrat N, Justes M, Cabre X, et al. Time spent in the emergency department and mortality rates in severely injured patients admitted to the intensive care unit: an observational study. J Crit Care 2012;27:58–65.
[17] Hung SC, Kung CT, Hung CW, Liu BM, Liu JW, Chew G, et al. Determining delayed admission to intensive care unit for mechanically ventilated patients in the emergency department. Crit Care 2014;18:485. http://dx.doi.org/10.1186/s13054-014-0485-1. [18] Howell MD, Donnino MW, Talmor D, Clardy P, Ngo L, Shapiro NI. Performance of severity of illness scoring systems in emergency department patients with infection. Acad Emerg Med 2007;14:709–14. [19] Schellevis FG, van der Velden J, van de Lisdonk E, van Eijk JT, van Weel C. Comorbidity of chronic diseases in general practice. J Clin Epidemiol 1993;6:469–73. [20] Brun-Buisson C, Doyon F, Carlet J, Dellamonica P, Gouin F, Lepoutre A, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults. A multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis. JAMA 1995;274:968–74. [21] Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2008. Intensive Care Med 2008;34:17–60. [22] Engoren M. The effect of prompt physician visits on intensive care unit mortality and cost. Crit Care Med 2005;33:727–32. [23] Sprivulis PC, Da Silva JA, Jacobs IG, Frazer AR, Jelinek GA. The association between hospital overcrowding and mortality among patients admitted via western Australian emergency departments. Med J Aust 2006;184:208–12. [24] Khan BA, Shakeel N, Siddiqui EU, Kazi G, Khan IQ, Khursheed M, et al. Impact of delay in admission on the outcome of critically ill patients presenting to the emergency department of a tertiary care hospital from low income country. J Pak Med Assoc 2016;66:509–16. [25] Harris S, Singer M, Rowan K, Sanderson C. Delay to admission to critical care and mortality among deteriorating ward patients in UK hospitals: a multicentre, prospective, observational cohort study. Lancet 2015;385(Suppl. 1):S40. [26] Cheng TM. Taiwan's new national health insurance program: genesis and experience so far. Health Aff (Millwood) 2003;22:61–76. [27] Chen LS, Wang YR. A conceptual framework for Taiwan's hospital clinical performance indicators. J Formos Med Assoc 2015;114:381–3. [28] Milea D, Azmi S, Reginald P, Verpillat P, Francois C. A review of accessibility of administrative healthcare databases in the Asia-Pacific region. J Mark Access Health Policy 2015;20:3. http://dx.doi.org/10.3402/jmahp.v3.28076 [eCollection 2015]. [29] McHugh MD, Rochman MF, Sloane DM, Berg RA, Mancini ME, Nadkarni VM, et al. Better nurse staffing and nurse work environments associated with increased survival of In-hospital cardiac arrest patients. Med Care 2016;54:74–80. [30] West E, Barron DN, Harrison D, Rafferty AM, Rowan K, Sanderson C. Nurse staffing, medical staffing and mortality in intensive care: an observational study. Int J Nurs Stud 2014;51:781–9. [31] Checkley W, Martin GS, Brown SM, Chang SY, Dabbagh O, Fremont RD, et al. Structure, process, and annual ICU mortality across 69 centers: United States critical illness and injury trials group critical illness outcomes study. Crit Care Med 2014;42: 344–56. [32] Clark K, Normile LB. Delays in implementing admission orders for critical care patients associated with length of stay in emergency departments in six mid-Atlantic states. J Emerg Nurs 2002;28:489–95. [33] Park SH, Blegen MA, Spetz J, Chapman SA, De Groot H. Patient turnover and the relationship between nurse staffing and patient outcomes. Res Nurs Health 2012; 35:277–88. [34] Carson SS, Stocking C, Podsadecki T, Christenson J, Pohlman A, MacRae S, et al. Effects of organizational change in the medical intensive care unit of a teaching hospital: a comparison of ‘open’ and ‘closed’ formats. JAMA 1996;276:322–8. [35] Blegen MA, Goode CJ, Spetz J, Vaughn T, Park SH. Nurse staffing effects on patient outcomes: safety-net and non-safety-net hospitals. Med Care 2011;49:406–14.
Please cite this article as: Hsieh C-C, et al, Impact of delayed admission to intensive care units on patients with acute respiratory failure, Am J Emerg Med (2016), http://dx.doi.org/10.1016/j.ajem.2016.09.066