Comments
Abstract
Rationale: The characteristics and outcomes of patients presenting with an acute exacerbation of chronic obstructive pulmonary disease (AECOPD) requiring intensive care unit (ICU) admission are poorly understood and there are sparse epidemiological data.
Objectives: The objectives were to describe epidemiology and outcomes of patients admitted to an ICU with COPD and to evaluate whether outcomes varied over time.
Methods: We studied adult ICU admissions across Australia and New Zealand between 2005 and 2017 with a diagnosis of AECOPD and used an admission diagnosis of asthma as comparator for trends over time. We measured changes in characteristics and outcomes over time using logistic regression, adjusting for illness severity using the Australian New Zealand Risk of Death model.
Results: We studied 31,991 admissions with AECOPD and 11,096 with asthma. Mean (standard deviation) age for AECOPD patients was 68.3 (11.2) years, with 35.4% mechanically ventilated. For patients with AECOPD, the percentage of deaths in an ICU was 8.7% and in a hospital was 15.4% of admissions, with the proportion of 69.2% discharged home and 5.6% discharged to a high-level care facility. During the study period, the proportion of ICU admissions with AECOPD per 10,000 admissions decreased at an annual rate of 2.0 (95% confidence interval [CI], 0.8–3.2; P = 0.009) but their admission rate per million population increased annually by 4.5 (95% CI, 3.7–5.3; P < 0.0001). There was a linear reduction in mortality for AECOPD but not for asthma admissions (odds ratio annual decline: AECOPD, 0.94 [0.93–0.95] and asthma, 1.01 [0.97–1.05]; P = 0.001) and an increase in AECOPD admissions discharged to home (odds ratio annual increase, AECOPD, 1.04 [1.03–1.05] and asthma, 1.01 [0.99–1.03]; P = 0.01). The reduction in mortality was sustained after adjusting for illness severity.
Conclusions: Across Australia and New Zealand, the rate of ICU admissions due to AECOPD is increasing but mortality rates are decreasing, with a corresponding increase in the home discharge rates.
Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) is a leading cause of morbidity and hospital admission (1). Moreover, patients with AECOPD severe enough to warrant an intensive care unit (ICU) admission are at increased risk of death (2–5). Of those that survive admission, many are unable to be discharged home after hospitalization (6, 7).
Although care for patients with AECOPD is increasingly successfully administered outside an ICU (8–10), more acutely ill patients with AECOPD still require admission to an ICU (9, 10). Patients with AECOPD may be in a hospital and in an ICU for prolonged periods; have predetermined limitations of treatment; have high mortality rates; and, in those who survive hospitalization, may be more likely to be discharged to a residential care facility rather than their own home (6).
There is limited knowledge about the outcomes of critically ill patients with AECOPD, as well as their epidemiological and physiological features (11). To provide relevant data, we aimed to investigate the characteristics, acute physiological features, invasive mechanical ventilation rates, and outcomes over time of critically ill patients admitted to Australia and New Zealand ICUs with a diagnosis of AECOPD. We used the diagnosis of asthma as a comparator so as to evaluate whether changes observed represented changes for all patients admitted to an ICU with airway obstruction or were specific to patients with AECOPD. The abstract of this article was presented at the Thoracic Society of Australia and New Zealand Annual Scientific Meeting on April 2, 2019, Broadbeach, QLD, Australia.
We performed a multicenter observational study using the Australia and New Zealand Intensive Care Society (ANZICS) Adult Patient Database (APD). The study was approved by the Alfred Hospital Human Research Ethics Committee, Melbourne, with a waiver of informed consent (number 3018/18).
Data were collected from the ANZICS APD on all adult patients admitted to 168 Australian and 13 New Zealand ICUs between January 1, 2005 and December 31, 2017 (12). Records were included if the patient was aged 17 years or older, a primary diagnosis was recorded, and there were completed data fields for hospital and ICU outcomes.
To limit inclusion of data from critically ill patients in whom COPD was a coexisting medical illness but not the primary precipitant underlying hospital and ICU admission, we only studied patients with an ICU admission diagnosis of AECOPD recorded using APD coding. ANZICS APD records for patients presenting with asthma were used as comparators for epidemiological and outcome trends (6, 13). Asthma was used as a comparator to identify whether changes over time represented changes in all patients admitted to an ICU with airway obstruction or were specific to AECOPD.
Before 2017, only invasive mechanical ventilation was collected as part of the ANZICS APD. As such, patients who only received noninvasive ventilation in the first 24 hours were categorized as not receiving invasive mechanical ventilation. In 2017, the minimum dataset was increased to collect additional information about those who received only noninvasive ventilation, but accuracy of data collection for this variable is yet to be determined.
We obtained data on patient demographics and physiological and biochemical parameters in the first 24 hours of ICU admission and the ventilation status at the time of the worst arterial blood gas in the first 24 hours. We also extracted data on treatment limitations on admission to an ICU (see online supplement). Finally, we obtained outcome data, including length of ICU and hospital admission, ICU and hospital mortality, and discharge destination, among survivors.
We initially assessed data for normality. Group comparisons were performed using chi-square tests for equal proportion, Student’s t test for normally distributed data, and Wilcoxon rank sum tests, with results presented as percentages (number), means (standard deviation), and medians (interquartile range), respectively. Changes in outcomes over time were determined using logistic regression, adjusting for illness severity using the Australian and New Zealand Risk of Death model (14), and year, with patients nested within site and site treated as a random variable.
Year was treated first as a categorical variable and then second as a continuous variable to determine annual change. To ascertain if changes in outcome over time differed between groups, we fitted an additional interaction term between group and year of admission to the model. Results from logistic regression models have been reported using odds ratio (OR) and 95% confidence interval (CI).
To account for survival bias, we further stratified results by survival status. We calculated population-level estimates for the projected proportion of ICU admissions due to AECOPD admissions using the annual adult population sizes of Australia and New Zealand (15, 16), with adjustment for the coverage of the APD, and presented as number of admissions per million persons per year. Trends over time were determined using linear correlation coefficients. All data were analyzed using SAS software, version 9.4 (SAS Institute Inc.), and a two-sided probability value (P) of 0.01 was used to indicate statistical significance.
Of 1,460,264 ICU admissions, 31,991 had an admission diagnosis of AECOPD (proportion of ICU admissions, 2.2%). In comparison, 11,906 patients had an admission diagnosis of asthma (proportion of ICU admissions, 0.8%). Using the population of Australia and New Zealand during the study period, the rate of admissions to ICUs with AECOPD was estimated to be between 100 and 150 admissions per million people per year.
Patients with AECOPD were older, had greater illness severity scores, and had increased risk of death when compared with patients admitted with asthma (Table 1). One-fifth of patients with AECOPD were admitted with some form of limitation of medical treatment (see Table 1).
| All Admissions (N = 43,897) | AECOPD Group (N = 31,991) | Asthma Group (N = 11,906) | |
|---|---|---|---|
| Baseline characteristics | |||
| Age, years, mean (SD) | 61.6 (16.9) | 68.3 (11.2) | 43.6 (16.5) |
| Sex, male | 45.3 (19,880) | 50.8 (16,253) | 30.5 (3,627) |
| APACHE III score, mean (SD) | 51.9 (22.1) | 56.4 (21.3) | 39.9 (19.7) |
| APACHE III standardized mortality rate | 0.76 (5,110/6,685) | 0.79 (4,931/6,245) | 0.41 (179/440) |
| ANZROD, % (IQR) | 7.8 (2.3–17.0) | 11.8 (6.4–21.4) | 0.9 (0.5–1.6) |
| ANZROD, mean (SD) | 11.6 (13.8) | 16.3 (14.2) | 1.6 (2.6) |
| Indigenous | 7.5 (2,701) | 7 (1,854) | 8.9 (847) |
| Source of hospital admission | |||
| Home | 75.5 (33,163) | 75.0 (23,991) | 77.0 (9,172) |
| Other hospital | 19.2 (8,420) | 20.0 (6,413) | 16.9 (2,007) |
| Other source* | 2.5 (1,093) | 2.8 (911) | 1.5 (182) |
| Not recorded | 2.8 (1,221) | 2.1 (676) | 4.6 (545) |
| Source of ICU admission | |||
| Emergency department | 66.7 (29,300) | 64.2 (20,554) | 73.5 (8,746) |
| Ward | 20.8 (9,120) | 23.1 (7,377) | 14.6 (1,743) |
| Other source† | 12.4 (5,447) | 12.6 (4,038) | 11.8 (1,409) |
| Not recorded | 0.7 (30) | 0.7 (22) | 0.7 (8) |
| Treatment limitations | |||
| Treatment limitation on ICU admission | 14.2 (6,254) | 19.2 (6,144) | 0.92 (110) |
| Admitted to ICU for palliative care | 0.35 (154) | 0.47 (149) | 0.04 (5) |
Compared with those with asthma, the mean inspired oxygen was higher in those presenting with AECOPD and the nadir arterial partial pressure of oxygen was lower (Table 2). The highest arterial partial pressure of carbon dioxide was greater in those admitted with AECOPD, as was the highest calculated arterial bicarbonate (see Table 2). The white cell count and blood glucose concentrations were less and the heart rate was slower in admissions with AECOPD (see Table 2).
| All Admissions (N = 43,897) | AECOPD Group (N = 31,991) | Asthma Group (N = 11,906) | |
|---|---|---|---|
| Biochemistry | |||
| Worst FiO2* | 0.42 (0.20), 33,368 | 0.46 (0.22), 24,977 | 0.41 (0.19), 8,391 |
| Lowest PaO2 (mm Hg)*, median (IQR), n | 72.0 (61–99), 33,298 | 68.0 (59–85), 24,943 | 94.0 (72–140), 8,355 |
| PaO2/FiO2 ratio, mean (SD) | 236 (124) | 220 (108) | 282 (153) |
| Highest PaCO2 (mm Hg)* | 57.9 (22.7), 33,327 | 61.2 (21.1), 24,952 | 48.2 (24.6), 8,375 |
| Lowest pH* | 7.31 (0.12), 33,397 | 7.31 (0.11), 24,992 | 7.32 (0.14), 8,405 |
| Highest HCO3† | 28.5 (7.64), 38,223 | 30.3 (7.4), 28,169 | 23.3 (5.6), 10,054 |
| Highest glucose (mmol L−1) | 10.9 (4.8), 38,394 | 10.7 (4.8), 28,118 | 11.3 (4.69), 10,276 |
| Highest creatinine (μmol L−1), median (IQR), n | 78.0 (61–105), 39,877 | 81.0 (62–114), 29,226 | 72.0 (60–880), 10,651 |
| White cell count (109 L−1) | |||
| Highest | 14.5 (10.1), 39,562 | 14.0 (10.1), 29,038 | 16.0 (9.9), 10,524 |
| Lowest | 12.0 (7.49), 33,933 | 11.6 (7.5), 25,086 | 13.3 (7.4), 8,847 |
| Highest hematocrit (%) | 40.0 (6.7), 36,139 | 40.2 (7.2), 26,542 | 39.5 (5.4), 9,597 |
| Physiological parameters | |||
| Respiratory rate (breaths per minute) | |||
| Highest | 28.9 (8.32), 41,018 | 28.9 (8.1), 29,933 | 29.1 (9.0), 11,085 |
| Lowest, median (IQR), n | 15.0 (12–18), 39,550 | 15.0 (12–18), 28,900 | 15.0 (12–18), 10,650 |
| Temperature (°C) | |||
| Highest | 37.1 (0.7), 41,089 | 37.1 (0.7), 30,005 | 37.2 (0.6), 11,084 |
| Lowest | 36.1 (0.7), 39,652 | 36.1 (0.7), 28,990 | 36.1 (0.65), 10,662 |
| Heart rate (beats per minute) | |||
| Highest | 114.0 (20.9), 41,192 | 111.0 (20.9), 30,060 | 122.0 (18.6), 11,132 |
| Lowest | 81.4 (16.5), 39,728 | 79.1 (15.7), 29,044 | 87.6 (16.8), 10,684 |
At the time of the worst arterial blood gas in the first 24 hours, one-third of admissions were receiving support with invasive mechanical ventilation (Table 3). Over the duration of this study, those presenting with asthma had very low mortality rates, whereas those presenting with AECOPD had an ICU mortality rate of 8.7%, with an almost twofold greater mortality rate after ICU discharge and before hospital discharge (see Table 3). AECOPD survivors remained in ICUs and hospitals for longer periods than survivors with asthma (see Table 3). Patients with AECOPD who received invasive mechanical ventilation in the first 24 hours were more likely to die in the hospital than those who did not receive invasive mechanical ventilation (18.3% vs. 13.9%; P < 0.0001; Figure 1A). Admissions with AECOPD and a limitation of medical treatment order were more likely to die, and the time to death was shorter than those without limitations of treatment (Table 4 and see Figure 1B).
| All Admissions (N = 43,897) | AECOPD Group (N = 31,991) | Asthma Group (N = 11,906) | |
|---|---|---|---|
| Ventilation | |||
| Invasive mechanical ventilation | 33.0 (14,460) | 35.4 (11,317) | 26.4 (3,143) |
| Mortality | |||
| ICU | 6.6 (2,897) | 8.7 (2,773) | 1.0 (124) |
| Hospital | 11.6 (5,110) | 15.4 (4,931) | 1.5 (179) |
| Length of stay (days) | |||
| ICU (n = 43,895) | 2.31 (1.21–4.26) | 2.61 (1.39–4.69) | 1.76 (0.94–3.21) |
| Hospital (n = 43,283) | 7.02 (3.98–12) | 8.15 (4.87–13.4) | 4.42 (2.63–7.64) |
| ICU-free days to Day 28 | 22.9 (7.1) | 22.1 (7.8) | 25.0 (4.2) |
| Hospital-free days to Day 28 | 16.4 (9.1) | 14.7 (9.2) | 21.1 (6.8) |
| Hospital for survivors (n = 38,431) | 7.16 (4.13–12.1) | 8.48 (5.28–13.6) | 4.41 (2.62–7.58) |
| Hospital for deceased (n = 4,852) | 5.46 (2.59–11.4) | 5.45 (2.58–11.4) | 5.74 (2.93–11.9) |
| Readmission to ICU within same hospital stay | |||
| Readmission to ICU (n = 43,897) | 2.9 (1,252) | 3.1 (982) | 2.3 (270) |
| Hospital outcome | |||
| Home | 75.5 (33,136) | 69.2 (22,122) | 92.5 (11,014) |
| Other hospital | 8.2 (3,582) | 9.5 (3,054) | 4.4 (528) |
| High-level facility, including nursing home, chronic care facility, or palliative care | 4.4 (1,949) | 5.6 (1,782) | 4.4 (528) |
| Other* | 0.3 (120) | 0.3 (102) | 0.2 (18) |
Figure 1. Time-to-event (death in a hospital) curves of admissions with acute exacerbation of chronic obstructive pulmonary disease, with patients categorized as (A) those receiving invasive mechanical ventilation in first 24 hours of ICU compared with those not receiving (P = 0.001) and (B) those with a treatment limitation and those without (P < 0.0001). Survival curves are compared using a log-rank test. ICU = intensive care unit.
[More] [Minimize]| Limitation of Treatment (N = 6,144) | No Limitation of Treatment (N = 25,847) | |
|---|---|---|
| Mortality, % (n) | ||
| Intensive care unit admissions | 13.6 (836) | 7.5 (1,937) |
| Hospital admissions | 25.6 (1,570) | 13.0 (3,361) |
The majority of admissions with AECOPD were discharged home; however, discharge to another hospital or to a high-level facility (nursing home, chronic care facility, or palliative care) was relatively common (see Table 3). The likelihood of readmission to an ICU following discharge from ICU within the same hospital admission was higher with AECOPD than asthma (see Table 3).
The raw number of ICU admissions due to AECOPD increased over time (from 1,768 in 2005 to 3,141 in 2017). To adjust for increases in ICU bed numbers and population over time, data are presented as the proportion of admissions due to AECOPD per ICU bed (reduced; Figure 2A) and per million population (increased; see Figure 2B).
Figure 2. The percentage of admissions due to acute exacerbation of chronic obstructive pulmonary disease (A) per ICU bed reduced over time (P = 0.009) but (B) per million population increased (P < 0.0001). ICU = intensive care unit.
[More] [Minimize]There was an annual reduction in mortality over time for admissions with AECOPD that was not evident for those with asthma (Figures 3A and 3B).
Figure 3. There was an annual reduction in (A) raw mortality (P < 0.0001) and (B) when adjusted for site and the Australian and New Zealand Risk of Death model over time for admissions with acute exacerbation of chronic obstructive pulmonary disease, with the odds ratio for annual decline in hospital mortality with acute exacerbation of chronic obstructive pulmonary disease (0.94; 95% confidence interval, 0.93–0.95) versus the odds ratio of mortality during hospital admission with asthma (1.01; 95% confidence interval, 0.97–1.05); P = 0.001. COPD = chronic obstructive pulmonary disease.
[More] [Minimize]There was a similar decline in mortality over time for admissions with AECOPD, with decreases in both the mechanically ventilated and nonmechanically ventilated cohorts (see Figures E1 and E2 in the online supplement). There was an annual decrease in the ICU readmission rate for AECOPD patients over time (OR, 0.98; 95% CI, 0.90–1.00), which was not apparent in patients admitted with asthma (OR, 1.02; 95% CI, 0.98–1.05).
The raw number and proportion of admissions to ICUs with either AECOPD or asthma and a limitation of medical treatment order increased severalfold over time (AECOPD, from 181 [9.3%] in 2007 to 967 [30.8%] in 2017; and asthma, from 3 [0.4%] in 2007 to 15 [1.3%] in 2017). This occurred despite age remaining relatively stable over time (see Figure E3). The likelihood of hospital mortality was increased for admissions with AECOPD and a treatment limitation (see Figure 1B). However, mortality for patients with AECOPD and a treatment limitation decreased annually over time at a similar rate to those without a treatment limitation (with a treatment limitation OR, 0.94; 95% CI, 0.93–0.95 and without a treatment limitation OR, 0.95; 95% CI, 0.93–0.97), with no interaction (P = 0.71).
There was a corresponding increase in the annual OR of being discharged home alive (Figure 4). These changes occurred without any increase in annual discharge to another hospital (see Figures E4 and E5) or any increase in annual discharge to a high-level facility, including nursing home, chronic care facility, or palliative care (see Figures E6 and E7).
Figure 4. The annual odds ratio of being discharged home alive after admission to an intensive care unit with acute exacerbation of chronic obstructive pulmonary disease increased over time. The severity-adjusted odds ratio of discharge home is shown: acute exacerbation of chronic obstructive pulmonary disease 1.04 (95% confidence interval, 1.03–1.05) versus asthma 1.01 (95% confidence interval, 0.99–1.03); P = 0.01. COPD = chronic obstructive pulmonary disease.
[More] [Minimize]In the largest examination to date of outcomes following ICU admission for AECOPD, we observed that the proportion of AECOPD admissions to an ICU in Australia and New Zealand was between 100 and 150 admissions per million per year and increased over time. In contrast, AECOPD admissions as a proportion of all ICU admissions were approximately 2%, and appeared to be decreasing over time. One in 5 AECOPD admissions required invasive mechanical ventilation. The ICU mortality approximated 1 in 11 admissions and hospital mortality was approximately 1 in 6 admissions. However, the observed unadjusted and adjusted likelihood of hospital death among these admissions has decreased. Moreover, the likelihood of returning home following an ICU admission has also increased, with almost two-thirds able to be discharged home. During the period studied, there was a significant increase in the number of patients admitted to an ICU with a limitation of treatment order.
Previous data describing the epidemiology and outcomes of patients admitted to an ICU with AECOPD are mostly limited to single-center studies (3–5, 17–20). The majority of multicenter studies of patients with AECOPD include all those admitted to a hospital (2, 11, 13, 21–26), of which only a small proportion required ICU-level care.
A recently published systematic review of 37 studies identified 12 studies evaluating outcomes in critically ill patients with AECOPD (i.e., those admitted to ICUs) (27). These studies included 2,776 patients with AECOPD with 805 deaths. Such aggregate mortality of 29.0% is twofold greater than observed in our study (27).
The largest prospective cohort study was conducted in the United Kingdom over 18 months across 92 ICUs in the United Kingdom and included 832 patients admitted to ICUs and high-dependency units with AECOPD, asthma, and asthma–COPD overlap, and was completed 10 years ago (13). The mean age of patients and the percentage with some form of limitation of medical treatment were similar to our cohort. However, 54% of patients received invasive mechanical ventilation, ICU mortality was 18.9%, and hospital mortality was 29.6%, twice that of our cohort (13). The largest multicenter ICU dataset was from Sweden and also used registry data (28). It included 1,009 patients with AECOPD (28). ICU mortality in this study was similar to our study at 7.3% but 30-day mortality was high at 26%.
Our findings imply that the proportion of patients admitted with AECOPD in income-rich countries approximates 2%, with an admission rate of approximately 100 to 150 ICU cases per million people per year. Moreover, the mortality rate for patients admitted to an ICU with AECOPD appears less than previously reported. Mortality has decreased and continues to decrease over time, and the rate of discharge home has increased and continues to increase over time.
This population-level information is important for decision-making for individual patients. Not only are outcomes improving but limitations of medical treatment among this cohort are common and have significantly increased. Although mortality is greater in those admitted with a treatment limitation, the trajectory of reduction in mortality is similar. These data imply that clinicians, families, and medical treatment decision-makers can discuss limitations of medical treatment when appropriate, without relegating patients to inferior care. They also imply that future trials in this cohort may need to include patients with a treatment limitation and, if using mortality as the primary outcome, given current trends, such studies should adjust sample size estimates for a greater number of survivors.
There are certain limitations to our study. Baseline lung function tests are not collected as part of this dataset, so adjustment for baseline severity of COPD cannot be performed. Data were only collected from the time of ICU admission; thus, patients who were never admitted to an ICU and only received care, including noninvasive ventilation, in either the emergency department or respiratory wards were not captured (26, 29). However, such selection bias is likely to have favored mortality being increased, rather than the reduction we observed, because those that rapidly respond to initial interventions will not require ICU admission.
We can only report on the proportion of admissions with AECOPD because APD data are collected from an index hospitalization but are not identifiable. Because we cannot identify whether an individual patient was previously admitted to an ICU at another hospital, the proportions per population we provide are only estimates of the period prevalence. Furthermore, this means that we cannot report readmission rates for separate hospital admissions, which may be an important prognostic factor (4, 30–32).
Recorded variables included invasive ventilation, which is the minimum number of admissions who received a period of invasive mechanical ventilation within the first 24 hours of being in an ICU. This variable will not capture those admissions who have mechanical ventilation commenced after 24 hours in an ICU. Before 2017, noninvasive data were not captured (i.e., patients who only received noninvasive ventilation were grouped with all other patients as not receiving invasive ventilation).
The ANZICS APD only allows one diagnostic code and interobserver variability in allocation of diagnostic codes is recognized (30). However, patients with asthma were used as an illustrative comparison for trends over time to evaluate whether admissions identified as AECOPD represented trends observed with other airflow limitation pathologies (13). Nonetheless, the inferences from our data should not be extrapolated to those with COPD as coexisting illness and another primary medical problem as the reason for ICU admissions. Data collection on limitations of medical treatment commenced in 2007 and only became mandatory from 2011, meaning that the proportion of admissions with a treatment limitation is likely to have been underestimated. Finally, the ANZICS APD only includes mortality during the index hospitalization. Although acute mortality is important, studies have reported that short-term outcomes for patients with AECOPD undergoing a period of invasive mechanical ventilation may be much better than longer-term outcomes (4, 33). Accordingly, mortality trends at longer time points (e.g., 12 mo) may not follow the hospital mortality trends we observed (31–33).
In a study including over 30,000 admissions with AECOPD to ICUs across Australia and New Zealand, we calculated an estimated an admission rate of 100 to 150 ICU admissions per million per year, which is increasing. Approximately one-fifth of admissions received invasive mechanical ventilation and a similar proportion had limitations of medical therapy, with approximately one-sixth of admissions not surviving to hospital discharge and one-third of survivors unable to be discharged home. Outcomes for patients admitted with AECOPD appear to be improving over time.
The authors and the Australian and New Zealand Intensive Care Society Centre for Outcome and Resource Evaluation management committee thank clinicians, data collectors, and researchers at the contributing sites (see online supplement); and Ms. Brianna Tascone for preparation of the manuscript and images.
| 1 . | Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med 2006;3:e442. |
| 2 . | Seneff MG, Wagner DP, Wagner RP, Zimmerman JE, Knaus WA. Hospital and 1-year survival of patients admitted to intensive care units with acute exacerbation of chronic obstructive pulmonary disease. JAMA 1995;274:1852–1857. |
| 3 . | Breen D, Churches T, Hawker F, Torzillo PJ. Acute respiratory failure secondary to chronic obstructive pulmonary disease treated in the intensive care unit: a long term follow up study. Thorax 2002;57:29–33. |
| 4 . | Gadre SK, Duggal A, Mireles-Cabodevila E, Krishnan S, Wang XF, Zell K, et al. Acute respiratory failure requiring mechanical ventilation in severe chronic obstructive pulmonary disease (COPD). Medicine (Baltimore) 2018;97:e0487. |
| 5 . | Afessa B, Morales IJ, Scanlon PD, Peters SG. Prognostic factors, clinical course, and hospital outcome of patients with chronic obstructive pulmonary disease admitted to an intensive care unit for acute respiratory failure. Crit Care Med 2002;30:1610–1615. |
| 6 . | Cydulka RK, McFadden ER Jr, Emerman CL, Sivinski LD, Pisanelli W, Rimm AA. Patterns of hospitalization in elderly patients with asthma and chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1997;156:1807–1812. |
| 7 . | Mulpuru S, McKay J, Ronksley PE, Thavorn K, Kobewka DM, Forster AJ. Factors contributing to high-cost hospital care for patients with COPD. Int J Chron Obstruct Pulmon Dis 2017;12:989–995. |
| 8 . | Lindenauer PK, Stefan MS, Shieh MS, Pekow PS, Rothberg MB, Hill NS. Outcomes associated with invasive and noninvasive ventilation among patients hospitalized with exacerbations of chronic obstructive pulmonary disease. JAMA Intern Med 2014;174:1982–1993. |
| 9 . | Walkey AJ, Wiener RS. Use of noninvasive ventilation in patients with acute respiratory failure, 2000-2009: a population-based study. Ann Am Thorac Soc 2013;10:10–17. |
| 10 . | Rochwerg B, Brochard L, Elliott MW, Hess D, Hill NS, Nava S, et al. Official ERS/ATS clinical practice guidelines: noninvasive ventilation for acute respiratory failure. Eur Respir J 2017;50:5–6. |
| 11 . | Molinari N, Chanez P, Roche N, Ahmed E, Vachier I, Bourdin A. Rising total costs and mortality rates associated with admissions due to COPD exacerbations. Respir Res 2016;17:149. |
| 12 . | Kaukonen KM, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012. JAMA 2014;311:1308–1316. |
| 13 . | Wildman MJ, Sanderson CF, Groves J, Reeves BC, Ayres JG, Harrison D, et al. Survival and quality of life for patients with COPD or asthma admitted to intensive care in a UK multicentre cohort: the COPD and Asthma Outcome Study (CAOS). Thorax 2009;64:128–132. |
| 14 . | Paul E, Bailey M, Kasza J, Pilcher D. The ANZROD model: better benchmarking of ICU outcomes and detection of outliers. Crit Care Resusc 2016;18:25–36. |
| 15 . | Australian Bureau of Statistics. Australian Demographic Statistics. 2018, cat no. 3101.0, ABS Canberra. 2018 [updated 2018; accessed 2019 Aug 20]. Available from: https://www.abs.gov.au/ausstats/[email protected]/mf/3101.0. |
| 16 . | Statistics New Zealand. National Population Estimates. 2018 [updated 2018; accessed 2019 Aug 20]. Available from: https://www.stats.govt.nz/topics/population. |
| 17 . | Vitacca M, Clini E, Porta R, Foglio K, Ambrosino N. Acute exacerbations in patients with COPD: predictors of need for mechanical ventilation. Eur Respir J 1996;9:1487–1493. |
| 18 . | Nevins ML, Epstein SK. Predictors of outcome for patients with COPD requiring invasive mechanical ventilation. Chest 2001;119:1840–1849. |
| 19 . | Ucgun I, Metintas M, Moral H, Alatas F, Yildirim H, Erginel S. Predictors of hospital outcome and intubation in COPD patients admitted to the respiratory ICU for acute hypercapnic respiratory failure. Respir Med 2006;100:66–74. |
| 20 . | Christensen S, Rasmussen L, Horváth-Puhó E, Lenler-Petersen P, Rhode M, Johnsen SP. Arterial blood gas derangement and level of comorbidity are not predictors of long-term mortality of COPD patients treated with mechanical ventilation. Eur J Anaesthesiol 2008;25:550–556. |
| 21 . | Tabak YP, Sun X, Johannes RS, Gupta V, Shorr AF. Mortality and need for mechanical ventilation in acute exacerbations of chronic obstructive pulmonary disease: development and validation of a simple risk score. Arch Intern Med 2009;169:1595–1602. |
| 22 . | Shorr AF, Sun X, Johannes RS, Derby KG, Tabak YP. Predicting the need for mechanical ventilation in acute exacerbations of chronic obstructive pulmonary disease: comparing the CURB-65 and BAP-65 scores. J Crit Care 2012;27:564–570. |
| 23 . | Berkius J, Sundh J, Nilholm L, Fredrikson M, Walther SM. What determines immediate use of invasive ventilation in patients with COPD? Acta Anaesthesiol Scand 2013;57:312–319. |
| 24 . | Valley TS, Sjoding MW, Ryan AM, Iwashyna TJ, Cooke CR. Intensive care unit admission and survival among older patients with chronic obstructive pulmonary disease, heart failure, or myocardial infarction. Ann Am Thorac Soc 2017;14:943–951. |
| 25 . | Roberts CM, Stone RA, Buckingham RJ, Pursey NA, Lowe D; National Chronic Obstructive Pulmonary Disease Resources and Outcomes Project implementation group. Acidosis, non-invasive ventilation and mortality in hospitalised COPD exacerbations. Thorax 2011;66:43–48. |
| 26 . | Toft-Petersen AP, Torp-Pedersen C, Weinreich UM, Rasmussen BS. Trends in assisted ventilation and outcome for obstructive pulmonary disease exacerbations. A nationwide study. PLoS One 2017;12:e0171713. |
| 27 . | Singanayagam A, Schembri S, Chalmers JD. Predictors of mortality in hospitalized adults with acute exacerbation of chronic obstructive pulmonary disease. Ann Am Thorac Soc 2013;10:81–89. |
| 28 . | Berkius J, Nolin T, Mårdh C, Karlström G, Walther SM; Swedish Intensive Care Registry. Characteristics and long-term outcome of acute exacerbations in chronic obstructive pulmonary disease: an analysis of cases in the Swedish Intensive Care Registry during 2002-2006. Acta Anaesthesiol Scand 2008;52:759–765. |
| 29 . | Hukins C, Wong M, Murphy M, Upham J. Management of hypoxaemic respiratory failure in a respiratory high-dependency unit. Intern Med J 2017;47:784–792. |
| 30 . | Heldens M, Schout M, Hammond NE, Bass F, Delaney A, Finfer SR. Sepsis incidence and mortality are underestimated in Australian intensive care unit administrative data. Med J Aust 2018;209:255–260. |
| 31 . | Hartl S, Lopez-Campos JL, Pozo-Rodriguez F, Castro-Acosta A, Studnicka M, Kaiser B, et al. Risk of death and readmission of hospital-admitted COPD exacerbations: European COPD Audit. Eur Respir J 2016;47:113–121. |
| 32 . | Suissa S, Dell’Aniello S, Ernst P. Long-term natural history of chronic obstructive pulmonary disease: severe exacerbations and mortality. Thorax 2012;67:957–963. |
| 33 . | Hannan LM, Tan S, Hopkinson K, Marchingo E, Rautela L, Detering K, et al. Inpatient and long-term outcomes of individuals admitted for weaning from mechanical ventilation at a specialized ventilation weaning unit. Respirology 2013;18:154–160. |
Author Contributions: F.B.: literature search and writing. D.P.S.: literature search and writing. Y.A.A.: writing and manuscript review. M.J.B.: figures and data interpretation. D.V.P.: figures, data analysis, and data interpretation. R.B.: data interpretation, verification of analytic methods, and manuscript review. M.E.F.: manuscript review and data interpretation. P.J.Y.: manuscript review. A.M.D.: literature search, study design, data interpretation, and writing.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
