Chest ACCP Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Garland, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Garland, A.
(Chest. 2004;126:1897-1904.)
© 2004 American College of Chest Physicians

Outcomes up to 5 Years After Severe, Acute Respiratory Failure*

Allan Garland, MD, FCCP; Neal V. Dawson, MD; Irene Altmann, MD; Charles L. Thomas, AB; Russell S. Phillips, MD; Joel Tsevat, MD, MPH; Norman A. Desbiens, MD; Paul E. Bellamy, MD; William A. Knaus, MD and Alfred F. Connors, Jr, MD, FCCP; for the SUPPORT Investigators{dagger}

* From the Case Western Reserve University School of Medicine (Drs. Garland, Dawson, and Connors, and Mr. Thomas), Cleveland, OH; The Everett Clinic (Dr. Altmann), Everett, WA; the Harvard Medical School (Dr. Phillips), Boston, MA; the University of Cincinnati Medical Center (Dr. Tsevat), Cincinnati, OH; the University of Tennessee College of Medicine (Dr. Desbiens), Chattanooga, TN; Kaiser Permanente (Dr. Bellamy), Woodland Hills, CA; and the University of Virginia Health Sciences System (Dr. Knaus), Charlottesville, VA. {dagger} The research was performed at the following institutions: MetroHealth Medical Center, Cleveland, OH; Duke University Medical Center, Durham, NC; Beth Israel Hospital, Boston, MA; UCLA Medical Center, Los Angeles, CA; and Marshfield Clinic, Marshfield, WI.

Correspondence to: Allan Garland, MD, FCCP, Division of Pulmonary and Critical Care Medicine, MetroHealth Medical Center, 2500 MetroHealth Dr, Cleveland, OH 44109; e-mail: agarland{at}metrohealth.org


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Conclusions
 References
 
Study objective: To use an existing database from a large cohort study with follow-up as long as 5.5 years to assess the extended prognosis of patients who survived their hospitalizations for severe acute respiratory failure (ARF).

Design, setting, and patients: Secondary analysis of an inception cohort of 1,722 patients with ARF requiring mechanical ventilation from five major medical centers who were entered into the prospective Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. The 1,075 patients (62.4%) who survived hospitalization had systematic follow-up of vital status for a median time of 662 days (interquartile range, 327 to 1,049 days; range, 2 to 2,014 days). Interviews performed a median of 5 months after hospital discharge assessed functional capacity and quality of life (QOL). The main outcome measure was survival after hospital discharge. Secondary measures were functional status and QOL. Cox proportional hazard regression identified factors influencing posthospital survival.

Results: The median survival time after hospital discharge for ARF was > 5.3 years. The posthospital survival time was shorter for those with older age, male gender, several preexisting comorbid conditions, worse prehospital functional status, greater acute physiologic derangement, and a do-not-resuscitate order while in the hospital, and for those discharged to a location other than home. Five months after hospital discharge, 48% of survivors needed help with at least one activity of daily living, and 27% rated their QOL as poor or fair. However, most of these impairments were present before respiratory failure occurred.

Conclusions: Extended survival is common among patients with ARF who require mechanical ventilation and who survive hospitalization. Among these patients, only a small fraction of the impairment in activity and QOL can be considered to be a sequela of the respiratory failure or its therapy. These findings are relevant to the care decisions for such critically ill patients.

Key Words: activities of daily living • acute lung injury • outcome assessment • quality of life • survival analysis


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Conclusions
 References
 
Acute respiratory failure (ARF), which frequently requires mechanical ventilatory support, is among the most common reasons for admission to an ICU.12 ARF is associated with substantial mortality and morbidity.3 While most research into outcomes among critically ill patients has assessed only short-term measures such as in-hospital mortality and complications,4 the human costs of severe illnesses extend well beyond the hospitalization.5 While three relatively small studies of ARF have reported survival rates of > 50% at 1 year or beyond,678 no large studies systematically following a cohort of such patients beyond 3 months have been published. The primary goal of this study was to use data from a large prospective cohort study with systematic follow-up as long as 5.5 years to assess the extended prognosis of patients with ARF requiring mechanical ventilation who survived their hospitalizations. We hypothesized that long-term survival would be consistent with the results of the smaller studies, and would be associated with patients’ chronic health, demographics, and severity of acute illness. In addition, we explored whether posthospital survival was related to in-hospital do-not-resuscitate (DNR) status, and hospital discharge location. Our secondary objective was to investigate functional capacity and quality of life (QOL) among hospital survivors.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Conclusions
 References
 
The data for this analysis come from the Study to Understand Prognosis and Preferences for Outcomes and Risks of Treatment (SUPPORT), a prospective, five-center study that was conducted in two phases, from June 1989 to January 1994. The details of the SUPPORT have been described previously.910 Aspects that are relevant to this study are summarized here. SUPPORT included 9,105 seriously ill hospitalized patients with any one of nine acute illnesses, who survived at least 48 h after study entry. ARF was one of the nine illnesses, with the following criteria: (1) respiratory failure due to pneumonia or ARDS, as determined by clinician judgment; (2) admission into an ICU or other specialty care unit; and (3) an acute physiology and chronic health evaluation (APACHE) III acute physiology score (APS)11 of ≥ 10, or ≥ 7 if the neurologic portion could not be evaluated due to pharmacologic sedation. Exclusion criteria for admission to the SUPPORT because of ARF were as follows: severe COPD; severe congestive heart failure (CHF); Pneumocystis carinii pneumonia; status asthmaticus; pulmonary embolism; immunologic lung disease; primary restrictive lung disease; primary hypoventilation; smoke inhalation; or having had a thoracotomy during the index hospitalization.12

The patients in this study were those entered into the SUPPORT for ARF who required mechanical ventilation within 48 h after study entry. Study entry could occur on meeting eligibility criteria any time after hospital admission. The data collected at study entry included age, gender, race, insurance status, yearly family income, the presence of 31 preexisting comorbidities, physical examination parameters, and laboratory values. Because < 5% of patients were of a race other than white or African American, race was dichotomized into white or minority. Insurance status was divided into the following three categories that were designed to capture socioeconomic status: (1) private insurance with or without supplemental Medicare; (2) Medicare only; or (3) no insurance, Medicaid only, or Medicaid and Medicare. Income at the time of study entry was divided into the following four groups: ≤ $10,000; $11,000 to $24,000; $25,000 to ≥ 49,000; and ≥ $50,000. The following seven categories were considered to be preexisting comorbid conditions: malignancy; diabetes mellitus; other pulmonary diseases; neurologic or psychiatric disorders; alcohol or IV drug abuse; CHF or myocardial infarction (MI); and a composite (necessitated by small numbers in each individual category) consisting of renal diseases, liver cirrhosis, HIV/AIDS, and other immunosuppressive disorders. Severity of illness was assessed on the day of study entry using the nonneurologic portion of the APS, the SUPPORT coma score, and the SUPPORT prognostic model estimates of survival to 60 and 180 days after study entry.1113 APS scores without its neurologic component ranged from 0 to 204, with higher values representing greater degrees of physiologic derangement. The SUPPORT coma score ranged from 0 (normal neurologic status) to 100 (deep coma). Chest radiograph (CXR) appearance within 48 h of study entry was dichotomized as having bilateral airspace infiltrates present or absent. Scores from the modified therapeutic intervention scoring system (TISS) [in which larger values indicate greater intensity of hospital resource use]14 were assessed prospectively at days 1, 3, 7, 14, and 25 of the study.

Hospital resource use was measured using length of stay (LOS), TISS, and hospital costs. Costs, expressed in 2003 US dollars, were determined by converting charges to costs using the cost/charge ratio of each hospital.151617 For the 0.9% of patients whose charge data were missing, costs were estimated using the product of the average TISS score and LOS, adjusted by hospital.18

Interviews of patients and surrogates were performed at days 3 and 180 after study entry. These assessed patient well-being in domains covering functional capacity and QOL, including the ability to perform activities of daily living (ADLs).9 Surveys performed at day 3 asked about the situation in the 2 weeks prior to hospitalization. The day-180 survey inquired about patients’ status at that later time. Functional capacity was measured with modified versions of the Duke activity status index (DASI)19 and the Katz index of activity of daily living (KADL).20 The DASI measures functional capacity on a scale that ranges from 11 (unable to walk indoors) to 33 (able to do vigorous exercise or aerobics). The KADL scores the number of ADLs, of a total of seven (ie, bathing, dressing, toileting, transferring, continence, feeding, and walking), with which the patient requires assistance. Global QOL (GQOL) was rated on a 5-point scale as excellent, very good, good, fair, or poor, producing a score of 1 to 5, with lower numbers reflecting better QOL. Statistical modeling of simultaneous patient and surrogate surveys allowed for the substitution of calibrated surrogate responses when patients were unable to participate in interviews.21 Survival to the end of the study was assessed in all patients by use of the National Death Index,22 allowing for follow-up times up to 2,014 days (or 5.5 years) after study entry.

Because individuals clearly differentiate the hospital phase of illness from what occurs after hospital discharge, this dichotomy guided the analysis. Outcomes of the hospital phase included mortality and resource use. Posthospital outcomes included survival, discharge location, and the well-being variables DASI, KADL, and GQOL.

Statistical Analysis
Unless otherwise indicated, values are presented as proportions, or median and interquartile range (IQR). Since none of the noncategoric variables had Gaussian distributions, we used the Wilcoxon signed ranks test for paired comparisons, and the Mann-Whitney U test for unpaired comparisons.

Posthospital survival was assessed with life table analysis and was modeled with multivariate Cox proportional hazards regression. Due to collinearity among severity-of-illness measures, models included only the SUPPORT coma score and the nonneurologic part of the APS. Because the ICU and hospital LOS were strongly correlated (r = 0.74; p < 0.001), only the latter was used as a model covariate. The independent variables used were age, gender, race, insurance status, income, the seven comorbidity categories, nonneurologic APS score, SUPPORT coma score, the lowest PaO2/fraction of inspired oxygen (FIO2) ratio in the first 48 h, prehospital well-being (ie, KADL, DASI, and GQOL scores), hospital LOS, DNR status during hospitalization, and discharge location after hospitalization. Because they are ordinal variables with a small number of values, the KADL and GQOL scales were converted into categoric variables. The linearity assumption of continuous covariates in proportional hazards regression was tested by graphic analysis, and when this assumption was violated, the continuous variable was divided into quartiles and replaced by indicator variables.23 The proportionality assumption of the Cox model was verified by log-minus-log plots and time-dependent covariates.23 Although there is no generally accepted method of assessing the ability of Cox models to discriminate between those who live and those who die, the c-statistic is such a measure for logistic regression models.24 A logistic model, in which all patients are followed up for the same length of time, is analogous to a Cox model at that same time point. Since the 1-year survival time is known for all patients in the SUPPORT, we estimated the discriminating ability of the Cox model by the c-statistic of the analogous logistic regression model of posthospital survival at 1 year after study entry.25 Statistical analysis was performed using a statistical software package (SPSS, version 10.0; SPSS, Inc; Chicago, IL). Two-sided p values of < 0.05 were considered to be significant.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Conclusions
 References
 
There were 1,729 ARF patients in the SUPPORT who required mechanical ventilation within 48 h after study entry. Seven of these patients had incomplete initial data, leaving 1,722 for analysis.

The baseline data are shown in Table 1 . The severity of respiratory impairment in this ARF population was high. The median PaO2/FIO2 ratio was 128. In 78% of patients, it was < 200, and in 94% of patients it was < 300. Half of the patients had bilateral infiltrates seen on CXRs. Two weeks prior to hospitalization, patients had substantial functional debility and impaired QOL. They required help with an average and median of two ADLs, and 79% of patients rated their QOL as fair or poor. These patients consumed large amounts of hospital resources (Table 2 ).


View this table:
[in this window]
[in a new window]

  		Table 1. Baseline Data*

 

View this table:
[in this window]
[in a new window]

  		Table 2. Hospital Outcomes and Resource Use*

 
Of the 1,722 patients, 647 (38%) died in the hospital. Table 3 shows survival rates at fixed intervals up to 1 year after study entry.


View this table:
[in this window]
[in a new window]

  		Table 3. Survival Rates in ARF Patients

 
Of 1,075 patients who survived their index hospitalization, the immediate posthospital location was ascertained for 1,060 patients (98.6%). Of these, 338 patients (31.9%) were discharged from the hospital to a location other than home. Of the hospital survivors, the in-hospital DNR status was known for 1,070 patients (99.5%). Of these, 138 patients (12.9%) had a DNR order during their hospitalization.

Posthospital Survival
The 1,075 patients who survived their hospitalization had a median follow-up time of 662 days after hospital discharge (IQR, 327 to 1,049; range, 2 to 2,014 days). The fractions of hospital survivors with follow-ups of > 1 to 5 years were 72.7%, 45.9%, 23.0%, 16.1%, and 6.0%, respectively. In addition to deaths, these numbers reflect the termination of data collection when the study ended, which was before most patients reached 5 years of follow-up.

During posthospital follow-up, 396 of the hospital survivors (36.8%) died. The median survival time after hospital discharge among ARF patients (Fig 1 ) was > 1,920 days (5.3 years). Survival declined steeply during the first year, after which it fell more slowly.



View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Posthospital survival in patients after severe ARF. Life table survival fraction (black curve) with 95% CIs (gray curves).

 
Cox regression analysis (Table 4 ) for the 1,055 hospital survivors with complete data for the independent variables (ie, 98.1% of the 1,075 patients [posthospital discharge location and in-hospital DNR status were unknown for 15 patients and 5 patients, respectively]) identified 10 variables that were associated with the hazard for posthospital death after ARF. In addition to increased hazards related to older age, male gender, worse physiologic derangement at study entry, and comorbidity, posthospital death was associated with worse prehospital functional capacity (ie, DASI score), having a DNR order during hospitalization, and being unable to go home after hospital discharge. The presence of chronic pulmonary disease was of borderline significance (hazard ratio [HR], 1.30; p = 0.075). Model fit, as estimated by the c-statistic of the analogous logistic regression model of posthospital survival at 1 year after study entry, was 0.80.


View this table:
[in this window]
[in a new window]

  		Table 4. Cox Proportional Hazards Model of Posthospital Death*

 
Posthospital Well-Being
There were 906 patients who had been discharged from the hospital and were still living 180 days after study entry, corresponding to a mean (± SD) time of 150 ± 26 days after hospital discharge. At that time point, KADL, DASI, and GQOL scores were obtained in 681, 427, and 678 of these patients, respectively. Approximately 5 months after hospital discharge, there was still substantial impairment of functional capacity and QOL, as follows: (1) 48% of patients needed help with at least one ADL; (2) the median DASI score was 20 (IQR, 15 to 25); and (3) 27% of patients rated their QOL as fair or poor, with another 37% rating it as only good. However, the deficits present at posthospital follow-up had largely been present before hospitalization. Only 24% of patients needed assistance with more ADLs on posthospital follow-up than before hospitalization. Similarly, only 11% of patients had interval worsening of their GQOL score. On the other hand, 61% of DASI scores were worse at follow-up. If consistent deterioration of well-being for a given patient is taken as numerical worsening in any of the three well-being variables while none of the others improved, then just 21% had consistent worsening from prehospitalization to posthospitalization. Correlation coefficients between prehospital and posthospital values for KADL, GQOL, and DASI scores were 0.47, 0.23, and 0.46, respectively, all of which were statistically significant.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Conclusions
 References
 
This study assessed survival and well-being among patients who survived hospitalization for severe ARF requiring mechanical ventilation. Our most important finding was that extended survival is common. Approximately half of these patients were still alive 5 years later. Our second finding was that, while functional capacity and QOL were significantly impaired approximately 5 months after hospital discharge, most of that impairment had been present prior to hospitalization. Thus, little of the long-term impairment can be considered a consequence of having experienced ARF or its therapy.

Before reviewing the prior literature, we must address the nature of our patient cohort, and the distinction between ARF and ARDS. Although the SUPPORT definition of ARF did not explicitly include a measure of gas exchange, the patients included in this report were unquestionably experiencing severe respiratory failure. Their median PaO2/FIO2 ratio was 128 (IQR, 85 to 190). All patients required mechanical ventilation, lasting a median of 8 days (IQR, 4 to 15) in the 985 patients in whom it was tracked. Although the SUPPORT was performed before the American-European Consensus Conference had formalized the case definition of ARDS,26 it is likely that many of our patients would have qualified as having ARDS by that definition, since the PaO2/FIO2 ratio was < 200 in 78%, half had bilateral infiltrates seen on their initial CXRs, and the most common alternative causes of respiratory failure, severe CHF, severe COPD, and other chronic lung diseases, were excluded. Even more of our patients likely satisfied the ARDSNet case definition of ARDS,27 which required a PaO2/FIO2 ratio of <300 (seen in 94% of our patients), bilateral infiltrates seen on CXR, and the absence of left atrial hypertension, and, as in the SUPPORT definition of ARF, excluded those patients with severe chronic lung diseases. Beyond any specific definitions, however, the literature is not clear on whether outcomes in ARDS differ from those of ARF more generally, after adjusting for severity of illness and other factors. Two studies2829 found that mortality in ARDS 1 month after onset was double that of non-ARDS patients. However, in a larger study,30 the 90-day survival rate for patients with ARDS was indistinguishable from that of ARF patients without ARDS. In addition, Davidson et al8 found that after adjusting for age, comorbidity, and severity of acute illness, ARDS per se did not influence the survival of patients with sepsis or trauma. Thus, it is reasonable to compare our results with prior ones assessing outcomes from patients with either ARF or ARDS.

Prior studies have assessed long-term survival after ARF or ARDS. The rate of survival to 1 year after ARDS onset was 56% in a cohort of 104 patients reported by Angus et al,7 compared with 47% here. Herridge et al6 observed that 89% of 109 survivors of ARDS were alive 1 year after hospital discharge, compared with 75% in our study. In another study,8 Kaplan-Meier analysis among 58 hospital survivors of sepsis-induced ARDS showed that 70% were alive 1,250 days after hospital discharge, compared to 59% in our study. Older age of the patients in the current study likely contributed to the lower survival rate in comparison with these previously published rates. However, in the largest prior study, Luhr et al30 reported on 1,231 ARF patients with a mean age of 63 years. They found that 59% of patients were alive 90 days after ARF onset, a value that is indistinguishable from the 58% in our patient cohort, who had a similar mean age of 59 years.

Additional perspective about posthospital survival after ARF was achieved by comparison with the following: (1) an age-matched, race-matched, and gender-matched sample of the general US population31; and (2) the 2,951 hospital survivors in the SUPPORT who required ICU admission and had received diagnoses other than ARF.12 The survival trajectory of the ARF patients was markedly different than that for the general population sample. This comparison group would have experienced death rates of 2.4% at 365 days after hospital discharge, and 13.3% at 2,000 days. This compares to 25.5% and 49.1%, respectively, for the ARF cohort. The group of SUPPORT patients without ARF had death rates of 35.0% at 365 days and 59.7% at 2,000 days after hospital discharge. To further assess how survival in this latter group compared to that of ARF patients, we generated a Cox survival model that incorporated the same covariates as those for the ARF survival model (omitting PaO2/FIO2 ratio, which was acquired only for ARF patients) with the addition of a categoric variable representing ARF status. This model verifies that ARF patients had better long-term survival rates than did this comparison group (HR, 0.83; 95% confidence interval [CI], 0.74 to 0.93).

This study has confirmed the results of others78 in associating posthospital mortality after ARF with older age, worse prehospital functional status, more physiologic derangement at onset, and comorbid conditions. There are few data on the relationship of comorbidities to posthospital survival after ARF or ARDS. Luhr and coworkers3032 addressed this topic using Cox survival models in two overlapping data sets from Scandinavia. In the first,30 concomitant immunosuppressed states conferred a higher hazard for 90-day mortality among ARF patients (HR, 1.22). In their later work,32 the presence of any of the five APACHE II comorbid states33 was associated with an HR of 1.31 among ARF patients without ARDS, but was not significantly associated with the hazard for death to 90 days among those with ARDS. In the current data set, diabetes mellitus, renal disease, cirrhosis, HIV or other immunosuppressed states, cardiovascular disease, and malignancy were associated with significant HRs for posthospital death, ranging from 1.30 to 2.22 (Table 4). Such associations do not seem surprising, as these conditions influence survival among the general population as well.

The associations with hospital discharge location and hospital DNR status have not been previously reported. One can speculate about explanations for the higher death rate among those who do not, or cannot, return home after hospitalization. The possibilities include that failure to return home is (1) an independent marker for physiologic or functional impairment, (2) due to poor care in the nursing home or rehabilitation setting, (3) an indication that they were discharged from the hospital prematurely, or (4) that the lack of family or other external support necessitating such a hospital discharge disposition is itself associated with worse long-term outcomes. Similarly, having had a DNR order while hospitalized (1) may be a marker, not included elsewhere in the model, for more severe illness at hospital discharge that was either preexisting or was acquired as a consequence of ARF, and/or (2) was associated with a posthospital DNR status that, in turn, relates to less aggressive care after hospital discharge.

Figure 1 demonstrates that during the first year after hospital discharge the death rate is steep, and then flattens out and approaches the slope of the general population. We explored this finding by comparing hospital survivors who died within 1 year with those who died > 1 year later. This analysis identified the possible contributors to this phenomenon, as the former patients were significantly more likely to have malignancy (31% vs 19%, respectively), to have had a DNR order while in the hospital (31% vs 11%, respectively), and to be less likely to have been discharged from the hospital to home (53% vs 71%, respectively).

Cross-sectional studies,673435363738394041 using various survey instruments, have assessed well-being among ARDS or ARF patients after hospital discharge. A number of studies42 have assessed health-related QOL after ARDS or ARF using the short-form 36, an instrument assessing multiple domains rated on a scale of 0 (worst) to 100 (best). These have reported scores of 55 to 60 in the physical functioning domain, and 50 to 60 in the general health domain at 6 to 15 months after the onset of symptoms.6363741 Rescaling our most analogous well-being variables at 180 days after onset into similar scales of 0 to 100, these findings compare with a mean of 45 for the DASI activity score, and 53 for the GQOL score. In a study of 55 ARDS patients, Hopkins et al37 observed that none required assistance with any ADLs at 1 year after the onset of ARDS. In comparison, 48% of our patients required assistance with one or more ADLs 6 months after the onset of ARDS. While this difference could be related to the differing follow-up intervals, or to the older age of our cohort (mean age, 59 years vs 46 years, respectively), it seems more likely to be related to the fact that 87% of our cohort required assistance with one or more ADLs prior to hospitalization. This parameter was not reported in the study by Hopkins et al.37 Indeed, only one prior study has related well-being after respiratory failure to that before respiratory failure. In 132 patients who were studied 1 year after the onset of ARDS, Angus et al7 found that the quality of well-being score correlated weakly with the prehospital Karnofsky performance status score (r = 0.27; p = 0.03). The most analogous of our correlations was between prehospital GQOL and DASI scores at 180 days after the onset of ARDS (r = 0.22; p < 0.001). Because of the emotional adaptation to disability that occurs,4344 these weak correlations between QOL score and functional status are not evidence against our finding of a strong relationship between prehospital and posthospital well-being.

The major strengths of this study are its size, length of follow-up, and prospective data collection. To our knowledge, this study is larger than any previous published prospective study of ARF patients, with the longest systematic follow-up as well. The largest prior prospective study30 of ARF assessed outcomes to 90 days after onset in 1,241 patients, compared with 1,722 in our study. The largest prior investigation8 of outcomes beyond 3 months after ARF had 144 hospital survivors. The longest prior comprehensive, systematic follow-up of a patient cohort was of 20 patients for a mean of 1.6 years,40 compared with a mean of 2.1 years in our study.

Some limitations of our findings result from the definition of ARF used in the SUPPORT study, which was completed before the development of the current case definitions of ARDS.2627 First, the inclusion criteria of ARF due to either ARDS or pneumonia were left to the judgment of the individual clinicians, rather than being predefined. Thus, it is difficult to judge how closely our patients who were given a clinical designation of ARDS or pneumonia (29% and 84%, respectively) corresponded to the current definitions. Although we chose not to incorporate the SUPPORT designation of ARDS and pneumonia into our analysis, their inclusion in the Cox regression model of posthospital survival had no significant effect (data not shown). Second, the definition of ARF in the SUPPORT excluded patients with severe CHF, severe COPD, and a group of other chronic or subacute lung diseases. While these exclusions make us unable to apply our current results to such patients, very similar exclusions were used in the ARDSNet case definition of ARDS.27 The long-term outcomes of patients in the SUPPORT with severe CHF and severe COPD have been reported previously.4546

Another potential limitation of the study is that care could have improved in the 9 years since the SUPPORT ended. This is unlikely in light of short-term survival, long-term survival, and long-term well-being among ARF patients in the SUPPORT, which is similar to those from more recent data sets. The hospital survival rate of 62.4% is similar to the rate of 60 to 70% in the ARDSNet study,27 and the 60% ICU survival rate found by Herridge et al.6 The 67.2% survival rate at 30 days is practically identical to the 28-day survival rate of 69.5% in the study by Angus et al.7


   Conclusions
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusions
References
 
This analysis of a large, prospective cohort with severe ARF requiring mechanical ventilation who survived to hospital discharge has shown that long-term survival among such patients is common. About half of these patients were alive 5 years later. Moreover, although half or more of the survivors had impaired functional capacity and QOL after hospital discharge, only a small fraction of those impairments can be considered to be sequelae of the respiratory failure or its therapy. These findings, along with the short-term survival rates of 60 to 70% reported here and in other studies,6727 are relevant to the care decisions for such critically ill patients.


   Footnotes
 
Abbreviations: ADL = activity of daily living; APACHE = acute physiology and chronic health evaluation; APS = acute physiology score; ARF = acute respiratory failure; CHF = congestive heart failure; CI = confidence interval; CXR = chest radiograph; DASI = Duke activity status index; DNR = do-not-resuscitate; FIO2 = fraction of inspired oxygen; GQOL = global quality of life score; HR = hazard ratio; IQR = interquartile range; KADL = Katz index of activity of daily living; LOS = length of stay; MI = myocardial infarction; QOL = quality of life; SUPPORT = Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment; TISS = therapeutic intervention scoring system

This research was supported by a grant from the Robert Wood Johnson Foundation.

Received for publication January 28, 2004. Accepted for publication June 23, 2004.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusions
 References
 

  1. Groeger, JS, Guntupalli, KK, Strosberg, M, et al (1993) Descriptive analysis of critical care units in the United States: patient characteristics and intensive care unit utilization. Crit Care Med 21,279-291[ISI][Medline]
  2. Vincent, JL, Sakr, Y, Ranieri, VM Epidemiology and outcome of acute respiratory failure in intensive care unit patients. Crit Care Med 2003;31,S296-S299[CrossRef][ISI][Medline]
  3. Rubenfeld, GD Epidemiology of acute lung injury. Crit Care Med 2003;31,S276-S284[CrossRef][ISI][Medline]
  4. Heyland, DK, Guyatt, G, Cook, DJ, et al Frequency and methodologic rigor of quality-of-life assessments in the critical care literature. Crit Care Med 1998;26,591-598[CrossRef][ISI][Medline]
  5. Vazquez Mata, G, Rivera Fernandez, R, Gonzalez Carmona, A, et al Factors related to quality of life 12 months after discharge from an intensive care unit. Crit Care Med 1992;20,1257-1262[ISI][Medline]
  6. Herridge, MS, Cheung, AM, Tansey, CM, et al One-year outcomes in survivors of the acute respiratory distress syndrome. N Engl J Med 2003;348,683-693[Abstract/Free Full Text]
  7. Angus, DC, Musthafa, AA, Clermont, G, et al Quality-adjusted survival in the first year after the acute respiratory distress syndrome. Am J Respir Crit Care Med 2001;163,1389-1394[Abstract/Free Full Text]
  8. Davidson, TA, Rubenfeld, GD, Caldwell, ES, et al The effect of acute respiratory distress syndrome on long-term survival. Am J Respir Crit Care Med 1999;160,1838-1842[Abstract/Free Full Text]
  9. SUPPORT Study. Study To Understand Prognoses and Preferences for Outcomes and Risks of Treatment; study design. J Clin Epidemiol 1990;43(suppl),1S-123S[CrossRef]
  10. The SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. JAMA 1995;274,1591-1598[Abstract]
  11. Knaus, WA, Wagner, DP, Draper, EA, et al The APACHE III prognostic system. Chest 1991;100,1619-1636[Medline]
  12. Murphy, DJ, Knaus, WA, Lynn, J Study population in SUPPORT: patients (as defined by disease categories and mortality projections), surrogates, and physicians. J Clin Epidemiol 1990;43(suppl),11S-28S
  13. Knaus, WA, Harrell, FE, Lynn, J, et al The SUPPORT prognostic model: objective estimates of survival for seriously ill hospitalized adults; study to understand prognoses and preferences for outcomes and risks of treatments. Ann Intern Med 1995;122,191-203[Abstract/Free Full Text]
  14. Keene, AR, Cullen, DJ Therapeutic intervention scoring system: update 1983. Crit Care Med 1983;11,1-3[ISI][Medline]
  15. Oye, RK, Landefeld, CS, Jayes, RL Outcomes in SUPPORT. J Clin Epidemiol 1990;43(suppl),83S-87S
  16. Chalom, R, Raphaely, RC, Costarino, AT, Jr Hospital costs of pediatric intensive care. Crit Care Med 1999;27,2079-2085[CrossRef][ISI][Medline]
  17. Shwartz, M, Young, DW, Siegrist, R The ratio of costs to charges: how good a basis for estimating costs? Inquiry 1995;32,476-481[ISI][Medline]
  18. Teno, JM, Hakim, RB, Knaus, WA, et al Preferences for cardiopulmonary resuscitation: physician-patient agreement and hospital resource use; the SUPPORT Investigators. J Gen Intern Med 1995;10,179-186[ISI][Medline]
  19. Phillips, RS, Goldman, L, Bergner, M Patient characteristics in SUPPORT: activity status and cognitive function. J Clin Epidemiol 1990;43(suppl),33S-36S
  20. Landefeld, CS, Phillips, RS, Bergner, M Patient characteristics in SUPPORT: functional status. J Clin Epidemiol 1990;43(suppl),37S-39S
  21. Wu, AW, Damiano, AM, Lynn, J, et al Predicting future functional status for seriously ill hospitalized adults: the SUPPORT prognostic model. Ann Intern Med 1995;122,342-350[Abstract/Free Full Text]
  22. National Center for Health Statistics. National death index. Available at: http://www.cdc.gov/nchs/r&d/ndi/ndi.htm. Accessed December 3, 2003
  23. Kleinbaum, DG Survival analysis: a self-learning text. 1996 Springer-Verlag. New York, NY:
  24. Hosmer, DW, Lemeshow, S Applied logistic regression. 1989 Wiley. New York, NY:
  25. van Houwelingen, NC Validation, calibration, revision and combination of prognostic survival models. Stat Med 2000;19,3401-3415[CrossRef][ISI][Medline]
  26. Bernard, GR, Artigas, A, Brigham, KL, et al The American-European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994;149,818-824[Abstract]
  27. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342,1301-1308[Abstract/Free Full Text]
  28. Roupie, E, Lepage, E, Wysocki, M, et al Prevalence, etiologies and outcome of the acute respiratory distress syndrome among hypoxemic ventilated patients. Intensive Care Med 1999;25,920-929[CrossRef][ISI][Medline]
  29. Bersten, AD, Edibam, C, Hunt, T, et al Incidence and mortality of acute lung injury and the acute respiratory distress syndrome in three Australian States. Am J Respir Crit Care Med 2002;165,443-448[Abstract/Free Full Text]
  30. Luhr, OR, Antonsen, K, Karlsson, M, et al Incidence and mortality after acute respiratory failure and acute respiratory distress syndrome in Sweden, Denmark, and Iceland: the ARF Study Group. Am J Respir Crit Care Med 1999;159,1849-1861[Abstract/Free Full Text]
  31. Centers for Disease Control and Prevention, National Center for Health Statistics. Life tables. Available at: http://www.cdc. gov/nchs/products/pubs/pubd/lftbls/life/1966.htm. Accessed November 16, 2003
  32. Luhr, OR, Karlsson, M, Thorsteinsson, A, et al The impact of respiratory variables on mortality in non-ARDS and ARDS patients requiring mechanical ventilation. Intensive Care Med 2000;26,508-517[CrossRef][ISI][Medline]
  33. Knaus, WA, Draper, EA, Zimmerman, JE APACHE II: a severity of disease classification system. Crit Care Med 1985;13,818-829[ISI][Medline]
  34. Schelling, G, Stoll, C, Vogelmeier, C, et al Pulmonary function and health-related quality of life in a sample of long-term survivors of the acute respiratory distress syndrome. Intensive Care Med 2000;26,1304-1311[CrossRef][ISI][Medline]
  35. Schelling, G, Stoll, C, Haller, M, et al Health-related quality of life and posttraumatic stress disorder in survivors of the acute respiratory distress syndrome. Crit Care Med 1998;26,651-659[CrossRef][ISI][Medline]
  36. Weinert, CR, Gross, CR, Kangas, JR, et al Health-related quality of life after acute lung injury. Am J Respir Crit Care Med 1997;156,1120-1128[Abstract/Free Full Text]
  37. Hopkins, RO, Weaver, LK, Pope, D, et al Neuropsychological sequelae and impaired health status in survivors of severe acute respiratory distress syndrome. Am J Respir Crit Care Med 1999;160,50-56[Abstract/Free Full Text]
  38. Davidson, RA, Caldwell, ES, Curtis, JR, et al Reduced quality of life in survivors of acute respiratory distress syndrome compared with critically ill control patients. JAMA 1999;281,354-360[Abstract/Free Full Text]
  39. McHugh, LG, Milberg, JA, Whitcomb, ME, et al Recovery of function in survivors of the acute respiratory distress syndrome. Am J Respir Crit Care Med 1994;150,90-94[Abstract]
  40. Cooper, AD, Ferguson, ND, Hanly, PJ, et al Long-term follow-up of survivors of acute lung injury: lack of effect of a ventilation strategy to prevent barotrauma. Crit Care Med 1999;27,2616-2621[CrossRef][ISI][Medline]
  41. Orme, J, Romney, JS, Hopkins, RO, et al Pulmonary function and health-related quality of life in survivors of acute respiratory distress syndrome. Am J Respir Crit Care Med 2003;167,690-694[Abstract/Free Full Text]
  42. Ware, JE, Snow, KK, Kosinski, M, et al SF-36 health survey manual & interpretation guide. 1993 Health Institute, New England Medical Center. Boston, MA:
  43. Lepiege, A, Hunt, S The problem of quality of life in medicine. JAMA 1997;278,47-50[Abstract]
  44. Winter, L, Lawton, MP, Ruckdeschel, K Preferences for prolonging life: a prospect theory approach. Int J Aging Hum Dev 2003;56,155-170[CrossRef][ISI][Medline]
  45. Jaagosild, P, Dawson, NV, Thomas, C, et al Outcomes of acute exacerbation of severe congestive heart failure. Arch Intern Med 1998;158,1081-1089[Abstract/Free Full Text]
  46. Connors, AF, Dawson, NV, Thomas, C, et al Outcomes following acute exacerbation of severe chronic obstructive lung disease: the SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments). Am J Respir Crit Care Med 1996;154,959-967[Abstract]



This article has been cited by other articles:


Home page
 ChestHome page
Y.-Y. Chen, A. F. Connors Jr, and A. Garland
Effect of Decisions to Withhold Life Support on Prolonged Survival
Chest, June 1, 2008; 133(6): 1312 - 1318.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
S. K. Aberegg
Two-Year Cognitive, Emotional, and Quality-of-Life Outcomes in Acute Respiratory Distress Syndrome
Am. J. Respir. Crit. Care Med., September 15, 2005; 172(6): 786 - 786.
[Full Text] [PDF]

Home page
 ChestHome page
A. Garland
Improving the ICU: Part 1
Chest, June 1, 2005; 127(6): 2151 - 2164.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Garland, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Garland, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS