American Journal of Respiratory and Critical Care Medicine

There is little information on long-term outcome after acute respiratory distress syndrome (ARDS). We measured quality-adjusted survival in the first year after ARDS in a prospective cohort (n = 200). All patients met traditional criteria for ARDS. Patients with sepsis and acute nonpulmonary organ dysfunction at presentation were excluded. The cohort was healthy before onset of ARDS as evidenced by high functional status (mean Karnofsky Performance Status index: 82.2/100 where ⩾ 80 = able to perform normal activities independently) and minimal comorbid illness (mean Charlson-Deyo comorbidity score: 0.32/17 where 0 = absence of chronic illness). We determined quality-adjusted life-years (QALYs) using the Quality of Well-being (QWB) scale (0 to 1 scale where 1 = optimal well-being), measured at 6 and 12 mo. Survival was 69.5 ± 5.0% at 1 month, fell to 55.7 ± 3.7% at 6 mo, and did not change at 12 mo, yielding a survival of 59 life-years in the first year per 100 patients with ARDS. QWB was low at 6 and 12 mo (0.59 ± 0.015 and 0.60 ± 0.015), yielding a quality-adjusted survival of 36 QALYs per 100 patients (sensitivity range: 21 to 46 QALYs). We conclude that ARDS developing in previously healthy patients is associated with poor quality-adjusted survival. These data are important for cost-effectiveness analyses and long-term care.

The acute respiratory distress syndrome (ARDS) is a rapidly progressive illness that is often fatal despite the provision of advanced technologic support in a modern intensive care unit (ICU). Consequently, the primary objective of most clinical ARDS trials is to evaluate whether interventions reduce short-term (e.g., 1-mo or hospital discharge) mortality. Examples of such studies include recent trials of surfactant, corticosteroids, and reduced volume ventilatory strategies (1-5). However, by focusing only on short-term survival, these trials fail to account for any lingering consequences of ARDS, such as ongoing risk of death beyond 1 mo or reduced quality of life in survivors.

Prior studies have suggested that ARDS survivors may indeed have reduced quality of life (6-10). There are also data suggesting survivors may be at increased risk of death for many months after hospital discharge (11). However, to adequately capture the consequences of ARDS and to assess the potential effects of new therapies or interventions requires integration of both the duration and quality of survival in a common metric—quality-adjusted survival. This is particularly important for cost-effectiveness analysis if effects are to be standardized across different interventions and diseases (12). To date, there are no studies documenting the quality-adjusted survival after ARDS.

In conjunction with a recent multicenter ARDS trial of inhaled nitric oxide therapy, we collected data on the quality-adjusted survival, measured as quality-adjusted life-years (QALYs), in the first year after the onset of ARDS. There were no differences in patient outcomes between treatment arms, and we therefore present the quality-adjusted survival for the entire cohort. Our goals were to provide a first estimate of quality-adjusted survival after ARDS (as observed in this large, multicenter cohort), to explore the extent to which quality-adjusted survival is associated with particular baseline characteristics, and to compare results in ARDS survivors with healthy and sick control subjects.

Patient Selection

Patients were enrolled in this study in conjunction with a Phase II clinical trial of inhaled nitric oxide therapy. The study was conducted on 200 patients at 35 academic, teaching, and community hospitals in the United States. The Phase II clinical results on the first 177 patients enrolled at 30 of the sites have been reported previously (13). The additional 23 patients included in this study were enrolled while data were being cleaned for an interim analysis by the data safety and monitoring board. The institutional review committee of each participating site approved the study, and informed consent was obtained from each patient or her legal representative. The inclusion criteria for the study have been described in detail elsewhere (13) and included onset of ARDS within the preceding 72 h, where ARDS was defined as a ratio of arterial oxygen tension to fraction of inspired oxygen (PaO2 / Fi O2 ) ⩽ 200, bilateral radiographic pulmonary infiltrates, and the absence of left atrial hypertension (by pulmonary artery occlusion pressure ⩽ 18 mm Hg or by clinical examination) (14). Patients were excluded if they presented with sustained hypotension or vasopressor requirement, severe head injury, sepsis, or multiple organ dysfunction syndrome (or a high risk of developing this syndrome, such as patients with ARDS secondary to a nonpulmonary infection source or severe immunocompromise). Importantly, these entry criteria were designed to select patients whose primary problem was ARDS.


Baseline data collected at enrollment included age, sex, severity of illness as measured by the Acute Physiology and Chronic Health Evaluation (APACHE) II score (15), and reason for admission to the ICU (medical, non–trauma-related surgical, or trauma). Information was collected on the hospital length of stay (LOS) and length of survival up to 1 yr from time of enrollment. We tracked patients using telephone numbers and addresses provided by the patient or family at hospital discharge. Patients were deemed lost to follow-up if we were unable to ascertain their survival status after multiple telephone calls.

We measured health-related quality of life (HRQL) using the Quality of Well-being (QWB) (16) scale version 7.0 at 6 and 12 mo after study enrollment. The QWB is a utility measure of HRQL recommended by the Panel on Cost-Effectiveness in Healthcare and Medicine for the calculation of QALYs (17). Although not used previously in ARDS survivors or survivors of other critical illnesses, it has been validated in patients with other pulmonary conditions such as cystic fibrosis and chronic obstructive pulmonary disease (COPD) (18, 19). The QWB was administered by one of three interviewers through a structured telephone interview. Each interviewer was trained in the administration and scoring of the instrument.

The QWB assesses HRQL across two dimensions—function and symptoms. Function is assessed separately across physical activity, social activity, and mobility (16). Patients are prompted with questions relevant to each domain regarding the 6 d before the interview. The results of the questions are integrated to produce a score using weights established from the preferences of a U.S. general population cohort. Importantly, the QWB assesses neither all components of quality of life nor the patient's own perception of the value of their current health state. Rather, the QWB estimates from a societal perspective the utility, or “worth,” of a given health state (i.e., given that a person's health state produces a particular array of symptoms and physical and social limitations, the QWB estimates how much less than optimal society perceives that state to be). The QWB score ranges from 0 to 1 where 0 equals death and 1 equals optimal well-being. The QWB score is multiplied by the time spent in a particular health state to calculate the number of QALYs.

We could not use the QWB to assess premorbid HRQL (because all patients were on mechanical ventilation and unable to complete a detailed interview). Therefore, we used an alternative approach designed to assess HRQL across the same two dimensions of function and symptoms. We measured functional status before the onset of acute illness with the Karnofsky Performance Status index, a simple, objective measure of functional status (20). The Karnofsky index ranges in units of 10 from 0 to 100 where 100 = no limitations and ⩾ 80 indicates the ability to carry on normal activities independently. Premorbid functional status was evaluated at the time of enrollment by asking the patient, if possible, or the patient's surrogate. Premorbid burden of chronic illness was assessed by the Charlson-Deyo comorbidity score, a widely used quantitative measure of chronic illness (range: 0 to 17, where 0 = absence of chronic illness) (21).

Control Groups

We compared the QWB scores from our cohort with those of two control populations: a cohort of patients with cystic fibrosis (n = 43) (22), for whom we had individual data, and a general population cohort (n = 1,356) (23), for whom we had age-specific mean values and distributions for persons living with and without five chronic conditions (arthritis, severe back pain, cataract, hypertension, and sleep disorder). The patients with cystic fibrosis were recruited from the Pittsburgh Cystic Fibrosis Center (22). The general population cohort were residents of Beaver Dam, Wisconsin, recruited for the Beaver Dam Health Outcomes Study (23).

Statistical Analysis

We compared baseline hospitalization and follow-up data between inhaled nitric oxide and control groups and between patients with and without postdischarge follow-up. We used unpaired t tests or Mann-Whitney U tests for continuous data and chi-square or Fisher exact tests for categorical data as appropriate. There were no differences in baseline characteristics or prospectively defined outcomes, including long-term survival and HRQL, between the inhaled nitric oxide and control groups. Therefore, we pooled the data from both treatment arms.

We examined QWB scores at 6 and 12 mo for associations with baseline characteristics using unpaired t tests or Spearman rank correlation as appropriate. We examined differences in scores at 6 and 12 mo using paired t tests and differences across diagnostic categories using one-way analysis of variance (ANOVA) with Bonferroni procedure for post hoc comparisons between categories. We also compared mean QWB scores between the ARDS cohort and the control cohorts by unpaired t tests and displayed age-dependent variation graphically.

The QWB is a rich source of reported symptoms. We explored this source in post hoc analysis both by examining overall symptoms and by classifying symptoms concerning the lower respiratory tract as probably related to ARDS; symptoms concerning the central nervous system, upper respiratory tract, and general constitutional symptoms as possibly related; and symptoms concerning other systems as unlikely to be related to ARDS (Table 1). Two reviewers (D.A. and G.C.) classified the symptoms independently with very good interrater agreement [kappa statistic (24): 0.86]. We examined differences by diagnostic category using ANOVA as previously described.


AllHospital Survivors (n = 132)p Value*
Long-term Survival Known Long-term Survival Unknown
Age, yr 48.6 ± 17.544.7 ± 15.039.8 ± 15.90.34
Sex, % male66.
APACHE II 17.1 ± 6.316.0 ± 6.417.7 ± 6.80.43
Premorbid Karnofsky score 87.9 ± 16.890.3 ± 14.088.6 ± 23.40.26
Charlson-Deyo comorbidity
 score  0.27 ± 0.570.22 ± 0.480.11 ± 0.420.15
Medical, % 5655.784.60.07
ICU length of stay
 (median, range)16, 0–9323.5, 0–8522, 0–850.79
Hospital length of stay
 (median, range)26, 0–11734, 0–10929, 0–930.53

*Between the two groups of hospital survivors.

Long-term survival was assessed at 6 and 12 mo. The long-term postdischarge survival status is unknown if no follow-up information could be obtained at either time-points.

Values are mean ± SD.

We assessed survival to 1 yr by the Kaplan-Meier method. We calculated quality-adjusted survival in the first year by multiplying the overall survival in the first year (as determined by the area under the Kaplan-Meier curve) by the average QWB score. This represented the best-case scenario because it assumed that quality of life while still in the hospital was equal to that measured later in the year. Because this assumption might overestimate in-hospital quality of life, we conducted a sensitivity analysis calculating separate quality-adjusted survival curves for each patient and varying the mean daily in-hospital QWB score from 0 to 0.4. We also calculated an upper limit on quality-adjusted survival by assuming that all patients in whom QWB had not been obtained after hospital discharge had survived to 1 yr with perfect QWB scores. We constructed individual curves assuming a straight line from hospital discharge to 6 mo and from 6 mo to 12 mo. For 13 patients who had missing 12-mo scores, we assumed their scores equaled those measured at 6 mo. We expressed results as the number of life-years and the number of QALYs accrued in the first year per 100 patients with ARDS.

We assumed statistical significance at p < 0.05, except where we performed Bonferroni corrections. We performed data management and statistical analyses using Excel (Microsoft Corporation, Redmond, WA; and DataDesk (Data Descriptions, Inc., Ithaca, NY; software programs.

The study cohort (n = 200) was young, predominantly male, with good functional status and low incidence of chronic illness at baseline (Table 1). Reasons for admission were medical (56%), surgical (nontrauma) (20%), and trauma (24%). Mortality at 28 d was 30.5% (n = 61). Of those enrolled, 16% (n = 32) and 22.5% (n = 45) were lost to follow-up at 6 and 12 mo. Those lost to follow-up after hospital discharge were not statistically significant different in baseline characteristics from those for whom we had follow-up to 1 yr (Table 1).

Survival estimated by Kaplan-Meier was 69.5 ± 5.0% at 1 mo and decreased to 55.7 ± 3.7% at 6 mo, with no change at 12 mo (Figure 1). The 12-mo survivors were younger (45.1 versus 56.2 yr, p < 0.05) and had a higher mean premorbid functional status (Karnofsky status: 90.6 versus 78.6, p < 0.05) and lower mean APACHE II scores (14.7 versus 18.8, p < 0.05) at enrollment than nonsurvivors. Survivorship in the first year was 59 life-years per 100 patients with ARDS. As shown in Figure 2, at 6 and 12 mo, the mean QWB was low and stable with narrow confidence intervals (0.59 ± 0.015 and 0.60 ± 0.015 at 6 and 12 mo, respectively, p = 0.48). Integrating the Kaplan-Meier survival with the mean QWB scores produced a quality-adjusted survival in the first year of 36 QALYs per 100 patients with ARDS (Figure 1). In the sensitivity analysis, this estimate fell further to 28, 24, and 21 QALYs assuming mean in-hospital QWB scores of 0.4, 0.2, and 0, respectively. Assuming perfect health (QWB = 1) and no postdischarge deaths in those lost to follow-up, the number of QALYs per 100 patients increased to 46. Age, sex, comorbidity, and APACHE II score were not associated with HRQL, as measured by the QWB, at 1 yr. However, there was a weak association between premorbid Karnofsky status and 12-mo QWB scores (r = 0.27, p = 0.03). There were no statistically significant differences in 6-mo or 12-mo QWB scores between medical, surgical, and trauma patients.

The mean QWB scores of the ARDS cohort at both 6 and 12 mo (p < 0.001) were significantly lower than a control population of patients with cystic fibrosis (0.76 ± 0.035) (Figure 2). Scores varied by age. Figure 3 shows the distribution of mean QWB scores by age for the patients with ARDS and the cystic fibrosis and general population control subjects.

Although the QWB is not traditionally disaggregated into its component subscores, post hoc analysis showed that the symptom component of the QWB score accounted for approximately 70% of the decrement in QWB from perfect health at both 6 and 12 mo (Figure 4). The most frequent symptoms at 6 and 12 mo are shown in Table 2. Although the most common categories of symptoms were general musculoskeletal and constitutional symptoms, respiratory symptoms were present in almost half of all patients at both 6 and 12 mo. There were no statistically significant differences in the distribution of symptoms that were probably, possibly, or unlikely to be related to ARDS between medical, surgical, and trauma patients.

Table 2.  SYMPTOMS AT 6 AND 12 mo*

Symptom GroupsProportion of Patients Expressing Symptom (%)
6 mo12 mo
Probably related to ARDS4743
 Lower respiratory tract4340
 Hoarseness/dysphonia20 5
Possibly related to ARDS6778
 Depression, anxiety, or insomnia5746
 Upper respiratory tract2819
Unlikely to be related to ARDS8073
 Ear and dental2416

*We classified all symptoms as probably, possibly, or unlikely to be related to ARDS using dual independent review with very good interrater agreement (kappa statistic: 0.86).

This study demonstrates that patients who appear in good health until presenting with primary ARDS are at significant risk of death for several months beyond the traditional endpoint of 28 d. Furthermore, the quality of life in those that do survive is markedly impaired. Integrating these two findings produces a low quality-adjusted survival in the first year after ARDS. These findings are both profound and disappointing considering the considerable advances in intensive care management for acute lung injury (4, 5) and observations that short-term mortality has been decreasing (25).

Although mortality beyond 28 d after critical illness has traditionally been considered negligible or unrelated to the acute illness (26), evidence to the contrary continues to accumulate (11, 27, 28). Quartin and coworkers showed a significant number of late deaths following sepsis and acute organ dysfunction that were distributed during the first year in an exponential decay, suggesting a direct relationship to the acute event (28). Davidson and coworkers showed similarly that patients discharged alive with ARDS, trauma, or sepsis had a high likelihood of death in the subsequent months (11). Davidson and coworkers, however, did not find that ARDS per se increased the risk of death over and above that posed by other components of the acute event (i.e., sepsis or trauma). Our study confirms the observation of a continued high likelihood of death for several months beyond hospital discharge in patients with ARDS. In our cohort, the increased risk of death appeared to persist for 6 mo.

This observation has a practical implication for future study design in ARDS. A therapy that has changed mortality at Day 28 may simply be delaying death by a few days or weeks without necessarily changing the overall mortality from the critical illness. Similarly, the beneficial effect of a therapy that decreases the incidence of late deaths from ARDS would not be captured at 28 d. Thus, we recommend that follow-up in ARDS trials be longer than 28 d. From our data, 6 mo appears to be appropriate.

Despite a 1994 consensus statement that “future outcome evaluation of intensive care should incorporate quality of life” (29), less than 2% of all ICU studies evaluate HRQL (30) and there are few studies of quality of life after ARDS. One U.S. study found a high frequency of pulmonary and psychological symptoms in ARDS survivors, and a German study noted survivors often showed signs of posttraumatic stress disorder (6, 7). There are also data from Utah demonstrating cognitive impairment and low HRQL in survivors at 1 yr (8).

The reason for low HRQL after ARDS is unclear. One possibility raised in earlier studies is that patients who develop ARDS often have preexisting chronic illness with low HRQL before the onset of ARDS. We were unable to assess premorbid HRQL but, unlike most prior studies of quality of life after ARDS, we studied a cohort of patients who were functionally highly independent and free of chronic comorbidities before ARDS. Thus, their average HRQL before the acute event was likely to have been typical of the general population. However, our ARDS cohort had worse scores than both healthy and chronically ill control cohorts. Prior studies suggest the mean QWB score for a healthy population of a similar age to that of our cohort is at least 0.8 (18). Our scores in patients with ARDS were also similar to those reported in a cohort of patients suffering from major human immunodeficiency virus (HIV)-related conditions (31). Thus, the post-ARDS QWB scores in our cohort were lower than one would expect if the patients had recovered fully, represented a considerable illness burden, and probably occurred as a direct result of their critical illness.

It is also possible that the decrement in HRQL was not due to ARDS itself but to other aspects of their critical illness, such as loss of limb from trauma. However, we excluded patients presenting with acute nonpulmonary organ dysfunction, we found no differences in HRQL between medical, surgical, and trauma subgroups, and respiratory symptoms featured prominently as a cause of the decrement in QWB scores. Furthermore, Davidson and colleagues demonstrated that patients with ARDS have worse HRQL than acutely ill, matched control subjects, measured both by a general health instrument and an instrument focusing on the impact of respiratory conditions on HRQL (9). Given that the late sequelae of ARDS include both a long, debilitating hospitalization, during which numerous short-term and long-term complications can occur, and a persistent restrictive pulmonary defect (10), it is not surprising that ARDS would result in a low, long-term HRQL.

This finding of low HRQL highlights important opportunities for improved care. For example, anxiety, depression and posttraumatic stress disorder commonly follow ARDS (6, 7) and a high index of suspicion during long-term follow-up may lead to earlier recognition and treatment. Excessive anxiety and depressive symptoms were reported by half of our patients at 1 yr and treatment of major depression may have improved quality-adjusted survival (32). Pulmonary rehabilitation and behavioral therapy appear to improve HRQL in patients with chronic obstructive pulmonary disease (COPD) (33) and may also be beneficial in ARDS. Many ARDS patients may receive care from different physicians at different time-points. For example, the care of ARDS patients while in the ICU is more likely to be provided by critical care or pulmonary physicians than by generalists (34). Yet, long-term follow-up may be the responsibility of the patient's primary physician. Ensuring adequate communication between caregivers about the potential for late sequelae will therefore be important if care is to be optimized.

Our study also highlights the importance of considering the effects of ICU therapies and interventions on quality-adjusted survival. Quality-adjusted survival integrates two of the most basic and important patient-valued and society-valued objectives—to prolong life and to preserve or enhance quality of life. Viewed from this perspective, therapies that selectively improve the quality of life in survivors could be as valuable as therapies that decrease mortality. Although ARDS interventions appeared to have beneficial physiologic effects in preliminary studies, they have often failed to decrease mortality in subsequent, large, multicenter trials, ending further investigation. Yet, if reduced lung injury resulted in less dyspnea and, consequently, improved HRQL, quality-adjusted survival would increase even in the absence of a change in mortality. Because mortality rates may be difficult to change in patients acutely ill with multiple life-threatening disorders, quality-adjusted survival may be more sensitive to changes in the severity of acute lung injury and subsequent lung function.

By measuring outcome in ARDS interventional trials using QALYs, we also facilitate cost-effectiveness comparisons with interventions in other diseases (12) and provide information about the impact of ARDS and its treatment to guide decisions by physicians, patients, and families (18). QALYs have been measured across a wide spectrum of medical conditions ranging from major depression (32) and schizophrenia (35) to myocardial infarction (36) and cancer (37). This study is the first to measure QALYs after ARDS and is one of only a few studies to measure QALYs after any critical illness (38). Although debate continues regarding the most appropriate choice of HRQL instrument and the relative importance of cost-effectiveness estimates in determining resource allocation (39), no argument exists as to the need to make such measurements. Our data demonstrate that this outcome measure is feasible and that it provides important insight into the course of ARDS.

A limitation of our study, and all prior studies of HRQL in critical illness, was the inability to directly measure premorbid HRQL. Further research into simpler HRQL instruments that can be administered to patients or patient surrogates with acceptable results is needed to surmount this problem. However, we did measure both premorbid illness and functional status, the same two dimensions measured by the QWB instrument. Our study cohort comprised patients selected for a randomized trial. As such, our study population may be skewed and unrepresentative of the overall ARDS population. Specifically, we selected a cohort of patients who were previously healthy and presented without either sepsis or other organ involvement. Davidson and colleagues observed that patients developing ARDS secondary to sepsis had worse quality of life than patients developing ARDS secondary to trauma (9). It is therefore likely that quality-adjusted survival is even lower in ARDS patients overall. Although 22% of patients were lost to follow-up at 1 yr, they were similar to those not lost to follow-up, the completeness of long-term follow-up was similar to that of other long-term follow-up studies of patients with ARDS (9, 10), and the impact on our findings, as suggested by the sensitivity analysis, was minimal. Our analysis of symptoms should be interpreted with caution because it was retrospective and relied on a previously unvalidated classification. Our failure to find many associations between baseline characteristics and subsequent HRQL may have been the result of a lack of power.

In conclusion, quality of life and quality-adjusted survival are low after ARDS. Short-term mortality significantly underestimates long-term mortality and does not capture the low quality of life in survivors. Measuring quality-adjusted survival after ARDS can be done within the context of a large multicenter clinical trial, facilitates cost-effectiveness comparisons across diseases and therapies, better captures patient-valued and society-valued outcomes, and may be more sensitive to the effect of therapies than short-term mortality.

The authors are grateful for the thoughtful comments and review by Mark Kamlet, Ph.D., of Carnegie Mellon University, Pittsburgh, Pennsylvania.

1. Weg JG, Balk RA, Tharratt RS, Jenkinson SG, Shah JB, Zaccardelli D, Horton J, Pattishall ENSafety and potential efficacy of an aerosolized surfactant in human sepsis-induced adult respiratory distress syndrome. JAMA272199414331438
2. Meduri GU, Headley AS, Golden E, Carson SJ, Umberger RA, Kelso T, Tolley EAEffect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: a randomized controlled trial. JAMA2801998159165
3. Stewart TE, Meade MO, Cook DJ, Granton JT, Hodder RV, Lapinsky SE, Mazer CD, McLean RF, Rogovein TS, Schouten BD, et al.. Evaluation of a ventilation strategy to prevent barotrauma in patients at high risk for acute respiratory distress syndrome. Pressure- and Volume-Limited Ventilation Strategy Group. N Engl J Med3381998355361
4. 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.
5. Amato MB, Barbas CS, Medeiros DM, Magaldi RB, Schettino GP, Lorenzi-Filho G, Kairalla RA, Deheinzelin D, Munoz C, Oliveira R, et al.. Effect of a protective-ventilation strategy on mortality in the acute respiratory distress syndrome. N Engl J Med3381998347354
6. Weinert CR, Gross CR, Kangas JR, Bury CL, Marinelli WAHealth-related quality of life after acute lung injury. Am J Respir Crit Care Med156199711201128
7. Schelling G, Stoll C, Haller M, Briegel J, Manert W, Hummel T, Lenhart A, Heyduck M, Polasek J, Meier M, et al.. Health-related quality of life and posttraumatic stress disorder in survivors of the acute respiratory distress syndrome. Crit Care Med261998651659
8. Hopkins RO, Weaver LK, Pope D, Orme JF, Bigler ED, Larson-Lohr VNeuropsychological sequelae and impaired health status in survivors of severe acute respiratory distress syndrome. Am J Respir Crit Care Med16019995056
9. Davidson TA, Caldwell ES, Curtis JR, Hudson LD, Steinberg KPReduced quality of life in survivors of acute respiratory distress syndrome compared with critically ill control patients. JAMA2811999354360
10. McHugh LG, Milberg JA, Whitcomb ME, Schoene RB, Maunder RJ, Hudson LDRecovery of function in survivors of the acute respiratory distress syndrome. Am J Respir Crit Care Med15019949094
11. Davidson TA, Rubenfeld GD, Caldwell ES, Hudson LD, Steinberg KPThe effect of acute respiratory distress syndrome on long-term survival. Am J Respir Crit Care Med160199918381842
12. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LBfor the Panel on Cost-Effectiveness in Health and Medicine. Recommendations of the Panel on Cost-Effectiveness in Health and Medicine. JAMA276199612531258
13. Dellinger RP, Zimmerman JL, Taylor RW, Straube RC, Hauser DL, Criner GJ, Davis KJ, Hyers TM, Papadakos PJEffects of inhaled nitric oxide in patients with acute respiratory distress syndrome: results of a randomized phase II trial. Inhaled Nitric Oxide in ARDS Study Group. Crit Care Med2619981523
14. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke KJ, Hudson L, Lamy M, Le Gall JR, Morris A, Spragg RThe American–European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med1491994818824
15. Knaus WA, Draper EA, Wagner DP, Zimmerman JEAPACHE II: a severity of disease classification system. Crit Care Med131985818829
16. Kaplan RM, Anderson JPA general health policy model: update and applications. Health Serv Res231988203235
17. Gold MR, Russell LB, Seigel JE, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996.
18. Orenstein DM, Kaplan RMMeasuring the quality of well-being in cystic fibrosis and lung transplantation: the importance of the area under the curve. Chest100199110161018
19. Kaplan RM, Atkins CJ, Timms RValidity of a quality of well-being scale as an outcome measure in chronic obstructive pulmonary disease. J Chronic Dis3719848595
20. Karnofsky DA, Abelmann WH, Craver LF, Burchenal JHThe use of the nitrogen mustards in the palliative treatment of carcinoma. Cancer11948634656
21. Deyo RA, Cherkin DC, Ciol MAAdapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol451992613619
22. Orenstein DM, Nixon PA, Ross EA, Kaplan RMThe quality of well-being in cystic fibrosis. Chest951989344347
23. Fryback DG, Dasbach EJ, Klein R, Klein BE, Dorn N, Peterson K, Martin PAThe Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors. Med Decis Making13199389102
24. Landis JR, Koch GGThe measurement of observer agreement for categorical data. Biometrics331977159174
25. Milberg JA, Davis DR, Steinberg KP, Hudson LDImproved survival of patients with acute respiratory distress syndrome (ARDS): 1983– 1993. JAMA2731995306309
26. Dellinger RPFrom the bench to the bedside: The future of sepsis research. Executive summary of an American College of Chest Physicians, National Institute of Allergy and Infectious Disease, and National Heart, Lung, and Blood Institute Workshop. Chest1111997744753
27. Stauffer JL, Fayter NA, Graves B, Cromb M, Lynch JC, Goebel PSurvival following mechanical ventilation for acute respiratory failure in adult men. Chest104199312221229
28. Quartin AA, Schein RM, Kett DH, Peduzzi PNfor the Department of Veterans Affairs Systemic Sepsis Cooperative Studies Group. Magnitude and duration of the effect of sepsis on survival. JAMA277199710581063
29. Predicting outcome in ICU patients. 2nd European Consensus Conference in Intensive Care Medicine. Intensive Care Med 1994;20:390–397.
30. Heyland DK, Guyatt G, Cook DJ, Meade M, Juniper E, Cronin L, Gafni AFrequency and methodologic rigor of quality-of-life assessments in the critical care literature. Crit Care Med261998591598
31. Kaplan RM, Anderson JP, Patterson TL, McCutchan JA, Weinrich JD, Heaton RK, Atkinson JH, Thal L, Chandler J, Grant IValidity of the Quality of Well-Being Scale for persons with human immunodeficiency virus infection. HNRC Group. HIV Neurobehavioral Research Center. Psychosom Med571995138147
32. Lave JR, Frank RG, Schulberg HC, Kamlet MSCost-effectiveness of treatments for major depression in primary care practice. Arch Gen Psychiatry551998645651
33. Toevs CD, Kaplan RM, Atkins CJThe costs and effects of behavioral programs in chronic obstructive pulmonary disease. Med Care22198410881100
34. Angus DC, Kelley MA, Schmitz R, Popovich JJCurrent and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population? JAMA284200027622770
35. Chouinard G, Albright PSEconomic and health state utility determinations for schizophrenic patients treated with risperidone or haloperidol. J Clin Psychopharmacol171997298307
36. Tsevat J, Duke D, Goldman L, Pfeffer MA, Lamas GA, Soukup JR, Kuntz KM, Lee THCost-effectiveness of captopril therapy after myocardial infarction. J Am Coll Cardiol261995914919
37. Goldhirsch A, Gelber RD, Simes RJ, Glasziou P, Coates ASCosts and benefits of adjuvant therapy in breast cancer: a quality-adjusted survival analysis. J Clin Oncol719893644
38. Kerridge RK, Glasziou PP, Hillman KMThe use of “quality-adjusted life years” (QALYs) to evaluate treatment in intensive care. Anaesth Intensive Care231995322331
39. Detsky AS, Naglie IGA clinician's guide to cost-effectiveness analysis. Ann Intern Med1131990147154
Correspondence and requests for reprints should be addressed to Derek C. Angus, M.B., Ch.B., M.P.H., Room 604 Scaife Hall, Critical Care Medicine, University of Pittsburgh, 200 Lothrop Street, Pittsburgh, PA 15213. E-mail:

Funded in part by INO Therapeutics, Inc., Clinton, NJ.


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record