| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Pulmonary and Critical Care Medicine (Drs. Dasgupta and Stoller), Epidemiology and Biostatistics (Mr. Mascha), Internal Medicine (Dr. Litaker), and the Section of Respiratory Therapy (Dr. Rice), Cleveland Clinic Foundation, Cleveland, OH.
Correspondence to: James K. Stoller, MD, FCCP, Department of Pulmonary and Critical Care Medicine, Desk A-90, Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: The features of the ReSCU include six private beds in a pulmonary inpatient ward staffed by nurses with special pulmonary expertise; 24-h respiratory therapy supervision; bedside and central noninvasive monitoring (ie, continuous pulse oximetry, end tidal capnometry, and ventilator alarms); and a multidisciplinary approach involving dietitians, physical therapists, occupational therapists, social workers, and speech pathologists. All ReSCU patients were cared for primarily by a pulmonary/critical care attending physician and fellow, with consultative input solicited as deemed necessary. The criteria for admission to the ReSCU included hemodynamic stability; absence of an arrhythmia requiring telemetry; and in the attending physician's judgment, the ability to benefit from the ReSCU.
Results: Between August 23, 1993, and August 31, 1997, 212 patients were admitted to the ReSCU. The median age was 68 years old; 55% were women; 86% were white; and 55% were transferred from the medical ICU. Underlying reasons for ventilator dependence were ARDS from a nonsurgical cause (33%), ARDS following surgery (18%), status post-cardiothoracic surgery (13%), status post-thoracic surgery (12%), and COPD (12%). The median length of ReSCU stay was 17 days (interquartile range, 10 to 29 days). Eighteen percent (n = 38) died during the hospitalization. Among the 174 survivors, complete ventilator independence was achieved in 127 patients (60% of the 212 patient cohort), 28 patients were ventilator dependent (13% of 212 patients), and the remaining 19 patients (9%) required partial ventilatory support. Univariate analysis regarding the association of baseline characteristics with death identified lower albumin and transferrin levels, increasing age, and the physician's estimate of lower weaning likelihood as significant correlates of death. In contrast, achieving complete ventilator independence was associated with a higher serum albumin level, a nonmedical ICU referral source, a cause of respiratory failure other than COPD, and a physician's estimate of higher weaning likelihood. To analyze the financial impact of the ReSCU, we assumed that ReSCU patients would have otherwise stayed in the medical ICU and compared the charges (ICU vs ReSCU) with, for a subset of patients, the true costs of ReSCU vs ICU care. Analyses of both charges and cost differences showed similar savings associated with ReSCU care ($13,339 per patient [charges] and $10,694 per patient [costs]).
Conclusions: We conclude the following: (1) the rate of achieving complete ventilator independence in the ReSCU was high; and (2) based on our achieving clinical outcomes, which are comparable to the most favorable rates reported in other series from ventilator units, we conclude that the ReSCU can be an effective and cost-saving alternative to the ICU for carefully selected patients.
Key Words: long-term ventilation • weaning • weaning predictors
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eligibility criteria for ReSCU admission included hemodynamic stability, the lack of need for invasive or continuous ECG monitoring, the presence of a mature tracheostomy, and in the attending pulmonary/critical care physician's judgment, the ability to benefit from the ReSCU. Hemodialytic support could be provided in the ReSCU. The patients already deemed to be ventilator-dependent could be admitted to facilitate the transition to an appropriate alternate-care setting, to optimize the patient's and caregivers' education regarding management of tracheostomy and both invasive and noninvasive ventilators, or to acutely decompress the ICU.
In order to analyze weaning success and predictive clinical features, this study considered patients who developed respiratory failure on the same hospitalization as the ReSCU stay and patients whose ReSCU admission had the intention of weaning. Patients who were unweanable before ReSCU admission or patients who were deemed ventilator-dependent prior to initial hospital admission were excluded from the analysis.
For all patients admitted to the ReSCU, the following data elements
were prospectively recorded: their demographic features; the type of
ICU from which the patient was admitted; their condition leading to
ventilator dependence; the duration of their hospital stay preceding
the ReSCU admission; the duration of their total ReSCU stay; the number
of days they required mechanical ventilation; their first day of
freedom from mechanical ventilation (defined as the first day after
which spontaneous breathing was maintained for 
48 h); their need
for readmission to an ICU; their admission and discharge body weight;
measurements of their nutritional status on ReSCU admission and
discharge (including serum albumin and transferrin level); their vital
status on hospital discharge; and their discharge destination. At the
time of ReSCU admission, the attending pulmonary/critical care
physician and/or the fellow physician was asked to predict the
likelihood that a patient would wean completely, using a 7-point
ordinal rating scale (where 1 signified impossible, and 7 meant
definitely successful).
The causes of respiratory failure leading to ReSCU admission were classified into nine categories as follows: (1) COPD; (2) ARDS with or without multiple organ failure in a nonsurgical patient; (3) ARDS with multiple organ failure in a surgical patient; (4) post-cardiovascular surgery (without ARDS); (5) post-thoracic surgery (without ARDS); (6) after any other surgery; (7) neuromuscular disease; (8) underlying lung disease other than COPD, eg, idiopathic pulmonary fibrosis; and (9) miscellaneous. In instances when more than one of the nine categories applied, the patient was categorized according to the proximate cause of the respiratory failure leading to ReSCU admission. For example, a patient who developed ARDS after coronary artery bypass surgery would be categorized as having ARDS of surgical cause rather than as having post-cardiovascular surgery.
Ventilatory outcomes were classified as follows: (1) complete
ventilator independence; (2) nocturnal ventilation only (with
tracheostomy); (3) noninvasive nocturnal ventilation support (bilevel
pressure ventilation or another noninvasive method); (4) partial
daily (and full nocturnal) mechanical ventilation; and (5) full (24
h/day) mechanical ventilation. The onset of complete ventilator
independence was defined as the first day followed by 48 h of
spontaneous breathing.
The primary outcome measures in this study were the rates of hospital mortality and ventilator independence at the end of the hospital stay. Time-to-event survival models were used to assess associations between these outcomes and baseline factors. Specifically, Cox proportional hazards survival models were used to assess the associations among each outcome and the patient's age, gender, race, source of referral to ReSCU, admission, albumin, transferrin, weight, ICU length of stay before ReSCU admission, cause of respiratory failure, and physician's initial wean estimate. Multivariable Cox models were attempted for each outcome. The likelihood ratio test p value is reported for each association, along with a risk ratio and 95% confidence interval (CI). The risk ratio gives the risk of either dying or becoming ventilator independent at a given point in time for a patient in the risk category compared to a patient in the reference category. Kaplan-Meier survival curves are also shown for several associations.
Costs were calculated using computer software (TSI; Transition Systems, Inc; Boston, MA), a decision support system that assesses direct and indirect costs for all clinical services. Calculations included all costs of care (eg, nursing, other allied health-care providers, laboratory tests, medications, etc) other than those for physician services.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
Table 3
presents univariate analyses regarding the association between baseline
variables and both mortality and weaning outcomes. Of the variables
examined on ReSCU admission, those associated with a higher rate of
hospital mortality were lower levels of serum albumin (p < 0.001)
and transferrin (p = 0.05), increasing age (p = 0.04), and
physician's estimate of lower weaning likelihood (p = 0.03).
Although multivariable Cox models were attempted for both weaning and
mortality outcomes, for neither outcome did more than one variable
simultaneously achieve statistical significance at a level of
p 0.05.
|
80
years old (63%). Similarly, features associated with ventilator
independence included a higher serum albumin level (p = 0.02), a
nonmedical ICU referral source (p = 0.02), a cause of respiratory
failure at ReSCU admission other than COPD (p = 0.04), and an
admitting physician's estimate of higher weaning likelihood
(p < 0.001). To assess the time course of weaning success in the ReSCU, Figure 4 depicts a Kaplan-Meier analysis of days to ventilator independence. The median time to achieving ventilator independence was 10 days. For the 212 eligible ReSCU patients, weaning success tended to plateau at about day 33, at 67% (95% CI, 60 to 73), so that few patients remaining in the ReSCU thereafter weaned successfully. By the time patients had spent 63 days in the ReSCU, the estimated percent who achieved ventilator independence increased only slightly to 71%. In interpreting these data, it must be remembered that this study is observational and that the curves reflect "out-migration" of patients whose poor weaning prospects and performance may have triggered their transfer to extramural extended care facilities.
|
|
70 years old. Although increasing age was
associated with higher mortality in this series (Table 3)
, weaning
success rates were nearly as high in older survivors (eg,
age 70 years old) as they were in younger patients (p = 0.13).
Specifically, as shown in Figure 6
, the rate of weaning success for patients 70 years old was 56%
(95% CI, 45 to 67) compared with 68% (95% CI, 59 to 77) for patients
< 70 years old.
|
Because the introduction of software in our institution that is used to calculate the true costs of care postdated the opening of the ReSCU, cost calculations were restricted to a convenience sample of 11 ReSCU patients. For these patients, the difference in the mean daily cost of their first 3 full ReSCU days (including costs of all allied health-care services, eg, physical therapy or other kinds of therapy) and the mean daily cost incurred during their last 3 full ICU days was $469. Extrapolated to the 4,834 ReSCU days occupied by the 212 patients in this series, this $469 difference equals a savings of $2,267,146 associated with the ReSCU ($10,694 per patient).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(1) Using the selection criteria that were applied, the rates of hospital survival and of achieving ventilator independence were high in this series and similar to the most favorable rates described in other reports1 4 5 6 7 8 9 (only two of which present larger patient cohorts; see Table 4 ).
|
(3) In contrast to published experience that advanced age was
associated with lower rates of ventilator independence, survivors
70 years old in this series achieved a similar rate of ventilator
independence as younger patients did (Fig 6)
.
(4) Features on ReSCU admission that were associated with a higher rate of hospital mortality include lower serum levels of albumin and transferrin, increasing age, and the physician's estimate of lower weaning likelihood. Features associated with a greater likelihood of ventilator independence include higher serum albumin level, referral from an ICU other than the medical ICU, a cause of respiratory failure other than COPD, and the physician's estimate of a higher weaning likelihood.
Under the assumption that ReSCU patients would have otherwise remained in the ICU had the ReSCU been not available, our calculations show that the availability of the ReSCU was associated with substantial per-patient savings ($10,694 to $13,339), whether they were calculated as actual costs of care or charges, respectively.
Our experience with the ReSCU has caused it to be favorably received at the Cleveland Clinic Hospital based on several features shown in this study: (1) availability of the ReSCU has allowed the ICUs to be decompressed by providing a setting in which appropriately selected patients can be managed safely; (2) ReSCU care has been associated with a high rate of ventilator independence; and (3) ReSCU care has incurred lower costs than the ICU. Though patient satisfaction was not examined in this study, most clinicians regard the ReSCU atmosphere of private rooms and health-care providers that are attuned to rehabilitation rather than to resuscitation as preferable for appropriate candidates.
Other results of our analyses have also provided important clinical lessons for our practice that we suspect will have general value. First, in contrast to some reports,1 2 3 our findings suggest that advanced age should not preclude aggressive support and weaning attempts in selected individuals. Also, the analysis of time to achieve ventilator independence (Fig 4) has helped to define important, clinically sensible temporal milestones for our practice in weaning. Recognizing that the results of our time-to-event analyses may certainly reflect local practice, we commend this type of analysis to others so that decisions about the need for long-term ventilator support can be based on observed outcomes.
As presented in Table 4 , outcomes for ReSCU patients in this series were similar to the most favorable rates of hospital survival and weaning independence in reports from other ventilator units.1 4 5 6 7 8 9 At the same time, the comparison of these rates among available series suggests substantial variations across units, with rates of hospital mortality ranging from as low as 8% to as high as 67%, and rates of ventilator independence ranging from as low as 32% to as high as 76%. In addition to possibly reflecting operational differences between units, it is likely that other sources of this variation include differences in patient characteristics (eg, medical vs surgical referral source, cause of respiratory failure, prevalence of COPD, etc) and selection criteria for admission to the various units. Indeed, a comparison of the performance of units like that performed for 13 ICUs by Knaus et al10 is hampered by a lack of standard severity-of-illness measuring scores across available series. Also, in interpreting the reported rates of weaning success, it is important to note whether the denominator used includes all patients considered or hospital survivors only (ie, excluding nonsurvivors). As an example of the impact of this difference in denominators, the 93.8% rate of weaning success among survivors reported by Latriano et al7 is reduced to 47% when the denominator of all patients (rather than only survivors) is used.
In order not to inflate the rates of weaning success reported in our series, our overall reported rate of 60% weaning success considers all 212 patients in this series (of whom 18% died). On the other hand, our inclusion in this series of only patients admitted to the ReSCU with the intention of weaning, even if the likelihood was deemed low, would be expected to enhance the weaning outcomes in this series, especially in view of our finding that the physician's estimate of greater weaning likelihood on ReSCU admission was associated with higher rates of achieving ventilator independence. Though not all of them have, some of the other studies have applied similar admission criteria to those used in our ReSCU.
Our high rate of survival among patients aged 80 years old
(63%) is almost identical to the rate (62%) reported by Scheinhorn et
al11
but is at variance with several other reports. For
example, in the series by Elpern et al,1
only 33% of 95
patients aged 60 to 90 years old who were on mechanical ventilation
> 3 days survived to hospital discharge, and only 16% remained alive
1 year later. In making this comparison, it must be recognized that the
cohort analyzed by Elpern et al1
considered all Medicare
patients rather than those deemed weanable. Similarly, in a
retrospective review of 282 patients aged > 80 years old, Swinburne
et al2
reported a 31% hospital survival and an only 9%
survival for such patients remaining on mechanical ventilation for
> 15 days. Finally, Cohen et al3
reported a 9% hospital
survival among 22 patients > 80 years old and observed that outcomes
were especially unfavorable in patients for whom the sum of "age plus
duration of mechanical ventilation" was a value > 100. Though it is
likely that these outcome differences reflect different selection
criteria for ventilator unit admission, a better understanding of
causes for these differences will require longitudinal tracking of
outcomes among a complete inception cohort of mechanically ventilated
patients, as well as a severity-of-illness adjustment.
In the context of prior reports that show that measures of nutritional status are significantly associated with survival,7 12 13 our finding that lower levels of serum transferrin and albumin on ReSCU admission were significantly associated with death and that a higher serum albumin level was associated with enhanced weaning are not surprising. On the other hand, our analysis of the change in the levels of serum albumin and transferrin over the course of ReSCU stay fails to show that improved nutritional measures were associated with improved outcomes (data not shown).
Several shortcomings of the current study warrant discussion. First, the lack of information about long-term survival in our series precludes comparison to rates of survival between 1 and 4 years, as reported by Gracey et al,8 or to 1-year survival rates, as reported by Latriano et al.7 Indeed, to address this need, examination of long-term survival by telephone inquiry and by analysis of survival databases (eg, the National Death Index and Equifax) is planned. Similarly, as with many other series, we lack standard measures of severity of illness that are needed to permit adjustment in comparing outcomes with other series. We elected not to include APACHE (acute physiology and chronic health evaluation) III scores in our analysis of outcome predictors because these scores characterize patients at the time of initial ICU admission. In the context that patients were admitted to the ReSCU a median of 25 days after initial ICU admission, it is reasonable to think that their clinical status had changed by the time of ReSCU admission.
Another shortcoming is that, although transfer to the ReSCU was arranged for most patients for whom requests were made, we neither tallied the number of requests for ReSCU transfer nor the number of ventilated patients in our hospital for whom the idea of ReSCU transfer might have been entertained. As such, we cannot calculate the rates of ReSCU admission.
Also, although our finding that a health-care provider's prediction of weaning likelihood is strongly associated with weaning success and agrees with findings by others, it could be argued that initial pessimism by the admitting physician may have caused a slackening of efforts to wean such patients. Because estimates from physicians uninvolved in the patient's subsequent management might be less susceptible to such bias, the fact that physicians rotated biweekly into the ReSCU means that patients were often managed by physicians other than those rendering the initial weaning prediction. Also, weaning in the ReSCU is facilitated by applying protocols by the supervising respiratory therapists, thereby lessening the risk that physician bias would affect weaning rates.
Finally, although our economic analysis is based on true cost measures and suggests that ReSCU care is less expensive then ongoing care in a traditional ICU, the comparison is based on several assumptions. First, because software for analyzing costs had become available at our institution only after the ReSCU opened, our assessment of costs is based on a small sample of patients and compares the mean costs of the initial 3 days in the ReSCU with mean costs of the final 3 days in the ICU (before ReSCU transfer) for this sample. Although the staffing ratio in the ReSCU and the avoidance of invasive monitoring and attendant testing (eg, frequently drawn arterial blood gases through an indwelling arterial line) virtually ensure lower ReSCU costs, our analysis assumes similar clinical outcomes of care, whether provided in the ReSCU or continued ICU stay. Although our observations suggest high rates of survival and weaning in caring for this group of ReSCU patients, analysis of costs and outcomes in a randomized trial (eg, ICU vs ReSCU care) would more clearly elucidate the relative clinical and cost benefits of the alternative clinical settings. Until such a trial is undertaken, we believe that results such as ours will provide an important clinical and cost endorsement for units like the ReSCU for appropriately selected patients.
In summary, our experience in caring for 212 patients over the first 4 years of ReSCU operation suggests that high rates of survival and of ventilator independence can be achieved and that the ReSCU is a clinically effective, cost-saving alternative to traditional ICU care for selected patients.
|
Footnotes |
|---|
Abbreviations: CI = confidence interval; IQR = interquartile range; ReSCU = Respiratory Special Care Unit
Received for publication July 28, 1998. Accepted for publication February 24, 1999.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. E. Nelson, A. F. Mercado, S. L. Camhi, N. Tandon, S. Wallenstein, G. I. August, and R. S. Morrison Communication About Chronic Critical Illness Arch Intern Med, December 10, 2007; 167(22): 2509 - 2515. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. El Solh and F. H. Ramadan Overview of Respiratory Failure in Older Adults J Intensive Care Med, December 1, 2006; 21(6): 345 - 351. [Abstract] [PDF]
|
|
|
|
|
|
J. E. Nelson, N. Tandon, A. F. Mercado, S. L. Camhi, E. W. Ely, and R. S. Morrison Brain dysfunction: another burden for the chronically critically ill. Arch Intern Med, October 9, 2006; 166(18): 1993 - 1999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. R. MacIntyre, S. K. Epstein, S. Carson, D. Scheinhorn, K. Christopher, and S. Muldoon Management of Patients Requiring Prolonged Mechanical Ventilation: Report of a NAMDRC Consensus Conference Chest, December 1, 2005; 128(6): 3937 - 3954. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Aboussouan, C. D. Lattin, and V. V. Anne Determinants of Time-To-Weaning in a Specialized Respiratory Care Unit Chest, November 1, 2005; 128(5): 3117 - 3126. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Hoffman, F. J. Tasota, T. G. Zullo, C. Scharfenberg, and M. P. Donahoe Outcomes of Care Managed by an Acute Care Nurse Practitioner/Attending Physician Team in a Subacute Medical Intensive Care Unit Am. J. Crit. Care., March 1, 2005; 14(2): 121 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A K Simonds Streamlining weaning: protocols and weaning units Thorax, March 1, 2005; 60(3): 175 - 182. [Full Text] [PDF]
|
|
|
|
|
|
D V Pilcher, M J Bailey, D F Treacher, S Hamid, A J Williams, and A C Davidson Outcomes, cost and long term survival of patients referred to a regional weaning centre Thorax, March 1, 2005; 60(3): 187 - 192. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Kollef Tracheostomy for Respiratory Failure: We Need More Answers Chest, January 1, 2004; 125(1): 7 - 9. [Full Text] [PDF]
|
|
|
|
|
|
J. K. Stoller, M. Xu, E. Mascha, and R. Rice Long-term Outcomes for Patients Discharged From a Long-term Hospital-Based Weaning Unit Chest, November 1, 2003; 124(5): 1892 - 1899. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Goldstone The pulmonary physician in critical care * 10: Difficult weaning Thorax, November 1, 2002; 57(11): 986 - 991. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Junker, J. E. Zimmerman, C. Alzola, E. A. Draper, and D. P. Wagner A Multicenter Description of Intermediate-Care Patients* : Comparison With ICU Low-Risk Monitor Patients Chest, April 1, 2002; 121(4): 1253 - 1261. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. R. MacIntyre Evidence-Based Guidelines for Weaning and Discontinuing Ventilatory Support : A Collective Task Force Facilitated by the American College of Chest Physicians; the American Association for Respiratory Care; and the American College of Critical Care Medicine Chest, December 1, 2001; 120(6_suppl): 375S - 396S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Scheinhorn, D. C. Chao, M. S. Hassenpflug, and D. R. Gracey Post-ICU Weaning From Mechanical Ventilation : The Role of Long-term Facilities Chest, December 1, 2001; 120(6_suppl): 482S - 484S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. VITACCA, A. VIANELLO, D. COLOMBO, E. CLINI, R. PORTA, L. BIANCHI, G. ARCARO, G. VITALE, E. GUFFANTI, A. LO COCO, et al. Comparison of Two Methods for Weaning Patients with Chronic Obstructive Pulmonary Disease Requiring Mechanical Ventilation for More Than 15 Days Am. J. Respir. Crit. Care Med., July 15, 2001; 164(2): 225 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M Confalonieri, M Gorini, N Ambrosino, C Mollica, and A Corrado Respiratory intensive care units in Italy: a national census and prospective cohort study Thorax, May 1, 2001; 56(5): 373 - 378. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
All percentages are expressed with 212 patients as the denominator.
*An additional 18 patients died after ReSCU discharge.
