Abstract
Long-term acute-care (LTAC) hospitals are facilities exempt from the Medicare prospective payment system and which provide care to patients suffering from prolonged critical illness. From August 1, 1995 to July 31, 1996, we studied the outcomes of 133 mechanically ventilated patients who were consecutively admitted to a large urban LTAC hospital from intensive care units (ICUs) of acute-care hospitals. Survival and functional status within 1 yr after the index admission were measured, and specific patient variables were used to develop a predictive model for survival at 1 yr. Of the 133 patients studied, 66 (50%) died prior to discharge. Of discharged patients, 70% had been successfully liberated from mechanical ventilation. One year after LTAC hospital admission, 103 (77%) of the patients had expired, typically after spending the majority of their days in acute care or long-term care facilities. Eleven 1-yr survivors (8%) were fully functional, whereas the remainder had significantly reduced functional status. Patients older than 74 yr, and patients older than 64 yr and not functionally independent before admission, had a 95% (confidence interval [CI]: 84% to 99%) 1-yr mortality; patients without these characteristics had a 56% (CI: 41% to 71%) 1-yr mortality (p < 0.001). We demonstrate characteristics predicting the poorest prognoses for patients requiring prolonged mechanical ventilation. These characteristics may be identifiable before transfer to an LTAC hospital.
Approximately 3% to 6% of patients admitted to an adult intensive care unit (ICU) require prolonged mechanical ventilation, either because of complex pulmonary or cardiac disease, or for other multisystem problems (1, 2). These patients consume a disproportionate amount of intensive care resources (3-5). Since 1983, when the Health Care Financing Administration created the long-term care exemption from the Medicare prospective payment system, numerous long-term acute-care (LTAC) hospitals have been created to care for chronically critically ill patients. LTAC hospitals provide the resources and personnel to meet the needs of patients with prolonged, complex medical problems. Most patients in LTAC hospitals are individuals who require prolonged mechanical ventilation after acute illness. They may also require continuous inotropic or vasopressor support, hemodialysis, total parenteral nutrition, or close airway monitoring. Some LTAC facilities also manage patients with advanced wounds. According to industry analysts, there are approximately 200 LTAC hospitals nationwide, with a capacity for 15,000 patients (6). The number of LTAC hospital discharges increased by 19.2% per annum in the period from 1900 to 1995 (personal communication, Tom McGinnis, SBC Warburg-Dillon-Read Equity Research, New York, NY). Reimbursed by private insurers, Medicaid, and Medicare under the Tax Equity and Fiscal Responsibility Act (7), LTAC hospitals in the United States generated revenues of over $3 billion in 1997.
Little is known about the role of LTAC hospitals in achieving survival or return of function for patients who suffer from prolonged critical illness. Information about these patients' experiences might assist clinicians in prognosticating for critically ill patients and their families, create a framework for discussions between physicians and their patients about decisions to continue prolonged mechanical ventilation; and inform the debate about use of health care resources for critically ill patients.
We therefore examined the outcomes of 133 patients requiring more than 14 d of mechanical ventilation in acute-care hospital ICUs who were transferred to a single LTAC hospital. We determined these patients' functional status prior to their acute-care hospital admission, and then followed their experiences over the year after their admission to the LTAC hospital. Over this 1-yr period, we assessed survival, the incidence of liberation from mechanical ventilation, the rate of return of function, and the amount of time spent in other care environments, such as acute-care and rehabilitation hospitals. We then developed a predictive model for survival at 1 yr, based on pretransfer patient characteristics.
The facility in which the study was conducted is a 75-bed LTAC hospital in an urban location. The hospital had a seven-bed intensive care unit (ICU) with full cardiovascular monitoring capabilities. Nurse:patient ratios in the ICU were 1:2, and on the wards were 1:4 during the day and 1:8 at night and on weekends. Full-time respiratory therapy, physical therapy, occupational therapy, and speech and swallowing therapy services were available for all patients. Three surgical suites were used for minor surgery such as wound debridement or permanent venous catheter insertion, but patients were transferred to acute-care hospitals for major surgeries. Discharge planning was managed by nurse or social-work case managers. Study patients (those requiring mechanical ventilation) accounted for 43% of patients admitted to the hospital. The remainder of the patients had other medically complex issues or were admitted for complex wound care.
All admissions to the study hospital represent transfers from acute- care hospitals. Patients were accepted regardless of the severity of their illness provided they were stable for transfer. Patients being ventilated with a positive end-expiratory pressure (PEEP) of > 10 cm H2O or a fraction of inspired oxygen (Fi O2 ) of > 0.60 were screened by the medical director of the study hospital. Patients receiving mechanical ventilation were required to have a tracheostomy. Patients were discharged after complex medical issues were resolved and they had either been safely liberated from mechanical ventilation or multiple attempts at liberation from mechanical ventilation had been unsuccessful.
Specialists in pulmonary and critical care medicine served as primary physicians for all patients; in-hospital nighttime coverage was provided by fellow trainees in pulmonary and critical care medicine who also assisted in the primary management of a portion of the patients in the hospital (53%). During the latter half of the study period, nurse practitioners assisted with primary patient management. Consultation services were available for most medical and surgical subspecialties, psychiatry, and clinical psychology.
Study patients include all mechanically ventilated patients transferred to the LTAC hospital between August 1, 1995 and July 31, 1996 from ICUs, respiratory care units, or stepdown units at 37 acute-care hospitals in the Chicago area. Any patient who had previously been an inpatient at the LTAC hospital or had been chronically mechanically ventilated prior to admission to the ICU of the acute-care hospital was excluded.
Patient characteristics. Medical records were retrospectively surveyed for age, gender, diagnoses leading to ICU admission at the acute-care hospital, length of stay (LOS) in the acute-care hospital ICU, and number of days of mechanical ventilation prior to transfer. Acute Physiology and Chronic Health Evaluation (APACHE II) scores were calculated according to the method described by Knaus and colleagues (8) from data recorded in the patient's medical record during the first 24 h after admission to the LTAC hospital. Glasgow Coma Scores (GCS) were obtained by nurses every 8 h. The highest GCS measured within the first 24 h at the LTAC hosptial was used in calculating the APACHE II score. Complete Apache II data were available for all but five patients. Interrater reliability was confirmed with kappa statistics and tests for correlation.
Acute and chronic comorbidities. Records from the acute-care hospital were reviewed for comorbidities. The acute comorbidities consisted of events occurring within 6 wk before admission to the LTAC hospital, including immunosuppression, recent myocardial infarction or cardiac arrest, congestive heart failure requiring vasopressor support, atrial or ventricular arrhythmias, bacteremia, acute hepatic failure, renal failure requiring dialysis, or the presence of skin ulcerations of Stage II severity or worse. Chronic comorbidities included a history of peripheral vascular disease, cerebrovascular accidents, coronary atherosclerotic disease, restrictive lung disease, metastatic cancer, acquired immunodeficiency syndrome, hepatic cirrhosis, or diabetes requiring medical therapy.
Definition of outcomes. Survival and discharge disposition from the LTAC hospital were recorded, as were also hospital and ICU lengths of stay (LOS). Documentation of do not resuscitate (DNR) orders, use of cardiopulmonary resuscitation (CPR) at death, withdrawal of mechanical ventilation or dialysis despite continued dependence on these modalities, the number of patients who were successfully liberated from mechanical ventilation prior to discharge, and the length of time required for liberation from mechanical ventilation were recorded. Successful liberation from mechanical ventilation was defined as the patient's independence of mechanical ventilation for seven consecutive days and nights regardless of outcome after that period.
Performance status before admission. Patients or families were surveyed by case managers at the time of admission, and were asked about the patients' performance status prior to admission to the acute-care hospital. The survey instrument, designed by the LTAC hospital staff, included six questions for assessing the patient's previous mental status, physical capabilities, ability to manage activities of daily living, and previous living arrangements, and a general statement of the patient's functional dependence or independence. This information was available for 91 (68%) of the study subjects.
Determination of 1-yr outcome and events after LTAC hospital discharge. All hospital survivors or their closest family members were contacted by a single investigator between May 26, 1997 and June 14, 1997 for a postdischarge follow-up evaluation of survival and functional status. A scripted telephone survey instrument was designed, based on the same questions used to assess performance status in the admission interviews. Patients or families also were asked about the number of subsequent hospital admissions and relative length of time spent at home, and in a nursing home, rehabilitation facility, or acute-care hospital since discharge from the LTAC hospital. If a patient or a family member could not be located (as was the case with five patients), survival was determined from Medicare records. Family members were asked how much time they spent with the patient. Follow-up success was 87% for functional status, 93% for location during the postdischarge period, and 100% for survival data.
Predictors of survival. We examined the relationship between 1-yr mortality with the following variables: acute and chronic comorbidities, APACHE II score, age, reported performance status prior to acute-care hospitalization, diagnosis leading to need for ICU care, and successful liberation from mechanical ventilation.
Statistical methods. All p values are two-sided. Ninety-five percent confidence limits are presented. Each analysis was performed on all available data. Relative risks (RRs) associated with individual predictors of mortality were assessed with chi-squared statistics. Significant predictors of mortality in contingency tables were examined together in a multiple logistic regression analysis. Likelihood ratio tests were used to choose the most parsimonious predictive model for 1-yr mortality (9). Comorbidities were dichotomized: (1) any patient who received hemodialysis or peritoneal dialysis at the previous hospital was considered to have renal failure; (2) previous performance status was dichotomized to “independent” or “dependent,” with patients needing any assistance with ambulation or activities of daily living being labeled dependent; (3) diabetes was considered present if the patient was receiving any oral hypoglycemic agent or insulin. Age was divided into roughly equal-sized categories (25 to 64, 65 to 74, and 75 yr old or older). The associations of predictors of death in contingency tables and multiple logistic regression analyses were reassessed, with death viewed as a time-dependent phenomenon; without exception, the logistic regression analyses revealed effects of similar direction and significance as those in the Cox proportional hazards model (10). All statistical analyses were performed with STATA 5.0 software (Stata Corporation, College Station, TX).
The project was reviewed and approved by the Institutional Review Board of the University of Chicago.
There were 133 patients eligible for the study. The characteristics of the patients are listed in Table 1. Seventy-seven percent of the patients were > 65 yr old, and acute and chronic comorbidities were common. The mean APACHE II score of 23 ± 6 (range: 11 to 44), mean alveolar–arterial oxygen gradient of 144 ± 87 mm Hg, and prevalence of 14% of patients requiring pharmacologic vasopressor support with dobutamine or at least 5 μg/kg/min of dopamine all indicated a high level of illness severity at the time of transfer to the LTAC hospital. The general diagnoses leading to initial ICU admission at the referring acute-care hospitals are listed in Table 2. Medical diagnoses accounted for 95% of initial admissions. Of note, was that 23 (18%) of these “medical” patients required a major surgical procedure during the course of their acute-care hospital ICU admission.
| Characteristic | Mean ± SD or % | |
|---|---|---|
| Age | 71 ± 12 | |
| Female | 52% | |
| APACHE II score | 23 ± 6 | |
| Glasgow Coma Score | 12 ± 3 | |
| Serum albumin, g/dl | 2.3 ± 0.6 | |
| Comorbidities | 3.2 ± 1.6 | |
| Requiring vasopressor support | 14% | |
| Alveolar–arterial oxygen gradient, mm Hg | 144 ± 87 | |
| Acute-care hospital ICU days prior to transfer, | ||
| median (range) | 25 (10–96) | |
| Days of MV prior to transfer, median (range) | 25 (9–123) |
| Acute lung injury | 42 (31%) | |
| Chronic lung disease | 22 (16%) | |
| Cardiac disease | 21 (16%) | |
| CNS disease | 20 (15%) | |
| Multisystem organ failure | 14 (10%) | |
| Neuromuscular disease | 7 (5%) | |
| Postoperative | 7 (5%) |
Mean length of stay was 63 ± 50 (SD) d. Sixty-six (50%) of the patients died while hospitalized at the LTAC hospital (Table 3). Of these patients, 14 (21%) received CPR at the time of death, whereas nine (14%) had life-sustaining care, in the form of dialysis or mechanical ventilation, withdrawn. Fifty percent of all patients were classified for exclusion from resuscitation (DNR). DNR orders were written 12 d (median) before death. Fifty-one patients (38%) were successfully liberated from mechanical ventilation; one patient required subsequent mechanical ventilation during his hospitalization. Among hospital survivors, 34 patients (51% of discharged patients) were discharged to a skilled nursing facility or nursing home, 13 (19%) were discharged back to acute-care hospitals, 12 (18%) went home, and eight (12%) went to rehabilitation hospitals. Forty-seven (70%) discharged patients had been liberated from mechanical ventilation. Of 20 patients discharged while receiving ventilation, one was discharged to home; the remainder were discharged to a skilled nursing facility.
| Event for Cohort | Event for Subgroups | n (%) | ||
|---|---|---|---|---|
| Expired at LTAC hospital | 66 (50%) | |||
| Care withdrawn | 9 (14%)† | |||
| Received CPR at death | 14 (23%) | |||
| Classified DNR | 67 (50%) | |||
| Liberated from MV | 51 (38%) | |||
| Discharged | 67 (50%) | |||
| Off ventilator | 47 (70%) | |||
| On ventilator | 20 (30%) |
Table 4 contains the aggregate 1-yr outcomes. Only 30 (23%) of the study patients were still alive at 1 yr. Most of these patients were living at home. Two patients still required mechanical ventilation, one in a skilled nursing facility and one at an LTAC hospital. Eleven patients (42% of 1-yr survivors, 8% of total study patients) were oriented, ambulatory, and independent at 1 yr. Of the 19 survivors who were independent prior to their acute illness, 13 returned to independent function.
| n (%)‡ | ||
|---|---|---|
| Residence | ||
| Home | 19 (73%) | |
| Nursing home | 6 (23%) | |
| LTAC hospital | 1 (4%) | |
| Mental function | ||
| Alert and oriented | 18 (69%) | |
| Alert but confused | 7 (27%) | |
| Unresponsive | 1 (4%) | |
| Physical Function | ||
| Independent | 11 (42%) | |
| Needed some assistance | 7 (27%) | |
| Totally dependent | 8 (31%) | |
| Mobility | ||
| Ambulatory | 17 (68%) | |
| Wheelchair | 3 (12%) | |
| Bedbound | 5 (20%) |
Patients' longitudinal experience over the 1-yr period after their admission to the LTAC hospital are presented as a Kaplan–Meier survival plot (Figure 1). In this plot, we have represented the location of each patient for every day of the year, as well as the survival of the cohort. Day zero represents the LTAC hospital admission and Day 365 represents the 1-yr follow-up period for the cohort. At 1 yr, 77% of the patients were dead, 14% were at home, and the remainder were in nursing homes or acute-care hospitals. The plot also shows that 23 (43%) of the 54 person-yr lived by our study patients were spent in the LTAC hospital. In comparison, roughly 14 person-yr (26%) were spent at home, 8 person-yr (15%) were spent in rehabilitation or in nursing homes, and 7 person-yr (13%) were spent in acute-care hospitals.
Fig. 1. One-year analysis of site of care and survival of 133 patients transferred from 37 acute-care hospital ICUs to an urban long-term acute-care facility. (n = 133; NH = nursing home/rehabilitation hospital; AH = acute-care hospital.)
[More] [Minimize]The significant results of contingency table analyses are shown in Table 5. We found that the RR of death by 1 yr was significantly increased for more elderly patients. For these analyses, the base population consisted of patients younger than 65 yr. Older age groups were then compared. Those aged 65 to 74 yr had an RR of 1.4 for dying at 1 yr, whereas those aged 75 yr or older had an RR of 1.7 of dying by 1 yr (p = 0.03, p < 0.001, respectively, for comparisons, p < 0.001 for trend). For those patients who were not independent before their acute-care hospital admission, the RR of death at 1 yr was 1.4 (p = 0.008). Renal failure (RR = 1.3, p = 0.03) and diabetes (RR = 1.2, p = 0.05) also increased the likelihood of death at 1 yr.
| Predictor | Absolute Risk without Characteristic | Absolute Risk with Characteristic | Relative Risk | p Value | ||||
|---|---|---|---|---|---|---|---|---|
| Age | ||||||||
| 25 to 64 yr | (Base case) | 53% (16/30) | 1.0 (—) | — | ||||
| 65 to 74 yr | 77% (40/52) | 1.4 (1.0–2.1) | 0.03 | |||||
| 75 to 100 yr | 92% (47/51) | 1.7 (1.2–2.4) | < 0.001 | |||||
| Dependent prior to admission† | 65% (35/54) | 89% (33/37) | 1.4 (1.1–1.7) | 0.008 | ||||
| Renal failure | 74% (29/112) | 95% (20/21) | 1.3 (1.1–1.5) | 0.03 | ||||
| Diabetes | 71% (55/77) | 86% (48/56) | 1.2 (1.0–1.4) | 0.05 |
Results of multiple logistic regression analysis done to examine the significant predictors of diabetes, dependence prior to admission, and three categories of age are shown in Table 6 (renal failure was not a significant predictor in this analysis). With these variables in combination, and as compared with an age of less than 65 yr, an age of 65 to 74 yr increased the odds of death by 1 yr by a factor of 3.7, and an age of 75 yr or older increased the odds by a factor of 16. The odds of death were 3.9 times greater for those subjects who were not independent before admission, and 3.2 times greater for those with diabetes.
| Predictor | Odds Ratio (95% CI) | p Value | ||
|---|---|---|---|---|
| Age | ||||
| 25–64 yr (base) | 1.0 (1.0–1.0) | — | ||
| 65–74 yr | 3.7 (1.1–12.7) | 0.04 | ||
| 75–100 yr | 15.6 (2.8–86.4) | 0.002 | ||
| Diabetes | 3.2 (1.0–10.3) | 0.05 | ||
| Dependent prior to admission | 3.9 (1.1–14.2) | 0.04 |
Combination of the highly significant predictors of age and prior independence is shown in Table 7. The cohort is split into two groups of roughly equal size: The “low risk” group consists of patients < 65 yr old and those ⩽ 75 yr old who were independent prior to admission, who collectively had only a 56% (confidence interval [CI]: 41% to 71%) 1-yr mortality. The “high risk” group includes all patients 75 yr old or older, combined with those 65 to 75 yr of age who had not been previously independent. This latter group had a 95% (CI: 84% to 99%) mortality at 1 yr (p < 0.001 for model).
| Age Group (yr) | ||||||
|---|---|---|---|---|---|---|
| Prior function | 25–64 | 65–74 | 75–100 | |||
| Independent | 44% (8/18) | 62% (13/21) | 93% (14/15) | |||
| Dependent | 67% (6/9) | 100% (12/12) | 94% (15/16) | |||
From these analyses, we noted that patients 75 yr or older, or 65 yr or older with impaired function prior to admission, had an estimated 5% 1-yr survival. Figure 2 shows the time these patients spent in various locations during the study year. These high-risk patients spent 71% of their time at the LTAC hospital, 13% of their time in nursing homes, 9% of their time in acute-care hospitals, and only 7% of their time at home.
Fig. 2. Site of care over 1 yr of observation for high risk (95% 1-yr mortality) subgroup, consisting of patients older than age 74 yr, or older than age 64 yr with poor functional status prior to initial hospital admission (n = 43).
[More] [Minimize]Patients requiring prolonged mechanical ventilation represent a substantial number of patients in the health care system, and LTAC hospitals have been proliferating as a more cost-efficient means of managing their care (11, 12). We found that patients requiring prolonged mechanical ventilation after acute illness and who were transferred from ICUs of acute-care hospitals to an LTAC hospital had only a 50% chance of being discharged from the latter facility. Discharged patients were unlikely to return home, and they experienced a 1-yr postdischarge mortality of 55%. Only 8% of our original patient cohort was fully independent at 1 yr.
Because patients were transferred to the LTAC hospital in our study early in the overall course of their hospitalization, when their illness severity was high, they may be representative of most patients requiring prolonged mechanical ventilation after acute illness. The survival of patients in our study was comparable to that of patients in previously reported series in which patients requiring prolonged mechanical ventilation were managed entirely in acute-care hospitals. As compared with the 50% hospital survival and 22% 1-yr survival in our patient group, Spicher and White (13) reported a 39.6% hospital survival and 28.6% 1-yr survival for a series of 250 patients with a mean age of 59.6 yr who required mechanical ventilation for > 10 d in 1987. In 1992, Gracey and colleagues (14) reported a 57.7% hospital survival and 38.7% 1-yr survival for a group of 104 patients with a mean age of 66.3 yr. Many of the patients in the study by Gracey and colleagues were postoperative patients with good functional status.
Comparison of outcomes in the present study with reported outcomes at other LTAC hospitals or other settings of long-term care are complicated by differences in case mix and illness severity. Units that have some selection for prognosis, rehabilitation potential, or illness severity (15-17) report hospital survival rates ranging from 69.7% to as high as 91.2%. The LTAC hospital in the present study had no selection criteria for prognosis or illness severity. Studies at two other units in which admission policies are similarly nonselective reported hospital survival rates of 47% (18) and 53% (11). To our knowledge, 1-yr survival, disposition, and functional status have not been previously reported from such nonselective units.
The poor short- and long-term outcomes of chronically critically ill patients are a point of substantial concern. Besides the need for continued improvement in their management and recognition of their unique needs, goals for their care may warrant closer examination. These individuals endure months of intensive medical management and personal discomfort. Their families experience a great deal of emotional and financial stress as well (19). This study suggests that some patients benefit from such efforts but many others do not. Further studies of the quality of life of these patients are warranted.
We have shown that distinct patient characteristics, identifiable before LTAC hospital admission, can predict poor outcome. In our population, advanced age, lack of independence before acute illness, diabetes, and renal failure were significant predictors of death at 1 yr. Patients older than age 75 yr or patients over age 65 yr who were not functionally independent before their acute-care hospital admission only had a 5% chance of survival at 1 yr. Other studies have also shown a relationship between age and increased mortality in patients requiring mechanical ventilation (13, 20-23). Poor functional status has also previously been shown to be a predictor of outcome both for non-ICU and ICU hospital admissions (13, 24-27).
This study had several limitations. The small number of patients limited the number of subgroups that we could study, and our models of 1-yr survival have not been validated with another population of similar patients. Also, preadmission functional status and postdischarge experience and functional status are either self- or surrogate-reported and are subject to unpredictable biases (28). Because data for preadmission functional status were not available for all patients, the reliability of this predictor of survival may have been affected. Larger studies, involving multiple institutions, should examine these concepts further. If the results of our study are supported, this type of information could be useful for chronically critically ill patients or their surrogates as they consider appropriate levels of aggressive care versus alternative measures that focus on comfort.
In summary, this study indicated that patients who require prolonged mechanical ventilation after acute illness have a very low 1-yr survival rate, and that few of these patients can manage independent living after a year. Patients with extremely poor prognoses may be identifiable before being transferred to an LTAC hospital.
The authors are indebted to Joan Scannell and Mike Waldman for assistance with data collection, and to Shannon Biggerstaff for her assistance with data and manuscript management.
Supported by an unrestricted grant from the Transitional Hospitals Corporation and by the Robert Wood Johnson Foundation.
| 1. | Stauffer J. L., Fayter N. A., Graves B., Cromb M., Lynch J. C., Goebel P.Survival following mechanical ventilation for acute respiratory failure in adult men. Chest104199312221229 |
| 2. | Daly R. F., Rudy E. B., Thompson K. S., Happ M. B.Development of a special care unit for chronically critically ill patients. Heart Lung2019914552 |
| 3. | Wagner D. P.Economics of prolonged mechanical ventilation. Am. Rev. Respir. Dis.140(Suppl.)1989S14S18 |
| 4. | Rosen R. L., Bone R. C.Economics of mechanical ventilation. Clin. Chest Med.91988163169 |
| 5. | Cohen I. L., Lambrinos J., Fein A.Mechanical ventilation for the elderly patient in intensive care: incremental charges and benefits. J.A.M.A.269199310251029 |
| 6. | Campbell, S. 1997. HCFA clamping down on long term acute care “hospitals within hospitals.” Health Care Strategic Management 13 (8). |
| 7. | Chan L., Koepsell T. D., Deyo R. A., Esselman P. C., Haselkorn J. K., Lowery J. K., Stolov W. C.The effect of Medicare's payment system for rehabilitation hospitals on length of stay, charges, and total payments. N. Engl. J. Med.3371997978985 |
| 8. | Knaus W. A., Draper E. A., Wagner D. P., Zimmerman J. E.APACHE II: a severity of disease classification system. Crit. Care Med.3111986818829 |
| 9. | Agresti, A. 1990. Categorical data analysis. In V. Barnett, R. A. Bradley, N. I. Fisher, J. S. Hunter, J. B. Kadane, D. G. Kendall, A. F. M. Smith, S. M. Stigler, J. Teugels, and G. S. Watson, editors. Wiley Series in Probability and Mathematical Statistics. John Wiley & Sons, New York. 22. |
| 10. | Collett, D. 1994. Modelling survival data in medical research. In C. Chatfield and J. V. Zidek, editors. Texts in Statistical Science, 1st ed. Chapman and Hall, London. 53–66. |
| 11. | Elpern E., Silver M. R., Rosen R. L., Bone R. C.The noninvasive respiratory care unit: patterns of use and financial implications. Chest991991205208 |
| 12. | Scheinhorn D. J., Artinian B. M., Catlin J. L.Weaning from prolonged mechanical ventilation: the experience at a regional weaning center. Chest1051994534539 |
| 13. | Spicher J. E., White D. P.Outcome and function following prolonged mechanical ventilation. Arch. Intern. Med.1471987421425 |
| 14. | Gracey D. R., Naessens J. M., Drishan I., Marsh M.Hospital and posthospital survival in patients mechanically ventilated for more than 29 days. Chest1011992211214 |
| 15. | Gracey D. R., Naessens J. M., Viggiano R. W., Koenig G. E., Silverstein M. D., Hubmayr R. D.Outcome of patients cared for in a ventilator-dependent unit in a general hospital. Chest1071995494499 |
| 16. | Scheinhorn D. J., Chao D. C., Tearn-Hassenpflug M., LaBree L. D., Heltsley D. J.Post-ICU mechanical ventilation: treatment of 1,123 patients at a regional weaning center. Chest111199716541659 |
| 17. | Rudy E. B., Daly B. J., Douglas S., Montenegro H. D., Song R., Dyer M. A.Patient outcomes for the chronically critically ill: special care unit versus intensive care unit. Nurs. Res.441995324331 |
| 18. | Bagley P. H., Cooney E.A community-based regional ventilator weaning unit: development and outcomes. Chest111199710241029 |
| 19. | Covinsky K. E., Goldman L., Cook E. B., Oye R., Desbiens N., Reding D., Fulkerson W., Connors A. F., Lynn J., Phillips R. S.The impact of serious illness on patient's families. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. J.A.M.A.272199418391844 |
| 20. | Knaus W. A.Prognosis with mechanical ventilation: the influence of disease, severity of disease, age and chronic health status on survival from an acute illness. Am. Rev. Respir. Dis.140(Suppl.)1989S8S13 |
| 21. | Seneff M. G., Wagner D. P., Wagner R. P., Zimmerman J. E., Knaus W. A.Hospital and 1-year survival of patients admitted to intensive care units with acute exacerbation of chronic obstructive pulmonary disease. J.A.M.A.274199518521857 |
| 22. | Cohen I. L., Lambrinos J.Investigating the impact of age on outcome of mechanical ventilation using a population of 41,848 patients from a statewide database. Chest107199516731680 |
| 23. | Kurek C. J., Cohen I. L., Lambrinos J., Minatoya K., Booth F. V. M., Chalfin D. B.Clinical and economic outcome of patients undergoing tracheostomy for prolonged mechanical ventilation in New York State during 1993: analysis of 6,353 cases under diagnosis-related group 483. Crit. Care Med.251997983988 |
| 24. | Davis R. B., Iezzoni L. I., Phillips R. S., Reiley P., Coffman G. A., Safran C.Predicting in-hospital mortality: the importance of functional status information. Med. Care331995906921 |
| 25. | Covinsky K. E., Justice A. C., Rosenthal G. E., Palmer R. M., Landefeld C. S.Measuring prognosis and case mix in hospitalized elders: the importance of functional status. J. Gen. Intern. Med.121997203208 |
| 26. | Chassin M. R.Costs and outcomes from medical intensive care. Med. Care201982165179 |
| 27. | Inouye S. K., Peduzzi P. N., Robison T. J., Hughes J. S., Horwitz R. I., Concato J.Importance of functional measures in predicting mortality among older hospitalized patients. J.A.M.A.370199811671193 |
| 28. | Epstein A. M., Hall J. A., Tognetti J., Son L. H., Conant L.Using proxies to evaluate quality of life. Med. Care271989S91S98 |
