| 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 Frances Payne Bolton School of Nursing (Drs. Douglas and Daly) and the School of Medicine (Dr. Gordon), Case Western Reserve University, Cleveland, OH; School of Nursing (Dr. Brennan), University of Wisconsin-Madison, Madison, WI; and the School of Nursing (Dr. Uthis), Chulalongkorn University, Bangkok, Thailand.
Correspondence to: Sara L. Douglas, PhD, School of Nursing, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106-4904; e-mail address: SLD4@po.cwru.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Design: This study was part of a larger prospective longitudinal descriptive study of posthospital outcomes for LTV patients.
Setting and participants: One hundred ninety-nine ICU patients admitted to a university medical center, Veterans Administration hospital, or small community hospital who required > 96 h of continuous in-hospital mechanical ventilation were enrolled.
Measurements and results: Descriptive statistics, logistic regression, and survival analytic techniques were used. The 6-month hospital readmission rate was 38%. Readmission occurred most often within days 1 to 60 days (mean, 39.2 days) posthospital discharge. Predictive variables for readmission were the following: length of the index hospital stay; length of the index mechanical ventilation; and the need for oxygen at hospital discharge. Using survival analysis, the age category of 66 to 71 years was statistically significant for the relative risk of readmission within the first 30 days of the index hospital discharge.
Conclusions: LTV patients should be considered at risk for hospital readmission. Further study examining the impact of closer follow-up in the first 60 days posthospital discharge is necessary in order to determine whether there is a more effective way of reducing the risk of readmission for LTV patients.
Key Words: aged • chronic disease • critical illness • hospital readmission • ICU • length of stay • long-term mechanical ventilation • mechanical ventilation
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 week ICU stay.7
A 30-day hospital readmission
rate of 11% and a 6-month readmission rate of 30 to 40% has been
reported for elderly patients.5
8
A 30-day hospital
readmission rate of 20.4% was reported for patients with
COPD,1
while a 3-month readmission rate of 24 to 45% was
reported for patients with congestive heart failure
(CHF).9
10 Another population that is potentially at high risk for hospital readmission comprises patients who have required long-term mechanical ventilation during their index hospital stay. These long-term ventilator (LTV) patients comprise approximately 3 to 6% of all ICU admissions while consuming a disproportionate amount of ICU resources.11 12 These patients are often elderly with a variety of underlying chronic conditions that complicate or exacerbate their acute illness.11 13 Approximately 35 to 50% of them die in the hospital, and most discharges are to an extended-care facility.7 11 14 15 16 After discharge from the hospital, these patients have a slow recovery of functional status,16 a high mortality rate,11 16 and a high cost of care.17
To date, and to our knowledge, data about readmission rates and post-hospital discharge outcomes for this group of patients have not been reported. The purpose of the present study was to document readmission rates for LTV patients and to examine factors that might identify patients who are at increased risk for readmission within 6 months after hospital discharge.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
5 days of mechanical
ventilation and those who have received 
4 days, and the 1993
Society of Thoracic Surgeons patient classification system for surgical
cardiac patients classified short-term mechanical ventilation as being
< 5 days.
The study was conducted at the following three sites: (1) a university
medical center; (2) a university-affiliated Veterans Administration
(VA) medical center; and (3) a small community hospital. All three
sites were located in Cleveland, OH. The study was approved and
reviewed annually by the institutional review boards of University
Hospitals of Cleveland and the VA Medical Center. Eligibility criteria
were the following: age 18 years; the ability to speak English; and
having received > 96 h (4 days) of continuous mechanical ventilation
while in the ICU. All diagnoses were accepted. Patients who required
mechanical ventilation at home before becoming hospitalized were not
eligible for enrollment.
From February 1997 to February 1999, research nurses screened all patients admitted to all ICUs at the three study sites to determine eligibility. The research nurses tracked eligible patients during their hospital stays, and patients (or their proxies) were approached for written informed consent once the patient was close to hospital discharge. Patients then were observed for 6 months posthospital discharge.
Prior to data collection, research nurses were trained in the use and
administration of all interview tools. Interrater reliabilities between
the research nurses was assessed; acceptable reliabilities of 90%
agreement as well as Pearson correlations of at least 0.80 (for
continuous variables) and 
’s of 0.70 (for categorical variables)
were established before data collection proceeded.18
19
Every 3 months throughout the data collection period, ongoing
interrater reliability also was assessed, and retraining occurred if
reliability fell below acceptable levels.
After hospital discharge, using a scripted telephone survey instrument, patients (or proxies) were asked monthly about their current living location and subsequent hospital readmission. The reliability of the use of proxies for such data has been established.20 21 In 95% of the cases, readmission was confirmed by using hospital computers to verify the dates and length of stays for readmission. This procedure has been used successfully in other research.11 Patients were then interviewed within 2 weeks of hospital discharge to assess their quality of life (QOL), functional status, and psychosocial status. Prior to the interviews, patients were given an orientation test22 ; for those deemed to be cognitively impaired, the QOL interviews were not conducted. Instead, the interviewer met with a patient proxy and obtained information only about the patient’s activities of daily living.
The instruments used for data collection purposes were as follows. The
acute physiologic and chronic health evaluation (APACHE) III was used
to obtain information about the index hospitalization severity of
illness. APACHE III scores were calculated within the first 24 h
of ICU admission. APACHE III is a severity-of-disease classification
system that predicts the risk of death. Possible scores range from 0 to
299, with higher numbers representing patients who are more ill.
The average admission APACHE III scores reported for 17, 440 ICU
patients was 50, and for most disease categories the probability of
in-hospital death was > 50% for APACHE III scores of
> 90.23
Accuracy in predicting death was found to be
90%.22
Knaus et al24
have reported
interrater reliability at 0.90. Because our study sample was a
purposive subgroup of the total population of ICUs, no attempt was made
to draw conclusions about differences or similarities between expected
and observed levels of mortality.
The Charlson weighted index of comorbidity (CWIC) was used to quantify information about the prehospital comorbid conditions. The CWIC is an instrument that uses a weighted index that takes into account the number and seriousness of comorbid disease.25 Weights (range, 1 to 6) are assigned to comorbid conditions, and the weighted scores are added to obtain a total score. Scores range from 0 to 37, with higher scores representing higher numbers and seriousness of comorbid conditions. Interrater reliability and concurrent validity have been established.25
The sickness impact profile (SIP)26 was used to measure QOL posthospital discharge. In an evaluation of the use of QOL instruments in medical research, the SIP was reported as the most widely used tool for measuring QOL.27 The SIP has the following two major dimensional subscales: physical and psychological. The SIP consists of 136 questions regarding physical functioning and psychosocial functioning for the individual on the day that it is administered. It has the following 12 specific subscales: ambulation; mobility; body care and movement; social interaction; communication; alertness behavior; emotional behavior; sleep and rest; eating; home management; recreating and pastime; and employment. The SIP measures health status by assessing the impact of sickness on daily activities and behavior.26 A SIP score is calculated from the dysfunction score attributed to each question. Total SIP and individual category dysfunction scores are expressed as a percentage of the sum of the weights of the affirmatively checked statements, divided by the sum of all factor weights under analysis.16 The greater the percentage reported, the greater the negative impact of sickness on the daily activities of living and QOL. The general population has an SIP score of approximately 5%. An SIP score of 20 indicates the need for substantial daily care, and a score of > 30 indicates the need for complete care.16 The SIP has been tested on patients with COPD, on ICU patients, and on patients with a variety of chronic illnesses.28 29 30 A test-retest reliability of 0.94 was reported, and internal consistency was reported from 0.81 to 0.97.26 For the present study, ongoing internal consistency for the SIP ranged from 0.89 to 0.96. Construct validity has been reported and convergent evidence shows that this scale correlates with the same trait as measured by other methods.26
Trained research nurses abstracted demographic and clinical data from
hospital charts. The primary variables of interest were the
following: age at index admission; gender; length of hospital stay;
length of ICU stay; length of mechanical ventilation for index
admission; presence of chronic conditions (eg, renal
disease, COPD, coronary artery disease [CAD], or diabetes), discharge
disposition, weaned status at discharge, need for oxygen at discharge,
discharge diagnosis-related group codes, ICD-9 codes, comorbid
conditions, and severity-of-illness APACHE III scores. An
investigator-designed data collection tool was constructed for the
purpose of abstracting these data from the charts. Interrater
reliability was assessed before data collection began and every 3
months throughout the data collection period. Pearson correlations
ranged from 0.79 to 1.0 for continuous variables, and values ranged
from 0.6 to 1.0 for categoric variables.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Refusals
Using 
2 analysis for categoric variables
and analysis of variance (ANOVA) for continuous variables, patients who
refused were compared to those who did not refuse on clinical and
demographic variables. There were no statistically significant
differences between the groups on any variables except for
precipitating factors for mechanical ventilation and gender. Sixty
percent of patients who refused to participate in the study had
respiratory distress as the major precipitating event for requiring
mechanical ventilation, compared to 44% for those who did not refuse.
In addition, 42% of those who did not refuse had operative ventilation
as the precipitating event for requiring mechanical ventilation
compared to 25% of those who refused consent. Of those subjects who
refused consent, a greater percentage were women (55.1%) than men
(44.9%).
A majority of refusals were given by patients (46.1%) or their families (42.1%). The relatively high refusal rate given by family members may represent family issues related to concerns about managing the patient’s posthospital discharge course of care. The major reason given for refusal was that the patient (or family) felt "too overwhelmed" or felt that the patient was "too ill" to participate. Given these reasons for refusals, we examined all appropriate variables to see whether refusers and nonrefusers had different outcomes. We examined APACHE III scores, hospital length of stay, age, need for discharge oxygen, tracheostomy or mechanical ventilation, and discharge disposition to see whether the groups differed statistically. There were no statistically significant differences between those who refused and those who did not on any of these variables, and we have no reason to think that gender and precipitating factors for mechanical ventilation differences had any impact on outcome. Given the hospital discharge variables that we were able to examine, there were no differences between these groups relevant to the severity of illness or their condition at discharge that would affect posthospital discharge outcomes.
Demographic and Clinical Characteristics
The demographic and clinical characteristics of the sample are
shown in Table 1
. Comparisons of the three hospital subjects were made on demographic,
clinical, and outcome variables. There were no statistically
significant differences on any outcome variables; however, there were
some in-hospital differences. Patients from the community hospital were
statistically different from VA medical center and university hospital
patients on the following variables: age; APACHE III score; length of
hospital stay; and total number of prehospital comorbid conditions.
Given the fact that community hospitals tend to have patients who are
older, the age and comorbid differences between subjects were not
surprising. In addition, given that the sole criterion for admission
into the study was length of mechanical ventilation, and that we
enrolled subjects from a variety of hospitals, we assumed that there
would be a heterogeneous population on clinical variables. Such
heterogeneous characteristics among LTV subjects have been documented
elsewhere.7
13
15
Finally, only 12 subjects (6%) in the
sample were from the community hospital setting, and their inclusion
did not distort outcome patterns as we found no differences in outcomes
by hospital setting. Therefore, subject data from all three hospitals
were combined for all future analyses.
|
Prior to the index hospitalization, 14.3% of the sample were living in an institutional setting (eg, extended-care facility or hospital), while 83.7% were living in a home setting (usually their own home). More than half of the patients (62%) were cared for in a medical critical-care unit (ie, either a medical ICU or a cardiac ICU). While only 28.6% of patients had COPD as one of their diagnoses in the medical records, 76.9% had CAD or CHF as one of their diagnoses. For patients with a diagnosis of diseases of the respiratory system, 46% had pneumonia or bronchitis, and 30% had acute respiratory failure. No single diagnosis or procedure (other than the need for mechanical ventilation) was present in all cases. The admitting diagnoses were varied: 16.8% were for cardiac surgery; 8.2% were for treatment of cancer; and 3% were for organ transplantation. Other reasons for hospital admissions included the following: infections; acute myocardial infarction; CHF; trauma; and digestive disorders. The most common causes precipitating mechanical ventilation were the following: respiratory insufficiency/failure (44.4%); and operative ventilation (41.8%). Greater than half of the sample (65.7%) had no advance directives, 35.3% of those readmitted had advance directives, and 33.7% of those not readmitted had advance directives.
Of the 199 patients who survived their in-hospital stay, 134 (67.3%) were discharged to an extended-care facility (nursing home, 89 patients; and rehabilitation center, 45 patients), 55 (27.7%) were discharged to home, and 11 (5%) were discharged to another hospital. The most common discharge diagnosis-related group code was 483 (28.6%; tracheostomy except for face, mouth, and neck diagnoses), followed by 475 (15.9%; respiratory system with ventilatory support) and 106/107 (6.2%; coronary artery bypass grafting). The most common primary ICD-9 diagnoses were coronary (26.4%) and pulmonary (26%).
Hospital Readmission
Of the 199 patients who lived to hospital discharge, 76 were
readmitted at least once to a hospital within 6 months of their index
hospitalization for a 6-month readmission rate of 38.2%. Twenty-nine
patients (38.2%) were hospitalized multiple times (range, 2 to 7
readmissions). These 76 patients had a total of 126 hospital
readmissions during the 6-month follow-up period. Demographic and
clinical characteristics for those patients who were readmitted
reflected the sample at large.
Figure 1 shows the frequency of readmission by month, with months 1 and 2 having the greatest number of readmissions. The 30-day readmission rate was 21.1%, and the 60-day readmission rate was 34.2%. The mean number of days after the index hospital discharge until the first readmission was 39.2 days (SD, 44.8 days) with a median of 19 days (range, 0 to 167 days). The average length of stay for the first readmission was 9.8 days (SD, 13.2 days) with a median of 5.0 days (range, 1 to 70 days). The most common reasons for readmission were the following: a new problem had developed (54.6%); the primary diagnosis had worsened (22.7%); and a secondary diagnosis had worsened (18.6%).
|
2 analysis, there was
no statistically significant association between discharge disposition
and readmission.
Clinical and demographic variables for patients who had been readmitted
were compared to those who had not been readmitted within 6 months of
their index hospitalization. ANOVA was used for continuous variables
(ie, age, APACHE III score, CWIC, length of hospital stay,
length of mechanical ventilation, length of ICU stay, functional status
at discharge, and Glasgow coma score on admission), and
2 was used to examine differences between
groups for categoric variables (ie, discharge disposition;
gender; race; presence of renal failure, COPD, CAD, or diabetes; and
need for oxygen at discharge). There was a statistically significant
difference between groups for the variable length of hospital stay
(p = 0.025). Those patients who had at least one readmission had a
longer mean length of stay for their index hospitalization (44.5 days;
confidence interval [CI], 36.6 to 52.5) than did those who had no
readmission (34.7 days; CI, 29.9 to 39.4).
Readmission and Mortality:
Of all patients who lived to their
index hospital discharge, 30.7% died within the 6 months posthospital
discharge follow-up period. For patients with at least one readmission,
the posthospital mortality rate within 6 months of their index hospital
discharge was 33.3% compared to 29.3% for those patients with no
readmission. The association between posthospital mortality and
readmission was not statistically significant. Using ANOVA and
2 analysis, we compared demographic and clinical
variables between patients who had a readmission with subsequent death
and patients who died without any readmission. There was a
statistically significant difference between groups for the variable
APACHE III score (p = 0.008). Patients who had a readmission with
subsequent death had a higher mean APACHE III score from their index
ICU admission (84.5; CI, 72.1 to 96.9) than did those who died without
any readmission (65.3; CI, 57.8 to 72.9). In addition, for those
subjects who were readmitted, there was no statistical association
between discharge disposition and mortality. Mortality rates were
39.5% for those discharged to a nursing home, 29.4% for those
discharged to a rehabilitation facility, and 25% for those discharged
to home.
Predictors of Readmission: Using logistic regression, we examined how well the following variables predicted readmission within the first 6 months of index hospitalization for our patient population. Variables included as covariates were the following: age; length of hospital stay; length of ICU stay; length of mechanical ventilation; functional status at discharge; admission APACHE III score; admission Glasgow coma score; cognitive status at discharge; renal dialysis; presence of diabetes; need for oxygen at discharge; and discharge disposition.2 6 7 11 14 16 31 32 While there were other variables relevant to patient condition at hospital discharge that could affect the risk for readmission (eg, loss of organ function or presence of decubitii7 ), we did not have adequate data for these variables and, thus, did not include them in our analyses. Using forward stepwise regression, the model that included the variables length of hospital stay, length of mechanical ventilation, and need for oxygen at hospital discharge was statistically significant in predicting readmission (p = 0.0025). The need for oxygen at discharge made the greatest unique contribution to predicting readmission (ß = 0.693), and length of hospital stay and length of mechanical ventilation each made the same unique contribution (ß = 0.03). With these three variables in the equation, correct classification for readmission and no readmission within 6 months of index hospital discharge occurred 62.8% of the time.
Using survival analytic techniques, we analyzed the relative risk for time to the first readmission. Using the Kaplan-Meier33 estimates of the survival time against the natural logarithm of the survival time indicated that the proportional odds model34 was appropriate to analyze the time to first readmission. Thus, the log-logistic parametric survival model with the proportional odds property was used to analyze the time to first readmission. This analysis was implemented using the statistical analysis system procedure PROC LIFEREG.35 A graphic representation of the hazard function for readmission was obtained using the parameter estimates of the log-logistic distribution. Ninety-five percent CIs for the hazard at 30 days and at 90 days were obtained from the life-table estimates of the hazard and its SDs implemented through the statistical analysis system procedure PROC LIFETEST.35
The age category of 66 to 71 years of age was statistically significant for relative risk of readmission within 1 month of index hospital discharge. Figure 2 shows the graphic representation of the log-logistic hazards for the four categories of age. While not statistically significant, the probability of readmission decreases over time, with the greatest risk of readmission being within the first 30 days after the index hospital discharge date. After 30 days, the hazard function curves flatten out with there being negligible change in the probability of readmission after 60 days.
|
Readmission and QOL: Patients residing at home prior to readmission had a significantly better SIP QOL score, better physical functional status, and better psychosocial functioning than did either the nursing home or rehabilitation facility groups (Table 2 ). Only patients deemed cognitively intact were interviewed for the SIP. The scores reported in Table 2 reflect the data from patients (no surrogates included). For the overall QOL score as well as for the functional status score, the mean scores for nursing home and rehabilitation facility subjects were worse than for the sample at large. We also found, when examining functional status subscales individually, that patients residing at home prior to readmission had better ambulation, mobility, and body care and movement scores than did either the nursing home or rehabilitation facility groups; these findings were statistically significant for the variable body care and movement, which relates to issues of bathing, dressing, and getting out of bed (Table 3 ).
|
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The relatively short average duration of time before readmission (39 days) suggests that patients develop a new related problem or continue to experience exacerbation of the illness state or condition that resulted in their prolonged stay during the original or index hospitalization. The length of stay once the patient is readmitted (10 days) is twice the national average for a medical-surgical hospital admission. This suggests that these patients have not regained clinical stability after the original discharge and that new problems or worsening of the underlying condition are likely to result in an additional prolonged stay in the hospital.
Third, this patient population has much higher SIP scores (indicating poorer QOL and functional status) than other patient populations. For example, the total SIP scores of 6.8 and physical function scores of 4.5 at hospital discharge have been reported for typical ICU patients,29 and total SIP scores of 42 and physical function scores of 47 have been reported at 30 days posthospital discharge for subjects with prolonged surgical ICU stays.16 The SIP and physical function scores reported for our subjects, regardless of whether or not they were readmitted, indicate that a high percentage of them have their QOL and functional status negatively impacted by their illness. These scores stand in stark contrast to those reported for other ICU patients. In addition, the SIP scores for patients discharged to a nursing home or rehabilitation facility placed them in the category of needing "complete care."16 While LTV patients have similar hospital severity of illness and length of stay as those of other samples of extended-stay ICU subjects, they had worse outcomes in terms of QOL, physical functioning, and body care and movement. While these worse outcomes did not appear to relate to readmission, further investigation into strategies to enhance QOL and physical functioning in all discharge dispositions might result in better outcomes and reduced readmission rates.
It is also of interest to note that there was no relationship between discharge disposition and readmission, despite the fact that patients discharged from the hospital to home appeared to be healthier or at least to have better indexes of functioning, as measured by the SIP. Thus, while placement in a nursing home did not contribute additional risk to the probability of readmission, neither did the presence of 24-h caregivers or professional nursing supervision in the nursing home do anything to lessen the risk. Finally, the identification of some risk factors for readmission provide some new information in targeting LTV patients who are at the highest risk for readmission. High index hospital lengths of stay (ie, > 40 days), the need for oxygen at hospital discharge, and the age cohort of 66 to 71 years can help in identifying and tracking specific subsets of LTV patients who might benefit from a closer post-hospital discharge follow-up regimen.
There are several limitations to the present study. First, because there are varying definitions of what constitutes long-term mechanical ventilation, it is difficult to compare the results of studies and to generalize from the results because the samples are not derived using the same criterion of mechanical ventilation days. A second limitation is our refusal rate of 16.5%. Consent rates of at least 70% are recommended in order to minimize threats to external validity.37 While our refusal rate is acceptable by these standards, a lower refusal rate would have resulted in a larger sample size and enhanced generalizability. Thus, while there were differences between patients who refused to participate and those who did not refuse (ie, gender and the precipitating cause for mechanical ventilation), we found no statistically significant differences between refusers and nonrefusers on any of our acuity or outcome variables. While it might be possible that there were differences between these groups on acuity or outcome, we found none, given the variables we were able to examine. A final limitation of the study relates to incomplete data regarding additional in-hospital variables that could relate to the risk for readmission (eg, decubitii, in-hospital events and procedures, loss of organ function). Complete data regarding these variables could, perhaps, result in additional variables that relate to the risk for readmission.
In summary, LTV patients should be considered at risk for readmission once they have been discharged from their index hospitalization. LTV patients who have had longer in-hospital lengths of stay (ie, > 40 days), who have had longer durations receiving mechanical ventilation, and who need oxygen at hospital discharge are at even greater risk for readmission and should be observed closely in the immediate posthospital discharge phase. Research examining specific posthospital discharge programs could help to identify additional factors related to the risk for readmission and could examine the effects of such an approach to care on readmission rates. The LTV patient population is a vulnerable one, not only in the hospital setting, but also in the posthospital setting.
|
Footnotes |
|---|
This study was funded by grant No. RO1-NRO4318 from the National Institute of Nursing Research.
Received February 3, 2000; revision accepted March 5, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Guentner, L. A. Hoffman, M. B. Happ, Y. Kim, A. D. Dabbs, A. B. Mendelsohn, and L. Chelluri Preferences for Mechanical Ventilation Among Survivors of Prolonged Mechanical Ventilation and Tracheostomy Am. J. Crit. Care., January 1, 2006; 15(1): 65 - 77. [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]
|
|
|
|
|
|
B. J. Daly, S. L. Douglas, C. G. Kelley, E. O'Toole, and H. Montenegro Trial of a Disease Management Program to Reduce Hospital Readmissions of the Chronically Critically Ill Chest, August 1, 2005; 128(2): 507 - 517. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Daly, S. L. Douglas, and C. G. Kelley Benefits and Challenges in Developing a Program of Research West J Nurs Res, April 1, 2005; 27(3): 364 - 377. [Abstract] [PDF]
|
|
|
|
|
|
S. L. Douglas and B. J. Daly Caregivers of Long-term Ventilator Patients: Physical and Psychological Outcomes Chest, April 1, 2003; 123(4): 1073 - 1081. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
