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


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Daly, B. J.
Right arrow Articles by Montenegro, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Daly, B. J.
Right arrow Articles by Montenegro, H.
(Chest. 2005;128:507-517.)
© 2005 American College of Chest Physicians

Trial of a Disease Management Program to Reduce Hospital Readmissions of the Chronically Critically Ill*

Barbara J. Daly, PhD, RN; Sara L. Douglas, PhD, RN; Carol Genet Kelley, PhD, RN; Elizabeth O’Toole, MD and Hugo Montenegro, MD

* From the School of Nursing (Drs. Daly, Douglas, and Kelley) and from the School of Medicine (Drs. O’Toole and Montenegro), Case Western Reserve University, Cleveland, OH.

Correspondence to: Barbara J. Daly, PhD, RN, School of Nursing, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106-4904; e-mail: bjd4{at}case.edu


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Recommendations
 References
 
Background: Patients requiring prolonged periods of intensive care and mechanical ventilation are termed chronically critically ill. They are prone to continued morbidity and mortality after hospital discharge and are at high risk for hospital readmission. Disease management (DM) programs have been shown to be effective in improving both coordination and efficiency of care after hospital discharge for populations with single-disease diagnoses, but have not been tested with patients with multiple-disease diagnoses, such as the chronically critically ill.

Study objectives: To test the effect of a DM program on hospital readmission patterns of chronically critically ill patients during the first 2 months after hospital discharge and to estimate the cost-effectiveness of the DM program.

Design: Randomized, controlled trial.

Setting: Academic medical center, extended care facilities, and participant homes.

Participants: Three hundred thirty-four consenting adults from one academic medical center who underwent > 3 days of mechanical ventilation and survived to hospital discharge.

Intervention: Two hundred thirty-one patients in the experimental group received care coordination, family support, teaching, and monitoring of therapies from a team of advanced-practice nurses, a geriatrician, and a pulmonologist for 2 months post-hospital discharge.

Measurements: Rehospitalization rate, time-to-first rehospitalization, duration of rehospitalization, mortality during rehospitalization, and associated costs.

Results: Patients who received DM services had significantly fewer mean days of rehospitalization (11.4; 95% confidence interval [CI], 9.3 to 12.6) compared with the control group (16.7 days; 95% CI, 12.5 to 21.0; p = 0.03). There were no other significant differences between experimental and control groups, although all measures of rehospitalization risk for the experimental group were in a positive direction. Total cost savings associated with the intervention were approximately $481,811 for the 93 subjects who were readmitted to the hospital.

Conclusions: Chronic critical illness may have a natural trajectory of continued morbidity following hospital discharge that is not affected by the provision of additional care coordination services. Nevertheless, given the high cost of rehospitalization and the additional burden it imposes on patients and families, interventions that can reduce the duration of rehospitalization are cost-effective and merit continued testing.

Key Words: chronically critically ill • disease management • long-term ventilation


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Recommendations
 References
 
The growing number of patients who survive a critical illness but require prolonged, intensive medical and nursing care, usually because of the need for long-term mechanical ventilation, has been recognized for several decades.123 Early descriptive work456 documented in-hospital mortality rates that ranged from 50 to 60%. Because of the high costs of care for these patients in the context of relatively poor outcomes, this population, termed the chronically critically ill (CCI), has been the subject of numerous investigations. Early studies78910 focused on the design of alternative care units, such as weaning centers, special care units, and, more recently, long-term acute care (LTAC) units.

Over the past decade, investigations have continued to examine long-term outcomes following prolonged stays in critical care units111213 and prolonged mechanical ventilation.1415161718 These descriptive, correlational reports have consistently documented the pattern of continuing morbidity and mortality among the CCI. A particular concern has been the high rate of hospital readmission in the first 6-month post-hospital discharge period, which is noted to be close to 40%.1419

CCI patients who survive lengthy stays in the ICU present many post-hospital discharge challenges to the health care teams who assume responsibility for their care. Extended care facility staff and primary care physicians are increasingly faced with assuming the responsibility for managing these complex patients, quickly gaining familiarity with the patient’s history and course of illness, and coordinating the continuing input from multiple medical specialists and health care providers. The primary purpose of our study was to evaluate the effect of adding disease management (DM) services to the usual care system on rehospitalization patterns among CCI patients in the first 2 months after their hospitalization.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Recommendations
 References
 
This study was a randomized trial of a DM program (DMP) for patients and caregivers following hospitalization and prolonged mechanical ventilation. Previous research1420 has established that patients who require mechanical ventilation beyond 72 h are at high risk of death or prolonged hospitalization with multiorgan dysfunction and continuing care needs beyond hospital discharge. There continues to be some debate in the literature about what duration constitutes long-term ventilation; investigators have used 2 days,31821 5 days,20 and 7 days.612 We chose 72 h in order to exclude patients who were simply slow to wean while still capturing those whose clinical problems were likely to entail a high risk of continued mortality and morbidity. Other inclusion criteria included an understanding of the English language, an absence of ventilator dependency prior to the index hospitalization, and a hospital discharge location < 80 miles from the study site. The patients who had received organ transplants and received case management from the transplant team, as well as hospice patients, were excluded from the study.

The study site was University Hospitals of Cleveland, a 950-bed tertiary care hospital associated with Case Western Reserve University. All of the patients who met the eligibility criteria during their hospitalization in any adult ICU of the study hospital were enrolled. Patients and family members were approached for consent to participate when it was clear that hospital discharge was expected within the next few days. The Institutional Review Board approval was obtained prior to data collection.

Between March 2001 and December 2004, research nurses screened all of the ICU patients for study eligibility. All of the study staff were trained in the use of the research instruments prior to data collection, and interrater reliability was monitored on a quarterly basis throughout the study. The retraining and refinement of data collection rules were done as needed if the correlations and percentage of agreements fell below acceptable levels.

Intervention
Patients were randomly assigned to the DM program (DMP) or usual care. After 18 months of the trial, the randomization scheme was changed from a 1:1 to a 1:4 (control/experimental ratio) block scheme, using computer-generated assignment, in order to maintain full caseloads for the intervention team. The 1:4 randomization scheme continued for the remaining 26 months of the study. The DM patients received case management services from an advanced-practice nurse (APN), who had access to a pulmonologist and a geriatrician for guidance and collaboration. The APN met with the patient and family several days prior to hospital discharge in order to review the hospital course, perform a baseline assessment of both the patient and the caregiver, discuss hospital discharge plans, and establish a plan of care. The APN consulted with the hospital care team about the patient’s care needs and completed a hospital discharge summary that included the post-hospital discharge plan of care for the patient, the patient goals, the presence of advance directives, and an assessment of family coping. This summary was sent on the day of hospital discharge to all of the relevant out-of-hospital health care providers (nursing home staff, family physician, consulting specialist, home care agency, etc).

All of the DMP patients were visited by the APN within 48 h of hospital discharge and a second time within the first week. Visits were made at least weekly for the next 3 weeks and then at least every other week for the last 4 weeks, for a minimum total of eight visits. A visit was also made whenever there was a transition in care location (eg, discharge to home, transfer from one level of care to another, or hospital readmission) to assure continuity in care. The patients who lived > 30 miles from the study site were contacted by telephone rather than through in-person visits. All of the contacts with patient, family, and health care providers were documented using data collection forms.

During the 8-week intervention, the APN performed case management activities, which varied greatly with patient condition, location, and presence of family support. Typical activities included attending team meetings at extended care facilities, helping families prepare for the patient’s eventual return home, emotional support for family members, counseling families about end-of-life options, coordinating services among multiple providers, arranging for needed follow-up care from specialists, and monitoring patient condition and medications. In addition to participating in overall project management, the geriatrician and pulmonologist were consulted by the APNs as needed for advice regarding medication management, treatment alternatives, and strategies addressing specific patient problems. This collaboration generally took place over the telephone and occurred once or twice a month.

Patients in the usual care group were interviewed within 2 weeks of hospital discharge for completion of study instruments, then at 1 and 2 months post-hospital discharge for data collection. The interviews were conducted in-person at hospital discharge and by telephone after hospital discharge. When patients and family members in the usual care group asked for advice or information from the interviewers, they were referred back to their primary care provider, extended care facility staff, or home care agency.

Instruments
Demographic data (age, gender, and race) were collected on both patients and caregivers, and descriptive clinical data (diagnosis-related group, reason for ventilation, and length of stay) were abstracted from patient medical records. Instruments included the acute physiology and chronic health evaluation (APACHE) III, the Medical Outcomes Study eight-item short form (SF-8), and the outcome and assessment information set (OASIS). Billing data were used to determine the standardized charges for post-hospital discharge use of services, such as nursing home, rehabilitation center, LTAC facility, and hospital readmission.

APACHE III is a well-established severity-of-illness classification tool that measures mortality risk using physiologic and chronic health data taken from the first 24 h of ICU admission. APACHE scores range from 0 to 299, with higher scores representing a higher risk of death.22

The SF-8 is the latest version of the original 36-item instrument designed to measure health status. It consists of eight items, each of which addresses a different dimension of quality of life, and yields two subscale scores (physical and mental). Scores on the physical subscale range from 14.1 to 64.0, with higher scores indicating a better physical health status. Scores on the mental subscale range from 6.8 to 70.4, with higher scores indicting a better mental health status. The SF-8 score was obtained within 2 weeks of hospital discharge and again at 2 months post-hospital discharge. The patients were also asked to complete the SF-8 in reference to their health status for the week prior to hospitalization. Alternative form reliability for summary scores has ranged from 0.85 to 0.90 for the 1-week recall version. For patients with one or more physical comorbidities, the mean (± SD) SF-8 physical scores have been reported as 48.12 ± 9.23, and mean SF-8 mental scores were 54.08 ± 6.54.23

All of the patients, regardless of discharge location, were assessed with the OASIS. The OASIS is a 79-item tool commonly used by home care agencies to assess the care needs and outcome measures of patients.24 Interrater reliability has been reported (Pearson’s r, 0.56 to 1.0). The 14 items assessing activities of daily living and instrumental activities of daily living (ADL/IADL) provide a total score ranging from 0 to 66, with higher scores indicating an increased dependency or need for help in performing these activities. Construct validity has been reported25 with functional domain items loaded onto one factor.

The Katzman Short Orientation-Memory Concentration Test was used to determine whether subjects were cognitively impaired prior to conducting the subject interviews. The test includes three orientation questions with the possible points ranging from 0 (normal) to 28. Scores > 6 correlate with dementia.26 Reported reliabilities range from 0.886 to 0.922.27

The resource use was assessed using standardized charges. Standardized charges are a measure for resource use that is both interpretable and standardized across settings and time. We utilized actual billing data from patients rehospitalized during the year 2001 to determine a "standardized charge" for 1 day of rehospitalization. This charge was used to calculate the total charge for each hospital readmission episode during the study period. This method has been used successfully in previous studies2028 by the investigators. We calculated the cost of the intervention by using the total salary of the APNs plus benefit costs.

Statistical Analysis
The comparisons between the experimental and control groups were done using analysis of variance for nonskewed continuous variables, Mann-Whitney U for skewed continuous variables (two-group comparisons), and {chi}2 for categorical variables. Time-to-first hospital readmission was compared using survival analytic techniques, and logistic regression was used to determine variables that predicted hospital readmission, as well as death, following hospital readmission. The sample size was calculated using power analysis that incorporated the following assumptions: level of significance = 0.05, nondirectional hypotheses, medium effect size, and desired power of 90%. Based on these assumptions, a sample size of 256 was needed.29


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Recommendations
 References
 
Figure 1 shows the distribution of the total sample, percentage eligible, refusals, and in-hospital deaths. During the 34-month enrollment period, 4,186 mechanically ventilated patients were assessed; of these, 1,041 patients (24%) required mechanical ventilation for at least 72 h, met the additional eligibility criteria, and, thus, were enrolled. The in-hospital mortality rate for this group was 37.7% (n = 393). An additional 314 (30.1%) of the 1,041 eligible patients were not randomized in the post-hospital discharge portion of the study because of refusal (n = 185), discharge to a location beyond the 80-mile requirement, lack of family care giver, non-English speaking, or transfer to hospice. This yielded a sample of 334 patients who were randomized to the experimental group (n = 231) or the control group (n = 103). The 28.5% refusal rate is consistent with that found in previous work with this population and was most often related to the patient’s or family member’s report of feeling "overwhelmed" and being reluctant to take on any additional burden of research.20 Although the two sample sizes were unequal, the key outcome variables (hospital readmission, survival, and mean number of days until first hospital readmission) were analyzed using nonparametric tests. Thus, there were no concerns related to violations of the assumptions (normal distribution or homogeneity of variance) as there would have been had we used parametric statistical tests.



View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Sample selection.

 
Table 1 displays the sample characteristics for the entire sample, and Table 2 shows the demographic and clinical characteristics of the 334 patients who were randomized to the intervention and control groups. The mean age of the entire sample was 62.9 years, with slightly more women (53.3%) than men and more Caucasians (61%) than other races. Although, on average, the patients had several chronic conditions (mean = 5.7; 95% confidence interval [CI], 5.5 to 5.9), almost all of the patients (89.3%) were living at home prior to the index hospital admission. Once admitted, half of the patients (49.5%) were cared for in a medical critical care unit (either a medical ICU or a cardiac ICU) and the other half in a surgical ICU or neurosurgical ICU. The two most common reasons for the initial episode of mechanical ventilation were respiratory insufficiency (47.7%) and routine postoperative ventilation (24.3%). The median hospital length of stay was 17 days (range, 3 to 126 days), and the median ICU stay was 11 days (range, 3 to 109 days).


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

  		Table 1. Sample Characteristics*

 

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

  		Table 2. Comparison of Demographic and Clinical Variables Between Experimental and Control Group Subjects*

 
Recognizing that patients who receive mechanical ventilation for routine postoperative reasons may differ in clinically important ways from patients who receive mechanical ventilation for other reasons, we examined several key variables by the four reasons for prolonged mechanical ventilation (routine postoperative care, respiratory insufficiency, airway protection, or postarrest). Although we found statistically significant differences between the categories of reasons for mechanical ventilation for the variables length of mechanical ventilation, APACHE, and mortality, we did not find differences in hospital readmission rates. Thus, the reason for mechanical ventilation did not have an association with the likelihood of hospital readmission, nor did any of the variables predict hospital readmission.

Most patients, as shown in Table 3 , were discharged from the hospital to an extended care facility, with only 76 patients (22.8%) able to return directly home. Discharge to a nursing home occurred more frequently (n = 122, 36%) than discharges to either an LTAC or rehabilitation facility (n = 68, 20% for both). At the time of hospital discharge, 57 of all patients (17%) still required mechanical ventilation for at least part of the day, and half of the patients (50%) were discharged from the hospital receiving oxygen. Although mean hospital admission APACHE III scores did differ between the groups (Table 2), there were no significant differences in hospital discharge disposition, ADL/IADL scores, or mental health status scores on the SF-8 between the experimental and the control patients. There was a statistically significant difference in the SF-8 physical health score between groups for those patients who were cognitively intact at hospital discharge, with the experimental group having a significantly lower mean.


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

  		Table 3. Comparison of Hospital Discharge Variables Between Experimental and Control Group Subjects*

 
Hospital Readmission
Of the 334 patients who lived to hospital discharge, 132 patients (39.5%) were readmitted at least once within the first 2 months of discharge from the index hospitalization. Ninety-one patients (68.9%) had only one hospital readmission, whereas 41 patients (31.1%) were hospitalized multiple times within the first 60 days (range, two to four hospital readmissions). These 132 patients had a total of 185 hospital readmissions during the study period. Twenty-five of the first hospital readmissions (19%) occurred within 48 h of hospital discharge. Of the readmitted patients, 25.8% (n = 34) died, whereas only 14.2% (n = 30) of the 211 patients who were not readmitted died (relative risk of 1.81; 95% CI, 1.61 to 2.61; p = 0.008).

Hospital Readmission for Experimental and Control Groups
There was no difference between the experimental and control groups in the relative risk for hospital readmission (see Table 4 ). There also was no statistically significant difference between the groups regarding the percentage of subjects having more than one hospital readmission during the 2-month study period. Although we did not obtain acuity data (APACHE) for hospital readmission, we did examine the primary reason for hospital readmission to see if there were differences between the experimental and control groups. There were no differences between groups on any of the clinical variables for hospital readmission except for "cardiac problems." More than twice as many of the hospital readmissions in the control group (38.9%) were for cardiac reasons (congestive heart failure, chest pain, etc) as compared with the experimental group (15.1%). This difference was statistically significant ({chi}2 = 4.57; 95% CI, 0.16 to 0.92; relative risk, 0.39; p = 0.03).


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

  		Table 4. Hospital Readmission Variables for Experimental and Control Groups*

 
There was a significant difference between the groups in the total number of days of rehospitalization. Summing all of the days spent in the hospital for all of the hospital readmissions, the experimental patients had 35% fewer total days (mean duration of stay, 11.4 days; 95% CI, 9.3 to 12.6 days) than the control group patients (mean, 16.7 days; 95% CI, 12.5 to 21.0 days; p = 0.03).

Death associated with hospital readmission was less frequent in the experimental group than in the control group (23.7% vs 30.8%, respectively), although the numbers in each group were small, and the difference did not reach statistical significance. There was also no difference in the risk of death following discharge from the readmission hospitalization. The mean number of days from the index hospital discharge date until the date of the first hospital readmission was slightly longer (mean, 15.9 days; 95% CI, 12.8 to 18.9 days) for the experimental group than for the control group (mean duration, 13.9 days; 95% CI, 9.4 to 18.5 days; p = 0.64). We also compared patients who were followed in-person (lived ≤ 30 miles from study site) with those who were followed long-distance by phone (lived > 30 miles from study site) to see if there were any differences in outcomes. We found no statistically significant differences between the groups for any of the hospital readmission variables.

The reason for hospital readmission was available from the medical records of those patients (n = 72) who were readmitted to the primary study hospital. Hospital readmissions were classified by research nurses as related to worsening of the original primary disease diagnosis (18%), worsening secondary diagnosis (29%), and/or development of a new problem (60%; 15 patients had two reasons for hospital readmission). There were no differences between the experimental and the control groups, tested with Fisher exact test. The most common clinical problem prompting hospital readmission was infection, and the most common type of infection was urinary tract.

Predictors of Hospital Readmission
Using logistic regression, we examined the influence of variables shown to predict hospital readmission in other studies1419303132 within the first 2 months of index hospitalization. Variables included as covariates were the following: age, gender, length of mechanical ventilation, functional status at hospital discharge (ADL/IADL scores from the OASIS), hospital admission APACHE III score, renal dialysis, presence of diabetes, and pressure ulcer at hospital discharge. The model with all of the variables in the equation was statistically significant in predicting the hospital readmission (p = 0.024) and the correct classification for hospital readmission, and no hospital readmission within 2 months of index hospital discharge occurred 62.3% of the time (R2 = 0.074 [Nagelkerke test]). Only one of the variables made a statistically significant contribution to the prediction of hospital readmission: functional status at hospital discharge (odds ratio, 1.13; 95% CI, 1.029 to 1.248; p = 0.011).

Next, we added the intervention variable to evaluate the impact of the intervention on hospital readmission. Adding the intervention variable to the model did not significantly increase the correct classification of hospital readmission. With the intervention variable in the model, correct classification for hospital readmission and no hospital readmission within 2 months of index hospital discharge occurred 65.2% of the time, an increase of only 2.9% over the model without the intervention variable included.

Hospital Readmission and Health Status
Using data from patient sources (no proxy data were included), we examined the physical and mental aspects of health status at hospital discharge of patients who were readmitted to the hospital. The physical SF-8 subscale variable was positively skewed, and the nonparametric Kruskal-Wallis test was used. As would be expected, patients residing at home prior to their first hospital readmission had higher (better) physical health scores than patients residing in all of the other types of facilities (p = 0.001). The mental SF-8 subscale variable was not highly skewed, and analysis of variance was used to compare mean mental health scores. Mental subscale scores behaved similarly to the physical scores, and those differences were also statistically significant (p = 0.002).

We then examined the physical and mental SF-8 subscale scores to determine whether there were differences between those patients readmitted to the hospital and those who were not readmitted. There were no statistically significant differences in the physical scores between those readmitted to the hospital (mean, 28.7; 95% CI, 27.2 to 30.3) and those not readmitted (mean, 30.4; 95% CI, 29.1 to 31.7; p = 0.116), nor in the mental subscale scores between those readmitted to the hospital (mean, 37.6; 95% CI, 35.3 to 39.9) and those not readmitted (mean, 38.1; 95% CI, 36.2 to 39.9; p = 0.766).

Cost Effectiveness
To examine the cost-effectiveness of reducing the hospital days for hospital readmissions, we calculated the total savings from 6.31 fewer days of hospitalization (for patients in the experimental group) minus the cost of the APNs who provided DM services. The salary plus benefit costs (fiscal year 2001, the same year used to obtain charges for services used) for two full-time APNs for 2.8 years (the duration of the intervention) was $420,000. Given the reduction in hospital days for the experimental group of 6.31 days per patient, at an average hospital charge of $3,415/d, the average savings were $21,548.65 per patient. Assuming that this savings was realized for each of the 93 rehospitalized intervention patients, the total reduction in hospital charges was estimated at $2,004,024. Approximating costs by using the study hospital’s Medicare cost-to-charge ratio of 0.45, rehospitalization cost savings were estimated at $901,810. Thus, after accounting for the salary costs of the APNs, the total savings were $481,811 for the 93 subjects who were readmitted or $5,180 per patient.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Recommendations
 References
 
Frequent hospital readmission in chronic illness is a ubiquitous phenomenon. Interventions designed to reduce the frequency of hospital readmission have been the focus of considerable study in this and other countries.31333435363738 Trials of interventions designed to reduce hospital readmissions typically have targeted patient groups who have single-disease diagnoses that lend themselves to the development of care paths and treatment algorithms, such as heart failure, stroke, and chronic lung disease.394041424344 The results of trials of education, case management, and close follow-up and home monitoring with these groups have been shown to have positive effects in reducing the risk of hospital readmission4445 and the combined risk of hospital readmission and mortality.46

Our results suggest that, first, patients who have survived a prolonged stay in an ICU are another group who are at high risk for hospital readmission, particularly within the first 2 months following hospital discharge. This finding is consistent with our previous investigations of this patient population.14 Second, although the frequency of hospital readmission was not altered by the intervention, some aspects of the hospital readmission pattern were modified by DM services. Third, because the number of costly rehospitalization days was significantly reduced, the provision of DM services was cost-effective.

DM programs have gained popularity as they have demonstrated the ability to improve outcomes and lower costs of common chronic conditions.4647 The three diseases most likely to be managed through DM programs are asthma, diabetes, and heart failure, and it has been estimated that as many as 25% of enrollees in managed-care insurance plans have access to DM programs.48 Although there has been some recognition that single-condition DM programs may be inadequate in improving care for the many people with multiple chronic conditions,49 there have been no other reports of applying DM principles to patients with multiple comorbidities, such as the CCI.

There are several possible ways in which the DM program we implemented may have led to the reduction in the duration of rehospitalizations and, possibly, in deaths associated with rehospitalization. Clarification of goals of care, resulting from the intensive support and counseling of patients and families in the period between initial hospital discharge and hospital readmission, was able to make the eventual transition from the readmission hospitalization back to an extended care facility more efficient. It is also likely that the improved communication and transmission of information between facilities and care teams accomplished by the APNs increased the efficiency of assessment, care planning and delivery during the rehospitalization. It may also have been that the close monitoring provided by the APNs was associated with more timely recognition of the need for rehospitalization, facilitating earlier intervention and, thus, a shorter hospitalization. However, the fact that the mean number of days between index hospital discharge and hospital readmission was slightly longer for the experimental group than control argues against this explanation.

Relatively simple, low-cost improvements in communication and coordination, recognized components of all DM programs,50 can have significant effects on care processes. By the time of hospital discharge, CCI patients have undergone multiple procedures, had exacerbations of many preexiting conditions, and have experienced repeated episodes of complications, such as sepsis and renal failure. Management of these complex patients is difficult when information is lacking or when it is available only after hours of sorting through volumes of medical records. In our study, providing a written summary of the lengthy and complicated illness course of the patient to each team of providers was almost universally identified as a significant help to the receiving care team. This is consistent with the work of van Walraven and Rokosh51 and van Walraven et al,52 who found a trend toward the reduced risk of hospital readmission among patients seen in follow-up by a physician who had received a written hospital discharge summary.

Finally, the ineffectiveness of the intervention in reducing hospital readmission rate and the predominance of a new problem as the reason for hospital readmission suggest that the hospital readmission rate is primarily a function of the natural pattern of chronic critical illness. As has been found in many other chronic conditions, prehospital functional status was the strongest predictor of rehospitalization.184042 This suggests that hospital readmission rate is a relatively insensitive outcome measure, although the frequency of urinary tract infection as a reason for hospital readmission might warrant targeting urinary tract infection prevention in post-hospital discharge quality improvement programs. In contrast, the length of stay for the hospital readmission, although in part a function of the condition that led to the hospital readmission, is apparently able to be shortened by the provision of the care coordination aspects of DM services. Additional trials of interventions should be tailored less toward reducing the hospital readmission rate and more toward preparing for and managing hospital readmissions efficiently, thus minimizing the human and financial costs.

Strengths and Limitations
The strength of our study was that it was a randomized trial and enrolled all of the eligible patients, regardless of their primary disease diagnosis or hospital discharge location. The intervention was implemented by experienced, skilled practitioners, but did not require any specialized training, nor did it require changes in the normal work schedule or pattern of the physician team members. As such, the results are applicable to the heterogeneous population of patients who become CCI and the clinicians who care for them.

Although the choice of 72 h of ventilation was intentionally liberal, in order to assure inclusion of all patients likely to experience prolonged ventilation, it probably did result in a sample that also included patients who did not share some of the more typical features of chronic critical illness. For example, relatively young patients with few comorbidities who require 3 or 4 days of ventilation either as part of treatment for a serious community-acquired pneumonia or severe asthma exacerbation can recover with few sequelae and relatively few post-hospital discharge care needs. The inclusion of patients like this who were at low risk for hospital readmission likely reduced the effect size of the intervention. More stringent enrollment criteria might have allowed a more sensitive analysis.

The study was limited in both the precision and comprehensiveness of measures of cost. Our results are also limited in generalizability in that the sample was drawn from one academic medical center, and the costs were calculated from charges in that facility. These rates are likely to be higher than in smaller community facilities. Given the high cost of care for patients with prolonged critical illness, a closer examination of all costs of care, including extended care facility costs, home care, and physician visits, is warranted.

Although we did not identify any demographic or clinical differences between the subjects who consented and those who refused, the high-refusal rate (28.5%) may be a threat to validity. However, because a majority of the refusals to participate were made by caregivers of patients too ill to consent who stated that they felt too overwhelmed to participate in research, it seems probable that those patients and caregivers most likely to benefit from a supportive intervention, such as the DMP, did not participate. Thus, the refusal rate, while of concern, may have resulted in underestimating, rather than overestimating, the effect of the intervention.


   Recommendations
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Recommendations
References
 
There are several avenues for the provision of DM for CCI. We recommend that managed care plans, the current major provider of these programs, consider expanding their diagnostic foci to include CCI patients. Hospital systems that include extended care facilities may also be able to improve their acute care bed utilization and facilitate transfers by implementing some aspects of a DM program. As the numbers of CCI patients continues to grow, there may be an additional need for freestanding DM programs, similar to the current model of independent geriatric case managers. The continued success in acute care of high-risk patients, such as the elderly and those with multiple preexisting chronic illnesses, is likely to be associated with growing numbers of patients who experience chronic critical illness. The additional development and testing of care delivery systems to address their complex needs will continue to be needed.


   Footnotes
 
Abbreviations: ADL = activities of daily living; APACHE = acute physiology and chronic health evaluation; APN = advanced-practice nurse; CCI = chronically critically ill; CI = confidence interval; DM = disease management; DMP = disease management program; IADL = instrumental activities of daily living; LTAC = long-term acute care; OASIS = outcome and assessment information set; SF-8 = Medical Outcomes Study eight-item short form

This study was supported by a grant from the National Institute of Nursing Research (NR-05207).

Received for publication October 10, 2004. Accepted for publication January 28, 2005.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Recommendations
 References
 

  1. Girard, K, Raffin, TA (1985) The chronically critically ill: to save or let die. Respir Care 30,339-347[Medline]
  2. Daly, BJ, Rudy, EB, Thompson, KS, et al Development of a special care unit for chronically critically ill patients. Heart Lung 1991;20,45-52[ISI][Medline]
  3. Gracey, DR, Gillespie, D, Nobrega, F, et al Financial implications of prolonged ventilator care of Medicare patients under the prospective payment system. Chest 1987;91,424-427[Free Full Text]
  4. Spicher, JE, White, DP Outcome and function following prolonged mechanical ventilation. Arch Intern Med 1987;147,421-425[Abstract]
  5. Douglas, P, Rosen, RL, Butler, PW, et al DRG payment for long-term ventilator patients. Chest 1987;91,413-417[Abstract/Free Full Text]
  6. Seneff, MG, Zimmerman, JE, Knaus, WA, et al Predicting the duration of mechanical ventilation. Chest 1996;110,469-479[Abstract/Free Full Text]
  7. Scheinhorn, DJ, Chao, DC, Stearn-Hassenpflug, M, et al Post-ICU mechanical ventilation: treatment of 1,123 patients at a regional weaning center. Chest 1997;111,1654-1659[Abstract/Free Full Text]
  8. Bagley, PH, Cooney, E A community-based regional weaning unit. Chest 1997;111,1024-1029[Abstract/Free Full Text]
  9. Bach, PB, Carson, SS, Leff, A Outcomes and resource utilization for patients with prolonged critical illness managed by university-based or community-based subspecialists. Am J Respir Crit Care Med 1998;158,1410-1415[Abstract/Free Full Text]
  10. Carson, SS, Bach, PB, Brzozowski, L, et al Outcomes after long-term acute care. Am J Respir Crit Care Med 1999;159,1568-1573[Abstract/Free Full Text]
  11. Heyland, DK, Konopad, E, Noseworthy, TW, et al Is it ‘worthwhile’ to continue treating patients with a prolonged stay (<14 days) in the ICU? Chest 1998;114,192-198[Abstract/Free Full Text]
  12. Lipsett, PA, Swoboda, SM, Dickerson, J, et al Survival and functional outcome after prolonged intensive care unit stay. Ann Surg 2000;231,262-268[CrossRef][ISI][Medline]
  13. Bashour, CA, Yared, JP, Ryan, TA, et al Long-term survival and functional capacity in cardiac surgery patients after prolonged intensive care. Crit Care Med 2000;28,3847-3853[CrossRef][ISI][Medline]
  14. Douglas, SL, Daly, BJ, Brennan, PF, et al Hospital readmission among long-term ventilator patients. Chest 2001;120,1278-1286[Abstract/Free Full Text]
  15. Chatila, W, Kreimer, DT, Criner, GJ Quality of life in survivors of prolonged mechanical ventilatory support. Crit Care Med 2001;29,737-742[CrossRef][ISI][Medline]
  16. Combes, A, Costa, MA, Trouillet, JL, et al Morbidity, mortality, and quality-of-life outcomes of patients requiring >14 days of mechanical ventilation. Crit Care Med 2003;31,1373-1381[CrossRef][ISI][Medline]
  17. Engoren, M, Buderer, NF, Zacharias, A long-term survival and health status after prolonged mechanical ventilation after cardiac surgery. Crit Care Med 2000;28,2742-2749[CrossRef][ISI][Medline]
  18. Chelluri, L, Pinsky, MR, Donahoe, MP, et al Long-term outcome of critically ill elderly patients requiring intensive care. JAMA 1993;269,3119-3123[Abstract]
  19. Nasraway, SA, Button, GJ, Rand, WM, et al Survivors of catastrophic illness: outcome after direct transfer from intensive care to extended care facilities. Crit Care Med 2000;28,19-25[CrossRef][ISI][Medline]
  20. Douglas, SL, Daly, BJ, Gordon, N, et al Survival and quality of life: short-term versus long-term ventilator patients. Crit Care Med 2002;30,2655-2662[CrossRef][ISI][Medline]
  21. Im, K, Belle, SH, Schulz, R, et al Prevalence and outcomes of caregiving after prolonged (>48 hours) mechanical ventilation in the ICU. Chest 2004;125,597-606[Abstract/Free Full Text]
  22. Knaus, W, Wagner, D, Draper, E, et al The APACHE III prognostic system: risk prediction of hospital mortality for critically ill hospitalized adults. Chest 1991;100,1619-1636[Medline]
  23. Ware, J, Kosinski, M, Dewey, J, et al How to score and interpret single-item health status measures: a manual for users of the SF-8 Health Survey 2001 QualityMetric. Lincoln, RI:
  24. Adams, CE, Biggerstaff, N Reduced resource utilization through standardized outcome-focused care plans. J Nurs Adm 1995;25,43-50[ISI][Medline]
  25. Madigan, EA, Fortinsky, RH Additional psychometric evaluation of the outcome and assessment information set (OASIS). Home Health Care Serv Q 2000;18,49-62[Medline]
  26. Boucher, L, Renvall, MJ, Jackson, JE Cognitiely impaired spouses as primary caregivers for demented elderly people. J Am Geriatr Soc 1996;44,828-831[ISI][Medline]
  27. Katzman, R, Brown, T, Fuld, P, et al Validation of a short orientation-memory-concentration test of cognitive impairment. Am J Psychiatry 1983;140,734-739[Abstract/Free Full Text]
  28. Katz, P, Showstack, J, Lake, J, et al Methods to estimate and analyze medical care resource use: an example from liver transplantation. Int J Technol Assess Health Care 1999;15,366-379[ISI][Medline]
  29. Elashoff, J nQuery advisor 2002 Statistical Solutions. Boston, MA:
  30. Weinberger, M, Oddone, E, Henderson, W Does increased access to primary care reduce hospital readmission? N Engl J Med 1996;334,1441-1447[Abstract/Free Full Text]
  31. Soeken, KL, Prescott, PA, Herron, DG, et al Predictors of hospital readmission. Eval Health Prof 1991;14,262-281[Abstract/Free Full Text]
  32. Librero, J, Piero, S, Ordinana, R Chronic comorbidity and outcomes of hospital care: length of stay, mortality, and readmission at 30 and 365 days. J Clin Epidemiol 1999;52,171-179[CrossRef][ISI][Medline]
  33. Westart, GP, Lagoe, RJ, Keskimaki, I, et al An international study of hospital readmissions and related utilization in Europe and the USA. Health Policy 2002;61,269-278[CrossRef][ISI][Medline]
  34. Formiga, F, Chivite, D, Manito, N, et al One year follow-up of heart failure patients after their first admission. QJM 2004;97,81-86[Abstract/Free Full Text]
  35. Camberg, LC, Smith, NE, Beaudet, M, et al Discharge destination and repeat hospitalizations. Med Care 1997;35,756-767[CrossRef][ISI][Medline]
  36. Ottenbacher, KJ, Smith, PM, Illig, SB, et al Characteristics of persons rehospitalized after stroke rehabilitation. Arch Phys Med Rehab 2001;82,1367-1374[CrossRef][ISI][Medline]
  37. Smith, DM, Giobbie-Hurder, A, Weinberger, M, et al Predicting non-elective hospital readmissions: a multi-site study. J Clin Epidemiol 2000;53,1113-1118[CrossRef][ISI][Medline]
  38. Naylor, MD, Brooten, D, Campbell, R, et al Comprehensive discharge planning and home follow-up of hospitalized elders. JAMA 1999;281,613-620[Abstract/Free Full Text]
  39. Goldberg, LR, Piette, JD, Walsh, MN, et al Randomized trial of a daily electronic home monitoring system in patients with advanced heart failure: the Weight Monitoring in Heart Failure (WHARF) trial. Am Heart J 2003;146,705-712[CrossRef][ISI][Medline]
  40. Tsuchihashi, M, Tsutsui, H, Kodama, K, et al Medical and socioenvironmental predictors of hospital readmission in patients with congestive heart failure. Am Heart J 2001;142,e7
  41. Babayan, ZV, McNamara, RL, Nagajothi, N, et al Predictors of cause specific hospital readmissions in patients with heart failure. Clin Cardiol 2003;26,411-418[ISI][Medline]
  42. Almagro, P, Calbo, E, de Echaguen, AO, et al Mortality after hospitalization for COPD. Chest 2002;121,1441-1444[Abstract/Free Full Text]
  43. Ottenbacher, KJ, Smith, PM, Illig, SB, et al Length of stay and hospital readmission for persons with disabilities. Am J Public Health 2000;90,1920-1922[Abstract/Free Full Text]
  44. Andersen, HE, Schultz-Larsen, K, Kreiner, S, et al Can readmission after stroke be prevented? Stroke 2000;31,1038-1045[Abstract/Free Full Text]
  45. Krumholz, HM, Amatruda, J, Smith, GL, et al Randomized trial of an education and support intervention to prevent readmission of patients with heart failure. J Am Coll Cardiol 2002;39,83-89[Abstract/Free Full Text]
  46. Akosah, KO, Schaper, AM, Havlik, P, et al Improving care for patients with chronic heart failure in the community. Chest 2002;122,906-912[Abstract/Free Full Text]
  47. Armstrong, EP, Sclar, DA Disease management: state of the art and future directions. J Clin Ther 1999;21,593-609[CrossRef]
  48. Welch, WP, Bergsten, C, Cutler, C, Bocchino, C, et al Disease management practices of health plans. Am J Manag Care 2002;8,353-361[ISI][Medline]
  49. Glynn, K, Patel, K Ensuring quality of highest-risk population care management in a teleworking environment. Case Manager 2004;15,61-64[Medline]
  50. Disease Management Association of America. Definition of disease management. Available at: http://www.dmaa.org/definition.html. Accessed June 6, 2004
  51. van Walraven, C, Rokosh, E What is necessary for high-quality discharge summaries? Am J Med Qual 1999;14,160-169[Abstract]
  52. van Walraven, C, Seth, R, Austin, PC, et al Effect of discharge summary availability during post-discharge visits on hospital readmission. J Gen Intern Med 2002;17,186-192[CrossRef][ISI][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Daly, B. J.
Right arrow Articles by Montenegro, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Daly, B. J.
Right arrow Articles by Montenegro, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS