18 Progress in Cardiovascular Nursing Winter 2008 www.lejacq.com ID: 6611
Heart failure (HF) is the leading cause of hospitalization for persons older than 65 years and often necessitates assistance from family caregivers.1,2 Approximately 40% of older adults are readmitted within
3 months postdischarge, resulting in significant health care costs.2,3 Indeed, more than one million hospitalizations attributed to HF cost the patient and
United States health care system over
$27 billion a year.1 Hospitalizations related to HF are complicated by the aging process, comorbidities, and psychosocial concerns that affect successful management.2 In a study of patient outcomes post– hospital discharge, Naylor and colleagues4 demonstrated that the use of advanced practice nurses (APNs) was effective in improving outcomes for patients with HF. The feasibility of adapting this approach is limited, however; costs for home visits by an APN are high, a shortage of APNs employed in home care exists, and funding for home care has decreased with the Medicare
Prospective Payment System.5
An alternative strategy for promoting self-management of HF is the use of electronic home monitoring
(EHM). EHM is a form of telemedicine in which medical/nursing management interventions are provided to individuals at a distance from the health care provider.6 The primary objective of this pilot study was to examine the effectiveness of postdischarge telemonitoring by an
APN on reducing subsequent hospital readmissions, emergency department
(ED) visits, and costs and increasing the time between discharge and readmission among older adults with HF.
Secondary objectives were to examine depressive symptomatology, quality of life, caregiver mastery, and social support for patients with HF.
Background
Older adults with HF face a high risk of early hospital readmission within 3 to 6 months of discharge.2,3
Upon hospital discharge, patients may encounter problems related to depressive symptoms, threatened quality of life, availability of informal and formal social support, and how they are monitored at home.2 Family caregiving is affected by how one copes with problems and resources available.
Researchers found that patients with cardiovascular disorders including
HF have a high prevalence of depressive symptoms or feelings of hopelessness, with rates reported from 24% to
>40%.7,8 Depressive symptoms were linked to increased hospital readmissions, number of comorbidities, impairment in activities of daily living, and the need for informal social support.7–9
HF has an enormous impact on the quality of life for patients because it is a chronic condition without a cure.
According to the National Academy on an Aging Society,10 persons with heart disease are less satisfied with their lives than the general population.
Heart failure (HF) is the leading cause of rehospitalization in older adults. The purpose of this pilot study was to examine whether telemonitoring by an advanced practice nurse reduced subsequent hospital readmissions, emergency department visits, costs, and risk of hospital readmission for patients with HF. One hundred two patient/caregiver dyads were randomized into 2 groups postdischarge; 84 dyads completed the study. Hospital readmissions, emergency department visits, costs, and days to readmission were abstracted from medical records. Participants were interviewed soon after discharge and 3 months later about effects of telemonitoring on depressive symptoms, quality of life, and caregiver mastery. There were no significant differences due to telemonitoring for any outcomes. Caregiver mastery, informal social support, and electronic home monitoring were not significant predictors for risk of hospital readmission. Further studies should address the interaction between the advanced practice nurse and follow-up intervention with telemonitoring of patients with HF to better target those who are most likely to benefit. Prog
Cardiovasc Nurs. 2008;23:18–26. ©2008 Le Jacq
From the College of Nursing, University of Akron, Akron, OH;1 the Department of Nursing, MetroHealth Medical Center, Cleveland, OH;2 and the Heart and
Vascular Center, Akron General Medical Center, Akron, OH3
Address for correspondence:
Karen A. Schwarz, PhD, RN, College of Nursing, University of Akron,
Akron, OH 44325-3701
E-mail: karen12@uakron.edu
Manuscript received June 13, 2007; revised August 10, 2007; accepted August 16, 2007
P i l o t S t u d y
Telemonitoring of Heart Failure Patients and Their
Caregivers: A Pilot Randomized Controlled Trial
Karen A. Schwarz, PhD, RN;1 Lorraine C. Mion, PhD, RN;2 Debra Hudock, MSN, RN;3 George Litman, MD3
Progress in Cardiovascular Nursing® (ISSN 0889-7204) is published Quarterly (March, June, Sept., Dec.) by Le Jacq, located at Three Enterprise Drive, Suite 401, Shelton, CT 06484. Le Jacq is an imprint of Blackwell Publishing, which was acquired by
John Wiley & Sons in February 2007. Blackwell’s programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Copyright ©2008 by Le Jacq. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in this publication are those of the authors and do not necessarily reflect those of the Editors or Publisher. For copies in excess of 25 or for commercial purposes, please contact Ben Harkinson at BHarkinson@bos.blackwellpublishing.com or 781-388-8511.
®
Winter 2008 Progress in Cardiovascular Nursing 19
How patients adapt to a therapeutic regimen is strongly influenced by how they live within their disability.11 In one of the few published studies that has examined quality of life in patients with HF that received post–hospital discharge case management, Pugh and colleagues12 reported that quality of life scores were more favorable among those who received intensive postdischarge collaboration with their providers than those who received usual care.
Furthermore, Goldberg and associates13 found that patients with HF who had access to in-home telemonitoring had a slightly greater improvement in quality of life than those who received standard care.
Caregiver mastery may be viewed as a personal resource that directly and indirectly affects the relationship between patient stressors and outcomes.
Caregiving mastery includes the expectation that one is capable of coping with problems.14 In a rare study about caregiving skills, Scott15 related that caregivers of patients with
HF described both positive and negative components of receiving technologic care. While caregivers found satisfaction from caregiving, they feared the responsibilities of learning to use new technology. This learning may be associated with caregiver mastery.
External resources in this study included informal social support provided by family and friends and formal social support from home health care and EHM by an APN. Although families provide 80% to 90% of informal social support,16 there is a dearth of studies about the relationship of informal social support and patient outcomes.
Schwarz and Elman’s17 findings suggested that informal social support significantly reduced the risk of hospital readmission in patients with HF.
Formal social support such as home care is a vital resource for patients with
HF because of the patient’s associated functional decline and need for education. Stewart and colleagues18 reported that those patients receiving a structured home visit from a cardiac nurse 7 to 14 days after discharge had fewer unplanned hospital readmissions.
Other researchers found that a comprehensive treatment program consisting of individualized home visits improved quality of life and decreased hospital readmissions and costs for patients with HF.19
An EHM system is typically monitored by an APN but does not replace direct home care visits. Rather, telemonitoring provides more frequent (eg, daily) surveillance of important clinical parameters that allows for more rapid intervention, which theoretically in turn would delay or prevent ED visits and hospital admissions.20 Some researchers reported that the home telecare intervention had significantly reduced hospital readmissions, ED visits, and costs postdischarge.20–22 In addition, Roglieri and colleagues’23 findings indicated that a comprehensive program of patient education and telemonitoring by phone significantly reduced hospital admissions and readmissions.
Using a daily EHM system,
Goldberg and coworkers13 found no significant differences in the overall time to death or first rehospitalization, however, and McManus24 found no significant differences in hospital readmissions that were attributed to telemonitoring. Although most of the previous studies were randomized control trials,13,20,21,24 they differed on patient age, technology used to monitor patients, means of data collection, and analyses of data.
Given the need to examine alternative methods to survey and monitor patients with HF in the home in a time of limited health care personnel, we examined the effectiveness of telemonitoring with an APN in a randomized clinical trial. The research hypotheses were as follows:
• Hypothesis 1: Hospital readmissions,
ED visits, and costs of care will be significantly lower for HF patients with EHM as compared with usual care.
• Hypothesis 2: Rates of depressive symptoms will be lower, but days to readmission and measures of quality of life and caregiver mastery will be significantly higher in the EHM group compared with usual care.
• Hypothesis 3: Caregiver mastery, informal social support, and EHM will significantly reduce the risk of hospital readmission for patients with HF.
Methods
Setting and Sample
This pilot study was conducted at a 537-bed tertiary teaching hospital in Northeastern Ohio. The study was reviewed and approved by the institutional review board at the participating hospital.
Potential participants for the study included patient/caregiver dyads who met the following criteria and routinely used the participating hospital.
The patients, aged 65 years or older, had a diagnosis of New York Heart
Association (NYHA) classification II,
III, or IV HF25 and were functionally impaired in at least 1 activity of daily living (ADL) or one instrumental activity of daily living (IADL), necessitating assistance of a family caregiver.
They received home care from the participating home care agency if it was ordered by their physician, had
Medicare eligibility and an operating telephone line, and were able to speak English. Classic symptoms of clinical HF are shortness of breath and fatigue, and abnormalities of systolic and diastolic dysfunction may coexist.
25 The principal investigator (PI) validated the diagnosis of systolic and/ or diastolic HF with chart review of the cardiologist’s impressions related to signs and symptoms of HF, ejection fraction and/or the echocardiography report after gaining oral consent from the patient before hospital discharge.
Exclusion criteria included planned discharge to a nursing home, inability to be interviewed because of physical illness, current use of a telemonitoring scale, inability to be contacted postdischarge, receiving regular infusions or dialysis, NYHA class I, independence in performing ADLs, no caregiver, use of hospice care, client of nonparticipating home health care
Progress in Cardiovascular Nursing® (ISSN 0889-7204) is published Quarterly (March, June, Sept., Dec.) by Le Jacq, located at Three Enterprise Drive, Suite 401, Shelton, CT 06484. Le Jacq is an imprint of Blackwell Publishing, which was acquired by
John Wiley & Sons in February 2007. Blackwell’s programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Copyright ©2008 by Le Jacq. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in this publication are those of the authors and do not necessarily reflect those of the Editors or Publisher. For copies in excess of 25 or for commercial purposes, please contact Ben Harkinson at BHarkinson@bos.blackwellpublishing.com or 781-388-8511.
®
20 Progress in Cardiovascular Nursing Winter 2008 agency, participation in another study, dementia, planned surgery, inability to speak English, planned hospitalization, and inability to stand on a scale.
Inclusion criteria for caregivers included being cognitively intact, having a familial relationship to the patient, and providing assistance with at least 1
ADL or 1 IADL. Overall, 562 patients were screened for eligibility (Figure).
Of these, 152 (27%) were eligible; 102
(67%) agreed to participate.
Sample size determination was based on the reported difference
(22%) in the proportion of patients readmitted in the control group vs the intervention group within 3 months of hospital discharge.19 The targeted enrollment was 84 patient/caregiver dyads using a power of 80% and a one-tailed test of significance with a set at .05. Of the 102 dyads enrolled,
84 (82%) completed the study; 40 in the usual care group and 44 in the intervention group. Attrition was equivalent between groups.
Procedure
Institutional review board approval was obtained, and cardiologists and internal medicine physicians gave written permission for their patients to be identified and enrolled. Potential participants were identified by the
HF care manager with the assistance of care managers in 4 hospital units.
While making daily rounds, the HF care manager informed potential participants about the study and gained oral permission for the PI to contact them before hospital discharge. Prior to discharge, the PI briefly explained the study to the patient and/or caregiver, provided a letter of explanation, and received oral consent for a chart review to verify whether they met study criteria.
Patients provided their phone numbers for contact upon discharge. Participants were randomized to usual post–hospital discharge care or to usual care with a telemonitoring scale by drawing from a preprepared, sealed envelope.
Participants were interviewed in their homes within 10 days of hospital discharge and 90 days later by trained registered nurses (RNs) who were not part of postdischarge care. To maintain inter-rater reliability, checks were conducted periodically throughout the study and maintained at >90% agreement. Written informed consent and Health Insurance Portability and
Accountability Act (HIPAA) authorization were provided by patients and caregivers at the first interview.
During data collection, patients received all standard treatments and services ordered by their primary physicians/ cardiologists. Participants randomized to the intervention group received the
Cardiocom EHM system (Cardiocom,
LLC, Chanhassen, MN) at the first interview, and the nurse removed the equipment 90 days later, at the second interview. The HF care manager trained the PI about use of the EHM system and ensured its availability.
The RN data collector was further trained by the PI about the equipment and taught the patient/caregiver dyad how to use the EHM system. The PI met with the HF care manager weekly to discuss technical issues with the equipment. On occasion, the PI called patients soon after placement of the scales to inquire whether they had any difficulty understanding instructions for its use.
EHM System
The Cardiocom EHM system was leased to the research team during the study period, and patients were not responsible for charges. The RN data
Hospitalized HF patients assessed for eligibility,
N=562 Excluded, n=460
Refused, n=50
Ineligible, n=410
Nursing home, n=117
Too ill, n=97
Telemonitoring use, n=52
Inability to contact, n=40
Dialysis, infusions, n=28
Independent ADL, n=24
No caregiver, n=18
Hospice care, n=10
Non-agency client, n=8
In other study, n=6
Dementia, n=3
Need for surgery, n=2
Non–English speaking, n=2
Planned hospitalization, n=2
Unable to stand on scale, n=1
Randomized,
n=102
Usual care, n=51 Usual care plus telemonitoring, n=51
Attrition
Death, n=7
Nursing home, n=3
Withdrawal, n=1
Attrition
Death, n=4
Nursing home, n=2
Withdrawal, n=1
Figure. Algorithm of study participants who were screened for eligibility.
Progress in Cardiovascular Nursing® (ISSN 0889-7204) is published Quarterly (March, June, Sept., Dec.) by Le Jacq, located at Three Enterprise Drive, Suite 401, Shelton, CT 06484. Le Jacq is an imprint of Blackwell Publishing, which was acquired by
John Wiley & Sons in February 2007. Blackwell’s programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Copyright ©2008 by Le Jacq. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in this publication are those of the authors and do not necessarily reflect those of the Editors or Publisher. For copies in excess of 25 or for commercial purposes, please contact Ben Harkinson at BHarkinson@bos.blackwellpublishing.com or 781-388-8511.
®
Winter 2008 Progress in Cardiovascular Nursing 21 collector placed a weight scale in the participants’ homes and connected via the telephone line to a computer system in the collaborating hospital.
The data-receiving computer was positioned in an office on the telemetry unit of the study hospital. The EHM system was programmed to measure weight on a daily basis. The display on the device asked the participants to answer “yes” or “no” to questions about shortness of breath, cough, fatigue, swelling, chest discomfort, urination, exercise, dizziness, medication use, and sodium intake. The computer stored each patient’s electronic health file and automatically displayed clinical variances when prescribed parameters exceeded predetermined ranges. Variances included failure to call daily, changes in symptoms, and weight outside prescribed parameters.
The Cardiocom Telescale (Cardiocom,
LLC, Chanhassen, MN) is accurate to ±0.1 lb and detects as little as 45 mL of fluid gain (D. Consentino, oral communication, February 8, 2002).
The HF care manager, an APN, was responsible for daily monitoring of parameters received electronically.
When participants had measurements outside of prescribed parameters, the monitoring nurse called the caregiver in the dyad to further assess the situation, provide education, and update the medication regimen. In addition, the
APN notified the primary physician or cardiologist about the patient’s status as needed.
Variables and Measures
Demographics such as age, sex, education, race, socioeconomic status, and perceived health of the patient and caregiver were measured by self-report at baseline. Hospital readmissions, defined as unplanned hospital readmissions for HF symptoms within 3 months post–hospital discharge, were collected by medical record review after 90 days postdischarge. Emergency department visits for HF were ascertained through medical record review after 90 days’ postdischarge. Days to readmission were used as a measure of risk for hospital readmission. Days to readmission, defined as the number of days between the date of initial hospital discharge and the first readmission to the hospital, was assessed through medical record review after 90 days’ postdischarge.
Physiologic health indicators, blood pressure, apical pulse, weight, and oxygen saturation were assessed by the
PI or research RN at baseline and 3 months later. Comorbidities and prescribed medications were abstracted from the medical record before hospital discharge and were confirmed at baseline. Use of home health care was documented with a computerized chart review after 90 days’ postdischarge.
Severity of HF was assessed subjectively by the PI or research RN using the
NYHA functional class25,26 at baseline and at 90 days’ postdischarge.
Functional status was measured as the ability to perform ADLs and
IADLs at baseline and 90 days’ postdischarge.
The ADL tool27 consists of 6 items (eating, dressing, bathing, transfers, incontinence, and toileting) and is scored from 0 (totally independent) to 2 (totally dependent). The
ADL Index has a Cronbach’s a of 0.82 for hospitalized older adults.28 Seven items from the IADL scale29 (eg, cooking, housekeeping) were scored similarly. The IADL scale has a high reproducibility coefficient of 0.96 and inter-rater reliability of 0.87.29 The
ADL and IADL sum score range is
0 to 26. Spector and Fleishman30 demonstrated the feasibility and validity of combining the 2 scales with a sample of functionally disabled older adults and reported that the correlation between the IADL scores and
ADL scores was 0.70. For this sample,
Cronbach’s a was 0.85 for the combined scale. Depressive symptomatology was measured using the Center for
Epidemiological Studies Depression
Scale (CES-D)31 at baseline and 90 days’ postdischarge. Participants rated
20 items on a 4-point Likert scale from 0 (“rarely”) to 3 (“most or all of the time”) with a possible range of 0 to 60. Higher scores indicate more depressive symptoms. The CES-D demonstrated excellent psychometric properties in community samples.31,32
In this sample, Cronbach’s a at time 1 and time 2 were 0.89.
Quality of life, defined as patient’s perceptions of the effects of HF on one’s life, was measured with 18 items from the Minnesota Living with HF questionnaire (MLWHF)11,33 at baseline and 90 days’ postdischarge. The
MLWHF measures individuals’ perceptions of the ways in which symptoms of HF have impacted their lives in the past month. Since the majority of patients were older and not employed and depression was assessed with the
CES-D, questions about working, sexual activities, and depression were eliminated from the original scale.
Eighteen items were rated on a 6-point
Likert scale from 0 (“no”) to 5 (“very much”), with a possible range of 0 to 90. A higher score indicated more symptomatic impact on one’s life. For this sample, Cronbach’s a coefficients at time 1 and time 2 were 0.91. Rector and Cohn33 reported that the MLWHF was a valid patient self-assessment of the therapeutic benefit of medical therapy.
Caregiver mastery, defined as a positive view of one’s ability to provide care, was measured with the mastery subscale from the Philadelphia
Geriatric Center Caregiving Appraisal
Scale (PGCCAS)34 at baseline and 90 days’ postdischarge. Six items assess the likelihood of caregiver uncertainty about how to provide care, reassurance that the patient is receiving proper care, feeling on whether they should be doing more for the patient, feeling that they are doing a good job of providing care, perceptions about capability of dealing with problems as they arise, and identifying the patient’s needs. Caregivers rated 6 items on a
5-point Likert scale from 1 (“never”) to 5 (“nearly always”), with a possible range of 6 to 30. Higher scores indicated greater mastery. Studies demonstrated moderate Cronbach’s a coefficients ranging from 0.61 to 0.73.34,35
For this sample, Cronbach’s a was
0.65 at time 1 and 0.70 at time 2.
Progress in Cardiovascular Nursing® (ISSN 0889-7204) is published Quarterly (March, June, Sept., Dec.) by Le Jacq, located at Three Enterprise Drive, Suite 401, Shelton, CT 06484. Le Jacq is an imprint of Blackwell Publishing, which was acquired by
John Wiley & Sons in February 2007. Blackwell’s programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Copyright ©2008 by Le Jacq. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in this publication are those of the authors and do not necessarily reflect those of the Editors or Publisher. For copies in excess of 25 or for commercial purposes, please contact Ben Harkinson at BHarkinson@bos.blackwellpublishing.com or 781-388-8511.
®
22 Progress in Cardiovascular Nursing Winter 2008
Informal social support, described as instrumental activities performed by families and friends, was measured with the tangible subscale from the Modified Inventory of Socially
Supportive Behaviors Scale (MISSB)36 at baseline. The tangible subscale reflects activities such as receiving a monetary loan. Caregivers rated 9 items on a 4-point Likert scale from
1 (“never”) to 4 (“very often”), with a possible sum score of 9 to 36.
Higher scores indicated more informal social support. Cronbach’s a coefficients ranged from 0.71 to 0.92 in studies of caregivers of patients who were recently hospitalized.17,32 For this sample, the Cronbach’s a coefficient was 0.90.
Cost of care was calculated for the
90-day period post–initial hospitalization.
Charges posthospitalization were calculated by tracking billing charges for rehospitalization, emergency department visits, and charges for usual home care from the provider of home health care. Costs of care for the EHM group included the former charges plus the additional monthly charge of renting the monitoring system.
Charges for usual home care were calculated by multiplying the standard charge data times the number of visits by the RN (at $155 per visit), home health aide (at $85 per visit), social worker (at $165 per visit), and physical therapist, occupational therapist, dietitian, or speech therapist (at
$140 per visit). Supply costs averaged
$38.50 per episode of care. Charges for
EHM were calculated for direct costs of placement of the Cardiocom unit
($165). Data for reimbursement for the telemonitoring specialist were not available and, therefore, not included.
In addition, out-of-pocket expenses for services posthospitalization were determined by calculating the number of physician office visits and instances of laboratory work and assigning a copay of $12.
Data Analyses
Descriptive and comparative analyses were performed using SPSS for windows, version 13 (SPSS, Inc,
Chicago, IL). Descriptive statistics, frequencies, and measures of central tendency and dispersion were used to describe the sample. Associations between variables were analyzed with
Pearson correlation coefficients for interval variables and the Spearman correlation coefficient for ordinal variables. Means were substituted for the relatively few areas of missing data.
The effectiveness of the intervention was examined by using an intentionto- treat analysis; a was set at ≥.05.
Outcomes were examined between the
2 groups using chi-squared likelihood ratio tests for categorical variables, t tests for approximately normally distributed variables, and Wilcoxon rank sum tests for skewed variables.
Subgroup analyses were conducted, comparing the intervention and usual care groups by risk status.
Survival analysis with Cox proportional hazard modeling37 was used to assess risk for hospital readmission by the number of days between discharge and first readmission. Cox proportional hazard modeling accommodates for the
Table I. Participant and Caregiver Characteristics at the Index Hospital Visit by Study Group (N=102)
Patient Intervention Group (n=51) Usual Care Group (n=51) P Value
Age, y 77.1±7.3 79.1±6.9 .17
Femalea 22 (43) 31 (61) .07
Whitea 41 (80) 42 (82) .56
Marrieda 36 (71) 28 (55) .25
High school graduate or highera 42 (82) 25 (49) .01
ADLb/IADLc 6.5±4.9 8.1±4.1 .08
NYHA class IIa 12 (24) 9 (18) .74
NYHA class IIIa 23 (45) 26 (51)
NYHA class IVa 16 (31) 16 (31)
Comorbidities 4.2±2.4 4.9±2.1 .14
Current medications 10.2±4.5 9.9±3.7 .79
Heart medications 5.6±1.9 5.4±2.0 .51
Cardiologista 46 (90) 41 (80) .16
Caregiver
Age, y 63.9±15.4 63.0±16.7 .76
Relationship .12
Spousea 32 (64) 22 (43)
Childa 13 (26) 16 (31)
Othera 5 (10) 13 (26)
Values are expressed as mean ± SD unless otherwise indicated. aValue is No. (%). bFunctionally impaired in
>1 activity. cFunctionally impaired in 1 activity. Abbreviations: ADL, activities of daily living; IADL, instrumental activities of daily living; NYHA, New York Heart Association.
Progress in Cardiovascular Nursing® (ISSN 0889-7204) is published Quarterly (March, June, Sept., Dec.) by Le Jacq, located at Three Enterprise Drive, Suite 401, Shelton, CT 06484. Le Jacq is an imprint of Blackwell Publishing, which was acquired by
John Wiley & Sons in February 2007. Blackwell’s programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Copyright ©2008 by Le Jacq. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in this publication are those of the authors and do not necessarily reflect those of the Editors or Publisher. For copies in excess of 25 or for commercial purposes, please contact Ben Harkinson at BHarkinson@bos.blackwellpublishing.com or 781-388-8511.
®
Winter 2008 Progress in Cardiovascular Nursing 23 censoring of information and accounts for the competing risk.17,37 The pool of potential predictors of risk for hospital readmission specific to the dyad included caregiver mastery, informal social support, and EHM. The multivariable model was derived using multiple model building techniques: backward elimination with a=.05 stay criteria, stepwise with a=.25 enter criteria, and a=.05 stay criteria to identify independent predictors of days to readmission.
Results
Participant Profile
One hundred two patients were originally enrolled in the study; 51 in the EHM group and 51 in the usual care group. The mean age of the entire sample was 78.1 years with a range of
65 to 94 years. Fifty-two percent (n=53) were women. Participant characteristics at the time of the baseline visit in the hospital are shown in Table I.
Education level was significantly higher for patients in the intervention group
(c2=18.5; P=.01). At baseline and 3 months later, there were no significant differences between groups for prescribed use of angiotensin-converting enzyme inhibitors, b-blockers, digoxin, or diuretics.
Twenty percent of the participants
(n=20) had implanted defibrillators.
Significant differences existed for use of defibrillators between the intervention
(n=14) and usual care (n=6) groups at baseline (c2=3.98; P=.05).
At 90 days’ postdischarge, differences for defibrillators were similar between the intervention (n=13) and usual care (n=6) groups (c2=2.53; P=.11).
Twenty percent of patients without defibrillators did not finish the study.
The number of hospital readmissions was similar between those with and without defibrillators (t=–1.3; P=.19).
Use of home care, total number of home care services, and informal social support were similar between intervention and usual care groups (P=.32,
.66, and .74, respectively).
The total sample reported having multiple comorbidities: hypertension
(51%), diabetes (50%), atrial fibrillation
(30%), myocardial infarction
(29%), stroke (13%), bypass surgery
(28%), chronic obstructive pulmonary disease (29%), and cancer (6%), and groups had similar comorbidities
(c2=12.13; P=.28). Systolic blood pressure at baseline ranged from 84 to 180 mm Hg with a mean of
127.56±19.21 mm Hg, and diastolic blood pressure ranged from 50 to 100 mm Hg with a mean of 69.3±9.74 mm Hg. Heart rate ranged from 40 to 110 beats per minute (bpm) with a mean of 71.39±11.4 bpm. Oxygen saturation ranged from 86% to 98% with a mean of 94.9±2.36%. Weight ranged from 85.6 \ to 372.5 lb with a mean of 179.14±49.53 lb. Caregivers reported providing assistance from 1 month to 27 years with a mean of
4.5±5.5 years.
Patient/caregiver dyads that completed the study (n=84) did not differ from those who did not (n=18) on demographic characteristics, physiologic health indicators, severity of illness, depressive symptomatology, quality of life, informal social support, caregiver mastery, or cognition.
Patients not completing the study were more dependent in ADLs and IADLs than those who completed it (10.8±4.1 vs 6.6±4.3; t=–3.67; P