Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Heart Lung. Author manuscript; available in PMC 2011 November 1.
Published in final edited form as:
PMCID: PMC3033014

Technology enhanced practice for patients with chronic cardiac disease Home Implementation and Evaluation

Patricia Flatley Brennan, RN, PhD, FAAN,1 Gail R. Casper, RN, PhD,1 Laura J. Burke, RN, PhD, FAAN,2 Kathy A. Johnson, RN, MS,1 Roger Brown, PhD,1 Rupa S. Valdez, MS,1 Marge Sebern, RN, PhD,3 Oscar A. Perez, MS,1 and Billie Sturgeon, RN, BSN1



This 3-year field experiment engaged 60 nurses and 282 patients in the design and evaluation of an innovative home care nursing model, technology enhanced practice (TEP).


Nurses in the TEP conditions augmented usual care with a web-based resource (HeartCareII) that provided patients with self-management information, self-monitoring tools, and messaging services.


Patients exposed to TEP demonstrated better quality of life and self-management of chronic heart disease during the first four weeks and were no more likely than patients in usual care to make unplanned visits to a clinician or the hospital. Both groups demonstrated the same long term symptom management and health status achievements.


This project provides new evidence that it is possible to purposefully create patient-tailored web resources within a hospital portal; that it is hard for nurses to modify their practice routines even with a highly-tailored web resource, and that the benefits of this intervention are more discernable in the early post-discharge stages of care.


Chronic cardiac diseases, including congestive heart failure and coronary artery disease, represent a set of complex health problems that feature a wide range of debilitating symptoms and a constant threat of morbidity. Care of this patient population has been changing dramatically over the past 20 years, characterized by shorter lengths of stay,1 a greater attention to home care,2,3 and the expanded use of technology.4,5 Effective management of chronic diseases requires intensive inpatient and community-based interventions, a high level of participation by patients and clinicians and the ability to rapidly respond to an ever-changing portfolio of symptoms. Home-based interventions, both human and technological, particularly those that afford early recognition and management of clinical problems, have been shown to be effective in reducing readmission rate.5 However, most intervention successes rest on evidence studies involving people with mild disease6 and have rarely demonstrated effects on the aspects that portent long-term gains, such as self-management. These changes are likely only when technological interventions are systematically designed to support, complement, and extend the professional nursing interventions. Creating such interventions requires user centered design strategies, workflow assessment and modifications, and high levels of engagement of patients. Here we report on a three-year field experiment assessing the impact of a technology enhanced practice (TEP) model of care on outcomes for patients with chronic cardiac disease. TEP represents an amalgam of professional nursing practice and computer systems purposefully built to support selected components of patient care. Professional nurses, in consultation with their patients, select which functions of the technology complement the aspects of care they are trying to deliver. In this project, we focused on home care nursing, and provided professional nurses and their patients with a web resource that included education, self-monitoring tools and a communication function. We have previously reported on the design and development processes of the technology and TEP,7,8 TEP training for the home care agency;9 facilitators and barriers to nurses’ implementing TEP;10 and patterns of participants’ use of the technology.11 This paper serves as a report of a field study evaluating the impact of TEP on select outcomes of home care patients with chronic heart disease. Our aim was to answer the question: “Does TEP lead to improvement or stabilization of patient outcomes including self-management of chronic heart disease, clinical status, quality of life, unplanned use of services and satisfaction with nursing care?”


Home care nurse management of patients with chronic cardiac disease has focused on three core elements: 1) education of patient and family (risk factors, medications and compliance, diet), 2) symptom monitoring (first by the nurse and then by the patient), and 3) close communication with health care providers. Advances in technology and increased use of the Internet by lay people make technology-based interventional research aimed at increasing patient self-management skills, improving quality of life and avoiding emergent hospitalizations possible.

Research in this area has taken advantage of available technology to support cardiac patients by addressing one or a combination of the described elements. Trans-telephonic monitoring,12 two-way telemedicine audiovisual system,13,14 and special-purpose monitoring devices like the HealthBuddy15 addressed symptom monitoring. Clinically meaningful and statistically significant improvement in blood pressure levels resulted from a one-year exposure to usual care that was enhanced by telephonic monitoring; however, the technological intervention stood separate from the clinical practice.16,17 Home monitoring also led to improved energy expenditures among older adults with high disease burden following cardiac surgery. The Med-eMonitor which prompted patients with congestive heart failure about diet, activity, medication taking, and requested answers for questions about symptoms, blood pressure and weight18 addressed both monitoring and education.

The first HeartCare study pulled together these three threads. We provided Internet-based information and support for patients experiencing the full recovery trajectory following coronary artery bypass graft surgery.19 Similarly, Westlake and colleagues20 demonstrated the effects of a Web-based education, including email capability to a clinical nurse specialist and other study participants, links to Web video content. This study, HeartCare II, builds on the successes of HeartCare I and other pioneers in this field, this time weaving the threads of education, symptom monitoring and communication to intuitive interactive web-based tools designed to help nurses teach patients to learn to manage their health. The result was a new model of nursing care: TEP.

Using technology to engage patients and nurse professionals in chronic cardiac disease management is challenging. For patients it requires developing a joint understanding of what self-management actually means through the trajectory of the disease process, not simply when the nurse is present.21 For nurses, the introduction and integration of technology requires a different way of thinking about the delivery of nursing care. Many home care nurses hold the belief that it is the “relational aspects of the nurse-patient relationship that hold the greatest significance for both nurses and elderly people…the therapeutic value of visits reaffirm their core values”22 ( p.86). Introducing information technology into this interpersonally rich dyadic relationship proved challenging.

We report on the effects of this novel practice model, TEP, on patient self-management, quality of life, health status, satisfaction with nursing care, and unplanned service use. The intention of this project was not to prescribe how the technology should be used, but rather to give the home care nurse a tool box that could be implemented specific to the patient and context, without disrupting established workflow or the relational aspects of care. As a result, there was no uniform exposure or dose, but rather a wide range of manifestations of use dependent on the context and on mental models of both patient and nurse.


Setting & Sample

This field experiment took place in the homes of patients recruited from a single home care agency of a large, integrated health care delivery system. This home care agency is geographically distributed across rural, suburban and urban areas in a Midwestern state. The agency is divided into offices, each staffed by professional nurses and home health aides, and serves about 25,000 patients annually. Patients referred for home care services are assigned to offices by the geographic location of their primary residence.

At the start of the study, six agency offices were matched into three pairs representing rural, suburban, and urban areas. Offices within each pair were randomly assigned to one of the two study conditions. Ten months into the 30-month study, the two urban offices merged for administrative reasons; the combined office was then assigned to the experimental condition. The merging of the two offices did not affect the demographic mix of participants.

All patients who were admitted to the home care agency for receipt of nursing care and deemed eligible for the study were assigned to the study condition to which that office had been assigned. Eligibility criteria included being diagnosed with an ICD-9 coded medical diagnosis indicating the presence of primary or secondary chronic cardiac disease, being clinically stable, the ability to read and write in English, having a working analog phone line, and living within a 100-mile radius of the central office for the home care agency. Exclusion criteria included attributes that impaired individuals’ ability to actively use information technology, including mental or sensory incapacity, and/or requiring in-home continuous professional care. There were 4,033 patients screened, 2,303 contacted, and 282 patients recruited into the study. The consent rate was comparable across all clinical sites. Figure 1, a modified CONSORT flow diagram, illustrates the recruitment, accrual and loss of participants throughout the study.

Figure 1
Flow Diagram of Participant Enrollment: Modified Version of CONSORT


Independent Variable

The independent variable is the nursing practice model. Two types of nursing practice models were compared in this study: usual care and technology enhanced practice (TEP). The models were similar in terms of scope, duration of contact, and expected outcomes; the models differed on the extent to which web technology supported the nurse and patient in the nursing care process.

Usual Care

Care is based on the nurse’s assessment of the patient’s needs, physician’s orders, and clinical guidelines. The patient’s medical diagnosis, acuity, physician’s orders, insurance coverage, and nursing care needs determine the number of home visits to be made, which typically ranges from one to nine. Nurses interact with patients through home visits and telephone calls for the specified number of visits typically spanning several weeks. Patient and family education are integral components of care. Nurses use an institutional set of care management initiative education materials to assist patients to adhere to medications, modify their life style, understand the disease, and recognize early manifestations of disease progression or complications. In addition to education, the practices include surveillance of cardiopulmonary symptoms, prompt response to a change in status and coordination of services. All tools used are paper-based.


The intervention, TEP, is a nursing practice model in which nurses selectively and deliberately employ information technologies to meet individual patient care goals. Nurses plan and provide nursing care supplementing usual practice with selected technology tools located on the HeartCare website, to provide care tailored to patients’ specific needs, abilities and illness trajectory. These tools resulted from work in the design phase of the study, through use of human factors techniques to solicit input from and analyze work of home care nurses. The specific technology tools developed either replaced paper forms or filled an identified gap in assessment or educational needs. The suite of technology tools, housed on the HeartCare website within the clinical partner’s clinical information system, could be accessed by both nurses and patients through an Internet connection. The interactive technology tools on the HeartCare website addressed education (patient education resources on a variety of topics: disease, medication and drug interaction information, plus food trackers), symptom monitoring (symptom checklist, weight tracker, blood pressure tracker, heart rate tracker), and communication (my goals, my journal, email, bulletin boards).

Outcome Measures and Instruments

Clinical Status

The SF-12 is a 12-item measure employed as an indicator of the patient’s physical and mental functional clinical status.23 A weighted additive model is used to calculate both a physical and mental health status, each scored on a scale from 0 to 100. A higher score is indicative of a better health state. A shortened version of the SF-36, it is based on the assumption that only one or two questionnaire items are necessary to estimate the average score on a concept for a population or to determine the overall prevalence of scores in either the best or the most impaired categories of eight of the most frequently measured health domains.24


The Self Care Heart Failure Index (SCHFI) is a 17-item measure of self-care in heart failure patients.25 Self-management was measured with a subscale from the Self-care in Heart Failure Index (SCHFI). SCHFI measures behaviors that maintain physiological stability, response to symptoms when they occur, and self-confidence. The 5-item self-maintenance subscale of this instrument addressing physiological stability is in close accordance with our viewpoint on the self management skills required of this patient population (alpha coefficient of .55; CI .43–.65). Participants are asked to respond on a 4-point Likert-type scale (never or rarely, sometimes, frequently, or always) to how often they perform the following activities that are important in chronic cardiac conditions: weigh daily, take part in physical activity, eat a low salt diet, check ankles for swelling, keep weight down. Summary scores are standardized on a scale from 0 to 100, with a higher score indicative of better self-management. Riegel and colleagues26 suggested that an 8-point increase in the SCHFI subscales is clinically meaningful.

Quality of Life

Quality of life was measured using The Multidimensional Index for Life Quality Questionnaire for Cardiovascular Disease (MILQ), a 35-item scale that covers 9 domains identified as critical for patients with coronary disease.27 Items are scored on a 7-point Likert scale. The score is calculated by doubling the mental health satisfaction subscore and adding it to the physical health satisfaction subscore. The range of this composite score is 12 to 84, with a higher score indicative of greater satisfaction with quality of life.27

Satisfaction with Nursing Care

This organization-specific survey is routinely administered by the home care-nursing agency for quality assurance purposes. Patients are asked to rate their satisfaction on 15 items related to nursing care on a phrase-anchored scale (range 1 to 5), e.g., nurse is knowledgeable about care, nurse explains medications, nurse gives information to manage care between visits. These scores are combined using an additive model and the summary score is then standardized on a scale of 0 to 100, with a higher score indicative of greater satisfaction with nursing care.

Unplanned Service Use

Unplanned service use included unexpected clinic or physician office appointments, emergency department visits, and unplanned hospitalizations. Use of any of these services was recorded dichotomously (yes, no) at each data collection point. Data were obtained through self-report and verified for a subset of the respondents through chart review.

Explanatory Variables

Demographics and descriptors

The study coordinator collected standard demographic and descriptive information about participants at baseline, including gender, age, race, ethnicity, educational level, experience with computers, and number of chronic illnesses (per patient self-report). The Specific Activities Scale (SAS), a 5-item measure of functional status under non-acute conditions,28 was employed as an indicator of disease severity. This scale categorizes the severity of the patient’s symptoms of heart failure from Class I (without limitation of physical activity) to Class IV (inability to carry on any physical activity without discomfort). SAS scores are described as more valid and reliable than scores estimated using the familiar New York Heart Association (NYHA) classification.28


Human subject approval was obtained for the study. A research nurse screened patients admitted to the home care agency on a daily basis. The study coordinator screened potential participants for eligibility using the agency’s electronic admission records. Potential patients were telephoned to verify eligibility, briefly explain the study and gain consent to meet with the patient in his/her home. At the home visit, the study was fully explained, consent was obtained and baseline data collected. Participants remained in the study for 24 weeks.

TEP Intervention

Training methods for nurses emphasized active participation and learning. Screen-capture videos of website tools, training outlines, tips on teaching elders, practice teaching scenarios, a pocket guide to the website tools, a guide mapping website tools to the agency’s home visit guidelines, and a CD with narrated webcam demonstrations of the website tools were employed. To assist nursing staff in tailoring TEP to their own practice style and patient needs, several support strategies were employed. This support included providing home care nurses with “first visit support” by a research team member in the patient’s home to reinforce TEP training, one-on-one and small group re-training approximately 12 months into the experiment, monthly staff meeting attendance by research liaison nurses to provide project updates and answer questions, and ongoing trouble-shooting by the research nurse and agency help desk with home care nurses to manage daily implementation issues. Bi-monthly newsletters were written for the experimental and comparison offices and provided project highlights, reminders, and motivational messages.

Unlike other studies of home care interventions where nurses follow a set protocol with every patient, this TEP allowed, even required, nurses to create the right mix of in-person nursing care and technology-delivered interventions. During home visits, nurses chose technology tools that addressed patient needs and goals and used the tools interactively with the patient, taught patients to use them independently between visits, and selectively monitored use between visits.

Patient participants in the TEP intervention group received either a web-computer or the modifications to their own computers that enabled them to access the HeartCare II resources. Patients received a written manual and pocket guide describing website access and use. Both nurses and patients could access the suite of technology tools, housed on the HeartCare section within the agency’s public portal, through a standard web browser and Internet connection.

Patients were encouraged to use technology tools on their own between nurse visits and after termination of home care nursing services for education and self-monitoring. Based on the available tracing data of logins, participants accessed the HeartCare Website from 0 to 314 times over the course of 24 weeks (mean 19, SD 40.6). Approximately 50% of participants continued to login to the HeartCare II website for four weeks; this declined to 33% at eight weeks while 15% continued to login for the 24-week duration of the study.11

Data Collection and Management

Exposure to the experimental conditions (TEP, Usual care) was treated as a dichotomous variable (using usual care as a reference). Data for explanatory (baseline only) and dependent variables were collected at six points in time: Baseline (in-home interview) and weeks 1, 4, 8, 12, and 24. A trained research nurse collected baseline data on a survey during in-home interviews. Data on surveys at subsequent time points were collected by one of two methods based on each participant’s preference: either by telephone interview by a trained research assistant or by mailed survey.

Data cleaning and Reduction

All data were cleaned, triple-entered, and evaluated for integrity and missingness. Little’s Missingness Completely at Random (MCAR) test with Bonferroni correction for multiple comparisons29 was used to check missingness. The MCAR test demonstrated that only two out of a possible 30 opportunities (5 scales over 6 data collection points) did not meet the assumption of MCAR. The two opportunities that failed MCAR were supported for the assumption of Missing At Random (MAR); thus we are confident that the missingness lacked any systematic pattern. We undertook an analysis strategy for treatment effects over time using a random-effects regression model for continuous measures and included the GEE for the binary unplanned service utilization.30

Additionally, data from some patients were missed at specific measurement time periods, with the result that some subjects provided data at some, but not all, data collection points. Laird31 has indicated that the random-effects model for longitudinal data provides valid inferences in the presence of ignorable non-response or, in our case, random dropouts. To assess the issue of ignorability of our dropout data, we used pattern mixture modeling29,32 and determined that no additional adjustments were necessary.


A total of 282 patients were enrolled over the 27-month recruitment period (Figure 1). The sample was composed of 84% Caucasian, 8% African Americans; the remaining 8% self-identified as Asian, Native American, Pacific Islander and multiple races. Most participants were non-Hispanic (95%), male (61%) and ranged in age from 28–93 years with a mean age of 64 years (SD = 12.7). Most were married (60%), and 80% reported living with someone. Over half of the study participants had at least a high school education, with 42% having 13–16 years of education and 19% having 17 or more years. Participants were randomized by home care office into Usual Care or TEP, and remained in the study for a 24-week period. The randomization process resulted in comparable participant groups in terms of major demographic and clinical status variables. See Table 2.

Table 2
Baseline profile of enrolled participants by treatment group

We employed a random-effects modeling strategy to evaluate the impact of the TEP intervention on the outcomes of self-management of heart failure, mental and physical health status, quality of life, unanticipated service use, and satisfaction with nursing care. We created models that accounted for scores on outcome measures at baseline and the key covariates of gender, age, and education. Parameters presented in Table 3 demonstrate that there were no main effects of TEP on the major outcome variables.

Table 3
General Linear Mixed Model results of study outcomes over time.

Figure 2 provides a graphical depiction of the major variables at the five post-intervention data collection points: week 1, 4, 8, 12 and 24 (note that the time intervals are not even). Figure 2 displays the patterns over time of adjusted means; the untransformed means compared at each point in time are displayed in Table 4. The TEP group is represented by the triangles and the Usual Care condition by circles. It is interesting to note that, similar to our earlier work with patients following acute cardiac events, the two groups reach similar endpoints but appear to have noticeable differences in the early post-discharge stage (time periods 2 and 3). Despite the appearance on the graph, none of the differences between TEP and Usual care are significant at any time point for the outcome variable of satisfaction with nursing care. Although the magnitude of the difference appears large, the maximum difference is only four points, which is not clinically meaningful or statistically significant. The range of scores for both groups for the entire study is less than 3 points (on a 100 point scale), indicating that satisfaction with nursing care was very stable over time for both groups. Additionally, although the difference in unplanned service use appears large, it is not significant. We now present the raw means and standard deviation of the key variables across time to better illustrate patterns of response.

Figure 2
Outcomes over time using adjusted means
Table 4
Untransformed mean and standard deviations of major outcomes across time

As a follow-up analysis we examined the fixed effect interaction of treatment by time using post-hoc multiple comparison tests (reporting both raw probabilities and Sidak adjustment). Comparing the treatment levels at each time point could aid in understanding the nature of the effect. It is well known that p-value adjustments need to be made when multiple tests are performed.33 Adjustments are usually made to preserve the family-wise error rate (FWER) of the group of tests. FWER is the probability of incorrectly rejecting at least one of the pair-wise tests. Table 5 depicts the mean TEP/usual care difference assessed using paired contrasts adjusted for the covariates.

Table 5
Mean TEP usual care difference paired-contrast adjusted for covariates.

These analyses reveal that there is a significant difference between TEP and Usual Care in terms of both physical health (weeks one and 4) and mental health (weeks one and 8). These results hold with the Sidak adjustment. The fixed time point exploration revealed no effect of TEP on quality of life or self-management at any of the time points, and no systematic effect of the intervention on service use or satisfaction with nursing care.


TEP is a model of nursing care that augmented the practice strategies available to the home care nurse by providing education, symptom monitoring and communication web-based tools that could be tailored to an individual patient’s needs and care plan. This model of care afforded improved outcomes and better self-management of chronic cardiac conditions for persons receiving home care nursing. These outcomes were most discernable in the early post-hospital period, with the strongest influence on the physical and mental function constructs. Our results were consistent with the impact of other technological interventions in that they 1) affected psychological and physical function rather than behaviors or knowledge3437 and 2) improved outcomes particularly in the early post-discharge period. Unlike Artinian’s work16 where participants had ongoing exposure to the intervention over an entire year, our participants had exposure to the full TEP intervention only during the first half of the observational period. We hypothesize that it may be unreasonable to expect distal effects to persist or arise simply due to the ability to access the HeartCare web server; the engagement of the home care nurse is critical to the TEP model of care.

The modest impact of TEP may be explained by the poor health status of our participants. Our participants were sicker than those in most other intervention studies,5,6,12,14 with 79.5% scoring at a level of 3 or 4 on the SAS. For these patients, TEP may have provided a greater sense of security and support, rather than an overall improvement in self-management. This finding in itself represents a significant challenge to contemporary thinking in home care nursing and in studies of consumer health technology38 where the intervention is designed to shift the responsibility for care more to the patient and off of the care providing system.

An interesting and relatively unexplored area is the interaction between technology resources and the cognitive burden of cardiac disease. Recent work by Jurgens and colleagues demonstrated that cardiac patients have difficulty understanding and interpreting their symptoms39. As we noted strong positive effects of the intervention on the SF-12 mental variable occurring concurrent to the home care intervention exposure, it may be that TEP supports the cognitive needs of the person by providing just-in-time coaching, while the home care nurse addresses remedial physical demands.

The intention of this project was not to prescribe how the technology should be used, but rather to give the home care nurse a toolbox that could be implemented specific to the patient needs and context, without disrupting established workflow. Thus, the degree of integration of the technology into the workflow was at the discretion of the nurse, and was influenced by the nurses’ perceptions of the patient acuity, the time available to incorporate it, and by the nurses’ own comfort with technology. Indeed, evidence from focus group explorations with the nursing staff suggested that nurses’ discretionary uses of the technology intervention in TEP was grounded in a large number of factors, including the nurse’s comfort with the technology, his or her appraisal of the patient’s skill and interest level, and general practice logistics.10 Because the trial was conducted in a real clinical setting, the intervention was affected by organizational changes and modifications in the basic technology. For example, the appearance and navigation of web resources were affected as the home care agency updated their clinical information system. Although the presence and functionality of tools and educational resources remained constant, the layout and navigation of webpages varied during the course of the study and may have influenced nurses’ engagement in integrating the technology tools into their care practices with the patients.

The intervention did not reduce unplanned service use. Our participants were similar to those of Woodend et al5 who found that 56% of the 121 patients in their study who had a primary diagnosis of heart failure had at least one ED visit during the first year. Unlike the findings from other studies of consumer technologies,15,40 participants in the TEP group showed a slight increase in health service use. However, the Sacks study addressed a generally healthy population, whereas our project is one of the few that addressed a seriously ill population.40 The increase in service use may in fact have resulted from patients’ greater recognition and willingness to act on symptoms.


TEP offers a promising intervention to support post-hospital care of persons with chronic cardiac disease. While not supplanting existing nursing services, TEP may better attend to the clinical and psychological needs of patients, augmenting education, symptom monitoring and communication opportunities. It is not likely that TEP will accelerate self-management, and determination of the appropriateness of this as a goal for care should be reexamined.

Table 1
Comparison of Usual Care and TEP


The authors gratefully acknowledge the support of NIH LM6249, Custom Computerized Home Care for Cardiac Disease, Patricia Flatley Brennan, RN, PhD

Analysis Appendix


Our analysis strategy for treatment effect used a random-effects regression model for longitudinal data. The model in terms of ni x 1 vector of responses across time, yi, for subject i, may be defined as:



  • yi = the ni x 1 vector of outcome measures for subject i,
  • Xi = a known ni x p design matrix,
  • β = a p x 1 vector of unknown population parameters,
  • Zi = a known ni x r design matrix,
  • vi = a r x 1 vector of unknown subject effects distributed N(0, Σv), and
  • εi = a ni x 1 vector of random residuals distributed independently as N(0, Σεi).

Our subjects were measured across five time periods of 1, 4, 8, 12 and 24 weeks. Since not all our outcome measures were continuous measures, we adapted the model to handle our binary outcome measure of service utilization using generalized estimating equations (GEE) for longitudinal data.1

In our study some patients were missed at specific measurement time periods, with the result that some subjects provide data at some, but not all, study time periods. Laird has indicated that the random-effects model for longitudinal data provides valid inferences in the presence of ignorable nonresponse, or in our case random dropouts.2 To assess issue of ignorability of our dropout data, we used pattern mixture modeling.3,4 In our analysis the pattern was defined as completers vs non-completers (partial respondents). The treatment effect across time varied by completion status, and effects were adjusted as suggested by Hogan and Laird.5. The adjustment is based on the averaged estimates for the effect parameter (b), which is equal to;


where πc represent the population weights for completers, with (1 – πc) for the dropouts. To obtain adjusted standard errors, the delta method as described by Hogan and Laird5 was used;






Note, under null model for completion (i.e., = binomial (πc))


A follow-up analysis examined the fixed effect interaction of treatment by time using post-hoc multiple comparison tests, as comparing the treatment levels at each time point could aid in understanding the nature of the effect. It is well-known that p-value adjustments need to be made when multiple tests are performed.6 Adjustments were made to preserve the family-wise error rate (FWER) of the group of tests. FWER is the probability of incorrectly rejecting at least one of the pair- wise tests.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. Splaine Wiggins M. The partnership care delivery model: An examination of the core concept and the need for a new model of care. J Nurs Manag. 2008;16:629–38. [PubMed]
2. Naylor MD. A decade of transitional care research with vulnerable elders. J Cardiovasc Nurs. 2000;14:1–14. [PubMed]
3. Ni H, Nauman DJ, Hershberger RE. Analysis of trends in hospitalizations for heart failure. J Card Fail. 1999;5:79–84. [PubMed]
4. Edwardson SR. Patient education in heart failure. Heart Lung. 2007;36:244–5. [PubMed]
5. Woodend AK, Sherrard H, Fraser M, Stuewe L, Cheung T, Struthers C. Telehome monitoring in patients with cardiac disease who are at high risk of readmission. Heart Lung. 2008;37:36–45. [PubMed]
6. DeBusk RF, Miller NH, Parker KM, Bandura A, Kraemer HC, Cher DJ, et al. Care management for low-risk patients with heart failure: A randomized, controlled trial. Ann Intern Med. 2004;141:606–13. [PubMed]
7. Or CKL, Casper GR, Karsh B, Brennan PF, Burke LJ, Carayon P, et al. Work system analysis of home nursing care and implications for patient safety. Proc Human Factors Ergonomics Soc. 2005:1052–1056.
8. Casper GR, Karsh B, Or CKL, Carayon P, Grenier AS, Brennan PF. 2005. Designing a technology enhanced practice for home nursing care of patients with congestive heart failure. In: Friedman C, Ash J, Tarczy-Hornoch P, editors. Biomedical and Health Informatics: From Foundations to Applications to Policy; AMIA Annu Symp Proc; 2005. pp. 116–120.
9. Kossman SP, Casper GR, Severtson DJ, Grenier AS, Or CKL, Carayon P, Brennan PF. Designing study nurses’ training to enhance research integrity: a macroergonomic approach. AMIA Annu Symp Proc; 2006. pp. 439–443. [PMC free article] [PubMed]
10. Johnson KA, Valdez RS, Casper GR, Kossman SP, Carayon P, Or CKL, Burke LJ, Brennan PF. Experiences of Technology Integration in Home Care Nursing. AMIA Annu Symp Proc; 2008. pp. 389–393. [PMC free article] [PubMed]
11. Casper GR, Brennan PF, Burke LJ, Nicolalde D. HeartcareII: Patients’ use of a home care web resource. Stud Health Technol Inform. 2009;146:139–43. [PubMed]
12. Benatar D, Bondmass M, Ghitelman J, Avitall B. Outcomes of chronic heart failure. Arch of Int Med. 2003;163:347–52. [PubMed]
13. Jenkins RL, McSweeney M. Assessing elderly patients with congestive heart failure via in-home interactive telecommunication. J Gerontol Nurs. 2001;27:21–7. [PubMed]
14. Jerant AF, Azari R, Nesbitt TS. Reducing the cost of frequent hospital admissions for congestive heart failure: A randomized trial of a home telecare intervention. Med Care. 2001;39:1234–1245. [PubMed]
15. Barnason S, Zimmerman L, Nieveen J, Hertzog M. Impact of a telehealth intervention to augment home health care on functional and recovery outcomes of elderly patients undergoing coronary artery bypass grafting. Heart Lung. 2006;35:225–33. [PubMed]
16. Artinian NT, Flack JM, Nordstrom CK, Hockman EM, Washington OG, Jen KL, Fathy M. 2007. Effects of nurse-managed telemonitoring on blood pressure at 12-month follow-up among urban African Americans. Nurs Res. 2007 Sep-Oct;56(5):312–22. [PubMed]
17. Barnason S, Zimmerman L, Schulz P, Tu C. Influence of an early recovery telehealth intervention on physical activity and functioning after coronary artery bypass surgery among older adults with high disease burden. Heart Lung. 2009;38(60):459–68. [PMC free article] [PubMed]
18. Artinian NT, Harden JK, Kronenberg MW, Vander Wal JS, Daher E, Stephens Q, et al. Pilot study of a web-based compliance monitoring device for patients with congestive heart failure. Heart Lung. 2003;32:226–233. [PubMed]
19. Brennan PF, Moore SM, Bjornsdottir G, Jones J, Visovsky C, Rogers M. Heartcare: An internet-based information and support system for patient home recovery after coronary artery bypass graft (CABG) surgery. J Adv Nurs. 2001;35:699–708. [PubMed]
20. Westlake C, Evangelista LS, Strömberg A, Ter-Galstanyan A, Vazirani S, Dracup K. Evaluation of a web-based education and counseling pilot program for older heart failure patients. Progr Cardiovasc Nurs. 2007;22:20–6. [PubMed]
21. Horowitz CR, Rein SB, Leventhal H. A story of maladies, misconceptions and mishaps: Effective management of heart failure. Soc Sci Med. 2004;58:631–43. [PMC free article] [PubMed]
22. McGarry J. Defining roles, relationships, boundaries and participation between elderly people and nurses within the home: An ethnographic study. Health Soc Care Community. 2009;17:83–91. [PubMed]
23. Resnick B, Nahm ES. Reliability and validity testing of the revised 12-item short-form health survey in older adults. J Nurs Meas. 2001;9:151–61. [PubMed]
24. Ware J, Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34:220–33. [PubMed]
25. Riegel B, Carlson B, Moser DK, Sebern M, Hicks FD, Roland V. Psychometric testing of the self-care of heart failure index. J Card Fail. 2004;10:350–60. [PubMed]
26. Riegel B, Lee CS, Dickson VV, Carlson B. An update on the self-care of heart failure index. J Cardiovasc Nurs. 2009;24(6):485–497. [PMC free article] [PubMed]
27. Avis ND, Smith KW, Hambleton RK, Feldman HA, Selwyn A, Jacobs A. Development of the multidimensional index of life quality. A quality of life measure for cardiovascular disease. Med Care. 1996;34:1102–20. [PubMed]
28. Goldman L, Hashimoto B, Cook EF, Loscalzo A. Comparative reproducibility and validity of systems for assessing cardiovascular functional class: advantages of a new specific activity scale. Circulation. 1981;64:1227–34. [PubMed]
29. Little RJA. Modeling the drop-out mechanism in repeated-measures studies. J Am Stat Assoc. 1995;90:1112–1121.
30. Liang KY, Zeger SL. Longitudinal data analysis using generalized estimating equations. Biometrika. 1986;73:13–22.
31. Laird NM. Missing data in longitudinal studies. Stat Med. 1988;7:305–315. [PubMed]
32. Hedeker D, Gibbons RD. Application of random-effects pattern-mixture models for missing data in longitudinal studies. Psychol Methods. 1997;2:64–78.
33. Hochberg Y, Tamhane AC. Multiple Comparison Procedures. New York: John Wiley & Sons; 1987.
34. Brennan PF, Moore SM, Smyth KA. The effects of a special computer network on caregivers of persons with Alzheimer’s Disease. Nurs Res. 1995;44:166–172. [PubMed]
35. Brennan PF. Computer network home care demonstration: A randomized trial in persons living with AIDS. Comput Biol Med. 1998;28:489–508. [PubMed]
36. Moore SM, Charvat JM, Gordon NH, Pashkow F, Ribisl P, Roberts BL, Rocco M. Effects of a CHANGE intervention to increase exercise maintenance following cardiac events. Ann Behav Med. 2006;31:53–62. [PubMed]
37. Zimmerman L, Barnason S. Use of a telehealth device to deliver a symptom management intervention to cardiac surgical patients. J Cardiovasc Nurs. 2007;22:32–7. [PubMed]
38. Demiris G. The diffusion of virtual communities in health care: concepts and challenges. Patient Educ Couns. 2006;62:178–88. [PubMed]
39. Jurgens C, Holke L, Byrnes J, Riegel B. Why do elders delay in responding to heart failure? Nurs Res. 2009;58:274–82. [PubMed]
40. Sacks N, Cabral H, Kazis LE, Jarrett KM, Vetter D, Richmond R, Moore TJ. A web-based nutrition program reduces health care costs in employees with cardiac risk factors: before and after cost analysis. J Med Internet Res. 2009;11:e43. [PMC free article] [PubMed]


1. Liang K-Y, Zeger SL. Longitudinal data analysis using generalized estimating equations. Biometrika. 1986;73:13–22.
2. Laird NM. Missing data in longitudinal studies. Stat Med. 1988;7:305– 315. [PubMed]
3. Little RJA. Modeling the drop-out mechanism in repeated-measures studies. J Am Stat Assoc. 1995;90:1112–1121.
4. Hedeker D, Gibbons RD. Application of random-effects pattern-mixture models for missing data in longitudinal studies. Psych Methods. 1997;2(1):64–78.
5. Hogan JW, Laird NM. Mixture models for the joint distribution of repeated measures and event times. Stat Med. 1997;16:239–258. [PubMed]
6. Hochberg Y, Tamhane AC. Multiple Comparison Procedures. John Wiley & Sons; New York: 1987.