View Entire Collection
By Clinical Topic
By Journal
By Specialty
By Category
Asthma
COPD
Diabetes – Summer 2012
Future of Nursing Initiative
Heart Failure - Fall 2011
Influenza - Winter 2011
Magnet Recognition
Nursing Ethics - Fall 2011
Nutrition
Pneumonia
Renal Disease
Stroke
Trauma - Fall 2010
Traumatic Brain Injury - Fall 2010
Fluids & Electrolytes
Background/Objective: Few studies have examined whether chronic heart failure (HF) outcomes can be improved by increasing patient engagement (known as activation) in care and capabilities for self-care management. The objective was to determine the efficacy of a patient activation intervention compared with usual care on activation, self-care management, hospitalizations, and emergency department visits in patients with HF.
Methods: This study used a randomized, 2-group, repeated-measures design. After consent was given, 84 participants were stratified by activation level and randomly assigned to usual care (n = 41) or usual care plus the intervention (n = 43). The primary outcomes and measures were patient activation using the Patient Activation Measure (PAM), self-management using the Self-Care of Heart Failure Index (SCHFI) and the Medical Outcomes Study (MOS) Specific Adherence Scale, and hospitalizations and emergency department visits. The intervention was a 6-month program to increase activation and improve HF self-management behaviors, such as adhering to medications and implementing health behavior goals.
Results: Participants were primarily male (99%), were white (77%), and had New York Heart Association III stage (52%). The mean (SD) age was 66 (11) years, and 71% reported 3 or more comorbidities. The intervention group compared with the usual care group showed a significant increase in activation/PAM scores from baseline to 6 months. No significant group-by-time interactions were found for the SCHFI scales. Although the baseline MOS Specific Adherence Scale mean was lower in the intervention group, results showed a significant group-by-time effect with the intervention group improving more over time. Participants in the intervention group had fewer hospitalizations compared with the usual care group when the baseline activation/PAM level was low or high.
Conclusion: This study supports the importance of targeted interventions to improve patient activation or engagement in HF care. Further work is needed related to HF self-management measurement and outcomes.
Heart failure (HF) has emerged as a common hospital discharge diagnosis among adults older than 65 years and a primary risk factor for hospital readmission within 60 days of discharge.1 Demographic projections, with improving survival from cardiovascular disease and the aging of our population, show a dramatic increase in new HF cases, resource use, and cost of treatment based on current trends in frequency, intensity, and scope of healthcare services.2
Chronic HF care places a tremendous burden on patients, their families, and society. Healthcare systems are being pressured to improve the delivery of HF care and to slow, if not reverse, escalating costs associated with chronic HF care and subsequent poor outcomes.3,4 Yet, among healthcare stakeholders, there is broad-based consensus that the care of individuals with HF is often fragmented and poorly coordinated across care providers and sites. Evidence-based practice guidelines related to optimal HF management strategies are not routinely being followed, and multiple barriers exist that impede patients' ability to engage in and manage their own healthcare/self-management.4-6
Given the complex and progressive nature of HF that often results in adverse events and poor outcomes (including costly hospital readmissions), it is important to identify interventions that assist those with chronic HF to practice effective self-management.4,5 Self-management strategies have become a core component and central theme for many HF disease management programs and chronic care models at organizational levels.7-9 However, efforts to improve chronic HF care must also focus on helping individuals become more informed about their illness and to actively engage in their own care (activation) and improve skills for self-management.
The concept activation is derived from Hibbard and colleagues10 and is defined as having the information, motivation, and behavioral skills necessary to self-manage chronic illness, collaborate with healthcare providers, maintain functioning, and access appropriate care. Hibbard and colleagues10,11 have also described the theory and measurement of activation. Activation theory was the basis for the study intervention.
Current evidence supports findings that outcomes for HF care not only depend on receiving better care while in the hospital but also increasingly depend on the quality and continuity of chronic care that a patient receives outside of the hospital.7,8,12 Identifying approaches that boost the impact of chronic disease management programs and self-management interventions provides a promising direction for improving the outcomes of chronic HF care. One such approach is tailoring interventions and self-management support to the individual patient's level of activation.
This study extends the work of Shively et al13 on the efficacy of behavioral change in HF.
Our first study13 demonstrated the efficacy of a behavioral self-management group intervention on the self-reported physical functioning aspect of health-related quality of life. We used the information-motivation-behavior components of the previous intervention but developed the current intervention based on activation theory and used an individualized delivery format. The primary objective of this study was to determine the efficacy of a patient activation (Heart PACT) intervention compared to usual care on patient activation, self-management, hospitalizations, and emergency department (ED) visits in patients at high risk of readmission/hospitalization for HF.
This study used a randomized, 2-group, repeated-measures design. After eligibility screening and consent, participants were randomly assigned to usual care or usual care plus the activation/Heart PACT intervention. A stratified blocked randomization approach based on the baseline Patient Activation Measure (PAM) level (low, medium, high) was used to ensure that patients were equally distributed between groups by activation level.11,14 Participants in the usual care group received general medical care and any HF-specific clinical care from a primary care provider. Participants in the intervention group received usual care plus a 6-month activation/Heart PACT intervention. Outcomes were measured at baseline, 3 months, and 6 months.
This was a single-site study conducted at the academically affiliated Veterans Affairs (VA) San Diego Healthcare System. The target population had documented clinical HF stage C,15 an incident hospitalization or ED visit for HF treatment within the previous 12 months, and was aged 18 years or older. Additional inclusion criteria required that the participant live in San Diego county, read and speak English, has telephone access, and has a primary care provider for routine medical care. Exclusion criteria included the inability to provide written consent, acute medical problems within the previous month, or considered by the investigators to be medically unstable. Also excluded were those enrolled in specialty HF care via the HF Program or telehealth or had long-term follow-up by cardiology after a hospital admission as well as severe medical problems, a life expectancy of less than a year, acute substance abuse or psychiatric problems, or homelessness. Although the Veteran population at the study site is typically male and white, efforts were made to recruit females and minorities. Enrollment started in September 2006 and ended in January 2009. All face-to-face assessments and intervention activities were completed by June 2009; medical record assessments for participant follow-up were completed by January 2010.
The appropriate institutional review board approved the study and all recruitment strategies. All participants provided written informed consent for both eligibility screening and the study. All participants were given a $10 payment at baseline and 3 months and $30 at the last assessment (6 months).
Each participant's medical record was reviewed after written informed consent and before randomization to confirm that inclusion criteria were met and to rule out any exclusion criteria. Participants were scheduled for a screening meeting within a month of enrollment and a baseline outcome assessment visit that included completion of the outcomes questionnaire packet, a blood sample for B-type natriuretic peptide, and a 6-minute walk. After completion of the approximate 2-hour baseline assessment, patients were randomly assigned to usual care or usual care plus the activation/Heart PACT intervention and given information regarding their follow-up schedules and study group assignment. The only component of this study that was experimental was the activation/Heart PACT intervention. Usual care was not manipulated in the study and consisted of routine medical care in primary care and specialty clinics (other than the HF Specialty Clinic) at the study site. All participants enrolled in the study were asked to repeat the same outcome assessments at 3 and 6 months (see Figure 1, Clinical Trial Design). Hospitalization and ED visits were collected by patient interview.
The activation/Heart PACT intervention was developed from our previous work with HF self-management16 and based on activation theory and measurement as described and defined by Hibbard and colleagues.10,17 The intervention was a 6-month program developed to enhance self-management and was given by 1 of 2 advanced practice nurses. The unique aspects of this intervention were that the intervention used activation theory and was tailored to each participant's activation level. Figure 2 shows the individualized/tailored goals for skills and behaviors according to baseline activation level. Table 1 shows intervention examples for each activation level. At each meeting/telephone call with patients in the intervention group, the goals and progress toward attaining these were discussed. At the first individual meeting, the intervention leaders assessed patients' level of activation as low, medium, or high. This baseline activation was assessed by both the patient's self-report score on the PAM11 and a brief interview. The tailored program focused on having individualized self-selected goals and moving the patient to a higher level of activation. Each participant met with the intervention nurses for 6 sessions, by telephone or in person. During these meetings, participants' individualized health behavior goals were discussed, progress toward goals was reinforced, barriers were addressed, and questions were answered. The most common goals chosen by participants in the intervention group were increasing activity (n = 17) and improving diet (n = 10). Intervention fidelity was evaluated by multiple methods including audiotaping of initial and randomly selected intervention visits, using a standardized manual, and keeping intervention logs to record activation ratings, plan of care, goal attainment, and content of each visit.
The practice policy for usual care at the study site requires an assigned primary care provider who assumes responsibility for preventive, acute, and chronic healthcare services consistent with established guidelines and procedures. Routine health maintenance care is every 6 months or annually as determined by the primary provider and individual needs of the patient. Primary care providers can be a physician, nurse practitioner, or physician assistant. Telehealth monitoring for chronic conditions such as diabetes, lung disease, and HF is available and managed by the primary care provider. All services are based on medical needs of the patient under the direction of the primary care provider.
The intervention group received a self-management toolkit (blood pressure cuff, weight scale, pedometer, HF self-management DVD, and educational booklet) at the first intervention visit. The usual care group received the self-management toolkit after the final 6-month assessment.
The primary outcome variables and measures were (1) activation using the PAM total score,10,11 (2) self-management using the 3 scale scores (maintenance, management, and self-confidence) from the Self-Care of Heart Failure Index (SCHFI)18 and the Medical Outcomes Study (MOS) Specific Adherence Scale,19 and (3) hospitalizations and ED visits as reported by participants.
We used the copyrighted 13-item PAM with permission from Insignia Health (Patient Activation Measure License Package, May 2007 (http://www.insigniahealth.com/ha/measure.html), to measure activation. The PAM is an interval-level, unidimensional, Guttman-like scale developed and tested by Hibbard and colleagues10,11 Hibbard and Mahoney17 describe the PAM as a measure of self-management self-concept, that is, "being in charge of one's own health" (http://www.insigniahealth.com/ha/measure.html). The authors describe 4 levels of activation, viewed as sequential across a hierarchical continuum: 1 (low level of activation)-believing the patient role is important (score <= 47), 2 (also a low level)-having the knowledge and confidence to take action (score = 47.1-55.1), 3 (medium)-taking action to maintain or improve health (score = 55.2-67), and 4 (high)-maintaining healthy lifestyle changes under stress (score >= 67.1).10,11 Permission to use the PAM and current scoring information must be obtained from Insignia Health. A sample item is "when all is said and done, I am the person who is responsible for managing my health condition(s)."
Each of the 13 PAM items is scored on a 4-point scale; the summed raw score is converted to a 0 to 100 activation score. The activation score is then used to identify the level of activation as provided on the licensed instrument instructions. Hibbard et al10 reported a Cronbach's [alpha] of .87 for the PAM; test-retest reliability was assessed, and construct and criterion validity were supported. For our study sample, the Cronbach's [alpha] for the baseline PAM was .88. The evaluation by Hibbard and colleagues11 suggests that the PAM is highly reliable at the individual patient level and a valid instrument for assessing activation and individualizing care. Hibbard and colleagues20,21 have assessed activation level as an approach to supporting self-management. The PAM questionnaire was used in our study to self-report activation and to document participants' activation throughout the intervention along the different points of the activation continuum. A brief clinical interview was also used by the intervention leaders to assess level of activation. This interview helped individualize the intervention/care plan to support beliefs, knowledge, confidence, and action or problem solving.
We used the copyrighted SCHFI, version 4, with permission to measure self-management.18,22 Riegel and Dickson23 describe self-care maintenance and management as vital elements of self-care. The SCHFI measures self-care maintenance as a process of maintaining health through positive health practices and HF self-management as a process of recognizing and evaluating HF symptoms, treating symptoms, and evaluating the treatments chosen. Confidence is thought to be a moderator of the relationship between self-care and outcomes.
The SCHFI is a self-report measure that has 3 scales: self-care maintenance (5 items; raw score range, 5-20), management (6 items; scale scored only if patient reports breathing difficulty or ankle swelling; raw score range, 4-24), and confidence (4 items; raw score range, 4-16). The 3 scale scores are treated as 3 variables and scored separately. Each item in the SCHFI is rated on a 4-point (1-4) Likert response scale. One sample item on the self-maintenance scale is, "How often do you eat a low salt diet?"18 Scores on each of the scales are standardized to 100; higher scores reflect better self-care. Riegel et al report that a score of 70 or greater has been used to indicate self-care adequacy and that benefit may occur at lower levels. A scale score change greater than a one-half SD is clinically relevant according to Riegel et al.22 The SCHFI has been widely disseminated and has reported psychometric testing in patients with HF.22 Riegel et al22 reported the following [alpha] coefficients for the original version scales: maintenance, .56; management, .59; self-confidence, .82. The [alpha] coefficient reliabilities for the scales in our study were as follows: maintenance, .50 (5 items); management, .42 (6 items); self-confidence, .88 (4 items).
The MOS Specific Adherence Scale was used to determine a patient's self-reported ability (over the previous 4 weeks) to adhere to specific treatment recommendations made by a medical provider (eg, follow a low-sodium diet) and frequency of carrying out the specific activity (eg, exercised regularly 3 or more times per week).19,24 The MOS Specific Adherence Scale varies with the kind of medical condition being examined and was developed to be specific to patients with heart disease.19 It has been used in research participants with various cardiovascular disorders, including HF.25-27 We used 8 items from the MOS Specific Adherence Scale that focused on common HF regimens (medication, low-sodium/low-fat diet, exercise, alcohol intake, tobacco use, monitoring daily weight/symptoms). Here is a sample item: "How often have you done each of the following in the past 4 weeks: weighed yourself every day to watch your fluid status?" The item response range is 0 (none of the time) to 5 (all of the time). The scores are transformed to a 0 to 100 scale. A 5-item general adherence scale19 had an [alpha] of .81. The Cronbach's [alpha] for our sample at baseline was .54.
We examined relationships between the PAM, SCHFI, and MOS instruments. There was no significant correlation between baseline PAM and MOS scores (r = 0.09, P = .44, n = 84). This indicates no overlap between the activation/PAM and self-management adherence/MOS measures. However, the items on the SCHFI maintenance and MOS scales are similar in content, and there was a significant correlation between these scales (r = 0.51, P < .001, n = 84), indicating a 25% common variance between these measures.
Healthcare use information was obtained from participants by interview at baseline, 3, and 6 months. Participants were asked to report any hospitalizations, ED visits, and other unscheduled visits including reason for visits and treatment. We also extracted hospitalization and ED visit data from the electronic VA national database, but these data do not capture non-VA healthcare. Therefore, we report only the interview data in this article.
We measured several other variables and potential covariates including sociodemographic and clinical characteristics (eg, B-type natriuretic peptide), such as medical diagnoses, cardiac functional class (New York Heart Association), comorbidities (Charlson Weighted Index of Comorbidity adapted by Katz et al28), ejection fraction, and medication profile. We also assessed health literacy (Short Test of Functional Health Literacy in Adults [STOFHLA]),29 functional capacity (6-minute walk), HF knowledge, health perception, healthcare services use, perceived control, anxiety and depression using self-report, patient interview, and/or review of medical records.
The analyses for the primary hypotheses consisted of mixed design 2 (group: usual care or intervention) x 3 (time: baseline, 3 and 6 months) analyses of variance with 1 between-subjects factor (group) and 1 within-subjects factor (time) for each dependent variable. Tests were conducted for main effects and interaction effects. The distributions for all dependent variables/outcome measures were checked for normality. The primary interest in the analyses was the group-by-time interactions, and because we expected linear effects, we evaluated the group-by-time linear contrasts. Initial analyses included all subjects who were randomized to treatment and completed all 3 data collection time points (intent to treat). A second set of analyses used activation level as an independent variable in 2 (group) x 3 (activation level: low, medium, or high) x 3 (time) analyses of variance.
Potential covariates were evaluated for differences between the groups at baseline. Age was the only covariate on which the groups differed.
Initially, analyses were conducted for evaluating whether missing data were missing completely at random (MCAR) using the missing value analyses module of SPSS version 1 7 (SPSS Inc, Chicago, Illinois). The distribution of missingness met the criteria for "missing completely at random" where Little's MCAR test was not significant ([chi]2689 = 749.22, P = .06).30 This allows the listwise dropping of subjects with missing data, and this is how the first set of hypotheses testing analyses was run. However, because the Little's MCAR test P value was close to .05, additional analyses were also run using the SPSS version 17 missing value analyses module using an iterative expectation and maximization procedure. The expectation and maximization module forms a missing data correlation matrix for the partially missing data, finds the conditional expectation of the missing data, then substitutes these expectations for the missing data. The maximization steps perform maximum likelihood estimation to generate imputed values.31,32 The additional hypothesis-testing analyses using imputed values for missing data are also reported here.
A total of 84 patients were stratified by activation level and randomly assigned to usual care (n = 41) or usual care plus the activation intervention (n = 43). Table 2 shows the sample characteristics at baseline. Participants were primarily male (99%), white (77%), and classified as New York Heart Association III (52%). The mean (SD) age was 66 (11) years; ages ranged from 42 to 89 years. The mean (SD) years of education was 14.8 (.3.1). Most (71%) of the participants reported 3 or more comorbid conditions. The most common etiologies for HF in this sample were ischemic cardiomyopathy (63%) and hypertension (52%). There were significant differences between groups at baseline on age. Participants in the usual care group were significantly older than those in the intervention group (69 vs 63 years, P = .02).
The total attrition rate for our study sample was 19%: 17% for the usual care group and 21% for the intervention group. Of the 84 total subjects enrolled, all 84 subjects completed baseline assessments; 77 subjects completed the 3-month assessments and 68 completed the 6-month assessments.
Tables 3 (PAM, MOS, SCHFI scores) and 4 (hospitalizations and ED visits) show the mean scores for the major outcome variables reported here. Table 5 shows the F test results, significance, effect sizes, and power values for these variables.
The intervention group compared with the usual care group showed a significant increase in activation/PAM scores from baseline to 6 months (significant group by time interaction, F = 3.73, P = .03) (see Figure 3). There was a significant group-by-activation/PAM level-by-time interaction (F = 3.89, P = .005). The intervention group improved more over time compared with the usual care group with the effect most clearly and strongly present for those with a medium baseline activation/PAM level (see Figure 4). The age covariate and missing values analysis did not show any additional effects for the PAM. There was still a significant group by activation/PAM level-by-time interaction (F = 3.85, P = .006) when age was entered as a covariate.
There was no significant group-by-time interactions for the SCHFI maintenance, management, or self-confidence scales and no significant interaction effects for group-by-PAM level-by-time interaction for the SCHFI scales. However, the age covariate did increase the 3-way effect size for the SCHFI maintenance scale scores, closer to a medium effect but still nonsignificant. Also, the 3-way pattern for SCHFI was the same as for the PAM with the medium PAM level pattern being the most clear. The missing values analysis did not show any additional effects for the SCHFI scales.
The baseline MOS Specific Adherence Scale mean was lower in the intervention group; however, there was a significant group-by-time effect (F = 7.48, P = .001) with the intervention group improving more over time than the usual care group (Figure 5). There were no significant 3-way interactions for the MOS Specific Adherence Scale, but the group-by-time patterns seemed to vary by PAM level. Even with age as a covariate, the group-by-time effect was still significant (F = 6.76, P = .002). The missing values analysis did not show any additional effects for the MOS.
The frequency of hospitalizations (self-reported at 3-month intervals) in the sample was as follows: baseline, 38%; 3 months, 23%; and 6 months, 26%. There was a significant 3-way interaction for hospitalizations (F = 2.57, P = .041). Participants in the intervention group had fewer hospitalizations compared with the usual care group when the baseline activation/PAM level was low or high. Participants in the intervention group compared with usual care with a medium level of activation showed more hospitalizations at 3 and 6 months.
The frequency of ED visits (self-reported for the previous 3 months) in the sample was as follows: baseline, 52%; 3 months, 35%; and 6 months, 24%. There were no group-by-time or group-by-PAM level-by-time interactions for ED visits.
The potential effect of age as a covariate is stated above. Health literacy was examined as a potential covariate. All participant health literacy/STOFHLA scores were in the "adequate" range. However, only 21% of participants at baseline and 26% at 3 and 6 months were able to complete the STOFHLA within 7 minutes. There were no significant differences in baseline STOFHLA scores between the usual care and activation intervention groups.
Participants reported the following rates of telehealth enrollment: baseline, 27%; 3 months, 29%; and 6 months, 29%. There was no significant effect of telehealth enrollment on activation/PAM scores, SCHFI scales, MOS scores, hospitalizations, or ED visits.
The intervention group showed increased perceived control scores at 6 months compared with the usual care group, that is, a significant group-by-time-by-PAM level interaction (F = 3.23, P = .015). This effect was most clear in those with a medium activation level. There were no significant interaction effects for anxiety or depression.
Our study is 1 of the first efficacy trials of an activation intervention in patients with chronic HF. The most important findings from this study were that patient activation could be improved through a targeted intervention and that the effect was more pronounced for those persons with a baseline medium level of activation. Age as a covariate did not change these findings. It is logical that activation/PAM scores were affected by the intervention because the intervention was based on activation theory and measurement.
Our activation intervention study supports the work by Hibbard and colleagues and demonstrates that an intervention focused on increasing patient activation can be successful. Hibbard et al20 showed that changes in participants' level of activation were accompanied by changes in some self-management behaviors in a controlled trial of 479 chronic disease patients randomized to either a chronic disease self-management intervention or control (ie, usual care) condition. Hibbard et al20 concluded that if activation (as measured by self-reported PAM score) was increased, a change in behavior would follow.20 Hibbard et al14 extended their work on activation and conducted a quasi-experimental study to determine the impact of tailoring individual care plans to the patient's level of activation as compared with a usual disease management approach. Results show a consistent picture that indicates the positive impact of tailored interventions with increased activation scores; improved clinical indicators, such as blood pressure and lipid levels; and decreased utilization rates (ED visits, hospitalizations, and office visits) in the intervention group compared with the control/usual care group. These studies suggest that activation is changeable and that increases in activation are followed by improvements in some self-management behaviors and health-related outcomes.14,20
Powell and colleagues33 reported the largest (N = 902) behavioral/educational trial of self-management in HF. The purpose of their Heart Failure Adherence and Retention Trial was to determine the efficacy of self-management counseling and HF education compared with HF education alone related to death and HF hospitalization outcomes. The intervention focused on self-management and problem-solving skills. The intervention was composed of eighteen 2-hour group meetings over a 1-year period. The results did not show an intervention effect on the primary outcomes of death and hospitalizations. Although the intervention targeted self-management, the outcomes measurement did not focus on self-management. This trial highlights the need for continued work on self-management interventions, measurement, and outcomes.
There is limited understanding of the effect of patient activation on self-management behaviors as well as a lack of evidence that better self-management behaviors translate to better outcomes in high-risk patients with HF. Our study did not explain what self-management behaviors are most affected by activation. However, our findings support the recommendations by Hibbard and colleagues20,21 to use the PAM in a clinical setting to plan interventions for improved self-management. A brief clinical interview as used in our study may complement the self-report PAM in clinical and research settings. Although clinical use is supported, there have been no published reports of PAM use and links between provider behavior and patient outcomes.
Our results indicate that our intervention made a greater impact on those participants with a medium level of activation. There may be a ceiling effect for those with high activation. Factors that may influence activation levels in those with baseline low activation must be considered, such as depression and negative self-perception and emotions (eg, feeling overwhelmed). Although our results did not show an effect of depression on the self-management outcomes, Hibbard et al20 found that those with depression were less likely to show improvement in activation or self-management behaviors. Hibbard et al's findings have implications for not only identifying depression as a potential barrier to improving activation but also treating the problem as a prerequisite to implementing interventions aimed at stimulating activation and behavior change in patients with complex chronic diseases. Hibbard and Mahoney17 indicate that activation is linked with emotions and recommend that working with individuals to improve health may require starting with improving one's self-efficacy and role as a self-manager.
Hibbard and colleagues10 surveyed a national probability sample (N = 1515) and found that patients with higher activation reported significantly better health as measured by the SF-8 and had significantly lower rates of healthcare visits, emergency department visits, and hospital stays. Our results are congruent with those of Hibbard et al for participants who had low or high activation levels but those with medium activation had more hospitalizations. The explanation for the pattern of hospitalizations is not clear. Possible reasons for our results of no significant activation effects on ED visits may be because of our small sample, low number of these outcomes in our sample, possible underreporting, and clinical practice improvements for HF.
Several studies have been reported related to self-management in HF. These studies are single studies with varied intervention components and outcome measures. Many studies examine outcomes of hospital readmissions and mortality. This study was 1 of the first to examine patient activation and self-management in HF using a randomized design.
The 2005 Cochrane Collaboration assessed the effectiveness of disease management interventions for care of HF patients after hospitalization (16 trials involving 1627 people). Interventions were classified as multidisciplinary (holistic approach by a team), case management (intense monitoring), or clinic (follow-up in HF clinic). Only 1 trial was of high quality. There was some weak evidence that intense monitoring after discharge might improve survival and reduce the number of readmissions. The authors concluded that there were insufficient data to form recommendations.34
The Centre for Reviews and Dissemination35 compared the effectiveness of disease management programs and usual care on mortality and hospitalizations. A total of 33 randomized controlled trials (n > 7400 people) were included, but only 10 studies were considered to be of high quality. The reviewers concluded that compared with usual care, the disease management programs were associated with a statistically significant reduction in HF-specific mortality and HF-related hospitalizations. Nurse-led interventions did not impact all-cause mortality. It may be that nurse-led interventions such as ours should be combined with comprehensive disease management programs for patients with HF. Further research is needed to establish the incremental benefit of the different elements and components of self-management programs to ensure that the stated intervention or program provides the intended clinical and economic benefits.36-38
The limitations for this study include the sample size, age and gender demographics, attrition, missing data, instrumentation issues, small number of hospitalizations and ED visits, and clinical practice changes during the course of the study. This was a small-efficacy study. Although acceptable observed power was present for most of the significant effects, some of the significant effects did show observed power less than 0.80. Therefore, caution should be used in interpreting even significant effects if they have lower power. The group difference in age is most likely a function of the relatively small sample size. The predominantly male sample is representative of the VA population but not the private sector. There may be gender differences in activation and self-management. Although the follow-up was for 6 months only, we experienced an attrition rate of 19%. This rate was higher than our previous study (13%) but not unexpected. Participants had multiple comorbidities, which may have contributed to attrition. Gasoline prices increased significantly during enrollment, and the economic downturn may have contributed to some attrition because of transportation issues. Participant attrition limits the representativeness of the sample in this study and has the potential to weaken the assessments of longitudinal data, such as hospitalizations. Missing data are common in randomized trials, and efforts were made to minimize this problem throughout data collection. Although a limitation, the missing data did not affect the results.
Findings regarding the impact of increased activation on self-management were inconclusive. Reasons may be because of the intervention or instrumentation and measurement issues. The activation and HF self-management constructs have only recently been explicated and operationalized. The PAM is described as a measure of self-management self-concept, which may only measure the perception or cognition of activation rather than serving as an indicator of actual behavior. The SCHFI maintenance and management scales are said to measure ability to maintain healthy behaviors and manage symptoms but were not sensitive to the intervention in this study. Future research must include measures in addition to self-reported self-management behaviors. Such measures or proxy measures for self-management may include a physiologic indicator of sodium levels; demonstration of medication adherence; or patient monitoring of weight, activity, or diet.
The instruments used for the major outcomes had acceptable psychometric properties as noted by the authors. The [alpha] coefficients for the MOS Specific Adherence Scale and the SCHFI were lower in our study than the authors reported. One reason may be because of the small number of scale items. Also, the SCHFI maintenance and MOS scales were related to a moderate degree. In some cases, test-retest may be a better reliability estimate. However, in this sample, we did not measure test-retest reliability because of subject burden, and we expected that some participants would change on the measure differentially over time.
Results for the effect of activation on self-management were somewhat different for the 2 measures of self-management. Group assignment and activation level did not significantly affect self-management as measured by the SCHFI scores. However, the intervention group showed more improvement in the MOS Specific Adherence Scale, a measure of self-management, than the usual care group. The MOS Specific Adherence Scale is very similar to the SCHFI maintenance scale. The reasons for these differences on the SCHFI and MOS measures are not known. There are inherent difficulties in examining relationships among patients with complex chronic diseases like HF and much of the research pertaining to self-management has focusedon patients with single or uncomplicated disease states and under circumscribed management regimens, or both. Patients with complex chronic conditions like HF are more likely to experience fluctuations in their health status and may have multiple self-management requirements.39 More work is needed on instruments that measure self-management in the face of changing circumstances with HF. Given the complexity of self-management tasks required of patients with complex chronic conditions such as HF, multiple measures (eg, observations of behaviors, self-report, clinical indicators) of self-management must be considered in future research.
There were changes in clinical practice over the course of our study that may have impacted the outcomes. Telehealth was implemented near the beginning of our study. To minimize any confounding effects, we enrolled patients who were in telehealth for a condition other than HF but not in telehealth primarily for HF. Currently, 28% of our 355 telehealth patients have chronic HF as 1 of their monitored diagnoses. As stated in the Results section, telehealth did not seem to be a confounder.
The other change in clinical practice about midway through enrollment was how patients with HF were followed after a hospitalization. The practice change required all patients discharged with a HF diagnosis to be seen at least once by the HF Specialty Program before returning to primary care. We did not enroll patients in the study until they returned to their primary care provider for follow-up.
Despite the limitations, this study is one of the first to test an activation intervention in a randomized trial and examine HF self-management outcomes. We also had participants with varied levels of activation (low, medium, and high) and balanced the study groups based on these levels. Our results support a causal link between the intervention and activation. However, the causal links between activation, self-management, and hospitalizations have not yet been established or supported by large, well-designed studies. The knowledge base is building regarding activation and self-management. Use of activation theory and assessment in the clinical setting should be considered for patients with chronic illness, such as hospital to home transitions and self-management programs.
* Patient activation can be improved through targeted intervention.
* The PAM (self-report) and a brief clinical interview may be useful in clinical settings.
* Activation level did not significantly affect SCHFI scores.
* The explanation for the pattern of hospitalizations is not clear.
* Further research needs to be done regarding the causal links between activation, self-management, and hospitalizations.
Alan Maisel, MD; Allen Gifford, MD; Marilynne Tseng, BSN, RN; Laureen Pada, MSN, RN; Justine Bono Foltz, MS, RN; Monica Wilson, MS, RN; Kathleen Dracup, DNSc, RN, FNP, FAAN; Paul Heidenreich, MD; Barbara Riegel, DNSc, RN; Merilee Pipkin Pearson; Stephanie Johnson; and Cyrene Benjamin are gratefully acknowledged.
1. Lloyd-Jones D, Adams R, Brown T, et al.. Heart disease and stroke statistics-2010 update: a report from the American Heart Association. Circulation. 2010; 121 (7): e46-e215. [Context Link]
2. Hunt SA, Abraham WT, Chin MH, et al.. Focused update incorporated into the ACC/AHA 2005 Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines Developed in Collaboration with the International Society for Heart and Lung Transplantation. J Am Coll Cardiol. 2009; 53 (15): e1-e90. [Context Link]
3. Krumholz HM, Currie PM, Riegel B, et al.. A taxonomy for disease management: a scientific statement from the American Heart Association Disease Management Taxonomy Writing Group. Circulation. 2006; 114 (13): 1432-1445. [Context Link]
4. Bodenheimer T, Chen E, Bennett HD. Confronting the growing burden of chronic disease: can the U.S. health care workforce do the job? Health Aff (Project Hope). 2009; 28 (1): 64-74. [Context Link]
5. Cameron J, Worrall-Carter L, Riegel B, et al.. Testing a model of patient characteristics, psychologic status, and cognitive function as predictors of self-care in persons with chronic heart failure. Heart Lung. 2009; 38 (5): 410-418. [Context Link]
6. Fonarow GC, Yancy CW, Heywood JT. Adherence to heart failure quality-of-care indicators in US hospitals: analysis of the ADHERE Registry. Arch Intern Med. 2005; 165 (13): 1469-1477. [Context Link]
7. Centers for Disease Control and Prevention. State of Aging and Health in America Report. Washington, DC: US Department of Health and Human Services; 2007. [Context Link]
8. Adams K, Corrigan JM, eds. Institute of Medicine. Priority Areas for National Action: Transforming Health Care Quality. Washington, DC: National Academic Press; 2003. [Context Link]
9. Wagner EH. Chronic disease management: what will it take to improve care for chronic illness? Eff Clin Pract. 1998; 1 (1): 2-4. [Context Link]
10. Hibbard JH, Stockard J, Mahoney ER, et al.. Development of the Patient Activation Measure (PAM): conceptualizing and measuring activation in patients and consumers. Health Serv Res. 2004; 39 (4, Part I): 1005-1026. [Context Link]
11. Hibbard JH, Mahoney ER, Stockard J, et al.. Development and testing of a short form of the Patient Activation Measure. Health Serv Res. 2005; 40 (6 Pt 1): 1918-1930. [Context Link]
12. Boutwell A, Griffin F, Hwu S, et al.. Effective Interventions to Reduce Rehospitalizations: A Compendium of 15 Promising Interventions. Cambridge, MA: Institute for Healthcare Improvement; 2009. [Context Link]
13. Shively M, Kodiath M, Smith T, et al.. Effect of behavioral management on quality of life in mild heart failure: a randomized controlled trial. Patient Educ Couns. 2005; 58 (1): 27-34. [Context Link]
14. Hibbard JH, Greene J, Tusler M. Improving the outcomes of disease management by tailoring care to the patient's level of activation. Am J Manag Care. 2009; 15 (6): 353-360. [Context Link]
15. Hunt SA, Baker DW, Chin MH, et al.. ACC/AHA guidelines for the evaluation and management of chronic heart failure in the adult: executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to revise the 1995 Guidelines for the Evaluation and Management of Heart Failure). J Am Coll Cardiol. 2001; 38 (7): 2101-2113. [Context Link]
16. Kodiath M, Kelly A, Shively M. Improving quality of life in heart failure patients: an innovative behavioral intervention. J Cardiovasc Nurs. 2005; 20 (1): 43-48. [Context Link]
17. Hibbard JH, Mahoney E. Toward a theory of patient and consumer activation. Patient Educ Couns. 2010; 78 (3): 377-381. [Context Link]
18. Riegel B, Carlson B, Moser DK, et al.. Psychometric testing of the Self-Care of Heart Failure Index. J Card Fail. 2004; 10: 350-360. [Context Link]
19. Hays RD. The Medical Outcomes Study (MOS) measures of patient adherence. http://www.rand.org/health/surveys_tools/mos/mos_adherence_survey.pdf. Accessed May 14, 2010. Rand Health:1-9. [Context Link]
20. Hibbard JH, Mahoney ER, Stock R, et al.. Do increases in patient activation result in improved self-management behaviors? Health Serv Res. 2007; 42 (4): 1443-1463. [Context Link]
21. Hibbard JH, Tusler M. Assessing activation stage and employing a "next steps" approach to supporting patient self-management. J Ambul Care Manage. 2007; 30 (1): 2-8. [Context Link]
22. Riegel B, Lee CS, Dickson VV, et al.. An update on the Self-Care Of Heart Failure Index. J Cardiovasc Nurs. 2009; 24 (6): 485-497. [Context Link]
23. Riegel B, Dickson VV. A situation-specific theory of heart failure self-care. J Cardiovasc Nurs. 2008; 23 (3): 190-196. [Context Link]
24. Hays RD, Kravitz RL. The impact of patient adherence on health outcomes for patients with chronic disease in the Medical Outcomes Study. J Behav Med. 1994; 17 (4): 347. [Context Link]
25. Chung ML, Lennie TA, de Jong M, et al.. Patients differ in their ability to self-monitor adherence to a low-sodium diet versus medication. J Card Fail. 2008; 14 (2): 114-120. [Context Link]
26. Tarlov AR, Ware JE, Greenfield S, et al.. The Medical Outcomes Study: an application of methods for monitoring the results of medical care. JAMA. 1989; 262: 925-930. [Context Link]
27. Wu JR, Moser DK, Chung ML, et al.. Objectively measured, but not self-reported, medication adherence independently predicts event-free survival in patients with heart failure. J Card Fail. 2008; 14 (3): 203-210. [Context Link]
28. Katz JN, Chang LC, Sangha O, et al.. Can comorbidity be measured by questionnaire rather than medical record review? Med Care. 1996; 34 (1): 73-84. [Context Link]
29. Baker DW, Williams MV, Parker RM, et al.. Development of a brief test to measure functional health literacy. Patient Educ Couns. 1999; 38 (1): 33-42. [Context Link]
30. Hill MA. SPSS Missing Value Analysis 7.5. Chicago: SPSS; 1997. [Context Link]
31. Schafer JL, Graham JW. Missing data: Our view of the state of the art. Psychol Methods. 2002; 7 (2): 147-177. [Context Link]
32. Tabachnick B, Fidell LS. Using Multivariate Statistics. 5th ed. Boston: Allyn & Bacon; 2007. [Context Link]
33. Powell LHP, Calvin JEJ, Richardson DP, et al.. Self-management counseling in patients with heart failure: The Heart Failure Adherence and Retention Randomized Behavioral Trial. JAMA. 2010; 304 (12): 1331-1338. [Context Link]
34. Taylor S, Bestall J, Cotter S, et al.. Clinical service organisation for heart failure [Systematic Review]. Cochrane Database of Systematic Reviews. 2005; 2: 2. [Context Link]
35. Centre for Reviews and Dissemination. Effectiveness of comprehensive disease management programmes in improving clinical outcomes in heart failure patients: a meta-analysis (structured abstract). Vol. 2. York, UK: University of York NHS Centre for Reviews and Dissemination; 2009. [Context Link]
36. Clark AM, Savard LA, Thompson DR. What is the strength of evidence for heart failure disease-management programs? J Am Coll Cardiol. 2009; 54 (5): 397-401. [Context Link]
37. Ditewig JB, Blok H, Havers J, et al.. Effectiveness of self-management interventions on mortality, hospital readmissions, chronic heart failure hospitalization rate and quality of life in patients with chronic heart failure: a systematic review. Patient Educ Couns. 2010; 78 (3): 297-315. [Context Link]
38. McAlister FA, Lawson FM, Teo KK, et al.. A systematic review of randomized trials of disease management programs in heart failure. Am J Med. 2001; 110 (5): 378-384. [Context Link]
39. Sevick MA, Trauth JM, Ling BS, et al.. Patients with complex chronic diseases: perspectives on supporting self-management. J Gen Intern Med. 2007; 22 (suppl 3): 438-444. [Context Link]