| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From Harvard Medical School (K.Z.B., O.A., P.H.S., S.E.L., T.J.M.), Harvard Pilgrim Health Care (K.Z.B., O.A., M.G.M., J.C., T.J.M.), Beth Israel Deaconess Medical Center (S.E.L.), and Brigham and Women’s Hospital (P.H.S.), Boston, Massachusetts.
Address correspondence and reprint requests to Thomas J. McLaughlin, ScD, Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care, 133 Brookline Avenue, 6th Floor, Boston, MA 02215. E-mail: thomas_mclaughlin{at}hms.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: We recruited hospitalized patients surviving an acute coronary syndrome with scores on the Hospital and Anxiety Depression Scale (HADS) indicating mild to severe depression and/or anxiety at 1 month postdischarge. Recruited patients were randomized into either an intervention or control group. Intervention patients received up to six 30-minute telephone-counseling sessions focused on identifying cardiac-related fears. Control patients received usual care. For both groups, we collected patients’ responses to the HADS and to the Global Improvement (CGI-I) subscale of the Clinical Global Impressions (CGI) Scale at baseline and at 2, 3, and 6 months postbaseline using Interactive Voice Recognition (IVR) technologies. We used mixed-effects analysis to estimate patients’ changes in CGI-I measures over the three time points of data collection postbaseline.
Results: We enrolled 100 patients, and complete CGI-I measures were collected for 79 study patients. The mean age was 60 years (standard deviation = 10), and 67% of the patients were male. A mixed-effects analysis confirmed that patients in the intervention group had significantly greater improvements in self-rated health (SRH) between baseline and month 3 than the control group (p = .01). Between month 3 and month 6, no significant differences in SRH improvements were observed between the control and intervention groups.
Conclusions: Study patients reported greater SRH improvement resulting from the telephone-based intervention compared with control subjects. Future research should include additional outcome measures to determine the effect of changes in SRH on patients with comorbid physical and emotional disorders.
Key Words: adjustment to chronic disease • cardiac disease • psychological distress • randomized controlled trial • self-rated health • telephone counseling
Abbreviations: ACS = acute coronary syndrome; ADL = activities of daily living; CAD = coronary artery disease; CGI = Clinical Global Impressions Scale; CGI-I = Global Improvement subscale of the Clinical Global Impressions (CGI) Scale; ENRICHD = Enhancing Recovery in Heart Disease Patients trial; HADS = Hospital and Anxiety Depression Scale; HTS = Healthcare Technology Systems; ICD-9 = International Classification of Diseases, Ninth Edition; IRB = Institutional Review Boards; IVR = interactive voice recognition; MI = myocardial infarction; SADHART = the Sertraline Antidepressant Heart Attack Randomized Trial; SRH = self-rated health.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Both pharmacological and psychosocial treatments for mood disorders among patients with CAD can improve patients’ prognosis and quality of life (22,34,37,38). Several studies also have demonstrated that psychosocial treatments focusing on social support (30,39,40), self-efficacy (41), affect (42), and coping style (30,43) appear to be components of an effective intervention for reducing depressive symptoms in cardiac patients. Although research results have not agreed on a mortality-reducing effect of depression treatment among cardiac patients, treatments to alleviate depression in patients with an acute coronary syndrome (ACS) may improve SRH and quality of life (44). Therefore, a key remaining question is whether psychosocial treatment for mental distress among patients with CAD might improve self-rated health.
The present study used a telephone-based counseling intervention to treat psychological distress among patients recently hospitalized for ACS. In contrast to traditional psychotherapy, this intervention focused not on underlying psychopathology, but on the patients’ adjustment to the effects of chronic illness (45). The goal is to help patients adjust successfully to illness by strengthening existing coping mechanisms and mobilizing resources (45). This longitudinal study examined whether a telephone counseling intervention improved patient SRH.
Objective
This study aimed to determine if the intervention led to SRH improvement among distressed patients with heart disease. We hypothesized that patients who received the intervention would experience greater SRH improvements over time than control patients.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Setting and Sources of Data
A major teaching hospital in Boston, Massachusetts, with large case-mixes of patient populations served as the study site. After approval by the Institutional Review Boards (IRBs) was obtained, patients who met specific inclusion criteria were recruited beginning in the fall of 2001 and ending in the summer of 2003.
Inclusion Criteria
Hospitalized patients were identified by medical chart review conducted by a cardiac research nurse to determine the presence of inclusion and exclusion criteria. Patients aged 35 years or older who had a primary diagnosis of ACS defined as unstable angina pectoris or acute myocardial infarction met the inclusion criteria for the study. Patients were excluded who had a history of substance abuse, mental health care, or antidepressant use in the 3 months preceding the index hospitalization.
Patients meeting these criteria were asked if they would be willing to be contacted once they returned home from the hospital. The study coordinator contacted consenting patients by telephone 1 to 2 weeks postdischarge and screened them over the telephone for depression and anxiety symptoms using the Hospital Anxiety and Depression Scale (HADS) (46). Patients with HADS scores indicating depression and/or anxiety symptoms were asked to participate in the study. During this screening telephone call, patients were randomized to one of the two study treatment arms.
Cohort Identification, Randomization, and Intervention
Eligible patients were randomized to either the intervention or usual care (control) group. The principal investigator of the study (T.J.M.) determined that patients would be randomized to either intervention or control status by using a coin flip. The study coordinator (J.C.) conducted the coin flip and assigned patients to a treatment arm when she contacted study participants by telephone and enrolled consenting participants. Participants could not be blinded to treatment assignment because patients in the intervention arm received telephone counseling, whereas the usual care patients did not. Sample size was determined based on power calculations to detect a difference of two units in HADS depression scores and indicated a required total sample of 68 patients. However, we enrolled 100 patients to control for possible loss to follow up at baseline. Enrollment occurred between September 2001 and August 2003.
Control Group
Control patients received a short booklet on coping with chronic illness and were instructed to contact their primary care physician if they experienced any warning signs of more significant depression. They were advised to continue to follow up with their primary care and specialist physicians as they normally would for ongoing care.
Intervention Group
Patients in the intervention group received up to a total of six telephone counseling sessions over a period of 8 weeks. The intervention addressed eight specific issues or fears, including loss of control, loss of self-image, dependency, stigma, abandonment, anger, isolation, and fear of death (45). All sessions were 30 minutes long and were administered by doctoral-level clinicians (a psychiatrist, a clinical psychologist, and/or an internist). Most patients were expected to complete six sessions, but they were allowed to participate in as few as three sessions, depending on the therapist’s assurance that all eight issues set forth during the counseling process were reviewed and that the treatment goal was reached, signaling the completion of treatment. Patients were also able to initiate telephone contact with the counselor and were provided with an emergency telephone number. A record of the sessions was kept to indicate issues reviewed, as well as to serve as a reference for the counselor as the patient progressed through the sessions. It is important to note that this intervention is not the same as traditional psychotherapy; rather, it helps patients identify and build their own capacity to manage the personal effects of chronic illness. It is goal-oriented, time-limited, and issue-focused in contrast to other forms of psychosocial treatment.
Measures
Data on study patients retrieved from the hospital medical records by the research nurse and study coordinator included age, gender, race, history of ischemic cardiac disease, dates of hospital admission and discharge, length of hospital stay, cardiac discharge diagnosis, cardiac procedure descriptors, and depression diagnosis (as assessed using the Primary Care Evaluation of Mental Disorders (47)). A data management and research organization with extensive experience in management of behavioral health data collected patients’ baseline scores and all follow-up data at months 2, 3, and 6. Patients in both treatment groups were asked to provide their responses to the HADS and the Global Improvement (CGI-I) (discussed subsequently) administered over the telephone using computer-assisted IVR technology by pressing telephone keys indicating their responses to each measure’s questions.
Indicator of Psychological Distress
Symptoms of depression and anxiety were assessed in study subjects using the HADS. Each HADS subscale contains 14 items rated on four-point Likert scales using seven questions to assess anxiety and seven questions to assess depression (46). Each subscale is scored on a scale from 0 to 21, in which patients with a score of 0 to 4 are considered to be "normal," 4 to 7 are considered to have "subclinical" depression, 8 to 10 have mild depression, 11 to 14 have moderate depression, and 15 to 21 have severe depression (46). In this study, patients were deemed eligible to participate if they had a score of seven or higher on either the depression or the anxiety subscale on the initial HADS screen conducted postdischarge. The criteria for choosing a cutoff on the HADS depression subscale or anxiety subscale were purposely lenient to be sure that all distressed patients were eligible to participate in the study.
Health Outcome Measurement
The primary outcome measure in this study was the CGI-I subscale of the Clinical Global Impressions (CGI) Scale (48), a standardized assessment tool that is widely used in clinical psychopharmacology trials as an outcome measure of SRH (49,50). The CGI-I is scored from one to seven, in which a score of one indicates a patient feels "very much better" than baseline and a score of seven indicates a patient feels "very much worse" than baseline. In this case, patients are asked to compare how their health is at the present compared with when they first began participating in the study (baseline). A score of four means that a patient feels "unchanged" from baseline and is also used in this article as the referent score to which changes are compared. Therefore, the CGI-I is actually a change score that assesses the patient’s improvement or worsening from baseline (50).
Statistical Analysis
Descriptive analyses on baseline demographic and other patient characteristics (e.g., age and gender) were performed. t tests and chi-squared tests were used to validate the randomization process and to test for comparability between the intervention and control arms.
Differences in CGI-I measures between the control and intervention groups were assessed at months 2, 3, and 6 using SAS PROC MIXED to model the repeated measures over the three time periods in the two treatment arms. Patient-level slopes and intercepts were modeled as random effects. A random-effects covariance structure using intercepts and slopes was used (51,52). This model controls for the correlated outcome measures (i.e., within-person correlation of responses) and allows each person to contribute different numbers of follow-up responses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
Table 1 presents baseline characteristics of study subjects by treatment status. The mean age was 60 (SD = 10), and 67% of the patients were male. Baseline patient characteristics such as age, cardiac condition, gender, length of stay, major depression diagnosis, and race were balanced between the two study groups. (It is important to note that patients were included in this study based on HADS scores, not based on the presence of a major depression diagnosis.) Therefore, these variables dropped out of the analysis. No baseline CGI-I score is available because CGI-I is a change score. However, Figure 2 highlights the changes in CGI-I score from an assumed CGI-I baseline score of 4.0. A regression line was used to illustrate the changes in CGI-I score at each follow-up time between month 2 and month 6.
|
|
The mixed-effects analysis (shown in Table 2 and Fig. 2) confirmed that patients in the intervention group had significantly better improvements in SRH over time (p = .01) relative to the control group. The average number of counseling sessions administered was four (SD = 2). There was no dose–response relationship between number of telephone sessions and SRH outcome. There was no significant difference in outcome based on type of clinician administering counseling.
|
Table 3 outlines changes in mean CGI-I scores in the two groups over time. Mean CGI-I scores at month 2 indicate that patients in the control group felt "unchanged" to "a little better" (with a mean score of 3.33), whereas intervention patients felt "a little better" to "much better" (with a mean score of 2.64, p < .01). This effect remains significant at month 3, when the mean intervention score is 3.25 and the mean control score is 2.44 (p = .01). Table 4 shows the proportion of "CGI-I responders" over time, indicating whether patients are feeling "much better" or "very much better" as compared with baseline (which corresponds to a CGI-I score of two or less). Patients in the intervention group had significantly more CGI-I responders by month 2 (54%) as compared with the control patients (27%, p = .02). Although by month 6 there were no significant differences in SRH improvement reported by the two study arms, intervention patients still reported higher absolute SRH improvements than controls.
|
|
Although not a focus of this paper, patient HADS scores for both the depression and anxiety subscales decreased over time, with significantly greater decreases in the intervention patients compared with control patients.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Recent clinical trials of cardiac patients such as the the Sertraline Antidepressant Heart Attack Randomized Trial (SADHART) and the Enhancing Recovery in Heart Disease Patients (ENRICHD) trials showed that depression treatment is effective in cardiac patients (49,54). Although the ENRICHD study did not find that treatment of depression in cardiac patients decreased cardiac-related mortality, potential design flaws in this study may have limited its ability to detect a beneficial effect of depression treatment on mortality. Other evidence suggests that "psychosocial interventions may be associated with benefits to patients with CAD over and above those achieved by medication and exercise—both in terms of improved quality of life as well as reduced mortality" (43).
Two recently published reports of telephone-based depression medication management programs demonstrated the effectiveness of telemedicine modalities of care in improving patient outcomes (55,56). The first of these studies, by Hunkeler et al (55), showed that telephone calls made by primary care nurses to depressed patients being treated with antidepressants in a primary care practice decreased depressive symptoms and increased patient functioning and satisfaction with care relative to patients who did not receive such calls. This study also demonstrated that this successful intervention did not require staffing changes and was thus feasible in a primary care population.
The second study, by Simon et al (56) demonstrated that telephonic antidepressant management in primary care populations improves depression outcomes. It also suggested that benefits of telephone treatment strategies are cost-effective and efficient because they decrease travel time for patients, increase access to treatment for patients with restricted mobility, and cost less for patients than office visits.
The present study extends the findings from these studies by demonstrating that a telephone-based psychosocial intervention can also be a successful strategy to improve patient SRH and level of mental distress (rather than telephone management that focuses principally on managing antidepressant use). Linden et al documented that psychosocial interventions decrease psychological distress, heart rate, cholesterol levels, and systolic blood pressure. Their work validated the benefits of psychosocial interventions in cardiac patients (43). We argue that telephone counseling interventions promise to be a more efficient, effective, and financially feasible means to decrease symptoms of mood disorders and improve the functioning of chronically ill patients than traditional, in-person, in-office counseling.
Furthermore, despite its importance as a predictor of morbidity and mortality, SRH has typically been assessed only in cross-sectional studies (3,9,57,58). Thus, this study is unique because it was a randomized, controlled trial that examined changes in SRH over time among patients identified as experiencing psychological distress persisting a few weeks after hospital discharge for life-threatening acute coronary syndrome. The CGI-I was used in this study as a measure of SRH because often, one global question of SRH is a reliable and valid measure of health status (4–6,59,60). A single item may better reflect patient health status and potential mortality than underlying medical conditions and risk factors (61). This type of single measure is quick and easy to administer, and it is especially useful when other measures of health outcomes are lacking (59).
However, although this study demonstrated the benefits of the intervention for distressed patients with ACS, a few limitations should be noted. This study has a relatively small sample size. However, a highly statistically significant difference (p = .02) in SRH improvement was detected by 2 months postbaseline.
It is worth noting that control patients also continued to improve over time such that at 6 months postbaseline, no significant differences were observed in the two study groups. Several mechanisms could explain this finding. First, patients adjust to chronic illness over time. Second, the changes could be the result of regression to the mean (62). When patients were first assessed 1 month after hospital discharge for cardiac illness, their experience of SRH may have been much more negative than it would have been under normal circumstances. As patients’ health improved, their comparison of their SRH to when they first were hospitalized would, on average, tend toward improvement over time. Furthermore, the intervention patients experienced more rapid and lasting improvements in SRH such that by the end of the study (6 months postbaseline), control patients never reached the level of CGI-I improvement that intervention patients reached by month two (control patients reached an average CGI-I score of 3 by month six, whereas the intervention patients achieved an average CGI-I score of 2.64 by month two).
Another concern is the fact that we used only one of the three subparts of the CGI measure, the CGI-I subpart, which addresses patient perception of health improvement, without including its Severity of Illness and Efficacy Index subscales. However, as we discussed earlier, a single global measure of health status has been repeatedly shown to be a valid, reliable, and effective means of measuring patient morbidity, morality, and quality of life (1,2,4,8,9).
An additional concern is that we did not have access to patient clinical outcome data such as heart-related surgeries and procedures, medications, rehospitalizations, and so on. However, because baseline patient demographic characteristics and cardiac conditions were balanced in the two study groups, clinical outcomes should have also been balanced in the two intervention groups.
Finally, it is uncertain to what extent this study is generalizable to patients with other mental and physical disorders. Although this is a concern in any clinical trial, it is reassuring to note that the results from this study are comparable to those of other studies, suggesting that treatment of mental and physical health disorders can improve patient perception of health status.
Despite any limitations, it is clear that the intervention had a significant effect on patient SRH among distressed patients hospitalized for cardiac disease. Potential convenience for both patients and providers, as well as the effectiveness of the intervention for patients who experienced a cardiac-related hospitalization, indicate that this treatment might be an effective add-on to disease management programs.
We acknowledge Leslie Griffin, Michael Johnstone, Elaine Polishuk, Stephen Soumerai, and Katherine Swartz for helpful contributions to an earlier draft of this paper.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Received for publication September 21, 2004; revision received January 26, 2005.
DOI:10.1097/01.psy.0000171810.37958.61
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  W. Linden, M. J. Phillips, and J. Leclerc Psychological treatment of cardiac patients: a meta-analysis Eur. Heart J., December 2, 2007; 28(24): 2972 - 2984. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
