| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Johns Hopkins University School of Medicine, Johns Hopkins Bayview Medical Center, Baltimore, Maryland.
Address correspondence and reprint requests to Roy C. Ziegelstein, MD, Department of Medicine, Division of Cardiology, Johns Hopkins Bayview Medical Center, Room 108B, 4940 Eastern Avenue, Baltimore, MD 21224-2780. E-mail: rziegel{at}jhmi.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Patients admitted with AMI underwent screening using the Beck Depression Inventory (BDI) administered by a research assistant. The cardiovascular nurse, medicine resident or intern, and attending cardiologist caring for the patient were then approached (blinded to the BDI results) and asked to assess, using a visual analog scale, whether the patient had symptoms that would warrant further evaluation for depression.
Results: BDI screening and at least one provider assessment were completed for 60 patients with AMI. A total of 18 of 60 patients (30.0%) had a BDI score of 
10. Symptoms of depression were considered not present in 24 of 32 patient assessments when the BDI was 10 (75% false-negatives). The mean BDI score of patients assessed as depressed by at least one provider (6.7 ± 6.3) was no different from the mean BDI score of patients assessed as not depressed (7.5 ± 7.2, p = .67). Overall, there was little correlation between BDI scores and provider assessments, and this was not influenced by provider type or provider gender.
Conclusions: Cardiovascular nurses and medicine residents and interns underrecognize depression in patients with AMI in the absence of formal screening. Formal screening for symptoms of depression should be considered part of routine AMI care.
Key Words: acute myocardial infarction • depression • Beck Depression Inventory • cardiovascular nurse • visual analog scale
Abbreviations: AMI = acute myocardial infarction; BDI = Beck Depression Inventory; CPK = creatine phosphokinase; D-VAS = depression–visual analog scale; DIS = Diagnostic Interview Schedule.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Depression may go unrecognized in patients hospitalized for AMI unless specifically assessed as part of usual care. Healthcare workers, patients, and families may feel that depressed mood is expected after AMI and therefore the condition may not be formally diagnosed or treated. Somatic symptoms that may be manifestations of depression such as fatigue, loss of appetite, sleep disturbance, and impaired concentration may not be recognized as being related to a mood disorder, because they are also seen in patients with heart disease who are not depressed or as a consequence of the hospitalization itself. Some of these symptoms may also occur as side effects of medications used in the treatment of AMI.
Depression may also go unrecognized in patients with AMI because those responsible for the care of the patient with AMI are typically cardiovascular physicians and nurses without expertise in, or a focus on, mental illness. Assessing for depression may be perceived as being too time-consuming during a hospitalization that typically lasts only a few days. The median hospital length of stay for patients in the United States admitted for an AMI was already less than 1 week almost a decade ago (6).
The significant adverse effects of comorbid depression on health outcomes in patients with AMI have led to a call for inhospital identification and treatment of patients with depression after AMI (7). Although formal screening for depression is not now part of routine AMI care, robust evidence on the relationship between depression and health outcomes after AMI (1–4) underscores the importance of identifying patients with AMI with depression or recognizing when a patient’s depressive symptoms might warrant further evaluation. The following study was performed to assess the accuracy of cardiovascular healthcare workers’ assessments of depression or significant depressive symptoms during usual care of patients with AMI in a university-based teaching hospital.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
240 mg/dL) on admission to the hospital. Patients were considered to have a history of coronary artery disease if they had a previously documented MI, a typical history of angina with a positive exercise test result, or angiographically documented stenosis of 50% or more. After informed consent was obtained, patients were interviewed by a research assistant 2 to 5 days after AMI using the Beck Depression Inventory (BDI) to assess for symptoms of depression. The BDI is a self-report instrument that has been used in the early post-MI period (1,2), with BDI scores of 10 or greater indicating at least mild depressive symptoms (8). The BDI was chosen as the screening instrument in this study because it is relatively easy to use and to score. Although the BDI is not by itself used to diagnose major depression, it has been used by many groups who have studied depression in patients hospitalized for AMI (1,2,4,9,10). The relationship between elevated BDI scores and post-MI mortality has also been demonstrated by several groups (2,4).
The patient’s attending cardiologist, internal medicine resident or intern, and primary nurse were asked to estimate their time in direct contact with the patient and to indicate whether they felt the patient was depressed based on their interactions with the patient. Because the goal of the study was to evaluate the ability of cardiovascular healthcare workers to recognize depression during usual care, these practitioners did not receive any specific training about the recognition or treatment of depression or mental illness as part of this study. Assessment of depression by interview or screening instrument was not specifically prohibited, although it is not typically part of the usual care of patients with AMI at our facility. Providers were blind to the results of the BDI. The cardiovascular healthcare workers were asked to indicate their assessment of their patient’s depressive symptoms by marking two lines on a depression–visual analog scale (D-VAS). They were instructed to mark a horizontal line to indicate the cutoff above which they felt that symptoms would warrant further evaluation for depression (i.e., not necessarily treatment for depression); this line is referred to as the "further evaluation line." They were then asked to mark a second horizontal line to indicate their rating of the patient’s symptoms of depression ("patient line"). The D-VAS evaluation score was transformed by dividing the score corresponding to the "patient line" by the score corresponding to the "further evaluation line." A value 1 indicated that the provider assessed the patient as having symptoms of depression warranting further evaluation.
Statistical Analysis
Comparisons between the patient’s BDI score and the provider’s assessment were made by dichotomizing the BDI total score at 10 (with 10 and above indicative of at least mild depressive symptoms) and the D-VAS score at 1. Dichotomized variables were compared by chi-squared analyses and continuous variables by t tests. The relationship of D-VAS score to the BDI score was analyzed by Pearson product-moment correlations. All values are presented as mean ± standard deviation, and a p value < .05 was considered significant. Agreement between raters was evaluated with the Cohen’s kappa statistic. Cohen’s kappa may be considered a measure of chance-corrected agreement (11). It can be used to test the proposition that agreement between two raters is better than what would be expected by chance. Kappa equals zero when the agreement equals that expected under independence and one under the condition of perfect agreement. With disagreement beyond what would be expected by chance, kappa can take on negative values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
10, or length of hospital stay. As shown in Table 1, the mean age of the patients was 66.5 ± 12.8 years and 40% were women. There was a high prevalence of all cardiovascular risk factors, including hypertension and hyperlipidemia (both 78%) and diabetes mellitus (50%). A total of 18 of the 60 patients (30.0%) had a BDI score of 10. The average BDI score of those patients with a BDI of less than 10 was 3.5 ± 3.0 and the average BDI score of those patients with a BDI 10 was 15.7 ± 5.2 (p < .01). BDI scores ranged from 0 to 29. A greater proportion of patients whose BDI was 10 were women compared with those whose BDI was <10 (61.1% versus 31.0%, p = .03). There were no other significant differences between BDI groups.
|
Provider Assessments
Of a total of 107 provider assessments, 49 were made by the primary cardiovascular nurse, 38 by the internal medicine resident, and 20 by the attending cardiologist. Twenty-five patients were assessed by only one provider, 23 by two providers, and 12 by all three providers. Nurses reported spending almost twice as much time with patients as did internal medicine residents and more than twice as much time as did attending cardiologists (383 ± 43 minutes for nurses versus 211 ± 49 minutes for residents versus 164 ± 68 minutes for attending cardiologists, p < .01). At least one provider assessed 26 of the 60 patients (43%) as having symptoms of depression that warrant further evaluation. Overall, providers identified potentially significant symptoms of depression as being present in 31 of 107 assessments (29%). There was considerable disagreement across providers, however. When comparing common assessments, agreement between nurses and internal medicine residents and between nurses and attending cardiologists was not significantly better than what would be expected by chance (kappas –0.15, p = .40 and –0.11, p = .68, respectively). Agreement between attending cardiologists and internal medicine residents was better than what would be expected by chance (kappa 0.58, p = .03).
Accuracy of Provider Assessments Compared With the Beck Depression Inventory
The mean BDI score of the 26 patients assessed by at least one cardiovascular healthcare provider as having symptoms that warrant further evaluation for depression (6.7 ± 6.3) was no different from the mean BDI score of the 34 patients assessed as not warranting further evaluation (7.5 ± 7.2, p = .67). Providers felt that symptoms of depression that warrant further evaluation were present in 23 of 75 patient assessments when the BDI score was <10 (30.7% false-positives). Symptoms of depression warranting further evaluation were considered not present by the providers in 24 of 32 patient assessments when the BDI score was 10 (75% false-negatives). The sensitivity of the provider assessments, using a BDI 10 as the standard, was 17.6% for cardiovascular nurses (3 of 17) and 33.3% for both the internal medicine residents (4 of 12) and attending cardiologists (1 of 3). The specificity for the assessments of cardiovascular nurses, internal medicine residents, and attending cardiologists was 62.5% (20 of 32), 73.1% (19 of 26), and 76.5% (13 of 17), respectively. Provider assessment scores were not significantly correlated with BDI total scores (items 1–21, r = 0.06, p = .54) or with either the cognitive/affective (items 1–14, r = 0.07, p = .50) or somatic subscales of the BDI (items 15–21, r = 0.04, p = .71).
Relationship Between Accuracy of Provider Assessments and Provider Type, Gender, or Self-Reported Time Spent With the Patient
There was no significant correlation between BDI scores and provider assessments of patients’ depressive symptoms. As shown in Figure 1, the correlation between continuous BDI scores and D-VAS ratios was 0.04 for cardiovascular nurses, 0.08 for internal medicine residents, and 0.30 for attending cardiologists (p = .78, 0.61, and 0.20, respectively). Accuracy did not differ by provider gender and there was no correlation between self-reported time spent with the patient and BDI score (not shown). The accuracy of physicians’ assessments compared with nurses’ assessments also did not differ by patient gender (not shown). The effect of gender was assessed because it has been reported that female providers engage in more psychosocial question asking (12) and may consider depression more often (13) than their male counterparts.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
10 has been associated with an increased risk of mortality in patients with AMI (2,4). Providers were asked to identify patients they felt warranted further evaluation for depression. This was based on the recommendation that patients with coronary artery disease who have significant symptoms of depression on a screening instrument should be further assessed using a structured interview (14). In our study, false-positives (i.e., patient considered by the provider to have symptoms that warrant further evaluation for depression, but BDI <10) occurred in 30.7% of assessments and false-negatives (i.e., patient not considered depressed by provider, but BDI 10) occurred in 75% of provider assessments. The sensitivity of providers’ assessments across provider groups ranged from 17.6% to 33.3% and the specificity ranged from 62.5% to 76.5%.
This study confirms that the prevalence of at least mild depressive symptoms is common in patients hospitalized with AMI. The prevalence of a BDI score 10 of 30% in this study is very similar to that reported by other groups (2,9). Although the prevalence of depressive symptoms warranting further evaluation as assessed by the provider (i.e., by D-VAS) was virtually the same (29% of assessments) as the prevalence of a BDI score 10 (30%), there was little correlation between the group of patients with a BDI 10 and the group of patients assessed by cardiovascular healthcare providers as meriting additional evaluation for depression. Indeed, the mean BDI scores of the patients assessed as depressed by providers was essentially the same as that of the patients assessed as not being depressed by provider assessments. In this study, the accuracy of provider assessments in the absence of specific screening (compared with the BDI) was not affected by provider type (i.e., nurse, resident, attending cardiologist), provider gender, or self-reported time spent with the patient. In addition, the relationship between provider assessments and the BDI did not improve when BDI cognitive/affective items and somatic items were analyzed separately.
Underrecognition of depression is a well-described problem in primary care (15–17). Although the extent of underrecognition of depression has not previously been reported in patients hospitalized for AMI, clinical experience and indirect evidence suggest that underrecognition is a problem in this setting. For example, in a study by Frasure-Smith et al. (1), of 35 patients hospitalized for AMI with depression based on the results of the Diagnostic Interview Schedule (DIS), only three (8.6%) were prescribed antidepressant medications. By contrast, the recognition of other cardiovascular risk factors like hypertension and cigarette smoking appears to be considerably better than the recognition of depression. A review of the medical records of 212 patients with a principal diagnosis of AMI who were discharged from our medical center between January 1, 2004, and October 31, 2004, showed that 150 (70.8%) had a secondary diagnosis of hypertension documented, a prevalence very similar to that reported in the recent National Registry of Myocardial Infarction 4 (18). This suggests that underrecognition of hypertension is very unlikely in the same setting in which underrecognition of depression appears quite prevalent. Review of these medical records also indicated that 97.9% of patients with AMI during this time period were screened for smoking history.
Failure to diagnose and treat depression may result in a missed opportunity to improve patients’ quality of life or to decrease their physical limitations. In addition, because patients with depression are less likely to follow recommendations intended to reduce cardiovascular risk (10), failure to identify depression during hospitalization for AMI may result in a lost opportunity to use compliance-enhancing interventions in these high-risk patients. Participation and retention in cardiac rehabilitation programs after hospital discharge is adversely influenced by comorbid depression (19–22), and failure to diagnose and treat depression soon after AMI may therefore decrease the likelihood that a patient will attend and benefit from a rehabilitation program.
This study has several important limitations. Although the BDI has several advantages as a screening instrument, it is not by itself a diagnostic tool. It is possible that the conclusions of this study would be different if the recognition of depression by cardiovascular healthcare providers had been compared with the results of a structured interview by a mental health provider using the diagnostic criteria for major depression, dysthymia, or bipolar disorder. It is also important to note that underrecognition of depression did not appear to be as much of a problem for attending cardiologists as for cardiovascular nurses or medicine residents or interns. In only two patients with a BDI 10 did an attending cardiologist fail to assess a patient as exhibiting symptoms that warrant further evaluation for depression. Before routine screening for depression in the care of patients with AMI can be recommended with the goal of improving outcome, studies should demonstrate that early identification and treatment of depression after AMI improves prognosis. To date, such studies are lacking and two major studies in this area have failed to show that the treatment of depression improves health outcomes in this setting (23,24). On the other hand, it seems reasonable to identify depression in patients with an AMI, because treating depression improves physical symptoms and quality of life in medical patients (25). This is particularly important, because depression may have an even greater effect on quality of life in patients with coronary disease than traditionally determined measures of cardiac function (5).
In summary, at least mild depressive symptoms are prevalent in patients with AMI. Depression is known to be associated with poor quality of life and high symptom burden in patients with cardiovascular disease (5) and with an increased risk of morbidity and mortality after AMI (1–4). Given these factors and the poor ability of cardiovascular healthcare providers to recognize depression in patients hospitalized with AMI, it is not surprising that the most recent American College of Cardiology/American Heart Association Task Force on Practice Guidelines for the treatment of AMI patients recommends that "the psychosocial status of the patient should be evaluated, including inquiries regarding symptoms of depression, anxiety, or sleep disorders and the social support environment" (26). Although the best manner to screen for depression in this setting remains to be defined, the findings of the present study suggest that standardized screening tests for depression may be a reasonable way to accomplish the new recommendations to evaluate the psychosocial status of patients with AMI.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This work was supported in part by a grant from the Charles A. Dana Foundation and in part by the Miller Family Scholar Program.
DOI:10.1097/01.psy.0000160475.38930.8d
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. A. Cully, M. Johnson, M. L. Moffett, M. Khan, and A. Deswal Depression and Anxiety in Ambulatory Patients With Heart Failure Psychosomatics, November 1, 2009; 50(6): 592 - 598. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. G. Smolderen, J. A. Spertus, K. J. Reid, D. M. Buchanan, H. M. Krumholz, J. Denollet, V. Vaccarino, and P. S. Chan The Association of Cognitive and Somatic Depressive Symptoms With Depression Recognition and Outcomes After Myocardial Infarction Circ Cardiovasc Qual Outcomes, July 1, 2009; 2(4): 328 - 337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Rutledge, V. Vaccarino, B. D. Johnson, V. Bittner, M. B. Olson, S. E. Linke, C. E. Cornell, W. Eteiba, D. S. Sheps, J. Francis, et al. Depression and cardiovascular health care costs among women with suspected myocardial ischemia: prospective results from the WISE (Women's Ischemia Syndrome Evaluation) Study. J. Am. Coll. Cardiol., January 13, 2009; 53(2): 176 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. POZUELO, G. TESAR, J. ZHANG, M. PENN, K. FRANCO, and W. JIANG Depression and heart disease: What do we know, and where are we headed? Cleveland Clinic Journal of Medicine, January 1, 2009; 76(1): 59 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Post-Myocardial Infarction Depression Clinical Pra AAFP Guideline for the Detection and Management of Post-Myocardial Infarction Depression Ann. Fam. Med, January 1, 2009; 7(1): 71 - 79. [Full Text] [PDF]
|
|
|
|
|
|
M. W. Ketterer, W. Knysz, S.J. Keteyian, J. Schairer, S. Jafri, M. Alam, A.J. Farha, and S. Deveshwar Cardiovascular Symptoms in Coronary-Artery Disease Patients Are Strongly Correlated With Emotional Distress Psychosomatics, May 1, 2008; 49(3): 230 - 234. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C Huffman Significant and increasing depressive symptoms are associated with increased risk of new cardiac events in people with a heart attack in the previous year Evid. Based Ment. Health, May 1, 2007; 10(2): 42 - 42. [Full Text] [PDF]
|
|
|
|
 |
|
 D. K. Moser, B. Riegel, S. McKinley, L. V. Doering, K. An, and S. Sheahan Impact of Anxiety and Perceived Control on In-Hospital Complications After Acute Myocardial Infarction Psychosom Med, January 1, 2007; 69(1): 10 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Ziegelstein and S. D. Miller Just a Spoonful of Sugar J. Am. Coll. Cardiol., December 5, 2006; 48(11): 2223 - 2224. [Full Text] [PDF]
|
|
|
|
 |
|
 J C Huffman, F A Smith, M A Blais, M E Beiser, J L Januzzi, and G L Fricchione Rapid screening for major depression in post-myocardial infarction patients: an investigation using Beck Depression Inventory II items Heart, November 1, 2006; 92(11): 1656 - 1660. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D C Haas Depression and disability in coronary patients: time to focus on quality of life as an end point Heart, January 1, 2006; 92(1): 8 - 10. [Full Text] [PDF]
|
|
|
|
 |
|
 Other articles noted Evid. Based Med., December 1, 2005; 10(6): 191 - 192. [Full Text] [PDF]
|
|
|
|
 |
|
 Minerva BMJ, June 4, 2005; 330(7503): 1338 - 1338. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
