Secondary Logo

Journal Logo

Scientific Review

A Systematic Review of the Diagnostic Accuracy of Depression Questionnaires for Cardiac Populations


González-Roz, Alba MA; Gaalema, Diann E. PhD; Pericot-Valverde, Irene PhD; Elliott, Rebecca J. BS; Ades, Philip A. MD

Author Information
Journal of Cardiopulmonary Rehabilitation and Prevention: November 2019 - Volume 39 - Issue 6 - p 354-364
doi: 10.1097/HCR.0000000000000408
  • Free

Depression in patients with cardiovascular disease (CVD) is associated with poorer prognosis and recovery, greater cardiac rehospitalizations, and lower participation in cardiac rehabilitation (CR) programs.1–3 Between 20% and 45% of CVD patients have either elevated depressive symptoms or a diagnosis of depression,4,5 with rates being higher in women relative to men.6,7

Given these profound health-related concerns, consensus has been reached in favor of screening for depression in patients with CVD and monitoring its effect on adherence to CR and its response to CR participation.8 Even subclinical depression involves increased risk of morbidity and mortality after cardiac events9; thus, identifying and treating patients with either low or clinical depression are essential. Despite the importance of screening for depression and given that CR settings offer a convenient opportunity for mental health screening, only 29% to 68% of CR programs evaluate depression,10 resulting in <15% of patients being routinely screened.11,12 Consequently, depression is underdiagnosed and undertreated in this population.13

Despite the many diagnostic tools available for detecting depression diagnosis, they are underutilized in the clinical setting. This is likely due to time constraints, limited knowledge of depression assessment, or lack of resources to treat or refer a high volume of patients, if screening is positive.14 However, there are many depression-screening instruments available, offering health professionals a time- and cost-efficient way of detecting depressive symptoms and identifying patients who might benefit from further assessment and treatment.

For the past several years, reviews on depression questionnaires for the cardiac population have contributed to the recognition of the importance of the assessment, referral, and treatment of depression.10,12,15 Italian, European, and American guidelines for CR have provided a comprehensive descriptive overview on the most commonly used depression questionnaires within the CR context. Of note is that these have yielded valuable information on which of them is more suitable in detecting depression in either hospitalized populations or outpatients. These prior studies have yielded practical recommendations for health providers and prompted calls for both examining the effect of depression screening on improvements in depression and the effect of CR on amelioration of depression. Despite these strengths, a limited number of assessment tools have been reviewed. In addition, those reviews did not provide an in-depth analysis of depression questionnaires specifically validated in cardiac populations. Considering that available depression measures are not always appropriate for cardiac patients, it is important to address this limitation. Furthermore, although valuable, these previous efforts did not provide evidence on the recommended cut scores for specific cardiac populations with different diagnoses (eg, stable angina, myocardial infarction [MI]). Given that previous studies have reported inadequate sensitivity and specificity of questionnaires when using cut scores obtained in noncardiac populations,16–18 reviewing evidence on depression questionnaires with good psychometric properties for this specific population is warranted.

Another important aspect of assessment tools is the capability to detect clinical changes over time brought on by an intervention (ie, internal responsiveness of the instrument).19 Monitoring depressive symptoms is relevant from both a research and clinical standpoint.20 Clinically, both the recurrent nature of depression and the health impacts of even subclinical levels of depression call for frequent assessment. This issue has not been addressed before and evidence on which screening instruments better capture changes in depression symptoms remains lacking.

The purpose of this systematic review was to (1) describe depression questionnaires for the cardiac population; (2) examine strengths and weaknesses; and (3) analyze existing evidence regarding accuracy in detecting depression and sensitivity to depression changes in the CR context. Because of the high prevalence of depression in CR patients and the increased interest in regular screening of CR patients, it is expected that this study will help guide CR professionals on both selection and use of depression questionnaires.



The literature search was carried out including articles up to May 2018, with no restriction on year of publication (see Figure). The literature search included the following databases: MEDLINE, PsycInfo, and PubMed. A combination of the following medical subject heading (MeSH) terms and Boolean connectors were used: “mood disorders” or “depression” or “depressive symptoms” or “depression diagnosis” or “depressed”, and “depression screener” or “depression questionnaire,” and “cardiac patients” or “cardiac rehabilitation.” We qualitatively reviewed study findings in the context of questionnaire characteristics, including questions, cut scores, initial validations, appropriate populations, and general concerns. Accuracy on detecting depression as well as responsiveness to changes in depressive symptoms was also examined. This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement21 and the study protocol was registered in the International Prospero Register of Systematic Reviews, PROSPERO (ID: CRD42017077131).

PRISMA (adapted) flowchart of the review process.


To minimize biases due to language or differences in CR program structure, the literature search was restricted to studies conducted in English-speaking North American countries (ie, United States and Canada). Only those studies comparing the performance of depression questionnaires versus a clinical diagnostic interview for detecting depression were included. Although we did not restrict the literature search to a specific set of depression questionnaires, case studies, abstracts, study protocols, and studies using tools for depression assessment that were not validated or were considered unreliable (eg, use of single questions) were excluded. When considering the objective of assessing the internal responsiveness of depression questionnaires (sensitivity to changes), studies including nontraditional or noncomprehensive CR were excluded from this review.


We reported the available depression screening instruments for cardiac patients and their recommended cut scores specific to a cardiac diagnosis. We provided data on its accuracy in detecting depression and capturing depression changes (internal responsiveness to change). Sensitivity (adequacy of questionnaires for correctly identifying depression cases), specificity (adequacy of questionnaires for correctly identifying cases with no depression), positive predictive values (PPV) defined as the percentage of positive cases with patients who are depressed, and negative predictive values (NPV) defined as the percentage of negative cases with patients who are not depressed were examined as a measure of accuracy in identifying positive cases.

Although there are no standard cut scores for considering screening instruments as having excellent or poor diagnostic accuracy properties, current guidelines for interpreting the diagnostic accuracy of depression questionnaires advocate using those with high sensitivity and NPV.22 Thus, ideally, a depression screening will have 100% sensitivity and NPV so that positive cases are not missed.

Screening questionnaires were compared in terms of their capability to detect changes in depression after completing CR (internal responsiveness to change). For that purpose, we calculated standardized effect sizes. This is a common and valid approach for estimating sensitivity to change19 and thus, we computed them by using the following formula: M2M1/SD1 (M2, mean at time 2; M1, mean at time 1; SD1, standard deviation at time 1).23 Effect size coefficients were interpreted on the basis of benchmarks proposed by Cohen.24 Coefficients ≤0.2 and in the range of 0.21 to 0.49 indicate a small effect; coefficients between 0.50 and 0.79 indicate a medium effect; and coefficients ≥0.8 indicate a large size effect.


Two independent reviewers assessed the methodological quality of the included studies focusing on diagnostic accuracy. The revised version of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2)25 was used as it is appropriate for assessing the quality of diagnostic accuracy studies. It contains 4 domains (patient selection, index test, reference standard, and flow and timing) that are rated in terms of risk of bias and applicability. Intraclass correlations were calculated to determine the interrater agreement between reviewers in each of the domains. Values lower than 0.50 indicate poor reliability, between 0.50 and 0.75 suggest moderate reliability, between 0.75 and 0.90 suggest good reliability, and values >0.90 reveal excellent reliability.26


Of the 8943 articles identified through the literature search, 94 were individually assessed against eligibility criteria after examining title and abstract information. This procedure led to 23 studies meeting the criteria and thus being included in this review (Figure).


We identified 11 articles that assessed the performance of depression questionnaires in accurately detecting depression among cardiac patients (Table 1). Participants in these studies were patients with a variety of cardiac diagnoses (eg, MI, coronary artery disease [CAD]). Questionnaires with the greatest validation in the CVD population were the Beck Depression Inventory-II (BDI-II) and the Patient Health Questionnaire (PHQ-9/PHQ-2). As shown in Table 1, studies differed in the cut scores used to screen for depression in the cardiac population. Sensitivity (ranging between 39% and 100%) and specificity (ranging between 54% and 94%) values indicated adequacy of available depression questionnaires for detecting depression cases in the cardiac population. Overall, PPV and NPV ranged between 12% and 63% and 87% and 100%, respectively.

Table 1 - Accuracy of Depression Questionnaires in Detecting Depression Diagnosis Among Cardiac Patients
Author (yr) Depression Screener Age, yr Sex (% Males) Population Cut Scores Sensitivity, % (95% CI) Specificity, % (95% CI) PPV, % (95% CI) NPV, % (95% CI) Criterion Standard for Depression Diagnosis
Bambauer et al (2005)27 HADS-D-14 items 60 67 MI or IHD
Aged ≥60 y
≥7 81 54 PRIME-MD (DSM-IV)
Elderon et al (2011)28 PHQ-2/9 n = 602a 82 CHD patients PHQ-2 (Y/N) ≥1
PHQ-2 (MC) ≥2
PHQ-2 (Y/N) +
PHQ-9 ≥1 +≥10
PHQ-2 (MC) +
PHQ-9 ≥2 + ≥10
PHQ-9 ≥10
90 (86-94)
82 (77-87)
52 (46-59)
50 (44-57)
54 (47-61)
69 (66-73)
79 (76-82)
91 (89-93)
92 (90-94)
90 (88-92)
45 (40-50)
52 (47-57)
63 (56-70)
63 (56-70)
61 (54-68)
96 (95-98)
94 (92-96)
87 (85-90)
87 (85-89)
88 (85-90)
Frasure-Smith and Lesperance (2008)29 BDI-II 60 80.7 CAD patients ≥14 91 78 SCID-I (DSM-IV)
Huffman et al (2010)30 BDI-II
62 80 Post-MI patients ≥16
Low and Hubley (2007)31 BDI-II
GDS (30-items)
63 75 MI or UA patients Major depression ≥10
Major depression ≥14
McGuire et al (2013)32 PHQ-2 (MC)
PHQ-9 (MC)
PHQ-10 (MC)b
61.06 69 ACS >0

McManus et al (2005)33 PHQ-2 (MC)
PHQ-9 (MC)
67 82 CHD patients ≥3
39 (30-47)
52 (44-61)
77 (69-84)
92 (89-94)
90 (88-93)
79 (75-82)

The Diagnostic Interview Schedule (DSM-III)
Moullec et al (2015)34 BDI-II 58 69 Cardiac
Total sample ≥10
Women ≥13
Men ≥10
<12 yr of education ≥11
≥12 yr of education ≥10
Comorbid AD ≥15
Noncomorbid AD ≥10
CAD ≥11
Non-CAD ≥10
Razykov et al (2012)35 PHQ-9 (MC)
PHQ-8 (MC)c
50-79 (82%) 82 CAD 10 54 (47-60)
50 (44-57)
90 (88-92)
91 (89-93)
61 (54-67)
60 (53-67)
88 (85-90)
87 (84-89)
Swardfager et al (2011)36 CES-D 64 79 CAD patients ≥16 93 78 SCID-I/NP
Thombs et al (2008)18 PHQ-2 (MC)
PHQ-9 (MC)
PHQ-2 (MC)+
PHQ-9 (MC)
n = 522d 82 Outpatients with CAD PHQ-2 (MC) ≥ 2
PHQ-9 (MC) ≥ 6
82 (76-86)
83 (78-87)
75 (69-81)
79 (76-81)
76 (73-79)
84 (81-86)
52 (46-57)
50 (45-55)
57 (51-62)
94 (92-95)
94 (92-96)
92 (90-94)
Abbreviations: ACS, acute coronary syndrome; AD, comorbid anxiety disorders; BDI-II, Beck Depression Inventory, II; BDI-II-cog, Beck Depression Inventory cognitive subscale; CAD, coronary artery disease; C-DIS, Computerized Diagnostic Interview Schedule; CES-D, The Center for Epidemiologic Studies Depression Scale; CHD, coronary heart disease; DISH, Depression Interview and Structured Hamilton; DSM, Diagnostic and Statistical Manual of Mental Disorders; GDS, Geriatric Depression Scale; HADS-D, depression subscale of the Hospital Anxiety and Depression Scale; IHD, ischemic heart disease; MI, myocardial infarction; NPV, negative predictive value; PHQ-2/9, Patient Health Questionnaire, 2 item or 9 item version; PHQ-2 (MC), multiple choice version of the Patient Health Questionnaire-2; PHQ-2(Y/N), yes/no version of the Patient Health Questionnaire; PPV, positive predictive value; PRIME-MD, primary care evaluation of mental disorders; SCID-I, structured clinical interview for DSM-IV Axis I Disorders; SCID-I/NP, Structured Clinical Interview for DSM-IV, Axis I Disorders, nonpatient edition; UA, unstable angina.
aNumber of participants aged 65 yr or older.
bPHQ-10 (MC) version contains the same items as the PHQ-9 but with the addition of an item assessing the extent to which depression interferes with daily living.
cPHQ-8 version contains the same items as the PHQ-9 except for item 9 (ie, suicidal thoughts).
dNumber of participants aged 50 to 69 yr.

Beck Depression Inventory-II

This instrument consists of 21 items assessing depressive symptoms, with higher scores indicating greater severity. It covers cognitive, affective, somatic, and vegetative symptoms and is one of the most commonly used in the CVD population. Despite being originally validated in psychiatric populations, samples of college students, and adolescents, studies with cardiac patients have demonstrated good reliability coefficients (ranging between 0.83 and 0.85),37,38 which are comparable with those using noncardiac samples.39 Overall, using different BDI-II cut scores seems to differentially impact the capability to detect depression cases. Three of 4 studies displayed good psychometric properties when using a cut score between 10 and 14 for patients with CAD (sensitivity ranging from 83% to 100%; NPV ranging from 99% to 100%).29,31,34 Using a cut score of ≥16 in post-MI populations led to correctly detecting 88.2% of the positive cases but incorrectly identifying 1.9% as depressed when they were not (specificity = 92.1%; PPV = 62.5%).30

Regarding the limitations of the BDI-II, some studies have indicated that the somatic symptoms in cardiac patients can interfere with the accurate diagnosis of depression and thus, an alternative 8-item cognitive scale of the BDI-II that excludes somatic items has been proposed.30 Cut scores between 3 and 4 demonstrated accuracy in detecting depression as shown by a high percentage of post-MI depressed patients being accurately detected (sensitivity = 85.3%; NPV = 97.5%). Another important limitation involves its skewed properties, which leads to low accuracy in detecting subclinical to mild depression.40

Patient Health Questionnaire

The PHQ-9/PHQ-2 is a self-report instrument based on the Diagnostic and Statistical Manual of Mental Disorders (Fourth edition) text revision criteria. The PHQ-9 and its shorter version (PHQ-2) were initially validated in medical patients. They examine the 2 core features of depression over the past 2 wk: anhedonia (ie, inability to experience pleasure) and depressive mood (ie, low mood and pathological sadness). Given that the PHQ-2/9 versions showed adequate reliability17,41 and construct validity41,42 for detecting depression in cardiac populations, the American Heart Association has recommended using the 2-item Patient Health Questionnaire (PHQ-2-yes/no version), followed by its 9-item multiple choice (MC) version for further assessment. Although the 9-item validation among noncardiac populations proposed a score of ≥10 as the optimal cut score for detecting depression,43 several studies indicate that this cut score may lead to <55% sensitivity28,33 in cardiac samples suffering from coronary heart disease, meaning that nearly half of the positive cases were not accurately detected.

Sparse evidence was found regarding specific cut scores for specific disease states. Relative to PHQ-2 (MC) cut scores of ≥218,28 and ≥3,33 adopting a PHQ-2 (MC/YN) cut score of ≥1 seems to enhance its sensitivity (90%-95.65% vs 39%-82%) but slightly decreases specificity (69%-71.43% vs 79%-92%) among CAD patients.28,32

Overall, better performance in detecting depression was found when using the PHQ-2 (either the yes/no or the MC version) compared with the PHQ-9. Adopting a PHQ-2 cut score between 1 and 2 resulted in good capability to detect depression in 3 of 4 studies (sensitivity ranging from 82% to 95.65%; NPV ranging from 94% to 96%).18,28,32 When using the PHQ-2, followed by the PHQ-9, between 50% and 75% of the depression cases were correctly identified, while 8% to 13% of the assessed patients were misclassified18,28 (sensitivity = 59%; specificity = 89%). Either deleting or adding 1 item did not lead to a significant improvement in detection rates in CAD patients.32,35 Specifically, adding 1 item (ie, functional impairment) to the PHQ-9 (MC) results in 96.65% sensitivity and 77.92% specificity, whereas removing 1 item (ie, suicide thoughts) results in 50% sensitivity and NPV of 87%.

Hospital Anxiety and Depression Scale

This self-rating tool was originally developed to identify anxiety and depression in medical outpatient populations. It mostly includes anhedonia-related items and, unlike many other depression questionnaires, does not include somatic items. This, however, misses the chance to detect clinically relevant depression symptoms within the cardiac population (eg, difficulty in concentration, difficulty in falling asleep).44

The Hospital Anxiety and Depression Scale (HADS-D) includes an anxiety and a depression subscale, each comprising 7 items with a score of ≥8 indicating depression. Although compelling research has suggested that this instrument be used in cardiac patients (eg, coronary care patients following acute MI),12,27,45 other studies have recommended using other depression questionnaires instead for patients with ischemic heart disease or older chronic heart failure patients due to the limited ability of HADS to distinguish between anxiety and depression, which may overestimate the number of depression cases.46 With regard to optimal cut scores for detecting depression, we found only 1 study examining the HADS-D accuracy properties in detecting depression. Using a cut score of ≥7 led to the highest accuracy in detecting depression in patients aged ≥60 yr with MI or ischemic heart disease (sensitivity = 81%; specificity = 54%).27

Geriatric Depression Scale

The Geriatric Depression Scale (GDS) is a yes/no format scale for screening depression that has been assessed in older patients (ie, aged ≥63 yr).31 The GDS does not include somatic symptoms and is available in 30-item, 15-item, and 5-item versions, with greater scores indicating higher depression severity. Albeit limited, evidence on the 30-item GDS suggests using a cut score of ≥14 in patients with post-MI or post–unstable angina for detecting the presence of depression.31 Adopting cut score leads to high accuracy in detecting depression cases (sensitivity = 100%; NPV = 100%). However, the validity of the GDS with older adults with cognitive impairment has been questioned because of patients with dementia tending not to report depressive symptoms accurately; hence, evidence suggests not using this tool in populations with mild to moderate dementia.47

Center for Epidemiologic Studies Depression Scale

This tool is a 20-item self-report rating scale originally developed to assess depression in the general population. Studies on the Center for Epidemiologic Studies Depression Scale (CES-D) have demonstrated good internal consistency coefficients in the general and clinical populations. For the cardiac population, a 3-factor structure has been proposed as valid and reliable for depression-screening purposes. Nonetheless, the overlap between the negative affect and the somatic symptoms factors alongside with low-factor loading of some items in each of the purposed dimensions has resulted in recommendations toward using other well validated instruments instead.48

Of the 2 studies assessing the diagnostic accuracy properties of the CES-D, a cut score between 10 and 16 has yielded good sensitivity (ranging from 77% to 93%) in CAD patients.33,36


A total of 12 studies examined changes in depression scores over the course of participation in CR (Table 2). The BDI-II showed small to high short-term internal responsiveness to change in patients endorsing a wide set of cardiac diagnoses (eg, coronary artery bypass graft surgery, CAD, MI, percutaneous transluminal coronary angioplasty, heart valve patients, and women with CAD), with Cohen d ranging between 0.10 and 0.96.50,53,55–57,60 Of note is that the only 1 study using an older version of the BDI in outpatients with CAD yielded poor sensitivity to change (r values ranged between 0.0 and 0.26), with no effect shown in the comparison treatment group.58 Although with limited evidence, the CES-D showed small to medium capability to capture changes in depression as indicated by coefficient effect sizes between 0.12 and 0.54.49 The PHQ-9 and the HADS-D showed small responsiveness to change indices (Cohen d effect sizes below 0.44).52,54,59 The only study assessing the Profile of Mood States questionnaire indicated low responsiveness to detect depression changes in MI patients.58 Finally, evidence on the internal responsiveness to change of GDS-15 scale indicated medium capability to detect depressive symptoms changes (0.5-0.66).51 Overall, results on the internal responsiveness to change were sparse with the most evidence available using the BDI-II or the HADS-D for examining pre- and post-treatment depression changes. Both measures generally performed similarly, with effect sizes mostly being between 0.08 and 0.96.

Table 2 - Changes in Depression Scores Over the Course of CR Participation
Author (yr) Depression Screening Instrument Population Treatment Responsiveness to Changea Conclusion
Beckie et al (2011)49 CES-D Women with PCI, CABG surgery, SA, or MI. Traditional CR vs tailored CR (participants exercised specifically with women in their cohort) Traditional group:
Pre-treatment to post-treatment: 16.5 ± 10.3 vs 14.3 ± 10.3; Cohen d (pre-treatment to post-treatment) = −0.21
6-mo follow-up:
15.2 ± 10.8; Cohen d (pre-treatment to 6-mo follow-up) = −0.12
Tailored group:
Pre-treatment to post-treatment: 17.3 ± 11.7 vs 11 ± 9.4; Cohen d (pre-treatment to post-treatment) = −0.54
6-mo follow-up: 13 ± 9.9; Cohen d (pre-treatment to 6-mo follow-up) = −0.37
At short-term follow-up, small to medium coefficient effects sizes. At long-term, small coefficient effect sizes.
Caulin-Glasser et al (2007)50 BDI-II CABG, PTCA, MI/CAD, and patients with HVD/other CR program (exercise training, education, behavior modification therapy) Women:
Pre-treatment to post-treatment (12 wk): 8.7 ± 6.7 vs 5.2 ± 4.7; Cohen d = −0.52
Pre-treatment to post-treatment (12 wk): 7.4 ± 6.3 vs 4.2 ± 4.9; Cohen d = −0.51
At short-term follow-up, medium coefficient effect sizes
Gary et al (2004)51 GDS-15 Patients with DHF Home-based CR program vs control group Home-based CR:
Pre-treatment to 12 wk: 6 ± 4 vs 4 ± 4
3 mo: 4 ± 4
Control group:
Pre-treatment to 12 wk: 5 ± 3 vs 7 ± 5
3-mo: 7 ± 5
Home-based CR:
Cohen d (pre-treatment to 12 wk) = −0.5
Cohen d (pre-treatment to 3 mo) = −0.5
Control group:
Cohen d (pre-treatment to 12 wk) = 0.66
Cohen d (pre-treatment to 3 mo) = 0.66
At both short- and long-term follow-up, medium coefficient effect sizes
Ghisi et al (2017)52 PHQ-9 CAD patients (MI, PCI, CAD, CABG, ConHD, HF) CR program United States:
Pre-treatment to post-treatment (3 mo): 4.67 ± 5.09 vs 3.55 ± 4.14
Cohen d = −0.22
Pre-treatment to post-treatment (3 mo): 3.58 ± 4.35 vs 3.61 ± 4.37
Cohen d = 0.007
At short-term follow-up, small coefficient effect size
Glazer et al (2002)53 BDI-II Cardiac patients (CAD, SA, CABG, MI) CR Pre-treatment to post-treatment (12 wk): 6.2 ± 5.1 vs 5.7 ± 4.4; Cohen d = −0.10 At short-term follow-up, small coefficient effect size
Grace et al (2008)54 HADS-D Cardiac inpatients CR Pre-treatment to post-treatment (18 mo): 4.6 ± 3.5 vs 4.3 ± 3.7
Cohen d = −0.08
At long-term follow-up, small coefficient effect size
Maniar et al (2009)55 BDI-II CAD patients CR program Pre-treatment to post-treatment:
Younger patients (<65 yr): 9.28 ± 8.5 vs 5.66 ± 6.9; Cohen d = −0.42
Older patients (≥65 yr): 7.66 ± 5.4 vs 5.07 ± 4.99; Cohen d = −0.48
Small coefficient effect sizes at short-term follow-up
McGrady et al (2009)56 BDI-II Cardiac patients: MI, CABG surgery, SA (PCI), ChrHF CR Baseline to post-treatment:
18.6 ± 7.3 vs 11.6 ± 7.8; Cohen d = −0.96
Large coefficient effect size at short-term follow-up
McGrady et al (2014)57 BDI-II Cardiac patients (MI, CABG, SA, CHF, other) Intervention group vs control group Total sample: Pre-treatment to post-treatment: 7.5 ± 6.1 vs 4.9 ± 5.1, Cohen d = −0.43 Small coefficient effect size at short-term follow-up
Oldridge et al (1995)58 POMS
MI patients CR POMS
Pre-treatment to post-treatment (8 wk):
9 ± 9.6 vs 6.6 ± 8.7; Cohen d (pre-treatment to post-treatment) = −0.25
12-mo follow-up:
5.9 ± 8.7; Cohen d (pre-treatment to 12-mo follow-up) = −0.32
Usual care:
Pre-treatment to post-treatment (8 wk):
8.7 ± 8.6 vs 7.9 ± 10.3; Cohen d (pre-treatment to post-treatment) = −0.09
12-mo follow-up:
5.7 ± 7.7; Cohen d (pre-treatment to 12-mo follow-up) = −0.35
Pre-treatment to post-treatment (8 wk):
4.2 ± 4.1 vs 3.4 ± 3.7; Cohen d (pre-treatment to post-treatment) = −0.19
12-mo follow-up:
3.1 ± 3.4; Cohen d (pre-treatment to 12-mo follow-up) = −0.26
Usual care:
Pre-treatment to post-treatment (8 wk):
3.9 ± 3.9 vs 3.9 ± 4.5; Cohen d (pre-treatment to post-treatment) = 0
12-mo follow-up:
2.9 ± 3; Cohen d (pre-treatment to 12-mo follow-up) = −0.25
Small coefficient effect size at both short- and long-term follow-up
Rouleau et al (2017)59 HADS-D Patients with STEMI, NSTEMI, SA, HVD, cardiomyopathy, ConHD Outpatient CR Pre-treatment to post-treatment: 3.43 ± 2.75 vs 2.23 ± 2.11; Cohen d = −0.44 Small responsiveness to change at short-term follow-up
Sanderson and Bittner (2005)60 BDI-II Women with CAD CR Pre-treatment to post-treatment: 9 ± 7 vs 6 ± 5; Cohen d = −0.43 Small coefficient size effects at short-term follow-up
Abbreviations: BDI-II, Beck Depression Inventory-II; CABG, coronary artery bypass graft surgery; CAD, coronary artery disease; CES-D, Center for Epidemiologic Studies Depression Scale; CHF, congestive heart failure; ChrHF, chronic heart failure; ConHD, congenital heart disease;
CR, cardiac rehabilitation; DHF, diastolic heart failure; GDS-15, Geriatric Depression Scale (15-item version); HADS-D, Depression subscale of the Hospital Anxiety and Depression Scale; HF, heart failure; HVD, heart valve disease; MI, myocardial infarction; NSTEMI, non-ST elevation myocardial infarction; PCI, percutaneous coronary intervention; PHQ-9, Patient Health Questionnaire (9-item version); POMS, Profile of Mood States; PTCA, percutaneous transluminal coronary angioplasty; SA, stable angina; STEMI, ST elevation myocardial infarction.
aData reported as mean ± standard deviation.


Overall, methodological review led to 3 studies being identified as at risk for bias27,33,36 (Table 3). Three studies did not provide enough information on the index test,27,33,36 the reference standard,33,36 or the patient selection.27 Except for the reference standard dimension (r = 0.23), interrater reliability was good to excellent as evidenced by high coefficient correlations between 0.56 and 1.00.

Table 3 - Quality Assessment Results for Studies Examining Diagnostic Accuracy of Depression Questionnaires for Cardiac Patients
Author (yr) Risk of Bias Applicability
Patient Selection Index Test Reference Standard Flow and Timing Patient Selection Index Test Reference Standard
Banbauer et al (2005)27 Unclear Unclear Low Low Low Low Low
Elderon et al (2011)28 Low Low Low Low Low Low Low
Frasure-Smith and Lesperance (2008)29 Low Low Low Low Low Low Low
Huffman et al (2010)30 Low Low Low Low Low Low Low
Low and Hubley (2007)31 Low Low Low Low Low Low Low
McGuire et al (2013)32 Low Low Low Low Low Low Low
McManus et al (2005)33 Low Unclear Unclear Low Low Low Low
Moullec et al (2015)34 Low Low Low Low Low Low Low
Razykov et al (2012)35 Low Low Low Low Low Low Low
Swardfager et al (2011)36 Low Unclear Unclear Low Low Low Low
Thombs et al (2008)18 Low Low Low Low Low Low Low
ICC 1.00 0.56 0.23 91a 100a 91a 100a
Abbreviation: ICC, intraclass correlation.
aPercentage of agreement is reported for the flow and timing and reference standard dimensions because due to excessively low variance (ie, 0), intraclass correlations could not be computed.


This systematic review advances knowledge concerning performance of depression questionnaires in accurately detecting depression in cardiac populations. It represents the first attempt to review evidence on multiple depression questionnaires specifically within the cardiac population. We highlight 4 major findings: (1) depression questionnaires show adequate psychometric properties; (2) the BDI-II and the HADS-D showed the highest sensitivity and NPV for detecting depressive symptoms in the cardiac population; (3) using specific cut scores for specific CVD states results in enhanced accuracy in detecting depression; and (4) among the most commonly used depression questionnaires, the BDI-II, HADS-D, CES-D, and GDS-15 seem to be the most responsive to depression changes.

Depression questionnaires showed adequate properties for identifying patients who should be referred to further assessment and treatment. In the CR setting, it is important to accurately identify individuals with depressive symptoms so that health providers can deliver, or refer patients to, depression-focused interventions. When using depression measurements for screening purposes, high sensitivity and NPV are advocated so that positive cases are not missed and provided with a more comprehensive assessment and/or treatment, if necessary.22 In addition, using tools with such characteristics allows clinicians to plan and monitor patient progress. This result is relevant due to the recurrent nature of major depression and the negative impairment this condition exerts over patient recovery.61,62 Depression not only affects physical outcomes but also entails social isolation, low perceived support, and poor health style habits (ie, increased smoking rates and sedentary lifestyle). Consequently, continued monitoring of depressive symptoms allows detecting early warning signs related to depression that might prompt intervention.

Good sensitivity properties and NPV were found for the HADS-D (sensitivity = 81%) and the BDI-II (sensitivity = 83%-100%, NPV = 98.1%-100%). Overall, slightly lower values have been shown for the PHQ-2 version (sensitivity ranges from 39% to 95.65% and NPV ranges from 94% to 96%). Regarding older populations, the GDS 30-item version showed optimal psychometric properties for detecting depression (sensitivity = 100% and NPV = 100%). Notwithstanding, these results need to be considered with caution as evidence on sensitivity and NPV values is limited and varies as a function of specific cardiac diagnosis.

Using specific cut scores in relation to specific cardiac diagnoses seems to improve sensitivity rates and NPV. As previously shown,63 considering cardiac diagnosis when interpreting depression questionnaires results is essential to avoid both misdiagnosis and mistreatment. So far, evidence on adequate cut scores by specific cardiac diagnosis remains limited and sparse and, thus, more research is warranted in this field before concluding that specific cut scores should be adopted for specific disease states.

Because of depressive symptoms overlapping those typically displayed by cardiac patients, questionnaires used within the general population might not discern between somatic complaints related to cardiac disease and nonspecific physical symptoms that are better explained by depression. Despite not being included in the present study due to not meeting the inclusion criteria, the Cardiac Depression Scale represents the sole tool without somatic items validated against a criterion standard diagnostic interview within this population. Of interest is that 2-item (DS-SF1) and 5-item (DS-SF2) shorter versions have been recently validated in Australian cardiac outpatients, with the latter being considered as a suitable option for depression screening when adopting a cut score of ≥15.64 Whether these questionnaires perform well in accurately detecting depression cases still needs to be examined.

Although limited, evidence on the internal responsiveness to change suggests that the BDI-II, the HADS-D, the CES-D, and the GDS-15 are the most sensitive to changes in depression over the course of CR. The BDI-II performed the best in capturing depression in the short-term (post-treatment), while the GDS-15 best captured changes in depressive symptoms in older populations (ie, ≥63 yr of age). So far, very few studies have examined how depression questionnaires perform in terms of detecting changes in depression considering specific cardiac diagnoses. Although 1 study showed that the CES-D performs adequately for detecting depression changes in coronary artery bypass graft patients, more research is warranted to strengthen the evidence before concluding that specific depression questionnaires should be used for specific cardiac populations. Comparisons across tools are difficult mainly due to few studies examining changes in depression after CR and variability in the populations examined (ie, wide range of cardiac diagnoses) and the range in times of follow-up.

Limitations of this systematic review merit consideration. First, this study informs on the accuracy of depression-screening questionnaires for detecting depression diagnosis but not subclinical depression. As this condition poses an elevated risk for cardiac patients, further studies should provide evidence on the appropriate depression questionnaires for identifying subthreshold depression. Second, the studies we reviewed were conducted in US and Canadian cardiac patients, precluding these findings from being generalized to different populations. Third, evidence on the diagnostic accuracy of specific cut scores by cardiac diagnosis was sparse, with most of the studies comprising males endorsing a paucity of cardiac diagnosis. This represents an important gap in the literature that warrants further investigation. Fourth, we evaluated internal sensitivity properties of depression-screening instruments by calculating effect sizes. As there is limited evidence on this issue, future studies need to assess whether specific depression questionnaires detect a clinically important change relating to the corresponding change in measures of depression diagnosis. Finally, we evaluated only responsiveness to change in the context of formal CR programs that included recommended core components (eg, physical activity, counseling on self-care),65 so those studies including non-formal CR treatments were excluded from this review.


The BDI-II and the HADS-D accurately detect depression. Briefer screening questionnaires such as the PHQ-2 should be used only when time/cost constraints exist as it may result in one-quarter of the assessed patients not being accurately identified. Using several different depression questionnaires for assessing depression might be a difficult task for those health care providers without experience or background; thus, the best option might be the consistent use of the BDI-II, both for screening and tracking change over time. The BDI-II is the depression questionnaire most frequently validated across cardiac populations, consistently showing sensitivity values >83% and NPV ≥98%. Among the most sensitive and specific depression questionnaires, the BDI-II has also been shown to perform well at detecting the presence and severity of suicidal ideation. As both the pessimism and suicidal ideation items of the BDI have been shown as predictors of eventual suicide,66 using this assessment tool can be useful for identifying patients in need of thorough psychiatric evaluation. It should be noted that other depression-screening instruments (eg, PHQ-9) also assess thoughts of self-harm. Regardless of the source, any report of thoughts of harm to self or others should prompt immediate evaluation and referral to crises services, as necessary. In the context of CR, the availably of more evidence-based data on the BDI-II responsiveness to change results in advocating its use to monitor depression over time.


The BDI-II has the most support as a valid useful measure for both detecting depression in cardiac patients and monitoring its change in response to CR. To avoid misdiagnosis and mistreatment, health professionals should adopt cut scores according to specific cardiac diagnosis. Despite the GDS having sound psychometric properties, there is a surprising lack of empirical research on its psychometric validation properties for the English-speaking North American populations.


This research was supported in part by National Institutes of Health Center of Biomedical Research Excellence award P20GM103644, from the National Institute of General Medical Sciences and Tobacco Centers of Regulatory Science award P50DA036114, and from the National Institute on Drug Abuse and US Food and Drug Administration. This work was also supported by a predoctoral grant from the Spanish Ministry of Economy and Competitiveness (MINECO) (reference: BES-2016-076663) awarded to Ms González-Roz. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration.


1. Wang JT, Hoffman B, Blumenthal JA. Management of depression in patients with coronary heart disease: association, mechanisms, and treatment implications for depressed cardiac patients. Expert Opin Pharmacother. 2011;12(1):85–98.
2. Meyer FA, Hugentobler E, Stauber S, Wilhelm M, Znoj H, von Kanel R. Depressive symptoms at discharge from rehabilitation predict future cardiovascular-related hospitalizations. Cardiology. 2015;131(2):80–85.
3. Zullo MD, Gathright EC, Dolansky MA, Josephson RA, Cheruvu VK, Hughes JW. Influence of depression on utilization of cardiac rehabilitation postmyocardial infarction: a study of 158 991 Medicare beneficiaries. J Cardiopulm Rehabil Prev. 2017;37(1):22–29.
4. Rutledge T, Reis VA, Linke SE, Greenberg BH, Mills PJ. Depression in heart failure a meta-analytic review of prevalence, intervention effects, and associations with clinical outcomes. J Am Coll Cardiol. 2006;48(8):1527–1537.
5. Huffman JC, Celano CM, Beach SR, Motiwala SR, Januzzi JL. Depression and cardiac disease: epidemiology, mechanisms, and diagnosis. Cardiovasc Psychiatry Neurol. 2013;2013:695925.
6. Möller-Leimkühler AM. Gender differences in cardiovascular disease and comorbid depression. Dialogues Clin Neurosci. 2007;9(1):71–83.
7. Shanmugasegaram S, Russell KL, Kovacs AH, Stewart DE, Grace SL. Gender and sex differences in prevalence of major depression in coronary artery disease patients: a meta-analysis. Maturitas. 2012;73(4):305–311.
8. Herridge ML, Stimler CE, Southard DR, King ML. Depression screening in cardiac rehabilitation: AACVPR Task Force Report. J Cardiopulm Rehabil Prev. 2005;25(1):11–13.
9. Bush DE, Ziegelstein RC, Tayback M, et al. Even minimal symptoms of depression increase mortality risk after acute myocardial infarction. Am J Cardiol. 2001;88(4):337–341.
10. Cahill MC, Bilanovic A, Kelly S, Bacon S, Grace SL. Screening for depression in cardiac rehabilitation: a review. J Cardiopulm Rehabil Prev. 2015;35(4):225–230.
11. Huffman JC, Smith FA, Blais MA, Beiser ME, Januzzi JL, Fricchione GL. Recognition and treatment of depression and anxiety in patients with acute myocardial infarction. Am J Cardiol. 2006;98(3):319–324.
12. Ceccarini M, Manzoni GM, Castelnuovo G. Assessing depression in cardiac patients: what measures should be considered? Depress Res Treat. 2014;2014:148256.
13. Meneghetti CC, Guidolin BL, Zimmermann PR, Sfoggia A. Screening for symptoms of anxiety and depression in patients admitted to a university hospital with acute coronary syndrome. Trends Psychiatry Psychother. 2017;39(1):12–18.
14. Lea P. Factors affecting nurses' intent to assess for depression in heart failure patients. Dimens Crit Care Nurs. 2014;33(6):320–326.
15. Compare A, Germani E, Proietti R, Janeway D. Clinical psychology and cardiovascular disease: an up-to-date clinical practice review for assessment and treatment of anxiety and depression. Clin Pract Epidemiol Mental Health. 2011;7:148–156.
16. Forkmann T, Vehren T, Boecker M, Norra C, Wirtz M, Gauggel S. Sensitivity and specificity of the Beck Depression Inventory in cardiologic inpatients: how useful is the conventional cut-off score? J Psychosom Res. 2009;67(4):347–352.
17. Stafford L, Berk M, Jackson HJ. Validity of the Hospital Anxiety and Depression Scale and Patient Health Questionnaire-9 to screen for depression in patients with coronary artery disease. Gen Hosp Psychiatry. 2007;29(5):417–424.
18. Thombs BD, Ziegelstein RC, Whooley MA. Optimizing detection of major depression among patients with coronary artery disease using the patient health questionnaire: data from the heart and soul study. J Gen Intern Med. 2008;23(12):2014–2017.
19. Husted JA, Cook RJ, Farewell VT, Gladman DD. Methods for assessing responsiveness: a critical review and recommendations. J Clin Epidemiol. 2000;53(5):459–468.
20. Stegenga BT, Kamphuis MH, King M, Nazareth I, Geerlings MI. The natural course and outcome of major depressive disorder in primary care: the PREDICT-NL study. Soc Psychiatry Psychiatr Epidemiol. 2012;47(1):87–95.
21. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097.
22. Strik JJ, Honig A, Lousberg R, Denollet J. Sensitivity and specificity of observer and self-report questionnaires in major and minor depression following myocardial infarction. Psychosomatics. 2001;42(5):423–428.
23. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care. 1989;27(3 suppl):S178–S189.
24. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Hilldale, NJ: Erlbaum; 1988.
25. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–536.
26. Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016;15(2):155–163.
27. Bambauer KZ, Locke SE, Aupont O, Mullan MG, McLaughlin TJ. Using the Hospital Anxiety and Depression Scale to screen for depression in cardiac patients. Gen Hosp Psychiatry. 2005;27(4):275–284.
28. Elderon L, Smolderen KG, Na B, Whooley MA. Accuracy and prognostic value of American Heart Association: recommended depression screening in patients with coronary heart disease: data from the Heart and Soul Study. Circ Cardiovasc Qual Outcomes. 2011;4(5):533–540.
29. Frasure-Smith N, Lesperance F. Depression and anxiety as predictors of 2-year cardiac events in patients with stable coronary artery disease. Arch Gen Psychiatr. 2008;65(1):62–71.
30. Huffman JC, Doughty CT, Januzzi JL, Pirl WF, Smith FA, Fricchione GL. Screening for major depression in post-myocardial infarction patients: operating characteristics of the Beck Depression Inventory-II. Int J Psychiatry Med. 2010;40(2):187–197.
31. Low GD, Hubley AM. Screening for depression after cardiac events using the Beck Depression Inventory-II and the Geriatric Depression Scale. Soc Indic Res. 2007;82(3):527–543.
32. McGuire AW, Eastwood JA, Macabasco-O'Connell A, Hays RD, Doering LV. Depression screening: utility of the patient health questionnaire in patients with acute coronary syndrome. Am J Crit Care. 2013;22(1):12–19.
33. McManus D, Pipkin SS, Whooley MA. Screening for depression in patients with coronary heart disease (data from the Heart and Soul Study). Am J Cardiol. 2005;96(8):1076–1081.
34. Moullec G, Plourde A, Lavoie KL, Suarthana E, Bacon SL. Beck Depression Inventory II: determination and comparison of its diagnostic accuracy in cardiac outpatients. Eur J Prev Cardiol. 2015;22(5):665–672.
35. Razykov I, Ziegelstein RC, Whooley MA, Thombs BD. The PHQ-9 versus the PHQ-8—Is item 9 useful for assessing suicide risk in coronary artery disease patients? Data from the Heart and Soul Study. J Psychosom Res. 2012;73(3):163–168.
36. Swardfager W, Herrmann N, Marzolini S, et al. Major depressive disorder predicts completion, adherence, and outcomes in cardiac rehabilitation: a prospective cohort study of 195 patients with coronary artery disease. J Clin Psychiatry. 2011;72(9):1181–1188.
37. Campbell IM, Burgess PM, Finch SJ. A factorial analysis of BDI scores. J Clin Psychol. 1984;40(4):992–996.
38. Tully PJ, Winefield HR, Baker RA, Turnbull DA, de Jonge P. Confirmatory factor analysis of the Beck Depression Inventory-II and the association with cardiac morbidity and mortality after coronary revascularization. J Health Psychol. 2011;16(4):584–595.
39. Arnau RC, Meagher MW, Norris MP, Bramson R. Psychometric evaluation of the Beck Depression Inventory-II with primary care medical patients. Health Psychol. 2001;20(2):112–119.
40. Di Benedetto M, Lindner H, Hare DL, Kent S. Depression following acute coronary syndromes: a comparison between the Cardiac Depression Scale and the Beck Depression Inventory II. J Psychosom Res. 2006;60(1):13–20.
41. Löwe B, Kroenke K, Gräfe K. Detecting and monitoring depression with a two-item questionnaire (PHQ-2). J Psychosom Res. 2005;58(2):163–171.
42. Hammash MH, Hall LA, Lennie TA, et al. Psychometrics of the PHQ-9 as a measure of depressive symptoms in patients with heart failure. Eur J Cardiovasc Nurs. 2013;12(5):446–453.
43. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–613.
44. Hare DL, Davis CR. Cardiac Depression Scale: validation of a new depression scale for cardiac patients. J Psychosom Res. 1996;40(4):379–386.
45. Martin CR, Lewin RJ, Thompson DR. A confirmatory factor analysis of the Hospital Anxiety and Depression Scale in coronary care patients following acute myocardial infarction. Psychiatry Res. 2003;120(1):85–94.
46. Burns A, Hofer S, Curry P, Sexton E, Doyle F. Revisiting the dimensionality of the Hospital Anxiety and Depression Scale in an international sample of patients with ischaemic heart disease. J Psychosom Res. 2014;77(2):116–121.
47. Burke WJ, Houston MJ, Boust SJ, Roccaforte WH. Use of the Geriatric Depression Scale in dementia of the Alzheimer type. J Am Geriatr Soc. 1989;37(9):856–860.
48. Harenberg S, Marshall-Prain N, Dorsch KD, Riemer HA. Factorial validity and gender invariance of the center for epidemiological studies depression in cardiac rehabilitation patients. J Cardiopulm Rehabil Prev. 2015;35(5):320–327.
49. Beckie TM, Beckstead JW, Schocken DD, Evans ME. The effects of a tailored cardiac rehabilitation program on depressive symptoms in women: a randomized clinical trial. Int J Nurs Stud. 2011;48(1):3–12.
50. Caulin-Glaser T, Maciejewski PK, Snow R, LaLonde M, Mazure C. Depressive symptoms and sex affect completion rates and clinical outcomes in cardiac rehabilitation. Prev Cardiol. 2007;10(1):15–21.
51. Gary RA, Sueta CA, Dougherty M, et al. Home-based exercise improves functional performance and quality of life in women with diastolic heart failure. Heart Lung. 2004;33(4):210–218.
52. Ghisi GLM, Santos CVA, Benaim B, et al. Severity of depressive symptoms pre- and postcardiac rehabilitation: a comparison among patients in Brazil, Canada, Colombia, the United States, and Venezuela. J Cardiopulm Rehabil Prev. 2017;37(3):182–190.
53. Glazer KM, Emery CF, Frid DJ, Banyasz RE. Psychological predictors of adherence and outcomes among patients in cardiac rehabilitation. J Cardiopulm Rehabil. 2002;22(1):40–46.
54. Grace SL, Grewal K, Arthur HM, Abramson BL, Stewart DE. A prospective, controlled multisite study of psychosocial and behavioral change following women's cardiac rehabilitation participation. J Womens Health (Larchmt). 2008;17(2):241–248.
55. Maniar S, Sanderson BK, Bittner V. Comparison of baseline characteristics and outcomes in younger and older patients completing cardiac rehabilitation. J Cardiopulm Rehabil Prev. 2009;29(4):220–229.
56. McGrady A, McGinnis R, Badenhop D, Bentle M, Rajput M. Effects of depression and anxiety on adherence to cardiac rehabilitation. J Cardiopulm Rehabil Prev. 2009;29(6):358–364.
57. McGrady A, Burkes R, Badenhop D, McGinnis R. Effects of a brief intervention on retention of patients in a cardiac rehabilitation program. Appl Psychophysiol Biofeedback. 2014;39(3-4):163–170.
58. Oldridge N, Streiner D, Hoffmann R, Guyatt G. Profile of mood states and cardiac rehabilitation after acute myocardial infarction. Med Sci Sports Exerc. 1995;27(6):900–905.
59. Rouleau CR, Toivonen K, Aggarwal S, Arena R, Campbell TS. The association between insomnia symptoms and cardiovascular risk factors in patients who complete outpatient cardiac rehabilitation. Sleep Med. 2017;32:201–207.
60. Sanderson BK, Bittner V. Women in cardiac rehabilitation: outcomes and identifying risk for dropout. Am Heart J. 2005;150(5):1052–1058.
61. Burcusa SL, Iacono WG. Risk for recurrence in depression. Clin Psychol Rev. 2007;27(8):959–985.
62. Chauvet-Gelinier J-C, Bonin B. Stress, anxiety and depression in heart disease patients: a major challenge for cardiac rehabilitation. Ann Phys Rehabil Med. 2017;60(1):6–12.
63. Wang Y-P, Gorenstein C. Assessment of depression in medical patients: a systematic review of the utility of the Beck Depression Inventory-II. Clinics. 2013;68(9):1274–1287.
64. Hare DL, Selvadurai LP, Wang JQ, et al. Psychometric assessment of the Cardiac Depression Scale Short Form in cardiac outpatients. Eur J Cardiovasc Nurs. 2017;16(3):249–255.
65. Balady GJ, Williams MA, Ades PA, et al. Core components of cardiac rehabilitation/secondary prevention programs: 2007 update: a scientific statement from the American Heart Association Exercise, Cardiac Rehabilitation, and Prevention Committee, the Council on Clinical Cardiology; the Councils on Cardiovascular Nursing, Epidemiology and Prevention, and Nutrition, Physical Activity, and Metabolism; and the American Association of Cardiovascular and Pulmonary Rehabilitation. Circulation. 2007;115(20):2675–2682.
66. Green KL, Brown GK, Jager-Hyman S, Cha J, Steer RA, Beck AT. The predictive validity of the Beck Depression Inventory suicide item. J Clin Psychiatry. 2015;76(12):1683–1686.

cardiac rehabilitation; depression; screening; sensitivity; specificity; systematic review

© 2019 Wolters Kluwer Health, Inc. All rights reserved.