Volume 104, Issue 9 , Pages 1194-1197, 1 November 2009
Screening for Depression and Suicidality in Patients With Cardiovascular Illnesses
Article Outline
The American Heart Association (AHA) and the American Psychiatric Association jointly recommend screening for depression in cardiology clinics. This includes screening for suicidality. It is not known how frequently patients disclose suicidal thinking (ideation) in this setting, and what proportion of those will turn out to have suicidal intent. Patients were screened for depression using a protocol identical to the one endorsed by the AHA in a cardiology community clinic in Elmhurst (Queens, New York). Depression was assessed using the Patient Health Questionnaire. Reports of suicidal ideation were immediately evaluated by a mental health professional. We determined the degree to which suicidal ideation was identified, the proportion of patients with suicidal intent of those reporting suicidal ideation, and the relation between depression and suicidal ideation in this setting. One thousand three patients were screened; 886 had complete Patient Health Questionnaire data. Of those, 12% (109 patients) expressed suicidal ideation. Four of those were hospitalized for suicidal intent (0.45% of all screened patients). Suicidal ideation and depression were correlated (point biserial correlation coefficient 0.478). In conclusion, suicidal ideation can and will be identified using the AHA depression screening recommendations, but only a very small fraction (0.45%) of screened patients will turn out to have suicidal intent. Discovery and stabilization of suicidal patients is an important benefit of the screening, but the fact that >12% of all screened patients will need to be immediately evaluated for suicidal intent has important implications for resource allocation to screening programs.
The American Heart Association (AHA), joined by the American Psychiatric Association, recently recommended screening for depression in patients with cardiovascular illnesses.1 The recommendations specifically mention the use of the short (depression screening) form of the Patient Health Questionnaire (PHQ).1, 2, 3, 4 Question 9 in the PHQ asks about thoughts of being better off dead or hurting oneself (suicidal ideation [SI]). Some patients who report SI will turn out, on further evaluation, to have suicidal intent (an active desire to kill oneself), which is a medical emergency. However, identification, and immediate handling, of SI that is discovered during screening is seldom addressed in recommendations related to screening for mental health symptoms in physically ill populations. The purpose of the present study was to examine the relation between use of a screening instrument (the PHQ) and discovery of SI and suicidal intent in an ethnically diverse sample of cardiologic patients. We also looked at the relation between depression, post-traumatic stress symptoms, and suicidality. The clinical program Interdisciplinary Cardiology Outpatient Program @ Elmhurst (I-COPE) uses AHA guidelines for screening,1 although the guidelines were published after the program has started. Because questionnaires were immediately reviewed and addressed by mental health professionals and cardiologists in the clinic, we are able to report not only the frequency of SI but also the likelihood that those patients would eventually be in need of an urgent psychiatric hospitalization.
Methods
Patients were screened from June 2005 to November 2007 in this ongoing program. The cardiology clinic of Elmhurst Hospital, a part of the New York City Health and Hospitals Corporation, is located in the borough of Queens in New York City. All patients receive services irrespective of insurance status. Elmhurst is one of the most ethnically diverse zip codes in the United States,5 with >40 nationalities represented. The primary cardiac diagnoses in the clinic are coronary artery disease and heart failure. Because this is a clinical program, there were no specific inclusion/exclusion criteria (all patients who presented to the clinic were screened). Screening was performed only on one day of the clinic (the day in which general cardiologic patients were seen) and was offered only one time. Institutional review board approval was obtained to review screened patients' medical charts.
All patients who registered for an outpatient visit received the PHQ depression screening questionnaire and the Impact of Event Scale (IES) as a part of a Health and Hospitals Corporation initiative to screen for mental health symptoms in patients with cardiovascular illnesses. Questionnaires were available in English and Spanish. These were handed to patients with a registration to the clinic visit, were handed back to the registration desk, immediately reviewed, and were available to the cardiologist at the time of the visit. Any positive suicidal screening result (defined as a score >0 for question 9 in the PHQ) and above-threshold scores on the IES or PHQ screening generated immediate evaluation by a mental health professional (a licensed psychologist or psychiatrist who was present in the clinic). The clinical mental health evaluation was comprehensive and not standardized. It included an assessment of suicidal intent. The clinician decided whether to arrange for an urgent referral for further evaluation in the psychiatric emergency room. Thus, potentially, patients who answered the suicide question in the negative could still be deemed “suicidal,” if suicidality was discovered during an evaluation that was triggered by a threshold score on any of the questionnaires.
Cardiologists and nurses in the clinic participated in special training on detecting suicidality in primary care. This included a presentation by the first author and case supervision in the clinic by a psychologist (who participated in the first few evaluations alongside the cardiologist). There was no additional evaluation of the cardiologists' performance in assessing suicidality after the formal training.
Demographic variables (gender, age, ethnicity, and country of origin) were obtained from the medical record. Patients were listed as African, Caucasian, East Asian, Hispanic, Southeast Asian, or “other.”
The short form of the PHQ2, 4 was used because it specifically targets depression, a major risk for cardiovascular morbidity and the specific focus of the AHA recommendation. The 9 items are scored from 0 (not at all) to 3 (nearly every day). The final item (number 9) asks about “thoughts that you would be better off dead or of hurting yourself in some way.” A score ≥10 has acceptable sensitivity and specificity for predicting depression; a score ≥20 is considered suggestive of “major depression.”1 For our purposes, a score ≥10 (the less restrictive threshold) led to an immediate further evaluation by the cardiologist and mental health professional in the clinic.
The IES6 is a 15-item self-report questionnaire that measures current subjective distress/post-traumatic stress related to a specific event. The IES was used because post-traumatic stress symptoms are also predictive of morbidity and mortality in patients with cardiovascular illnesses.7, 8, 9, 10 The IES consists of 2 subscales measuring experiences of avoidance and intrusion related to the traumatic event. Good reliability and validity have been reported.11 Building on experience,7, 8 an IES score ≥18 led to immediate further evaluation.
Analyses used SPSS 12.0 (SPSS, Inc., Chicago, Illinois). Point biserial correlations were used to examine the relation among depression, post-traumatic stress, and suicidality. Receiver operating characteristic curve analysis was used to examine the degree to which depression or post-traumatic stress was a good predictor of suicidal thoughts in this cohort. Statistical analyses were supervised by Dr. Weatherley and Dr. Feaganes from Momentum Research, Inc. (Durham, North Carolina).
Results
One thousand three patients were screened from June 2005 to November 2007. Because this is a retrospective analysis of data from a clinical program, selection bias and nonparticipation rates were minimal. Nevertheless, 10% of patients in the clinic spoke neither Spanish nor English; those patients were not captured by the screening. Table 1 presents demographic characteristics of the sample. One hundred seventeen patients started but did not complete the 2 questionnaires, and another 109 did not complete the IES (for 226 patients who did not complete the IES). Although not formally assessed, it seems that a frequent reason for not completing the questionnaires was a language difficulty; the IES was administered after the PHQ and its questions are a bit more complex; hence, several patients who did answer the PHQ did not answer the IES.
Table 1. Demographic characteristics (n = 1,003)
| Age | 61.0 ±14.0 |
| Men | 625(62%) |
| Women | 378(38%) |
| Hispanic | 504(50%) |
| Southeast Asian | 229(23%) |
| Caucasian | 114(11%) |
| East Asian | 89(9%) |
| African | 53(5%) |
| Other | 6(<1%) |
| Unknown | 8(<1%) |
Of patients who completed the PHQ, 12% (109 patients) reported SI. Our cardiologists, although they were trained to assess for suicidality independently, did not suspect suicidality in any of the other patients who answered the PHQ “suicide” question in the negative. In addition, there was no instance in which an evaluation triggered by a threshold score on the IES or PHQ identified a suicidal patient who was not already identified by the PHQ suicide question (all suicidal patients of those evaluated were captured by the suicide screen). Table 2 presents the ethnic breakdown of patients who reported SI. Four of these patients (3.7% of those who disclosed SI and about 0.45% of the entire screened cohort) were urgently taken to the psychiatric emergency room because of suicidal intent. All 4 were eventually admitted to a psychiatric unit. During the same time that the mental health screening took place, 1% of clinic patients were urgently admitted for acute cardiovascular reasons.
Table 2. Suicidal ideation/ethnicity
| Ethnic Group | PHQ Threshold, Suicide Item |
|---|---|
| Hispanic | 12% |
| Southeast Asian | 18% |
| Caucasian | 6% |
| East Asian | 11% |
| African | 15% |
Table 3 presents point biserial correlation coefficients for the correlation between IES subscale and total scores, PHQ adjusted (minus the suicide question) score, and presence of SI (any positive score on the PHQ suicide question) or suicidal intent (confirmed after interview). Correlations were all significant, but many were not very large. The largest correlation (0.478) was found to be between adjusted PHQ scores and SI, reaffirming the known relation between depression and SI. All measurements were also significantly correlated with suicidal intent, but the coefficients were in a far lower range, all <0.2.
Table 3. Correlations between suicidal ideation and suicide intent, and adjusted Patient Health Questionnaire scores, Impact of Event Scale total scores, and Impact of Event Scale subscales
| IES Total | IES Avoidance | IES Intrusion | Adjusted PHQ | |
|---|---|---|---|---|
| Suicide ideation (yes/no) | ||||
| Point biserial correlation | 0.364⁎ | 0.297⁎ | 0.375⁎ | 0.478⁎ |
| Significance (2-tailed) | 0.000 | 0.000 | 0.000 | 0.000 |
| Suicidal intent (yes/no) | ||||
| Point biserial correlation | 0.131⁎ | 0.110⁎ | 0.142⁎ | 0.158⁎ |
| Significance (2-tailed) | 0.000 | 0.002 | 0.000 | 0.000 |
⁎Correlation is significant at the 0.01 level (2-tailed). |
Four patients were found to have suicidal intent on further evaluation. None of these patients were under current psychiatric care at the time of the screening, although 2 of the 4 received previous psychiatric treatment. None of those patients were currently taking psychiatric medications. Three of those 4 patients were men and were 29 to 55 years of age. Three were Hispanic and 1 was Caucasian. Note, however, that chi-square test showed that there were no significant ethnic differences between patients with SI on PHQ item 9 (p = 0.053; Table 2). Scores on question 9 for patients with suicidal intent were 1 (2 patients), 2 (1 patient), and 3 (1 patient). Therefore, the presence of suicidal intent was not related to the degree of the initial report of SI.
Receiver operating characteristic curve analysis (Figure 1) established that PHQ scores (depression) were better predictors of SI than IES scores (post-traumatic stress). Within the IES subscales, the intrusion scale was more predictive of SI than the avoidance subscale.
Figure 1.
Receiver operator curves for predicting SI from adjusted PHQ scores (green), IES total scores (blue), IES avoidance subscale scores (yellow), and IES intrusion subscale score (purple). Diagonal segment are produced by ties.
To our knowledge, no patient actually attempted suicide during the study period. However, we had no formal mechanism to ascertain that this is the case—except for patients who were identified as having suicidal intent (all of those were followed).
Total PHQ scores for patients with suicidal intent were 18 to 24, and their IES scores were 27 to 53. A PHQ score ≥18 (representing about 8% of all responders) would have identified all patients with suicidal intent in this cohort. A positive answer on the PHQ SI item represents 12% of responders and would have identified all of those patients, too. The IES threshold that would have identified all patients with suicidal intent was ≥27 in this sample, but it did not perform as well as a predictor because about 22% of responders reached that threshold.
Discussion
This study is the first to examine suicidal risk using the newly published AHA recommendations1; our findings suggest that SI is common in ethnically diverse cardiologic patients, but that suicidal intent is not.
It is important to use a consistent nomenclature about suicidality.12 Although our limited assessments do not permit a detailed use of the recently proposed nomenclature,13 we do distinguish between “suicidal ideation,” which is used in this report to denote thoughts about dying and killing oneself, and “suicidal intent,” which we use to denote an intent to do so (beyond an abstract thought). The PHQ asks only about SI.
SI was reported by 12% of patients who completed the PHQ, but suicidal intent was far less common (only 0.45% of the screened population). SI is probably more common in medically ill patients,14 but those thoughts about dying do not necessarily indicate suicidal intent. Interestingly, although all patients with suicidal intent reported SI (score ≥1 on question 9 of the PHQ), the “severity” of those initially reported thoughts was not related to the actual threat of imminent suicide. Consistent with AHA recommendations, this suggests that any degree of SI should be addressed immediately because it can reveal suicidal intent. Our results indeed establish that the cutoff used for the score on question 9 would have to be set at the lowest threshold (≥1) to detect all suicidal intent. Our study failed to suggest a more restrictive cutoff.
The discovery of suicidal intent in 0.45% of the responders is, we believe, clinically important. In this clinic as a whole, about 1% of patients were admitted urgently for other (cardiovascular) causes, and, hence, our questionnaire identified almost 1/3 of all urgent admissions during this period. Those patients' lives may well have been saved by the screen. Expenses incurred to ensure that all screening instruments are evaluated immediately are not minimal, however. Because only a very small minority (3.5%) of patients who initially reported SI were deemed to need urgent care after evaluation, a “multitiered” system may be warranted, in which initial evaluations of a positive screen are done by a nurse and not necessarily by a mental health professional.
Depressive symptoms were better predictors of SI than post-traumatic symptoms. Patients with suicidal intent had IES scores that were shared by about 22% of all screened patients, whereas their level of depression was shared by only about 8% of patients, making depression a more specific predictor of suicidal intent.
Three of 4 patients with suicidal intent were Hispanic. Caucasians were a minority in this clinic, and because of the very small number of patients with suicidal intent, the observation that Hispanics had a disproportionately large share in patients with suicidal intent probably needs further scrutiny. Indeed, it is not consistent with national prevalence data that show less suicidality in Hispanics.15
Our study has several limitations. The very small number of patients who were contemplating suicide (patients with suicidal intent) is a limiting factor in many studies addressing suicidality. We are not able to confirm without a doubt that no patients had a suicidal attempt after the screening. A few suicidal patients may have been missed because a formal assessment for suicidality by cardiologists, although encouraged, was not ascertained after the training that the cardiologists received. However, patients in this program received a mental health evaluation if they met any of 3 relatively nonrestrictive thresholds—a post-traumatic stress score threshold, a depression score threshold, or any positive answer to the PHQ suicide question. Although it is still possible that an acutely suicidal patient would have been missed using those criteria, we believe that it is not likely. Measurements selected for screening were given only in their English and Spanish versions, which were not the primary languages of all patients. Ideally, patients would be able to complete these measurements in their primary language, likely increasing the effectiveness of the screening process.
Use of AHA-recommended screening in a cardiology clinic has the potential to identify patients with suicidal intent and, presumably, save their lives—an important benefit of the screening. However, to not miss those patients, adequate resources for immediate evaluation have to be present. Asking about suicidal intent (rather than only SI) in the screening could be studied as a way to restrict the number of patients who need to be immediately evaluated after screening.
Acknowledgment
The authors acknowledge with gratitude the help received from staff members at Elmhurst Medical Center's cardiology clinic. The authors would also like to acknowledge the visionary leadership of the New York's Health and Hospital Corporation (HHC), which established the I-COPE program.
References
- . Depression and coronary heart disease: recommendations for screening, referral, and treatment: a science advisory from the American Heart Association Prevention Committee of the Council on Cardiovascular Nursing, Council on Clinical Cardiology, Council on Epidemiology and Prevention, and Interdisciplinary Council on Quality of Care and Outcomes Research: endorsed by the American Psychiatric Association. Circulation. 2008;118:1768–1775
- . The PHQ-9: a new depression and diagnostic severity measure. Psychiatr Ann. 2002;32:509–521
- . Validation and utility of a self-report version of PRIME-MD: the PHQ Primary Care Study. JAMA. 1999;282:1737–1744
- . The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606–613
- . In Queens, it's the glorious 4th, and 6th, and 16th, and 25th. The New York Times. July 4, 2006;B1, B5
- . Impact of Event Scale: a measure of subjective stress. Psychosom Med. 1979;41:209–218
- . A prospective study of posttraumatic stress symptoms and nonadherence in survivors of a myocardial infarction (MI). Gen Hosp Psychiatry. 2001;23:215–222
- . Posttraumatic stress, nonadherence, and adverse outcome in survivors of a myocardial infarction. Psychosom Med. 2004;66:521–526
- . Early post-transplant medical compliance and mental health predict physical morbidity and mortality one to three years after heart transplantation. J Heart Lung Transplant. 1999;18:549–562
- . Posttraumatic stress symptoms and predicted mortality in patients with implantable cardioverter-defibrillators: results from the prospective living with an implanted cardioverter-defibrillator study. Arch Gen Psychiatry. 2008;65:1324–1330
- . Impact of event scale: psychometric properties. Br J Psychiatry. 2002;180:205–209
- . Beyond the tower of Babel: a nomenclature for suicidology. Suicide Life Threat Behav. 1996;26:237–252
- . Rebuilding the tower of Babel: a revised nomenclature for the study of suicide and suicidal behaviors (Part 2: suicide-related ideations, communications, and behaviors). Suicide Life Threat Behav. 2007;37:264–277
- . Major depression, physical illness, and suicidal ideation in primary care. Psychosom Med. 2003;65:501–505
- . Suicide and ethnicity in the United States. Rep Group Adv Psychiatry. 1989;128:1–131
Data analysis for this study was partly supported by Grant MH-071249 from the National Institute of Mental Health (Bethesda, Maryland) to Dr. Shemesh. The Interdisciplinary Cardiology Outpatient Program @ Elmhurst is funded by the Health and Hospital Corporation of New York, Queens, New York.
PII: S0002-9149(09)01259-4
doi:10.1016/j.amjcard.2009.06.033
© 2009 Elsevier Inc. All rights reserved.
Volume 104, Issue 9 , Pages 1194-1197, 1 November 2009
