The American Heart Association science advisory on depression and coronary heart disease: An exploration of the issues raised

Three key questions

Key question 1: What’s the accuracy of screening instruments for depression in cardiovascular care populations? To answer this question, a case-finding method such as a questionnaire (eg, BDI or PHQ) must be compared with a structured interview by mental health personnel as the “gold standard” for diagnosis (ie, the truth).

To estimate the accuracy of clinical diagnosis by general practitioners, Mitchell et al pooled data from 50,371 primary care patients across 41 studies in Europe or the United States to evaluate general practitioners’ ability to make an unassisted diagnosis of depression (ie, without specific help from severity scales, diagnostic instruments, education programs, etc).³² They reported a sensitivity of about 50% and a specificity of 81% when the prevalence of depression was about 20%. In other words, general practitioners missed about half the cases of depression when no case-finding tool was used. These researchers pointed out that a low prevalence of depression favors identification of nondepressed cases (false-positive diagnoses), whereas a high prevalence favors diagnosis of depression (true-positive diagnoses).³²

Cardiologists focused on treatment of ACS are probably less likely to make a clinical diagnosis of depression and may attribute emotional symptoms to rapidly evolving cardiovascular events. The simple 2-question PHQ-2 case-finding instrument takes only a few minutes to administer and is recommended for use by primary care physicians or cardiologists to evaluate patients at high risk for depression or who manifest symptoms suggestive of depression.⁷

In the United States, most patients with depression are cared for in primary care venues. The AHA advisory recommends referring ACS patients who screen positive on a PHQ to a professional qualified in the diagnosis and management of depression.¹ Enhanced care—using outreach, monitoring, adjustment of therapy, and psychiatric backup—produces significant improvement in depression.³³

Key question 2: Is treatment of depression in cardiovascular care patients effective in improving depression? Cardiac outcomes? The evidence for benefit of SSRI therapy for depression detected at the time of ACS is consistent and was discussed above.^4,17–21 However, the antidepressant effect of SSRIs is modest in placebo-controlled trials. Most such trials excluded patients who were taking antidepressant drugs when screened, and many patients recruited for antidepressant clinical trials had no previous episodes of depression, had relatively mild symptoms of short duration, and were recruited by screening (ie, they were not seeking treatment for depression).³⁴ Patients with brief and short episodes are likely to remit spontaneously and to respond to psychotherapy or supportive care.³⁴ Moreover, in most antidepressant trials, both the intervention and “control” groups received elements of enhanced depression care, which will reduce the apparent benefit of antidepressant drugs. For instance, because the primary goal of SADHART was to evaluate the safety of sertraline in patients with ACS, monitoring for adverse effects included six or seven visits during 16 weeks of follow-up plus six or seven phone calls, providing several elements of enhanced depression care, including face-to-face education, frequent follow-up, support by a case manager (research coordinator) with a mental health background, and support from a psychiatrist or psychologist.¹⁷ Randomization to sertraline or placebo in SADHART was blinded, so the frequent contact and support was the same in both groups.

Whether SSRI treatment for depression will improve survival and cardiovascular outcomes is not established by adequately powered randomized trials. Most randomized (SADHART, CREATE) or nonrandomized (ENRICHD) studies suggest that SSRIs reduce cardiovascular events, but only ENRICHD produced a statistically significant result. SSRIs—certainly sertraline and citalopram—are not associated with significant cardiovascular adverse effects, even during ACS, when the cardiovascular system is unstable and multiple drugs are being started and titrated. Patients who do not improve significantly during antidepressant drug therapy or psychotherapy have a two-to threefold increase in cardiovascular events compared with patients who do improve substantially.³⁵

Key question 3: Is systematic screening for depression more effective than usual care for identifying patients with depression? Facilitating treatment of depression? Reducing depressive symptoms? Improving cardiac outcomes? Pignone et al conducted a literature review and meta-analysis on behalf of the USPSTF to clarify whether screening adults for depression in primary care venues improves recognition, treatment, and clinical outcomes.³⁶ They reviewed randomized trials conducted in primary care settings that evaluated the effect of screening for depression on identification, treatment, or health outcomes, including trials that examined integrated, systematic support for treatment after identification of depression. The meta-analysis suggested that screening and feedback of screening results reduced the risk of persistent depression. Stronger effects were observed with programs that integrated interventions to improve recognition and treatment of depressed patients and that incorporated quality improvements into clinic systems as compared with programs that provided only screening and feedback.

Screening patients with CHD, especially those with ACS, is more effective than usual care (no screening). Indeed, many depressed patients go undetected unless identified by screening. Not surprisingly, those identified by screening tend to have milder depression. For example, half the participants in SADHART had a HAM-D score less than 18 (≥ 25 is considered severe). However, a considerable number identified by screening had above-average HAM-D scores but either viewed their symptoms as appropriate for their medical condition or were simply denying their psychiatric illness. If patients had not been screened, their depression would not have been detected or treated. Even among SADHART participants with baseline HAM-D scores less than 18, those whose depression failed to respond to sertraline/placebo had twice the 7-year mortality rate as did those whose depression remitted.¹⁶

The Prevention of Suicide in Primary Care Elderly: Collaborative Trial (PROSPECT) examined the impact of a care management intervention on suicidal ideation and depression in older primary care patients and reported outcomes over a 2-year follow-up.³⁷ PROSPECT screened 9,072 patients in 20 primary care practices to find 599 study participants aged 60 years or older with major or minor depression (not all had cardiovascular disease). Participants were randomly assigned to either usual care or the PROSPECT intervention, which consisted of services rendered by trained care managers who offered algorithm-based recommendations to physicians and helped patients adhere to treatment during the 24-month trial. Compared with patients receiving usual care, those receiving the intervention had a greater likelihood of receiving antidepressant drugs and/or psychotherapy (85%–89% vs 49%–62%) and had a 2.2-fold greater decline in suicidal ideation over 24 months. Treatment response started sooner in the intervention group and continued to improve at 18 and 24 months; no appreciable increase in treatment response was observed in the usual-care group during the same period. Among patients with major depression, a significantly greater percentage of the intervention group achieved remission compared with the usual-care group at 4, 8, and 24 months. Patients with minor depression had favorable mental health outcomes regardless of treatment assignment. Sustained collaborative care maintained high utilization of depression treatment, reduced suicidal ideation, and improved the outcomes of major depression during 2 years of follow-up.³⁷

No randomized trial of SSRIs has been designed to show an effect on mortality or cardiovascular events. ENRICHD has produced the strongest evidence that an SSRI can reduce mortality or cardiovascular outcomes. The study was designed with approximately 85% power to detect a 25% to 30% reduction in the primary end point (death or MI) as a result of CBT; as noted above, 2,481 patients (1,834 with depression) were randomized to usual care or CBT within 28 days of MI.⁴ Depression significantly improved with CBT, but the rate of death or MI did not: during mean follow-up of 29 months, death or MI occurred in 299 patients in the CBT group versus 300 in the usual-care group.⁴ A post hoc analysis specific to the 1,834 ENRICHD participants who had depression aimed to determine the effects of antidepressant drugs on morbidity and mortality.¹⁸ The protocol required patients in the intervention group with scores of 25 or higher on the 17-item HAM-D, or those who had less than a 50% reduction in BDI scores after 5 weeks of treatment, to be referred to study psychiatrists for consideration of pharmacotherapy. Study psychiatrists met with patients who were being treated with antidepressant drugs to monitor medication use. Unless contraindicated or previously ineffective or poorly tolerated, sertraline was the first antidepressant used. Using a time-dependent multivariable Cox proportional hazards regression model to adjust for baseline depression score and cardiac risk factors, the researchers found SSRI use to be associated with a statistically significant 43% lower risk of death or nonfatal MI. Like other SSRI studies, ENRICHD found SSRIs safe for post-MI patients and to possibly, but not certainly, reduce death and MI.¹⁸