Hopeful Heart: Hope for Improving QoL in Depressed HF Patients

Ileana L. Piña, MD, MPH; Bruce L. Rollman, MD, MPH


April 01, 2019

This transcript has been edited for clarity.

You can download free screening instruments from the Hopeful Heart Study.

Ileana L. Piña, MD, MPH: Hello. I'm Ileana Piña, from Wayne State University and the Detroit Medical Center, and I am here in New Orleans at the American College of Cardiology (ACC) meeting, where there is a lot of buzz. The exhibition center is full of energy and a lot of good work is being presented. I'm happy to have with me Dr Rollman from University of Pittsburgh Medical Center (UPMC). What is your title?

If I had to boil it down to two words, it would be: Depression kills.

Bruce L. Rollman, MD, MPH: Director of the Center for Behavioral Health and Smart Technology and professor of medicine at the University of Pittsburgh.

Piña: Tell us about the Hopeful Heart study and the funding that made this possible.

Hopeful Heart Study

Rollman: We conducted the first clinical trial ever to look at the impact of treating both systolic heart failure (HF) and depression together versus doctor as usual care, and also versus care management for HF alone.[1]

Piña: Can you tell us more about blended care?

Rollman: It's based on the chronic care model which has been shown to be effective for treating HF, diabetes, hypertension, chronic obstructive pulmonary disease (COPD), and other chronic diseases, including depression and anxiety. But never before has it been used in an HF population.

Our team at the University of Pittsburgh published a paper in JAMA some years ago looking at the impact of treating post-coronary artery bypass graft (CABG) depression.[2] We found that these patients recovered faster, had less pain, and—significantly—it saved a median of about $2000/year after bypass surgery if the depression was recognized and treated. Our treatment method was all done over the phone with medical nurses contacting patients and recommending lifestyle changes, socialization, adequate sleep, quitting smoking, physical activity, and medications if they were interested.

We found that those patients recovered significantly. We had a usual-care control group, but I always wondered if they improved simply because nurses were calling them. There may be some nonspecific benefits. So, we recently finished the Hopeful Heart trial and opened the blind. It was a $7.5 million–dollar trial funded by the National Heart, Lung, and Blood Institute. We had zero funding from the device or drug industry; this was all funded by US taxpayer dollars. We had a team of steady nurses who screened HF patients for depression at eight Pittsburgh-area hospitals.

Piña: Why did you pick HF with reduced ejection fraction?

Rollman: It's a great question. I'm an internist by training, and when I attend on the medical wards I see a lot of what I call chart lore, where people are labeled as having HF but it's unclear if they have HF with preserved function or if they have HF at all. It could be COPD. So, we required participants to have an EF at or below 45%. We did not want people to say we had a mix of HF patients and that maybe people with COPD or some other condition were accidentally put in the mix.

We enrolled a total of 756 patients. All of them had an EF of ≤ 45%, and the mean EF was 28%.

Piña: This is the real McCoy.

Rollman: Yes, and people also had to have New York Heart Association (NYHA) classification II, III, or IV HF. They had to be somewhat symptomatic.

Evaluation for Depression

Piña: How did you determine whether they were depressed or not? What instruments did you use?

Rollman: In 2008, the ACC and the American Heart Association (AHA) published a joint guideline advocating routine screening of cardiac patients for depression using the Patient Health Questionnaire (PHQ-2) followed by the PHQ-9.[3]

Piña: Why the two instruments?

Rollman: When the ACC/AHA guideline was published in Journal of the American College of Cardiology and Circulation jointly, they did not recommend how far apart to do the screenings. We needed to consent people in person according to our investigational review board, and the inpatient stay was the only time our team met the patients face to face. The PHQ-2 screening questions are: Do you feel down, depressed, or hopeless or tired with little energy on most days in the past 2 weeks? We screened people with the PHQ-2 in the hospital but did not want to label them depressed in the hospital because being in the hospital and having an HF exacerbation (eg, little energy, trouble sleeping) could be the reason they are symptomatic. Then we called people 2 weeks after they were discharged and screened with a PHQ-9. We found that 46% of patients screened positive for depression in the hospital, but when we called those people 2 weeks later, 36% scored a 10 or above on the PHQ-9, indicating a clinically significant level of depression.

Piña: That is a number that you use as sort of the dividing line. How does that compare to the Beck Depression Inventory (BDI)? I look at the BDI and I think > 14 is very significant depression.

Rollman: I'm a very practical guy; the BDI is about 17 items and there are different versions, whereas the PHQ-9 is nine items. The PHQ-9 is a more current way to screen for depression. It's been popularized by the Robert Wood Johnson Foundation and MacArthur Foundation. They both had initiatives on screening and treating depression. The PHQ-9 is the most popular tool used in current clinical practice. It's a shorter response burden and can be administered over the telephone. It's easy to remember the 5-point cut-offs: < 5 for minimal depression; 5-9 for mild depression; 10-14 for moderate depression; 15-19 for moderately severe depression; 20-27 for severe depression. The maximum score is 27.

Piña: Those are the people you randomized.

Rollman: Yes. In addition, because we knew at the start that some of these patients would die, be readmitted, or develop other issues, we also enrolled a nondepressed comparison cohort of HF patients. We had two controls. The nondepressed comparison patients screened negative for depression in the hospital and then 2 weeks later they scored < 5. They were truly not depressed.

Piña: How long was your follow-up time after the randomization?

Rollman: Our primary outcome point was at 12 months. Our main outcome measure was health-related quality of life. Mood symptoms, readmission, and cardiac function (measured with the Kansas City Cardiomyopathy Questionnaire) were secondary measures. We also studied healthcare costs and mortality.

Hopeful Heart Results

Piña: What were your results?

Rollman: Depression was really common. Almost half of the patients in the inpatient setting reported symptoms, and a third of those had persistent elevations in mood symptoms 2 weeks after discharge home. Number two, the patients who were depressed compared with the nondepressed controls had similar ejection fraction at baseline. Both were 28%.

The rates of diabetes, hypertension, post-myocardial infarction, bypass, and use of beta-blockers were the same. But when they self-reported their health, they were more likely to be in NYHA class III or IV HF, and they rated their function lower on the Kansas City Cardiomyopathy Questionnaire. Essentially, the only major difference was mood symptoms. Over the course of follow-up, we found that we were able to support our primary hypothesis: that patients randomized to the blended model had significantly improved health-related quality of life as measured by the SF-12 mental health composite score compared with patients randomized to their doctor's usual care.

As I mentioned in the beginning, we wanted to have the attention care controlled. I can tell you now, conclusively, that patients randomized to the blended model also had significantly reduced levels of mood symptoms compared with patients randomized to the collaborative care for HF alone as well as to usual care. In those cases, our primary hypotheses measures were supported. On the other hand, to date we have not seen a reduction in hospital readmission. The overall incidence of readmission to the hospital was roughly 90% among the depressed and 75% among the nondepressed.

Piña: Was that statistically significant?

Rollman: It was kind of close. Some of the patients were hospitalized 10 or 20 times, and a problem with the Kaplan-Meier analysis is that it only captures one.

Piña: You should do the repeated-events type of analysis.

Rollman: Exactly, and we should look at the median number because there might be differences there. Now, you only die once, so with the mortality data, we had about a 14.5% versus 10.5% incidence of all-cause mortality.

Piña: It's still pretty encouraging, though. What's next?

Next Steps for Hopeful Heart

Rollman: We need to do a couple of things. We need to delve down. We only unblinded the data a couple of weeks ago.

Piña: Are you looking at gender differences?

Rollman: That is important in studies of heart disease and depression. Women are twice as likely as men to experience depression or anxiety disorders, so we looked at that too. We powered the trial to look at gender differences.

Piña: You have half and half. You have enough women in here that you should be able to analyze it. We are going to be looking forward to your next study and I may ask you back again.

Final Thoughts

Piña: What do you want to tell doctors?

Rollman: The bumper sticker is: Depression kills. If I had to boil it down to two words, it would be: Depression kills. It's really important to consider the ACC/AHA guidelines about screening your patients for depression, particularly if they are not getting well. If you have them on maximal therapy for HF, they are on the right angiotensin-converting enzyme (ACE) inhibitors, or maybe they have an automatic implantable cardioverter-defibrillator and they are still reporting really poor quality of life, it's very important to consider that this person may have comorbid depression or anxiety.

There are a number of HF care management programs around the country, such as Veterans Affairs, UPMC, Highmark, Intermountain, and Kaiser. They have collaborative care programs for HF, but to the best of my knowledge, they do not routinely screen for depression. I think for our cardiologists and primary care doctors in the room, screening could be routinely assigned to the nurse care managers in these HF programs.

Piña: It's very easy to give the questionnaire in the office. We were doing the BDI years ago in the clinic. In our world, about 35%-40% of HF patients are depressed. We are writing an American Heart paper about medication adherence. Another idea for you is to see what happened with medication adherence, because depression can have a large impact on medication adherence.

Rollman: We will be looking at that. Some of your viewers may also be paid on pay-for-performance, in which case you are looking at the rates of ACE inhibitors and beta-blockers. Again, if your patients are not on the right medications, could it be because they have comorbid mental illness issues?

Piña: Or they are not taking them. A whole issue for us has been adherence.

I want to thank you for coming today, and I want to wish you a lot of luck because this is the kind of work that you do not get a pharmaceutical or device company to pay for. Kudos to the NHLBI for taking this on, and I wish you luck on the next. I hope you finish off this dataset. We've got a lot of work inside this dataset to do.

Rollman: If your readers and viewers come to our website, they can download some of our study instruments, like the PHQ. These are all free and linked to the papers.

Piña: I'm going to be downloading them for my clinic. Thanks again.

Rollman: Thank you very much.

Piña: Thank you to our viewers. I hope you found this very interesting. I certainly did. You can put this right into practice in your own clinic and your patients may appreciate it much more. This is Ileana Piña, signing off from New Orleans at the ACC. Have a great day.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.