Polypharmacy in Palliative Care for Advanced Heart Failure: The PAL-HF Experience

Abstract

Background:

Palliative care (PC) in advanced heart failure (HF) aims to improve symptoms and quality of life (QOL), in part through medication management. The impact of PC on polypharmacy (>5 medications) remains unknown.

Methods and Results:

We explored patterns of polypharmacy in the Palliative Care in HF (PAL-HF) randomized controlled trial of standard care versus interdisciplinary PC in advanced HF (N=150). We describe differences in medication counts between arms at 2, 6, 12, and 24 weeks for HF (12 classes) and PC (6 classes) medications. General linear mixed models were used to evaluate associations between treatment arm and polypharmacy over time. The median age of the patients was 72 (62–80) years, and 47% were female while 41% were Black. Overall, 48% had ischemic etiology, and 55% had an ejection fraction </= 40%. Polypharmacy was present at baseline in 100% of patients. HF and PC medication counts increased in both arms, with no significant differences in counts by drug class at any time-point between arms.

Conclusion:

In a trial of patients with advanced HF considered eligible for PC, polypharmacy was universal at baseline and increased during follow-up with no effect of the palliative intervention on medication counts relative to standard care.

Lay Summary

Palliative care aims to improve symptoms and quality of life for people with serious illness. In advanced heart failure, palliative care helps patients manage symptoms in part through medications, which may increase the total number of medications prescribed. This is known as polypharmacy (>5 medications). The effects of polypharmacy on patient outcomes in advanced heart failure is unknown. This study showed that the number of medications prescribed in the PAL-HF trial increased in both the intervention and standard care groups. Despite increasing medications, symptoms and quality of life improved, which may be beneficial for patients with advanced heart failure.

Polypharmacy, defined by the World Health Organization (WHO) as use of ≥ 5 medications daily,¹ is common in advanced heart failure (HF).² In non-HF cohorts, polypharmacy has been associated with worse outcomes, such as increased falls, dementia, delirium, forgetfulness, and lower patient-reported quality of life (QOL),² particularly in geriatric populations. In advanced HF, the concept of polypharmacy as an undesirable aspect of medical care is complicated by the pressing need to follow guidelines³ and a lack of agreement on what forms of therapeutic de-escalation are appropriate and safe.^4,5 In the context of palliative care (PC), the objectives of treatment are more focused on improving QOL rather than prolonging survival. However, the interaction of guideline-directed medical therapy and PC goals in this population is largely unexplored.

We performed a post-hoc analysis of the Palliative Care in HF (PAL-HF) trial to describe the patterns of medication use and to evaluate the effects of the PAL-HF intervention on the amount of polypharmacy present.

Methods

The PAL-HF trial is a previously reported randomized controlled trial designed to test the efficacy of a multidisciplinary PC intervention on QOL in patients with advanced HF.⁶ The study protocol was approved by the institutional review board, and all participants gave informed consent. Patients admitted to a single center with a primary diagnosis of acute decompensated HF with advanced disease were randomized to standard care (N=75) versus standard care and an interdisciplinary PC intervention (N=75). The intervention included hospital and clinic visits by a PC nurse practitioner who addressed multiple domains of QOL, including physical symptoms, psychosocial and spiritual concerns, and advance care planning. While medical therapy de-escalation was not a specific goal of the PAL-HF intervention, the assessment included a review of medications to maximize use of medications that improve QOL and reduce medications deemed no longer beneficial or consistent with patients’ goals.

Polypharmacy was measured by medical record review of prescribed and discontinued drugs and was present at baseline for all patients. We described the counts for classes of PC and HF medications using the mean and standard deviation. We used general linear mixed models to evaluate the differences in medication use by drug class between treatment versus standard care at 2, 6, 12, and 24 weeks.

Results

The median age of the patients was 72 (62–80) years, and 47% were female while 41% were Black. Overall, 48% had an ischemic etiology, and 55% had an ejection fraction </= 40%.

At 2 weeks following hospital discharge, the mean count for HF medications in the intervention arm was 6.7 (5.2) versus 6.5 (5.0) in the control arm (p=0.84), and the mean count for PC medications was 3.6 (3.4) versus 3.0 (2.9) (p=0.25), respectively. At 6 months, the mean count for HF medications in the intervention arm was 6.9 (4.3) versus 7.4 (5.8) in the control arm (p=0.58), and the mean count for PC medications was 4.1 (5.7) versus 4.9 (6.7) (p=0.49), respectively (Table 1). Overall, medication counts increased for both the intervention and control groups from baseline to 6 months across all HF and PC medication classes (Figure 1). From week 2 to 24, the mean count for HF medications changed from an average of 6.7 to 6.9 in the intervention arm versus 6.5 to 7.4 in the control arm. The mean change for PC medications was 3.6 to 4.1 in the intervention arm and 3 to 4.9 in the control arm (Table 1).

Table 1.

Change in HF medication counts over time by treatment arm.

Outcome	Control	Intervention	Estimate (CIs) Intervention vs Control	P-Value
Evidence-Based HF Medications	M (SD)	M (SD)
Week 2	6.5 (5.0)	6.7 (5.2)	−1.19 (−3.83, 1.45)	0.37
Week 6	4.9 (3.6)	5.6 (4.3)	0.23 (−2.64, 3.10)	0.25
Month 3	5.0 (3.4)	5.9 (4.3)	0.40 (−2.25, 3.05)	0.25
Month 6	7.4 (5.8)	6.9 (4.3)	−2.29 (−8.34, 3.76)	0.58
Palliative Care Medications	M(SD)	M(SD)
Week 2	3.0 (2.9)	3.6 (3.4)	0.26 (−2.63, 3.15)	0.25
Week 6	2.4 (2.8)	2.8 (3.3)	0.58 (−2.35, 3.52)	0.19
Month 3	2.5 (2.8)	2.8 (3.3)	−0.18 (−3.21, 2.85)	0.55
Month 6	4.9 (6.7)	4.1 (5.7)	−1.54 (−4.82, 1.74)	0.49

CI, confidence interval; HF, heart failure; M, mean; SD, standard deviation.

Figure 1.

Changes over time in palliative care and heart failure medications.

When individual classes of HF and PC medications were evaluated in both arms, the most prevalent prescribed HF medications were beta-blockers, statins, and diuretics (Table 2). The most common PC medications were opioids, antidepressants, and over-the-counter medications for symptom management, such as non-steroidal anti-inflammatory drugs (Table 2).

Table 2.

HF, PC, and OTC medication counts by drug class in the treatment arms (n=150).

	Control (N=72)	Intervention (N=70)	P-value
Palliative Care	M (SD)	M (SD)
Opioids	2.1 (3.1)	1.9 (2.5)	0.73
Anxiolytics	0.3 (1.1)	0.3 (0.8)	0.25
Antidepressants	0.4 (0.8)	0.3 (0.6)	0.51
Antipsychotics	0.1 (0.4)	0.1 (0.4)	0.98
Benzodiazepines	0.1 (0.2)	0.1 (0.4)	0.93
Heart Failure (Evidence-Based)
ACE/ARB/ARNI	0.4 (0.7)	0.4 (0.5)	0.65
Beta-blockers	0.7 (0.8)	0.7 (0.8)	0.73
Ca Channel blockers	0.3 (0.6)	0.2 (0.4)	0.52
Antiarrythmics	0.2 (0.5)	0.2 (0.6)	0.91
Digitalis glycosides	0.0 (0.2)	0.1 (0.3)	0.62
Anti-anginals, Nitrates	0.7 (1.2)	0.6 (0.9)	0.98
Statins (all subclasses)	0.8 (1.0)	0.7 (0.9)	0.53
Platelet inhibitors	0.0 (0.0)	0.0 (0.0)	1.0
Anticoagulants	0.4 (1.0)	0.4 (0.8)	0.57
Aldosterone Antagonists	0.0 (0.0)	0.0 (0.0)	1.0
Diuretics	1.4 (1.7)	1.6 (1.2)	0.23
Electrolyte supplements (K+, Mag, Ca+, etc)	1.1 (1.6)	0.9 (1.3)	0.55
Over-the-Counter
Herbal supplements	0.1 (0.4)	0.1 (0.3)	0.51
Other symptom therapy for palliative care	1.6 (3.0)	1.2 (2.1)	0.50
Other/antiemetic/antacid/vitamin	0.7 (1.4)	0.9 (1.6)	0.53
Other pain/Nonsteroidal anti-inflammatory drugs	7.7 (15.7)	6.1 (12.9)	0.40

The values are expressed as the mean (SD), unless otherwise indicated.

ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; HF, heart failure; OTC, over-the-counter; PC, palliative care; SD, standard deviation.

Discussion

In the PAL-HF trial, polypharmacy as defined by the WHO was universal, and the number of prescribed medications actually increased over time. Importantly, this result occurred in the context of improved patient-reported QOL among those in the intervention arm.⁶

Historically, PC studies have found that achieving the key goals of PC, such as ameliorating symptoms, reducing the frequency of symptom exacerbations, and improving comfort, often necessitates a burgeoning number of medications.² In geriatric populations, these interventions may have competing effects on patient QOL and on the quality achieved by the palliative interventions. In that context, it is encouraging that in the PAL-HF trial the aim to improve symptom management and QOL as a primary outcome of care was achieved despite overall increases in medication counts.

Striving for “appropriate polypharmacy,” that is, negotiating and designing prescribed medications together with patients, and individualizing the regimen to address each patient’s clinical context and multiple comorbidities can be both safe and effective.⁷ While such approaches increase medication burden, they may be advantageous to patients’ symptom experience, reducing the frequency or intensity of daily symptom episodes. And yet, in contrast to the PAL-HF trial, previous studies have been unable to demonstrate improvements in clinical outcomes, such as QOL.

Clinically and practically, what does the medication count infer? Outside of PC, studies in similar populations of patients with advanced HF report similar rates of polypharmacy.^2,8 More recently, guideline recommendations to prescribe increasing numbers of medications to achieve quality metrics has made de-escalation of polypharmacy a growing challenge.³

As noted in a review of the Get With The Guidelines-HF® registry from 2008–2013,³ adding medications to an evidence-based HF regimen is indicated in >25% of all patients.² In addition, a systematic review reported that the most pressing need in PC for HF is a focus on medication management to ameliorate symptoms and improve QOL, and yet, the effect of this type of intervention on overall polypharmacy has not been previously reported.

In the setting of advanced HF, which represents a confluence of multimorbidity with frequent, severe symptom exacerbation, these findings suggest that a multidisciplinary intervention with frequent patient-clinician communication, medication review, and individual persistent attention to polypharmacy may be beneficial. Ongoing attention to “appropriate polypharmacy” in medication management may effectively optimize drugs, such as diuretics, that serve two ends, addressing both PC symptoms (shortness of breath) and underlying physiologic shifts in advanced HF (fluid overload). Additionally, 6-month counts of PC meds reinforce our concept of “appropriate polypharmacy” in that the use of more PC meds is not necessarily better. Thus, the PC intervention may have allowed more judicious use of PC meds that were less well used or overly used in the control arm.

Anecdotal experience in the PAL-HF trial with attempts to stop statin therapies in the PC arm suggests that even when therapies have a minimal chance of providing patient benefit, the act of de-escalation may be perceived by patients as a signal that the care team is “giving up.” Whether and to what extent de-escalation of different classes of medications commonly used in advanced HF provides discernably better patient outcomes is currently poorly defined.

Cumulatively, these observations suggest that PC interventions in the PAL-HF trial may have mitigated the burden of polypharmacy.

Limitations

Several caveats should be considered in interpreting the results presented in this report. First, the sample size was relatively small, and patients were enrolled at a single academic center. Second, the duration of follow-up was short, and longer term follow-up of the patterns of medication use in these patients remains undefined. Third, we do not know how much selection for PAL-HF enrollment affected the patterns of medication use relative to comparably ill advanced HF patients not considered for PC. Finally, we did not collect health insurance information to understand how different levels of drug coverage affected the number and type of medications used in this trial.

Conclusions

Polypharmacy was universal at enrollment into the PAL-HF trial and the number of medications prescribed increased in both the PAL-HF intervention and standard care groups over time. Optimizing medications for symptom management regardless of medication count, known as “appropriate polypharmacy,” may be an important component in interventions to improve QOL for patients with advanced HF.

Highlights.

• Palliative care (PC) interventions designed for advanced, complex chronic illness aim to improve symptoms and QOL, in part through medication management.

• Few PC interventions report the effect of the intervention on patterns of polypharmacy, or the relationship of those patterns with QOL.

• In the PAL-HF trial, polypharmacy was universal at enrollment, and the number of medications prescribed increased in both the intervention and standard care groups over time.

• Though medications increased in both arms, patient-reported QOL improved among those in the intervention arm compared with the control arm.

• Appropriate polypharmacy may be an important contributor to QOL for patients with HF receiving PC.

Patient Significance.

• Palliative care aims to improve symptoms and quality of life for people with serious illness, including advanced heart failure.

• Palliative care helps patients manage symptoms in part through medications, which may increase the total number of medications prescribed.

• This study showed that despite increasing medications, the palliative care intervention improved symptoms and quality of life in people participating in the intervention, which may be beneficial for patients with advanced heart failure.

Data Availability Statement

Due to the sensitive nature of the questions asked in this study, survey respondents were assured raw data would remain confidential and would not be shared. De-identified data are available with permission in the National Institutes of Health (NIH) data sharing repository.