Introduction

Delirium affects up to 50% of hospitalized older adults and is estimated to cost the US health care system $150 billion per year.^1,2 Delirium is associated with poor outcomes, including prolonged hospitalization, institutionalization, dementia, and death.^3-5 Agitation due to delirium is often treated with antipsychotic medications, such as haloperidol or olanzapine,⁶ and guidelines consider them the medications of choice.⁷ However, antipsychotics are associated with an increased risk of death from respiratory and cardiac causes.^6,8 Alternative approaches to managing the behavioral manifestations of delirium are, therefore, greatly needed.

Alpha-2 receptor agonists reduce catecholamine outflow resulting in anxiolysis and sedation.^9-12 Both clonidine and dexmedetomidine are utilized to manage agitation or hyperactivity due to delirium but are associated with hypotension and bradycardia.^13,14 Furthermore, dexmedetomidine use is often restricted to the ICU.¹¹ Clonidine must be administered multiple times per day at varying doses to provide sedation without hemodynamic compromise,¹⁵ and abrupt discontinuation may be associated with rebound sympathetic phenomena.^16-18

Guanfacine is an oral alpha-2 receptor agonist commonly used to treat attention-deficit/hyperactivity disorder. Recently, guanfacine has been used off-label for the management of irritability, impulsivity, and agitation in hospitalized older adults with dementia.¹⁹ The mechanism by which guanfacine exerts its effects are not fully understood; however, it is thought to reduce sympathetic outflow at post-synaptic alpha-2A receptors in the dorsolateral prefrontal cortex.^19,20 Guanfacine has a much higher affinity to the alpha-2A subtype versus alpha-2B compared with clonidine.²¹ This mechanism, in addition to guanfacine’s activity in the locus coeruleus, is thought to result in enhanced working memory, attention, and arousal control.^21,22

In 2021, based on a case series describing its use in delirium,²³ the inpatient geriatrics consult service at our institution began administering guanfacine to manage the behavioral manifestations of delirium in older adults in an effort to reduce or prevent the use of antipsychotics. The dosing was typically started at 0.5 mg twice daily based on the case series referenced. There was no algorithm developed at this time, and dose changes, monitoring, and discontinuation of guanfacine was at the discretion of the provider. The objective of this study was to estimate guanfacine’s effect on delirium resolution, need for other psychoactive medications, and clinically relevant safety events in hospitalized older adults experiencing agitated delirium.

Methods

Results

Demographics and Clinical Characteristics

Ninety-six patients were ordered guanfacine during the study period, and 56 (58%) were included in the final analysis (Figure). Table 1 shows demographic data, and Table 2 provides details of hospital course. Mean age was 80.5 ±7.1 years, and most patients were male (40/56, 71%) and white (53/56, 95%). Most patients were admitted to a surgical service (32/56, 57%), and hypertension was the most common comorbidity (44/56, 79%). The median hospital length of stay was 17 (8.5 to 25.8) days, and 54/56 (96%) patients received a geriatrics consult. Patients were discharged to a variety of situations as seen in Table 2, and 4 (7.1%) patients expired during their hospitalization.

View Figure

• Download as Powerpoint
• Download Figure

TABLE 1Baseline demographics and clinical characteristics of the study group

View Fullscreen

TABLE 2Characteristics of hospital course among the study group

View Fullscreen

Discussion

This is the first cohort study to describe the administration of guanfacine to manage the behavioral manifestations of delirium in hospitalized patients who were all 65 years of age or older. We found that guanfacine administration at a dose of 0.5 to 2 mg per day was associated with a significant decrease in the incidence of delirium by 21% over 72 hours. Furthermore, the need for antipsychotics decreased by 2.2 OE per day; however, this finding was not statistically significant, possibly due to the relatively small size of the subgroup in which we assessed efficacy. We also found a decrease in the number of patients requiring antipsychotics after administration of guanfacine that trended toward statistical significance. The lack of a protocol precluded the routine discontinuation of antipsychotics, and future studies should incorporate a discontinuation algorithm if the initial response to guanfacine is favorable in an effort to reduce risks associated with polypharmacy.²⁸ Bradycardia or hypotension occurred in about a third of patients and required intervention in 12% of all study participants. Based on other studies of guanfacine referred to in the following discussion, this was a larger than expected outcome for bradycardia and hypotension. Future prospective studies should evaluate whether this is clinically significant and if steps could be taken to mitigate this risk.

Three other studies describe the administration of guanfacine to manage the behavioral manifestations of delirium in adults. The first was a case series of 7 patients that included both ICU patients who were being weaned off a dexmedetomidine infusion (n = 4) and non-ICU (n = 3) patients.²³ They documented that delirium and concomitant psychoactive medication use decreased within 3 days of guanfacine initiation,²³ and our finding of a significant 21% decrease in delirium over 72 hours is consistent with this. In this study, no patient developed clinically relevant hypotension or bradycardia.²³ This was a small sample size of just 7 patients, which may not accurately reflect a larger population.²⁹

The second study was a retrospective chart review and included 105 adult ICU patients (median age 59 years) who were being weaned off dexmedetomidine.¹³ The reported median dose was 1.5 (1 to 2.2) mg/day. At the time of guanfacine initiation, 41% of patients screened positive for delirium using the CAM-ICU. Dexmedetomidine was discontinued within 48 hours in 58% of patients and in 71% of patients within 72 hours. Hypotension, defined as a systolic blood pressure <90 mmHg, MAP <65 mmHg, or need for vasopressors, occurred in 8% of patients. Bradycardia, defined as a heart rate <60 beats per minute, occurred in 2% of patients.¹³ Eleven (10%) patients required vasopressors after developing hypotension, but these patients were critically ill.¹³

The third study was a retrospective chart review and included both ICU (n = 95; 64%) and non-ICU (n = 54; 36%) patients with a mean age of 58 years.²⁹ The majority (86%) of patients had hyperactive delirium. The guanfacine doses administered in the present cohort (daily and maximum doses of 1.0 [0.5 to 1.0] mg/day) were similar to those reported in this study of a mean initial dose of 1.82 mg/day with a mean maximum dose of 2.61 mg/day. These authors reported a 25% reduction in acute sedative use at 71.6 ± 39 hours,²⁷ which again is consistent with the 34% reduction in antipsychotic use at 72 hours that we documented. However, we cannot attribute these changes directly to guanfacine, and further work is needed to distinguish the effect of guanfacine from natural resolution of symptoms during a hospital stay. Hypotension, defined as a systolic blood pressure <90 mmHg requiring intervention, was not observed, and 2 (1%) patients developed bradycardia, defined as a heart rate <50 beats per minute.²⁹ In contrast, approximately 13% of our patients developed hypotension that needed intervention, and 41% had at least 1 instance of bradycardia. This difference may be explained not only by our study participants’ older age, differing diagnostic profiles, and high rates of concomitant antihypertensive and heart rate controller use, but also, it may be an artifact of our requiring only a single abnormal measurement to define either condition. Jiang et al²⁹ had the more stringent criterion of requiring 2 consecutive abnormal measurements to define either hypotension or bradycardia. A future opportunity includes proactively and temporarily reducing heart rate controllers or antihypertensives prior to guanfacine initiation.

Our study has several limitations. As the parameters needed for a power analysis were not available, this is an exploratory analysis, and significance should be interpreted in this context. This was a descriptive study that will be used to determine a sample size calculation for a prospective study as safety data in this patient population were not available prior to commencement. The retrospective design and lack of a comparator group limited our ability to determine if guanfacine affected delirium incidence or if it would have resolved via natural progression and whether adverse events such as hypotension or bradycardia were drug-related or associated with this older population. The single-center design means our results may not be generalizable to other settings. More than 90% of our population was white, and more than 70% male, which limits generalizability to other populations. Last, the involvement of the geriatric service, which focuses on nonpharmacologic interventions for delirium prior to medication intervention, may have introduced provider or selection bias into our findings.

In order to adjust for the above limitations, some of these authors are conducting another retrospective study of patients who received a geriatric consult for delirium comparing patients who received guanfacine with those who did not. In addition, plans are underway to conduct a prospective safety trial, which will help inform the risk of bradycardia and hypotension. A future efficacy trial comparing guanfacine to standard of care (antipsychotics) will help inform if there is a reduction in outcomes such as falls, pneumonia, or death. The study that informed the boxed warning for antipsychotics had an average duration of treatment of 10 weeks,³⁰ which may be longer than the duration used for symptoms of delirium. Studies show that inadvertent continuation of antipsychotics is common upon discharge,³¹ which could lead to longer term use involving further harms.

Part of the explorations of future studies should include dose finding. This could determine if guanfacine follows a similar pattern to clonidine wherein higher doses cause more stimulation of alpha1-adrengeric receptors and maintenance of blood pressure, whereas lower doses maintain alpha2-adrengergic receptor selectivity and reduce blood pressure.³² Until further studies are conducted, guanfacine could be considered if standard treatment of behavioral manifestations of delirium fail.

Conclusion

Guanfacine may reduce the behavioral manifestations of delirium in hospitalized patients who are 65 years of age or older albeit with some adverse effects. Future prospective controlled studies are needed to better evaluate the effectiveness and safety of guanfacine in this older population and to validate our findings.