Chlorpromazine as Treatment for Refractory Agitation Associated with Pediatric Delirium

Abstract

OBJECTIVE

Delirium and agitation can be devastating and prolong the length of hospitalization. As part of our continuous improvement efforts, we implemented the use of intermittent chlorpromazine therapy to target refractory agitation associated with hyperactive or mixed delirium (RAA-D). The purpose of this study was to evaluate the effectiveness of chlorpromazine on RAA-D and delirium symptoms as well as any adverse effects in critically ill children.

METHODS

Retrospective chart review was conducted for children admitted to the pediatric intensive care unit who were treated with chlorpromazine for RAA-D from March 2017 to January 2019. The primary end point was to determine differences in Cornell Assessment for Pediatric Delirium (CAPD) and State Behavioral Scale (SBS) scores 24 hours before and after chlorpromazine administration. The secondary end points were the 24-hour cumulative dosing of narcotic and sedative agents before and after chlorpromazine administration and adverse events associated with chlorpromazine use.

RESULTS

Twenty-six patients were treated with chlorpromazine for RAA-D; 16 (61.5%) were male with a median age of 14.5 months (IQR, 6–48). The mean CAPD (n = 24) and median SBS (n = 23) scores were significantly lower 24 hours after chlorpromazine use when compared to baseline scores, 12 vs 8.9 (p = 0.0021) and 1 vs −1, (p = 0.0005) respectively. No significant adverse effects were observed.

CONCLUSIONS

Chlorpromazine use in critically ill children with RAA-D was helpful for managing symptoms without adverse events. Further investigation is needed to evaluate the use of chlorpromazine to treat RAA-D to avoid long-term use of an antipsychotic.

Introduction

Delirium is a neuropsychiatric syndrome characterized by an acute change in the patient's neurocognitive baseline with disturbances in awareness and cognition and triggered by a new medical condition rather than a preexisting disorder.¹ Clinical presentation of delirium varies based on the 3 subtypes: hyperactive, hypoactive, or mixed delirium. Although it is usually a temporary state, delirium can lead to longer pediatric intensive care unit (PICU) and hospital length of stay (LOS), as well as increased mortality.² The pathophysiology of PICU delirium is complex, and it appears to occur as a result of predisposing and precipitating factors, including younger age, male sex, preexisting cognitive impairment, and/or developmental delay. Importantly, pharmacologic agents frequently used in the PICU, such as benzodiazepines and narcotics, are strong predisposing factors for delirium, and minimizing overuse of these agents has shown to be beneficial in preventing delirium.^3–5

Previous studies have reported an incidence of delirium in the PICU ranging from 12% to 65%.² Traube et al⁶ reported an incidence of 17% in 1547 pediatric critically ill patients; 78% of them developed delirium within the first 3 days of admission to the PICU. Most cases were described as hypoactive delirium (47%), whereas mixed and hyperactive were 45% and 8%, respectively. Delirium in critically ill children results in increased morbidities and potentially long-term cognitive impairment, necessitating the need for early identification and treatment.^6,7

In order to minimize the devastating effects of PICU delirium, standard management can be used which includes identifying the causes of delirium, providing non-pharmacologic interventions, and lastly judiciously using pharmacologic treatment with antipsychotics.^2,8,9,10 Even though there are no US Food and Drug Administration (FDA)–approved antipsychotic medications to treat delirium in adults or children, antipsychotic agents have been shown to be helpful in reducing agitation, confusion, sleep-wake disturbance, and shortening the duration of delirium symptoms.^8,11,12 Common antipsychotic agents used include haloperidol, risperidone, and quetiapine; however, there are possible adverse effects, such as agitation, hypotension, extrapyramidal symptoms (EPS), prolonged QTc interval, and anticholinergic symptoms. Common drug-to-drug interactions that can be of concern in the PICU include azole antifungals, antiepileptics, and antiarrhythmics, such as amiodarone. These adverse events and drug-to-drug interactions may limit their use in critically ill children.

The purpose of this study was to evaluate the effectiveness of chlorpromazine on RAA-D and delirium symptoms as well as any adverse effects in critically ill children.

Materials and Methods

Design and Population. This study was a retrospective chart review of critically ill children (ages <18 years) admitted to the PICU who were treated with intermittent intravenous (IV) chlorpromazine for RAA-D per our chlorpromazine administration protocol (Figure). Participants were identified based on the history of receiving chlorpromazine for delirium management while in the PICU. The study collected data from a single 19-bed PICU in an academic tertiary hospital from March 2017 through January 2019 following implementation of an ICU bundle including a delirium protocol, sedation protocol, and early mobilization protocol.⁸

Figure.

Delirium protocol: pharmacologic treatment.

End Points. The primary end point was the average 24-hour score for Cornell Assessment for Pediatric Delirium (CAPD) and the median State Behavioral Scale (SBS) score 24 hours before and after the first intermittent IV chlorpromazine administration. The CAPD score was assessed by nursing every 12 hours and SBS scores every 2 hours in patients on mechanical ventilation (MV). The secondary end points were the 24-hour cumulative dosing of narcotic and sedative agents before and after chlorpromazine administration and the presence of any adverse effects, which included anticholinergic side effects, oversedation, hypotension, QTc prolongation, EPS (such as acute dyskinesias and dystonia), and neuroleptic malignant syndrome.

Delirium Protocol Description. In December 2013, we implemented a delirium protocol as part of an ICU bundle.⁸ Patients were screened for delirium twice daily using the CAPD tool. The delirium assessment occurs during several hours to evaluate for fluctuating symptoms of delirium. A CAPD score of 9 or greater represents a positive delirium screen; diagnosis of delirium was then confirmed by a PICU physician or pediatric psychiatrist. The type of delirium was determined by assessing the patient's psychomotor activity and level of alertness using our PICU standard of care sedation assessment tools (sedation/analgesia algorithm for mechanically ventilated children)⁸ in combination with the CAPD scored items. If the patient's sedation score was consistent with deep sedation or unresponsiveness (SBS –2 or –3), then the CAPD was not performed. For all other sedation scores, the CAPD was performed, and if the score was consistent with delirium, then the delirium type was determined by evaluating the individual item scores (ranging from never to always) and consistency with degree of alertness and psychomotor activity. Items 1 through 4 and 7 through 8 on the CAPD tool were used to help determine if the patient had hypoactive, hyperactive, or mixed delirium. Ultimately, the identified delirium type was confirmed by a PICU physician or pediatric psychiatrist.⁸

In this guideline, antipsychotic agents (haloperidol, risperidone, or quetiapine) were used to treat hypoactive, hyperactive, or mixed delirium when nonpharmacologic strategies had been unsuccessful. Subsequently, our delirium pharmacologic guideline was updated in 2017 to include the use of short-term chlorpromazine on an as-needed basis for agitation refractory to sedation dose escalation in children with hyperactive or mixed delirium. This update was implemented as part of our quality improvement initiatives in an effort to avoid excessive use of sedative agents and the need to initiate maintenance antipsychotic therapy, thereby avoiding the associated harmful side effects and interactions with antipsychotic use.

The decision to treat delirium and RAA-D with pharmacologic agents was made by an interdisciplinary team, including a pediatric psychiatrist, PICU attending, pediatric clinical pharmacist specialist, and PICU nurse practitioner. The team evaluated the patient's clinical characteristics, disease process, and medications prior to deciding which pharmacologic agent would be more effective.

Chlorpromazine Protocol Administration. Chlorpromazine 0.4 mg/kg/dose IV (50 mg maximum single dose) was administered every 6 hours as needed to treat refractory agitation associated with hyperactive or mixed delirium for a maximum of 72 hours. All doses of chlorpromazine for IV use were prepared and diluted in the pharmacy and administered on an IV infusion pump during a minimum of 10 minutes to reduce the risk of hypotension. As per the delirium protocol, baseline and daily electrocardiograms were obtained while receiving chlorpromazine to assess for prolongation of the QTc interval.

Data Collection. The clinical assessment scores (CAPD performed every 12 hours and SBS performed every 2 hours) were obtained and recorded by nurses according to the ICU bundle. Data collection included the following: demographic information, primary and secondary diagnoses and severity at the time of admission, Pediatric Index of Mortality (PIM) score, length of MV, hospital and PICU LOS, CAPD and SBS scores, antipsychotic agents, cumulative doses of sedative and narcotic agents (hydromorphone, fentanyl, morphine, midazolam, and dexmedetomidine continuous infusions and intermittent IV lorazepam) that are commonly used in the PICU for MV patients. Clinical assessment scores, such as CAPD and SBS, were assessed for each patient. The individual mean score was collected for CAPD for 24 hours prior to and 24 hours after chlorpromazine use, while the median score was used for SBS and collected during the same time frame. Benzodiazepines were converted to IV lorazepam equivalents (mg/kg/day)^13,14 and opioids were converted to IV morphine equivalents (mg/kg/day).^14–16

Statistical Analysis. Data were analyzed using descriptive statistics for frequencies and ranges. To detect any differences in CAPD and SBS scores and sedation agent cumulative dosing between 24 hours before and after chlorpromazine treated critically ill children, paired t-test and Wilcoxon signed-rank test were used. All continuous variables were checked for normal distribution using the normality test and histogram. Analyses were performed with SAS version 9.4 (SAS Institute, Cary, NC).

Results

A total of 26 patients were treated with chlorpromazine for RAA-D. Sixteen patients (61.5%) were male, with a median age of 14.5 months (IQR, 6–48; Table 1). Median hospital LOS was 29 days (IQR, 19–80.9) median PICU LOS was 20.1 days (IQR, 9.4–78). Duration of MV was 9.2 days (IQR, 3.8–12), and PIM was −4.6 (IQR, −4.9 to −3.3; Table 1). Seven patients (27%) developed delirium at early onset (defined as ≤72 hours of admission), whereas 19 patients (73%) developed at delayed onset (defined as >72 hours of admission; Table 1). Of the 26 patients with RAA-D, 11 (42%) were concurrently treated with a traditional antipsychotic agent: 5 patients (46%) received quetiapine, 2 (18%) received quetiapine and haloperidol, and 4 (36%) received risperidone (Table 2). Of these 11 patients, 3 (27%) received chlorpromazine within 72 hours of admission and 8 (73%) after 72 hours of admission (Table 2). The as-needed number of doses of chlorpromazine administered ranged between 1 and 12, with a median of 6 doses in 72 hours (Table 3). One patient was treated with chlorpromazine and a traditional antipsychotic (quetiapine) and received chlorpromazine for an additional 72 hours (Table 3).

Table 1.

Baseline Characteristics of Study Patients

Characteristic	Total (N = 26)
Male sex, n (%)	16 (61.5)
Age, median (IQR), mo	14.5 (6–48)
Weight, median (IQR), kg	8.8 (5.9–18.1)
Admitting diagnosis, n (%)
Respiratory failure, pneumonia	9 (34.5)
Encephalitis	4 (15.3)
Congenital heart disease	3 (11.5)
s/p cardiac arrest; pneumonia	3 (11.5)
Respiratory failure, pulmonary hypertension	2 (7.7)
Medulloblastoma	1 (3.9)
Gastroschisis, volvulus	1 (3.9)
Respiratory failure, pulmonary hemorrhage, ECMO	1 (3.9)
Respiratory failure, streptococcal septic shock, VV ECMO	1 (3.9)
s/p heart/lung transplant	1 (3.9)
Delirium onset, n (%)
Early onset (≤72 hr of admission)	7 (27)
Late onset (>72 hr of admission)	19 (73)
Hospital LOS, median (IQR), days	29 (19–80.9)
PICU LOS, median (IQR), days	20.1 (9.4–78)
Length of MV, median (IQR), days	9.2 (3.8–12)
PIM score, median (IQR)	−4.6 (−4.9 to −3.3)
PIM ROM, median (IQR), %	1.0 (0.7–1.5)

ECMO, extracorporeal membrane oxygenation; LOS, length of stay; MV, mechanical ventilation; PICU, pediatric intensive care unit; PIM, Pediatric Index of Mortality; ROM, risk of mortality; s/p, status post; VV, venovenous

Table 2.

Characteristics of Patients Treated With Chlorpromazine + Traditional Antipsychotic

Characteristic	Total (n = 11)
Male sex, n (%)	9 (81.8)
Age, median (IQR), mo	22 (7–201)
Admitting diagnosis, n (%)
Respiratory failure, pneumonia	1 (9)
Encephalitis	3 (27.3)
s/p cardiac arrest; pneumonia	3 (27.3)
Respiratory failure, pulmonary hypertension	2 (18.2)
Respiratory failure, pulmonary hemorrhage, ECMO	1 (9)
Respiratory failure, streptococcal septic shock, VV ECMO	1 (9)
Traditional antipsychotic treatment, n (%)
Quetiapine	5 (46)
Risperidone	4 (36)
Haloperidol	2 (18)
Chlorpromazine treatment onset, n (%)
Early onset (<72 hr of admission)	3 (27)
Late onset (>72 hr of admission)	8 (73)
PICU LOS, median (IQR), days	18.3 (1.9–78)
Length of MV, median (IQR), days	9.5 (0–78)
PIM ROM, median (IQR), %	1.0 (0.7–1.5)

ECMO, extracorporeal membrane oxygenation; LOS, length of stay; MV, mechanical ventilation; PICU, pediatric intensive care unit; PIM, Pediatric Index of Mortality; ROM, risk of mortality; s/p, status post; VV, venovenous

Table 3.

Chlorpromazine Dosing

Chlorpromazine Dosing (n = 26)	Median
Starting chlorpromazine, median (IQR), mg/kg	0.4 (0.39–0.402)
Cumulative chlorpromazine dose 24 hr, median (IQR), mg/kg	2.1 (0.8–4.0)
Chlorpromazine doses 72 hr, median (IQR), dose	6 (1–12)

There was a median reduction in cumulative 24-hour dosing in opioids and dexmedetomidine after chlorpromazine use, 6.7 to 3.8 mg/kg/day and 1.5 to 1.3 mcg/kg/hr, respectively, although this was not statistically significant (p > 0.05; Table 4). In addition, no change in cumulative 24-hour dosing was detected in benzodiazepines after chlorpromazine use (p > 0.83; Table 4). The average 24-hour CAPD scores were significantly lower after chlorpromazine treatment compared with before chlorpromazine (12 vs 8.9; p = 0.002; Table 5). The 24-hour median SBS scores were significantly lower after chlorpromazine use compared with before chlorpromazine administration (1 vs −1; p < 0.001; Table 5). The difference in CAPD and SBS before vs after chlorpromazine in patients who also received traditional antipsychotic therapy was less than in patients who only received chlorpromazine. Only 3 patients (n = 11) had a CAPD score <9 after chlorpromazine and traditional antipsychotic treatment. No serious adverse effects (e.g., prolongation in QT segment, EPS, hypotension, or neuroleptic malignant syndrome) were reported in the study group.

Table 4.

Cumulative Dosing of Sedative and Narcotic Agents

Medications	Before Chlorpromazine	After Chlorpromazine	p value
Total morphine IV equivalents, median (IQR), mg/kg/day (n = 22)	6.7 (2–11.4)	3.8 (0.5–9)	0.21
Lorazepam cumulative equivalent IV, median (IQR), mg/kg/day (n = 21)	0.3 (0.015–1.97)	0.3 (0.015–0.86)	0.83
Dexmedetomidine IV, median (IQR), mcg/kg/hr (n = 23)	1.5 (0.7–1.8)	1.3 (0.086–1.73)	0.11

IV, intravenous

Table 5.

Cornell Assessment for Pediatric Delirium (CAPD) and State Behavioral Scale (SBS) Scores

CAPD and SBS Scores	Before Chlorpromazine	After Chlorpromazine	p value
Delirium, mean ± SD* CAPD score (n = 24)	12.0 ± 4.5	8.9 ± 3.2	0.002
Sedation, median (IQR)† SBS score (n = 23)	1 (−1 to 2)	−1 (−1 to 0)	<0.001

* Frequency missing n = 2 in the CAPD score.

^† Frequency missing n = 3 in the SBS score.

Discussion

Critically ill children are at high risk for delirium in the PICU with many predisposing factors. In our patient population, the mean CAPD and median SBS scores were significantly lower 24 hours after chlorpromazine use (p < 0.001) compared with baseline scores, and no significant adverse events were observed. In a previous study, we presented our delirium treatment protocol and found antipsychotic treatment with haloperidol, quetiapine, and risperidone improved delirium symptoms without significant adverse events.¹² Additionally, we found that patients with early-onset delirium (≤72 hours of admission) not responsive to non-pharmacologic strategies had a more effective response to antipsychotic therapy than patients with late-onset delirium.¹² This is important because many critically ill children who develop protracted delirium have refractory agitation. Currently, there is a paucity of research to support a standard management approach for RAA-D and guide clinicians' treatment choices.

Hui and colleagues¹⁷ recognized this treatment dilemma and described 3 main approaches when targeting RAA-D: 1) escalation of the dose of the same antipsychotic; 2) rotation to another antipsychotic; or 3) addition of a second pharmacologic agent (combination therapy). Although their research is with delirium in adults with advanced cancer, the same theoretical approach can apply to pediatric delirium. Each approach has potential advantages, such as the following: dose escalation maximizes the dose response curve; antipsychotic rotation uses the possible benefit of different spectrum of coverage; and combination therapy can potentially have fewer side effects by using the lowest dose of each agent. Disadvantages of each approach include the following: patients can be refractory to escalating dosages; there is the potential for different side effects from antipsychotic rotation; and there are logistical challenges when administering 2 agents in combination therapy.

Few IV antipsychotic options exist for treatment of hyperactive and mixed delirium. Haloperidol and chlorpromazine are the only typical antipsychotic agents available in IV formulation, and atypical antipsychotic agents are only available in oral and intramuscular formulations. Chlorpromazine, like haloperidol, has a high affinity for dopamine (D₂) receptors but with lower potency and less risk of EPS; therefore, they are a potential reasonable alternative drug. Chlorpromazine is an FDA-approved typical antipsychotic that acts as an antagonist on many different postsynaptic receptors, including dopaminergic, serotonergic, histaminergic, α₁/α₂, and muscarinic M1/M2. The pediatric FDA-approved indications include severe behavioral problems and short-term treatment of hyperactivity in children who show excessive motor activity with accompanying conduct disorder.^18,19 Despite its sedative effects, chlorpromazine has been infrequently used as therapy for difficult to sedate pediatric patients in ICU.²⁰

In this study, intermittent IV chlorpromazine was used to treat severe agitation associated with delirium in patients' refractory to escalating doses of sedation. Short-term use of chlorpromazine was safe and effective in treating both hyperactive and mixed delirium symptoms, and its associated refractory agitation. These results follow the extant evidence in the adult literature in which chlorpromazine has found to be efficacious in reducing delirium symptoms.^21,22 In a retrospective study with adults with delirium in an ambulatory palliative care center (n = 167), 128 received haloperidol and 39 needed a second antipsychotic. Most patients (91; 77%) had symptom improvement with haloperidol. However, 39 (23%) continued to have delirium (23 were hyperactive and mixed delirium) and 5 had an adverse event.²³ After haloperidol failure, 37 received chlorpromazine, with a success rate of 33%. In comparison, 50% (n = 13) of our patients responded to chlorpromazine alone. No adverse events were seen with chlorpromazine; however, 1 of 2 patients who received haloperidol experienced a dystonic reaction following administration.

Although we hypothesized that the use of chlorpromazine may limit the use of long-term oral antipsychotic therapy, 42% of the cohort received chlorpromazine and a traditional antipsychotic agent. In addition, the pre–post difference in CAPD and SBS score was less than that in patients who only received chlorpromazine. However, these patients were severely ill with acute neurologic pathology or cardiopulmonary disease and long-term MV and LOS.

Although there is a limited amount of data on the use of chlorpromazine to treat delirium in adults, there is even less in the pediatric population. A 2007 retrospective study (n = 84) described delirium cases seen by a child and adolescent psychiatry consultation service and reported that 1 patient received treatment with chlorpromazine without providing specific information on the case.²⁴ Additionally, as part of a review paper on infant delirium, Silver et al²⁵ identified chlorpromazine as an agent used to treat delirium in pediatric critical care settings. However, to our knowledge there are no other published articles specifically describing its use in pediatric delirium as part of a protocol, guideline, or research study.

This is the first study to introduce the use of chlorpromazine to treat RAA-D. A protocolized approach was used for the assessment, diagnosis, and treatment of delirium in patients with RAA-D. These preliminary data suggest chlorpromazine may be an effective short-term agent to treat RAA-D in critically ill children and potentially avoid long-term antipsychotic use in this high-risk population.

Limitations. There are several limitations to this study. The study is a retrospective single-center study with a small sample size and without a matched control group. RAA-D patients were defined as having persistent elevated CAPD (>9) for 24 hours or greater, with physical signs of agitation, and poor response to escalation in narcotics and sedatives. However, the time frame to initiation of chlorpromazine varied by provider, which can introduce selection bias. The exclusion of reported pain scores is also a limitation, but the delirium protocol includes pain assessment and treatment as an expected priority and optimized before escalating sedation to treat agitation or delirium. Finally, pharmacologic recommendations for hyperactive or mixed delirium in our delirium protocol included the use of chlorpromazine as needed in an IV formulation or quetiapine for oral maintenance therapy. The decision to start maintenance with or without intermittent IV therapy was expected to be a multidisciplinary decision including psychiatry, pharmacy, and the critical care team and based on the patient's acuity, disease process, comorbidities, and potential drug interactions. However, adherence to the protocol recommendations varied, particularly early in the study due to provider preference and comfort with antipsychotic therapy.

Conclusion. Chlorpromazine use in critically ill children with refractory agitation associated with hyperactive or mixed delirium was helpful for managing symptoms without adverse events. Further large-scale investigation is needed to evaluate the use of chlorpromazine to treat RAA-D and avoid long-term use of antipsychotic in this high-risk population.

ABBREVIATIONS

CAPD

Cornell Assessment for Pediatric Delirium

EPS

extrapyramidal symptoms

FDA

US Food and Drug Administration

ICU

intensive care unit

IV

intravenous

LOS

length of stay

MV

mechanical ventilation

PICU

pediatric intensive care unit

PIM

Pediatric Index of Mortality

RAA-D

refractory agitation associated with hyperactive or mixed delirium

SBS

State Behavioral Scale