Management of Hyperactive Delirium in the Pediatric Intensive Care Unit: Case Series of Three Young Children

Abstract

Children in the intensive care unit (ICU) are at high risk of developing delirium, given their underlying disease processes, the adverse effects of treatments and medications, and the stressful, abnormal environment. If prevention and nonpharmacologic measures to treat delirium are unsuccessful, atypical antipsychotics are considered, although they are not approved by Food and Drug Administration for the treatment of pediatric delirium and could have significant adverse side effects. This case report presents three pediatric patients with hyperactive ICU delirium that risked life-threating complications who were successfully treated with short courses of atypical antipsychotic medications.

Introduction

Delirium is a complex and multifactorial phenomenon that until recently went largely unrecognized and underdiagnosed in the pediatric population. ¹ It is defined by the American Psychiatric Association as an acute and fluctuating change in awareness and cognition and occurs in the setting of serious medical illness. This condition can be further subdivided into categories based on symptomatology, such as hypoactive, hyperactive, and mixed. Hypoactive delirium is the most common form; symptoms consist of somnolence, difficulty arousing, and disassociation from the environment. Children with hyperactive delirium often exhibit erratic behavior, hyperactivity, purposeless movements, and inconsolability.

Children in the intensive care unit (ICU) are at particular risk, given their underlying disease processes, the adverse effects of treatments and medications, and the stressful, abnormal environment that deregulates sleep–wake cycles. Pediatric delirium affects 20 to 30% of ICU infants and children, with a higher prevalence in mechanically ventilated patients but is more difficult to diagnose in younger and developmentally delayed patients. ^{2 3} Pediatric ICU delirium leads to increased mechanical ventilation days, hospital length of stay, health care costs, neurodevelopmental regression, and mortality. ⁴ Although less common than hypoactive delirium, hyperactive delirium poses a greater challenge in the ICU, as its symptoms can place children's safety at risk.

The treatment of delirium is, fundamentally, the treatment of its underlying causes. ⁵ Pediatric ICU delirium treatment algorithms have been evolving over the past decade. They provide systematic approaches for management including: recognizing underlying disease and iatrogenic factors, environmental modifications, and occasionally pharmacologic management. ⁴ After nonpharmacologic measures to prevent or treat delirium, atypical antipsychotics are used to treat the intolerably distressing or life-threatening symptoms of hyperactive delirium. ⁵

Atypical antipsychotic medications are used cautiously for the following reasons: (1) use of antipsychotics is off-label for pediatric delirium; (2) neurologic, cardiac, and metabolic side effects can be severe and, at times, irreversible ⁶ ; and (3) adult data have not demonstrated significant improvement in the duration of delirium with the use of both typical and atypical antipsychotics. ⁷ Some studies suggest that the use of atypical antipsychotics may be associated with increased mortality in older adult ICU patients, and high-dose off-label use could be associated with increased risk of unexpected death in children and adolescents. ^{8 9}

In light of these concerns, α-2 agonists (dexmedetomidine and clonidine) are alternate sedatives with emerging evidence of decreased rates of delirium. ^{1 10} A randomized controlled trial in critically ill adults demonstrated that dexmedetomidine reduced the incidence of delirium. ¹¹ One meta-analysis in children demonstrated that dexmedetomidine reduced the incidence of delirium but increased time to extubation and discharge. ¹²

A randomized controlled trial of quetiapine (an atypical antipsychotic) as treatment for ICU delirium in adults indicated benefits, including reduced agitation and delirium duration. ¹³ There have been pediatric case series, as well as a retrospective safety study, which suggest efficacy of quetiapine for pediatric ICU delirium without serious adverse events. ^{10 14 15 16} Other antipsychotic medications (risperidone and olanzapine) have been less well studied for pediatric ICU delirium, although both have the advantage of being easy to administer orally. A retrospective chart review of 19 children younger than 36 months old diagnosed with delirium found that the use of olanzapine ( n  = 16) or risperidone ( n  = 3) significantly decreased symptoms of delirium without significant adverse side effects. ¹⁷ A case report of a 6-month-old infant treated with risperidone reports similar results. ¹⁸

In July 2017, a 2-month-old infant with a complex congenital heart condition was transferred to the Pediatric Intensive Care Unit (PICU) of University of Virginia (UVA) for evaluation for surgical management. He was receiving risperidone (0.25 mg per dose) twice daily for previously diagnosed ICU delirium. Management of this infant prompted further consideration of our evolving evaluation and treatment algorithms of pediatric ICU delirium at UVA. Here, we present three subsequent cases in the UVA PICU of hyperactive delirium that posed risk of serious complications for which we treated with short courses of atypical antipsychotic medications. These cases have served as a framework for the evolving approach to delirium management in our PICU.

Case 1

Case 1 is an infant male with CHARGE (coloboma and cranial nerve defects, heart defects, atresia of the choanae, retardation of growth and development, genital underdevelopment, ear abnormalities and sensorineural hearing loss) association and cardiac defects, including unbalanced atrioventricular canal and aortic arch hypoplasia, requiring single-ventricle palliation. He had his stage-2 palliation at 4 months of age, and then struggled with chronic respiratory insufficiency, pulmonary hypertension, and chylothorax. Following cardiac catheterization at 5 months of age, he required increased vasoactive, respiratory, and sedative support. The infant acutely developed fluctuating mental status ranging from extreme fussiness to somnolence and apathy. Attempts to wean his sedation resulted in agitation and hypoxia. He was unable to attend to his parents' faces and voices, as he was previously able. An episode of agitation characterized by head thrashing resulted in an unplanned extubation and a bradycardic arrest. Despite infusions of morphine (maximum of 0.09 mg/kg/h), midazolam (0.06 mg/kg/h), and dexmedetomidine (1 mcg/kg/h), as well as scheduled lorazepam (0.6 mg/kg q6h), and multiple bolus doses of morphine and midazolam, he remained poorly sedated. He screened positive for delirium using the Preschool Confusion Assessment Method for the Intensive Care Unit (psCAM-ICU). ¹⁹ Child Psychiatry was consulted and agreed with the diagnosis of acute multifactorial delirium.

The infant's parents were counseled regarding the morbidity of ongoing delirium, and its contribution to his chronic critical illness. Once nonpharmacologic measures were exhausted, a short course of risperidone was initiated. His electrocardiogram (EKG) was checked to exclude prolonged QTc interval (the measurement made on an EKG from the start of the Q wave to the end of the T wave, corrected for heart rate). He was given a single dose of risperidone 0.13 mg at 0200 hours (recommended dosing for delirium is 0.1 to 0.2 mg once daily at bedtime). Following the dose, both providers and parents noted that he was less agitated, more lucid, and easier to console; his midazolam drip was able to be weaned, and he required fewer bolus doses of sedation. In the 24 hours preceding, the initial risperidone dose, he received 10 boluses of morphine (0.1 mg/kg) and 18 boluses of midazolam (0.1 mg/kg) in addition to his continuous infusions; in the following 24 hours, he received only two boluses of morphine and four boluses of midazolam. He received an additional dose of risperidone 0.13 mg the following night, midazolam infusion was weaned further, and he required no additional sedative boluses that night. The infant remained critically ill and was unable to be extubated; however, his delirium resolved. He did not require additional risperidone for 2 weeks following this episode.

During his hospital stay, he required two additional one-time dose of risperidone 0.13 mg for delirium, identified with the psCAM-ICU screening tool. His sedative infusions and boluses decreased the days following his doses of risperidone. The infant did not display signs of neuroleptic malignant syndrome, QTc prolongation, dystonia, akathisia, or dyskinesia.

Case 2

Case 2 is a toddler female who had a posterior mediastinal mass compressing the mainstem bronchi caused by mycobacterium avium complex (MAC) infection.

During her 19-day intubation period, achieving appropriate sedation and comfort was difficult, and she was persistently agitated, tachycardic, and hypertensive. She received dexmedetomidine (maximum dose 1.5 mcg/kg/h), hydromorphone (switched from fentanyl secondary to tachyphylaxis; maximum dose 0.03 mg/kg/h), midazolam (maximum dose 0.21 mg/kg/h), and cisatracurium (maximum dose 2 mcg/kg/min) infusions. She required intermittent boluses of these medications, as well as ketamine and pentobarbital. She was also started on a low-dose naloxone infusion for treatment of possible opioid-induced itching that may have contributed to her agitation. She was transitioned off continuous infusions to methadone, lorazepam, and clonidine intermittent enteral dosing periextubation.

Six hours after extubation, she became delirious. Her symptoms included inattention, reaching for items that were not present and nonverbal inappropriate moaning. She developed hallucinations and insomnia, and neither staff nor her mother was able to console her. She screened positive for delirium using the psCAM-ICU. Her Withdrawal Assessment Tool-Version 1 (WAT-1) scores, which are used in all patients at risk of iatrogenic withdrawal syndrome, were also consistently above 4, indicating that she suffered from iatrogenic withdrawal, though these symptoms did not improve with enteral medications (specifically opioids and clonidine) aimed at minimizing withdrawal symptoms. Her symptoms persisted and worsened. She had multiple risk factors for delirium, including multiple tubes and drains, bilateral wrist restraints during intubation, benzodiazepine and opioid exposure, neuromuscular blockade, long-term steroid use, and intermittent exposure to ketamine. She was treated with 1-mg intravenous (IV) haloperidol, a first generation antipsychotic, due to its rapid onset (typically within seconds) to minimize the risk of injury from attempts to remove numerous medical devices.

Other than sedation, she did not suffer from any side-effects of haloperidol, including tremor, hyper- or hypotension, anticholinergic effects, neuroleptic malignant syndrome, or extrapyramidal movements. There was no QTc prolongation on EKG prior to or 48 hours after administration. To avoid the continued use of first generation antipsychotics given the multiple potential side effects, risperidone (0.25 mg nightly) and melatonin (3 mg nightly) were started enterally. She completed a 5-day course of nightly risperidone and demonstrated significant improvement in mental status and attention. Her tachycardia and hypertension, which initially attributed to iatrogenic withdrawal, were resolved completely following treatment with risperidone, suggesting that these symptoms were more likely a result of her delirium. Her previously elevated WAT-1 scores decreased to appropriate levels following the initiation of treatment with antipsychotics.

Case 3

The third case is a 6-month-old male with hypoplastic left heart syndrome (HLHS) who had received his stage-2 palliation. After an uneventful postoperative course, he was transferred to a hospital closer to home to work on oral feeding skills. The patient was readmitted to UVA following a code event precipitated by respiratory failure with a difficult intubation.

Upon arrival to the PICU at UVA, the patient remained intubated and sedated on morphine (0.14 mg/kg/h) and midazolam (0.13 mg/kg/h) infusions for several days while attempting to achieve adequate diuresis prior to extubation. He also received scheduled lorazepam (0.2 mg/kg in every 6 hours), methadone (0.2 mg/kg in every 6 hours), gabapentin (12.5 mg/kg twice daily), and clonidine patch (0.1 mg), as well as melatonin, as needed nightly, all of which he had been on prior to intubation. Over a 48-hour period, the patient became increasingly agitated and inconsolable, and bolus doses of morphine (0.1 mg/kg) and midazolam (0.1 mg/kg) were ineffective leading to the use of pentobarbital for sedation. He stopped responding to his mother, developed lack of focus, and he was unable to track caregivers. He screened positive for delirium using the psCAM-ICU tool.

He was given doses of risperidone 0.2 mg at 0200 hours, and then again at 2330 hours the following evening. There was no QTc prolongation on EKG prior to or 48 hours after administration. Additional interventions made to address delirium included weaning midazolam boluses, incorporating melatonin scheduled nightly, increasing daytime stimulation while decreasing nighttime stimulation. Developmental and Behavioral Pediatrics was consulted, and agreed with the delirium diagnosis. After the brief course of risperidone, many of the patient's symptoms indicative of delirium improved significantly. He was more consolable and able to focus, track, and follow. He responded to his mother's presence once again. He did not require boluses of morphine or midazolam for over 72 hours, and he was able to be weaned from, both his midazolam (to 0.08 mg/kg/h) and morphine infusions (to 0.07 mg/kg/h). No further risperidone doses were necessary and the patient was successfully extubated shortly thereafter.

Discussion

Pediatric delirium is common and increasingly diagnosed in the ICU setting, but it can also be misinterpreted as inadequate sedation, pain, and iatrogenic withdrawal. ^{1 20} In February 2018, subsequent to the cases described in this series, the PICU at UVA established a guideline for the management of pain, agitation, and delirium in critically ill neonatal and pediatric patients. The goals are to achieve prompt and effective pain management to attain the lightest sedation clinically indicated, and to recognize, prevent, and treat delirium. These cases helped us to clarify this framework which first requires that care providers are proficient with standardized assessment tools for sedation, withdrawal, and delirium.

Due to inherent communication limitations in infants and young children, accurate diagnosis relies on symptom observation rather than patient interview, and it is imperative to understand infant and early childhood orientation and cognition and how pain, withdrawal, and delirium disrupts them. The state behavior scale (SBS) is used to target and assess the sedation levels of infants and young children being mechanically ventilated, with the goal being a score of −1 or 0 depending on the child's developmental level and medical condition. As previously described, the WAT-1 scores are used to identify patients who might be suffering from iatrogenic withdrawal.

Several tools are used to assess for the presence of delirium. In 2016, the psCAM-ICU was developed and found to be a valid and reliable screening tool for delirium for children 6 months to 5 years of age. It requires the presence of inattention, the most fundamental feature of delirium, as well as objective and interactive patient assessments to suggest the diagnosis of delirium. ¹⁹ In the PICU of UVA, delirium screening was initiated using the psCAM-ICU if under 5 years of age or the pCAM-ICU if over 5 years on every shift. ²¹ The Cornell Assessment of Pediatric Delirium (CAPD) is a validated assessment tool that relies more heavily on observed patient responses rather than direct interaction. Its validation population included large proportions of children under 2 years old and those who are developmentally delayed. ^{22 23} The CAPD was not used for the patients in this case series, but it is also appropriate for their age range and would have been considered if the patients were developmentally delayed. Patients who screen positive are then evaluated by the ICU physician staff, assisted by the Developmental Pediatrics and Pediatric Psychiatry teams as necessary based on team preference or if there is concern for acute brain injury.

There are several modifiable factors in the care of children in the ICU that may minimize risk for delirium. One critical piece to examine is the child's environment, specifically what can be done to mimic a daytime routine such as that experienced outside of the hospital setting by adequately signaling to the child that it is daytime or nighttime, minimizing the use of restraints, optimizing positioning, encouraging family involvement, and having familiar objects available. ⁴ We have focused on improving sleep hygiene by dimming lights at night, minimizing nighttime interruptions and delaying daily X-ray studies from 4 a.m. until 6 a.m., and considering the use of melatonin during the bedtime routine if sleep onset is challenging. The use of educational and rehabilitation therapy services, child life, and pet therapy early in a patient's hospital course, as soon as the patient is medically able, help to optimize a patient's environment. In complement with environment, it is necessary to continue to provide appropriate and adequate nutrition, preferably enterally, as well as continue, to assess other physiologic considerations that could be contributors to delirium, such as hydration status, blood pressure, metabolic or electrolyte disarray, or infectious etiologies. Several medications have been implicated in the development or exacerbation of delirium, most notably benzodiazepines and opiates but also steroids and certain antibiotics. ^{1 24} Therefore, it is crucial to provide adequate sedation and pain control without over sedation.

In chronically critically ill patients, particularly those who are intubated for long periods of time, it is challenging and often impossible to avoid medications and other risk factors that predispose patients to delirium. Our guideline encourages the consideration of antipsychotic therapy if hyperactive delirium persists after all potential underlying causes have been adequately addressed and clinically modifiable factors have been exhausted. Haloperidol is a first-generation antipsychotic that is available for treatment intravenously and has a rapid onset of action. However, there is a significant risk for dystonic reactions and hyperpyrexia, therefore use should be limited to acute psychosis or severe delirium that requires urgent treatment. ²⁵ Although quetiapine is the atypical antipsychotic more broadly described in the treatment of pediatric delirium, it's method of administration as an immediate-release tablet, and concentration per tablet makes it at times challenging to administer in appropriate doses to young patients. Similarly, olanzapine is also typically available as a tablet, though intramuscular injection exists. Risperidone is another atypical antipsychotic agent that has been used successfully to treat delirium in critically ill patients in the PICU, ²⁶ although its side effect profile is less desirable in that there is an increased risk for extrapyramidal symptoms, tardive dyskinesia, and elevated prolactin levels (clinical evaluation of QT/QTc interval prolongation and proarrhythic potential for nonantiarrhythmic drugs—questions and answers; Guidance for industry U.S. Food and Drug Administration). ²⁷ It is available in tablet form and oral solution, making administration simpler in infants and young children.

In our PICU, the critical care nursing and physician staff has taken on primary responsibility for diagnosing and treating PICU delirium, as these are patients with complex, often competing physiologic issues. When necessary, and in particular if a trial of pharmacological therapy fails, we consult with the Pediatric Psychiatry and Developmental and Behavioral Pediatrics services for their expertise in the diagnosis and management of delirium and clinical experience with atypical antipsychotics for other indications. As pediatric delirium is being increasingly recognized and treated in the pediatric ICU setting, it is not surprising that pediatric intensivists have taken the lead in establishing guidelines for its use. ⁴

Our three cases demonstrate resolution of hyperactive delirium from the careful use of antipsychotics in children who pose a danger to themselves. In each case, after a short course of antipsychotics the patients were able to attend and respond more effectively to their surroundings and caregivers, and the patients stopped being a threat to their own safety. Multiple doses of sedatives and analgesics were proving ineffective in controlling dangerous behaviors, and they might have been contributing to the delirium. Pharmacologic treatment of delirium was able to effectively control behaviors, such as head thrashing and extremity flailing. Similar results are described in the literature for pediatric delirium. ²⁸ Finally, our patients did not demonstrate any of the known concerning side effects associated with the use of antipsychotics or its withdrawal, in part because we limited treatment courses to a maximum of 3 days.

Conclusion

Following these cases, education and guideline institution, our team is increasingly aware of the signs and symptoms of delirium and able to modify environment and sedation choices with delirium minimization as a goal in addition to, first and foremost, appropriate pain control and sedation. We are now examining how these interventions have impacted the use of opioids and benzodiazepines in patients at risk of developing delirium. Further multicenter studies are needed to establish the efficacy of and dosing strategies for antipsychotics in infants and children experiencing ICU delirium. As pediatric ICU delirium guidelines develop, these cases are illustrative of its manifestations and how short courses of antipsychotics can benefit intubated patients with hyperactive delirium.