A case of infant delirium in the neonatal intensive care unit

Abbreviations

1 Introduction

Delirium is a manifestation of underlying acute cerebral dysfunction characterized by a disturbance in awareness and cognition in the presence of an underlying medical condition [1]. Delirium is a well-studied phenomenon in the adult intensive care setting and has been linked to adverse outcomes, including more self-extubations, longer duration of mechanical ventilation, prolonged hospital stays, elevated healthcare costs, and increased mortality [2]. It has also been described in the pediatric intensive care setting, but has only recently been recognized as a clinical entity in infants. Few studies have described the signs, symptoms, and management of delirium in this unique population [3–5]. We describe a case of a 6-month-old, former 32-week gestation, male infant diagnosed with delirium during a prolonged mechanical ventilation course in the neonatal intensive care unit (NICU) and the management of his delirium with the atypical antipsychotic risperidone.

2 Case description

A 6-month-old 4.7 kg male infant, born at 32 weeks gestation secondary to maternal prolonged premature rupture of membranes, required elective intubation for cardiac catheterization to evaluate his pulmonary hypertension. The infant had been stable on nasal cannula with 2-3 liters flow of 100% oxygen prior to the procedure, but remained critically ill after the procedure on a mechanical ventilator and was unable to be extubated.

Two days following intubation, the infant was noted to be agitated with “elevation in heart rate, decreased [oxygen] saturations, increased temperature, and restless motion.” His agitation was initially treated with unscheduled, as-needed bolus doses of intravenous morphine at 50 micrograms per kilogram (mcg/kg) and lorazepam at 0.1 milligrams per kilograms (mg/kg). Scheduled diazepam was added eight days into the patient’s mechanical ventilation course at 0.1 mg/kg, and a midazolam infusion of 1 micrograms per kilogram per minute (mcg/kg/min) was added on day nine due to persistent agitation described by his primary nurse as “not improved with suctioning, repositioning, diaper change, positive touch, or decreasing stimulation”. By day 19 of mechanical ventilation, the patient was on continuous infusions of fentanyl (1 mcg/kg/hr), midazolam (3 mcg/kg/min), and dexmedetomidine (2 mcg/kg/hr) in addition to scheduled gabapentin (5 mg/kg every 8 hours), phenobarbital (5 mg/kg/day), methadone (0.15 mg/kg every 6 hours), and diazepam (0.15 mg/kg every 6 hours). Additionally, the patient received up to 19 boluses of fentanyl, midazolam, and dexmedetomidine in a 24-hour period for management of agitation. At this time, the patient’s resuscitation status was changed to “Do not attempt resuscitation” (DNAR) due to concern that there had been no progress in weaning ventilation settings or decreasingsedation.

The patient’s mother expressed concern that her son was restless, irritable, and inconsolable. She noted that his symptoms gradually worsened over the course of two weeks following intubation. He fought against the endotracheal tube and seemed to have purposeless movements in the form of head shaking and arm waving. He no longer made consistent eye contact, stopped visually tracking, and did not respond to his mother’s voice. Child psychiatry was consulted on day 19 of mechanical ventilation due to concern for superimposed delirium in the setting of increased agitation unresponsive to high doses of sedative and analgesic medications and a change in baseline mental status and developmental milestones.

The child psychiatrist identified the patient’s symptoms as consistent with infant delirium that was likely multifactorial in nature: underlying illness, including acute-on-chronic respiratory failure, and the use of multiple deliriogenic medications, specifically benzodiazepines, opiods, and barbiturates. The primary recommendation was to wean the deliriogenic medications, preferentially the midazolam infusion. The recommendation to additionally start a neuroleptic agent was based on the concern that the patient’s agitation and associated ventilation challenges would make weaning of sedation medications difficult. A trial of low-dose, enteral risperidone was started on day 20 of mechanical ventilation at 0.01 mg/kg every 12 hours in an effort to decrease delirium-associated agitation and allow for removal of offending agents.

Within 48 hours of initiating risperidone, the patient’s nurses and family members remarked on a significant improvement in mental status and social interactions. Four days after initiating risperidone, the patient was weaned from the midazolam infusion and phenobarbital completely, and received significantly fewer bolus doses of midazolam, fentanyl, and dexmedetomidine: one, three, and one respectively (Fig. 1). The day prior to discontinuing the midazolam infusion, the patient’s dose of risperidone was increased to 0.01 mg/kg in the morning and 0.02 mg/kg in the evening to promote sleep at night and facilitate the continuation of sedative and analgesia weaning. During the following five days on this dosing regimen, the patient remained off his infusion of midazolam, received no bolus administrations of midazolam, fentanyl, or dexmedetomidine, and weaned from his fentanyl infusion completely (Fig. 2).

OPEN IN VIEWER

Fig.1

Effect of risperidone on use of sedation boluses. Day 0 represents the day risperidone was initiated, with Day -1 serving as the day prior and Day 1 as the day after. No midazolam blouses were administered after Day 5; fentanyl, Day 7. ◊ Midazolam, □ Fentanyl,▴ Dexmedetomidine.

OPEN IN VIEWER

Fig.2

Effect of risperidone on continuous sedation infusions. By day 4 the midazolam infusion was discontinued; fentanyl, day 9. Dexmedetomidine was the last infusion to be weaned since this medication has not been shown to be deliriogenic. ◊ Midazolam, □ Fentanyl,▴ Dexmedetomidine.

Electrocardiograms were completed on days one and four of risperidone with no evidence of prolonged QT interval. Additionally, the patient did not display any stereotypical behaviors associated with neuroleptics such as lip smacking. The patient’s family reported continued improvement in mental status, noting that the patient was more awake and interactive, showed a social smile at times, and appeared more comfortable and aware of his surroundings. The patient remained on risperidone for a total of 32 days. Over that time, the patient was taken off all sedation and, after 2 months of stability, the DNAR status was reversed.

While the core symptoms of delirium are universal to all age groups: impaired consciousness and awareness, inability to focus or sustain attention, impairments of the sleep-wake cycles, and disturbance of thought processes, there are clear challenges to making the diagnosis in the very young [6]. Infants cannot be asked about orientation or delusions, they have a tendency to regress when ill, and their symptoms – crying, restlessness, and inconsolability – are often nonspecific. The history provided by our patient’s mother was key in identifying the change in mental status. Without this information, the assumption had been that patient’s agitation was due to refractory pain and discomfort, leading to escalation of sedation and analgesics. The fact that delirium was considered at all was likely due to implementation of pediatric delirium screening in other units of the hospital, which had led to increased awareness of infant delirium among clinicians.

There are several delirium screening tools recently validated for the pediatric population, [6]. Since not all of these tools can be used in the infant, toddler, and non-verbal populations, as some require patient participation, there has been a push to take a more developmentally-informed, observational approach to screening and assessing delirium in children [5]. The Cornell Assessment for Pediatric Delirium is a screening tool aimed to identify critically ill children of all ages and developmental stages who were at risk for delirium using specific developmental anchor points. A double-blinded, single-center study found this tool to be valid, rapid, and easily applied by nursing staff [5]. This tool has been implemented in our pediatric intensive care units to screen children as young as two-months-old, but has not been adopted in the NICU due to lack of data in the premature population.

Increased awareness of the differential diagnosis of infant agitation in the neonatal intensive care population is critical for the appropriate management of these infants. Studies in the pediatric intensive care setting have found significant associations between pediatric delirium and developmental delay, need for oxygen, use of mechanical ventilation, and deeper sedation levels [7, 8]. Our patient had several of these risk factors: critical illness, developmental delay, prolonged mechanical ventilation, and escalating use of sedatives and analgesics.

Delirium can be managed both non-pharmacologically and pharmacologically. Ideally, identification and removal of the causative factor should be the first line intervention. Environmental interventions by nursing and families including repeated reassurance and reorientation may help to manage the behavioral and cognitive symptoms [9]. When the underlying causes cannot be removed or symptoms are leading to harm, antipsychotics can be considered. The decision to start an antipsychotic in an infant is not an easy one. Antipsychotics have been prescribed off-label in infants and children to relieve agitation and perceptual disturbances associated with delirium, but their safety and efficacy in this young population is not well studied yet. In this patient, the risk of not treating the delirium was felt to outweigh the risk of a brief course of an antipsychotic.

We chose an atypical antipsychotic rather than a typical agent such as haloperidol due to the lower risk for extrapyramidal side effects and QTc prolongation. Risperidone was selected because it comes in a liquid formulation and can be dosed in small, reliable quantities [9]. Risperidone is an atypical antipsychotic with mixed serotonin-dopamine antagonist activity, which is thought to improve negative symptoms of psychoses and reduce the incidence of extrapyramidal side effects. Alpha-1, alpha-2 adrenergic, and histaminergic receptors are also antagonized with high affinity [10, 11]. Our experience suggests that recognition and treatment of delirium in the NICU can alter the course of a patient’s hospitalization for the better, but as it is only one case, further studies are needed to elucidate the diagnosis and pharmacologic management of delirium in both term and preterm infants.

Financial disclosure

The authors have no financial relationships relevant to this article to disclose.

Conflict of interest

The authors have no conflicts of interest relevant to this article to disclose.

Funding source

P.B.S. received financial support from the National Institutes of Health (HHSN-275201000003I). C.M.C. received financial support from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (5U10HD040492-10) and the National Heart, Lung, and Blood Institute (1R01 HL1057020A1).