Abstract
Advances in cardiac surgery over the last three decades have translated to a shift in the epidemiology of congenital heart disease mortality from childhood to adulthood. As a result, patient and family–centered care with therapies designed to improve neurodevelopmental, functional, and quality-of-life outcomes in children with heart disease are crucial. There is no question that sedation, analgesia, and mechanical ventilation will always be central aspects of care for children with critical heart disease. However, minimizing the risks associated with these therapies while optimizing patient safety and both short- and long-term outcomes is a practical and achievable goal for all providers caring for these children. There is a strong interplay between sedation, sleep, delirium, and rehabilitation—a child who is heavily sedated and restrained is at high risk for sleep disturbance and delirium and is unable to participate in rehabilitation therapies. Conversely, a child who is awake during the day and engages in rehabilitation in the acute phase of his or her recovery will be more likely to maintain a normal circadian rhythm and less likely to transition to delirium due to decreased administration of sedative medications. When immobilization with sedation is medically necessary (generally only for the most severely ill children), goal-directed and titrated-sedation approaches are needed to optimize outcomes. In the following sections we provide an overview of sedation, analgesia, delirium, sleep, and rehabilitation as they relate to the pediatric cardiac patient.
Key Words
cardiac surgery, sedation, analgesia, delirium, sleep, rehabilitation, pediatrics
Advances in cardiac surgery over the last three decades have translated to a shift in the epidemiology of congenital heart disease mortality from childhood to adulthood. As a result, patient and family–centered care with therapies designed to improve neurodevelopmental, functional, and quality-of-life outcomes in children with heart disease are crucial. There is no question that sedation, analgesia, and mechanical ventilation (MV) will always be central aspects of care for children with critical heart disease. However, minimizing the risks associated with these therapies while optimizing patient safety and both short- and long-term outcomes is a practical and achievable goal for all providers caring for these children. There is a strong interplay between sedation, sleep, delirium, and rehabilitation—a child who is heavily sedated and restrained is at high risk for sleep disturbance and delirium and is unable to participate in rehabilitation therapies. Conversely, a child who is awake during the day and engages in rehabilitation in the acute phase of his or her recovery will be more likely to maintain a normal circadian rhythm and less likely to transition to delirium due to decreased administration of sedative medications. When immobilization with sedation is medically necessary (generally only for the most severely ill children) goal-directed and titrated-sedation approaches are needed to optimize outcomes. In the following sections we provide an overview of sedation, analgesia, delirium, sleep, and rehabilitation as they relate to the pediatric cardiac patient.
Pain and Sedation Management in the Cardiac Intensive Care Unit
Pediatric patients with cardiac disease often experience pain and anxiety. Despite substantial advances in perioperative, surgical, and critical care, they remain at high risk for adverse events such as multiorgan dysfunction, need for extracorporeal membrane oxygenation and cardiac arrest. Low cardiac output syndrome (LCOS), a period of inadequate oxygen delivery, can occur and usually peaks 8 to 12 hours postoperatively and is associated with increased morbidity and mortality. In attempts to circumvent the severity of LCOS, sedation and analgesia have been effectively used to decrease metabolic demand, myocardial oxygen requirement, and the stress response. However, this practice has become skewed over time, resulting in many patients who are either deeply sedated or comatose for prolonged periods of time.
Prolonged MV and intensive care unit (ICU) stay, along with the development of tolerance, withdrawal, posttraumatic stress syndrome, and delirium, have been reported in critically ill children with longer duration of and higher exposure to sedatives. Moreover, the US Food and Drug Administration (FDA) recently issued an FDA Drug Safety Communication ( https://www.fda.gov/Drugs/DrugSafety/ucm532356.htm ) reporting that prolonged (>3 hours) or repetitive exposure to certain groups of drugs (anesthetics and sedatives) in children less than 3 years of age “may affect the development of children’s brains.” Due to this the labeling of 11 common anesthetics and sedatives, including volatile agents (sevoflurane) and intravenous (IV) agents (barbiturates, benzodiazepines, propofol, and ketamine), have been modified to include this neurodevelopmental warning. The basis for this warning is largely dependent on evidence from animal and in vitro studies, with the available clinical evidence admittedly difficult to interpret. As such, this warning concludes that “additional high quality research is needed to investigate the effects of repeated and prolonged anesthesia exposures in children, including vulnerable populations.”
As the care provided in the cardiac intensive care unit (CICU) continues to evolve, using evidence-based medicine, different aspects of the patient assessment, presentation of patient data, and development of patient care plans will be challenged. The foundation to begin this discussion is based on the following conclusions:
- •
The routine assessment of both pain and sedation level is recommended by the Pediatric Cardiac Intensive Care Society.
- •
Pain, sedation, and delirium assessments, along with the current disease state and patient-specific psychosocial goals, should be integrated and discussed on medical rounds daily using an interdisciplinary approach.
- •
Thoughtful decisions regarding sedation and analgesia should be made, including drug choice and titration, providing a personalized patient approach, recognizing the risks of associated delirium, withdrawal, and possible long-term cognitive outcomes.
Pain Assessment
Pain is “an unpleasant sensory and emotional experience associated with actual or potential tissue damage,” that is largely subjective in the assessed intensity or importance to each patient. Unfortunately, the course of critical illness in the CICU involves numerous sources of pain via necessary therapies such as placement of invasive lines, chest tubes, or Foley catheters; requirement of an endotracheal tube, mask, or nasal cannula; frequent repositioning; and wound care. Due to variation in psychosocial maturity, the emotional component of pain can be greatly magnified in pediatric patients and confused with symptoms of agitation. This emotional component is not restricted to the patient and may extend to the family, thereby changing perceptions of satisfaction with the ICU experience. Moreover, the routine assessment of pain or sedation level in pediatric patients can be complex given variations in cognitive, verbal, and motor skill development, and even regression of previously learned skills. Clinical signs such as restlessness, agitation, grimace, or increased muscle tone in nonverbal pediatric patients may indicate undersedation, pain, fear, or even separation anxiety. Though purely objective assessment tools are preferred, the pain and sedation level evaluation commonly employs observational components based on age and development.
There are two general approaches for pain assessment, using either self-report or observational scales. Self-reporting pain scales are by nature more objective, but their usefulness can be limited by age and development. These tools use a linear display of “no pain” to “worst pain” using phrases and numbers or pictures, like the numeric rating scale (NRS), verbal numeric scale, and visual analogue scale, or the Wong-Baker “faces” pain scales, and can be altered for ethnicity and age (Oucher Scale). Though a preferred tool for assessment of pain, the NRS cannot be consistently used in children less than 4 years of age. Observer-rated pain tools include the FLACC (acronym for face, legs, activity, crying, and consolability) and COMFORT-B scales. The FLACC is a pain assessment tool used in infants and children whereby five variables are observed or assessed and scored, with a maximum score of 10 points (associated with discomfort). The COMFORT-B scale combines assessment of both pain and sedation level using two physiologic and seven behavioral variables for estimation of pain and agitation among infants and children. Although it may be difficult to differentiate pain from agitation and vice versa in the very young using the COMFORT-B scale, targeting levels and using a treatment algorithm has been associated with decreased ICU length of stay, decreased duration of MV, less withdrawal, and lower sedative exposure in ventilated pediatric patients.
Sedation Assessment
The term sedation often is used to encompass components of anxiolysis, amnesia, and analgesia. Sedation assessment scales measure arousal or the level of consciousness (LOC). Historically these assessments have been largely subjective with descriptors such as “unresponsive,” “lethargic,” “obtunded,” “calm,” “restless,” “distressed,” “agitated,” or “combative.” Monitoring LOC in the CICU provides early recognition of worsening neurologic status related to the critical illness, oversedation, or delirium. Routine monitoring also encourages targeted sedation whereby patients are maintained at the most appropriate LOC based on needs and disease state, the goal being a patient who is comfortable and either calm or slightly sedated. This level of sedation allows for more accurate clinical assessment, caregiver interaction, and possibly quicker liberation from MV. Pediatric tools using both objective and observational components to assess LOC have become the cornerstone of CICU care. The Richmond Agitation-Sedation Scale (RASS) has 10 possible levels of arousal (−5 to +4) defined by three clearly defined steps of patient interaction to elicit a response ( look or observe the patient, talk to the patient, touch the patient). Patients are scored as being alert and calm (RASS 0), agitated (+1 to +4), responsive to voice (RASS −1 to −3), responsive to touch (RASS −4), or comatose (RASS −5). The RASS has been successfully employed in the pediatric cardiac and medical ICUs for routine monitoring, sedation targeting, and delirium monitoring. The State Behavioral Scale (SBS) is similar to the RASS but has six levels of arousal (−3 to +2). The SBS score is determined by patient observations for respiratory drive, cough, response to stimulation, attentiveness, tolerance to care, consolability, movement after comforting, and pain assessment using the NRS. The SBS has also been successfully used in the ICU setting associated with nurse-driven sedation protocols and delirium.
Nonpharmacologic Management
A major strength of recent pediatric sedation regimens is the importance placed on a patient’s baseline psychologic health to include recognition of disorders such as anxiety, depression, or traumatic stress before the critical illness. Another strength is the reliance on comfort measures provided by caregivers through touch and verbal reassurance in conjunction with optimal analgesia when appropriate. The CICU can quickly overwhelm children, challenging them to use often immature coping mechanisms to deal with pain and anxiety from both surgery and necessary procedures. Family presence during procedures has been shown to not only preserve quality of care but also provide enhanced comfort to the patient and decrease pharmacologic needs. The environment of the CICU can either be part of the solution or remain part of the problem for our fragile patients. Moving forward, we may find just how important a regular routine, sleep hygiene, and family presence are to not only the general well-being of our patients, but also how they experience or interpret their circumstances, modulating their responses to pain and anxiety.
Pharmacologic Management ( Table 21.1 )
Management of pain and anxiety is of the utmost importance to the medical team and the family. Finding the right balance of nonpharmacologic and pharmacologic therapies will be an ongoing challenge in the ICU setting. CICU clinicians must manage (1) targeted sedation, (2) increasing patient wakefulness and attention, (3) early extubation following surgical interventions, (4) early mobility, and (5) greater parental involvement. Understanding the associated pharmacokinetic changes during critical illness when using analgesics and sedatives is extremely important. Many patients in the CICU depend on continuous infusions for sedation and analgesia, for which awareness of the context-sensitive half-time will need to be considered. The context-sensitive half-time or half-life is the time required for the plasma drug concentration to decrease by 50% following discontinuation of the infusion. The context-sensitive half-time depends on drug distribution and therefore cannot reliably be predicted by the elimination half-life alone. Some drugs that usually have a short context-sensitive half-time (in which the drug concentration and clinical effect quickly resolve) will demonstrate noncharacteristic action when infusions are continued for long periods due to saturation of both central (blood) and peripheral (adipose tissue) compartments.
| Drug Name | Bolus Dose | Infusion Dose | Cardiovascular Effects | Properties |
|---|---|---|---|---|
| Nonopioid Analgesics: COX-1 and COX-2 Inhibitors → Analgesia, Antiinflammation | ||||
| Acetaminophen | 15 mg/kg PO/PR/IV q6h 7.5 mg/kg IV (<10 kg) | NA | None | ↑Analgesia, ↑antipyretic, ↓↓antiinflammatory Overdose: ↑NAPQI requiring N -acetylcysteine |
| Nonsteroidal Antiinflammatory Drugs (NSAIDs): Nonselective COX Inhibitors → Analgesia, Antiinflammation | ||||
| Ketorolac | 2-16 y: 0.5 mg/kg IV q6h × 5 d (maximum single dose 30 mg) <2 y: 0.25 mg/kg IV q6h × 3 d | NA | None | NSAID ↑Analgesia, ↑antipyretic, ↑antiinflammatory Overdose/↑duration: renal insufficiency |
| Opioid: µ-, κ-, δ- Receptor Agonists → Analgesia | ||||
| Fentanyl | 1-2 mcg/kg IV | 1-20 mcg/kg/h | Bradycardia | ≈100× potency of morphine Rigid chest |
| Hydromorphone | 5-10 mcg/kg IV | 3-4 mcg/kg/h | Minimal when used as single agent | ≈10× potency of morphine Faster onset (5 min) than morphine |
| Methadone | 0.05-0.15 mg/kg PO/IV q4-8h | NA | Bradycardia Dysrhythmia (prolonged QT) | Equipotent to morphine |
| Morphine | 0.05-0.2 mg/kg IV/IM/SC | 0.01-0.03 mg/kg/h | ++Histamine ↓SVR, ↓MAP | Active metabolite, adjust in ARI Pruritus, hallucinations |
| Remifentanil | 1-3 mcg/kg | 0.4-1 mcg/kg/min | Bradycardia | Equipotent to fentanyl Constant context-sensitive half-time |
| Tramadol | 1-2 mg/kg/dose PO q4-6h (maximum single dose 100 mg) | NA | None | FDA recommend use in patients at over 12 y of age ≈1/10 potency of morphine Minimal respiratory depression |
| GABA A Receptor Agonist → Sedation | ||||
| Barbiturate | ||||
| Pentobarbital | 1-2 mg/kg/dose q3-5 min to desired effect (maximum total dose 100 mg/dose) | 0.5-5 mg/kg/h | Significant CV depressant | Deep sedation, ↓ICP, antiepileptic |
| Thiopental | 4-6 mg/kg IV | 5-10 mg/kg/h | Significant CV depressant | Deep sedation, ↓ICP, antiepileptic |
| Benzodiazepine | ||||
| Midazolam | 0.025-0.1 mg/kg IV/IM 0.5-1 mg/kg PO/IN | 0.05-0.3 mg/kg/h | Minimal | Active metabolite, adjust in ARI |
| Lorazepam | 0.02-0.1 mg/kg q4-8h | NA | Minimal | Contains propylene glycol |
| Phenol → Directly Potentiates GABA A Activity Leading to Hypnosis, Amnesia, and Sedation | ||||
| Propofol | 0.5-2 mg/kg IV | 25-350 mcg/kg/min | ↓SVR, ↓MAP ↓Contractility | Minimal respiratory depression PRIS (↑dose, ↑duration) Antiemetic, antipruritic, antiepileptic FDA approval: infusion for <48 h |
| Alpha (α)-adrenergic Receptor Agonist → Analgesia, Sedation | ||||
| Dexmedetomidine (α 2 ) | 0.1-1 mcg/kg IV over 10-20 minutes | 0.2-1.5 mcg/kg/h *Doses as high as 2.5 mcg/kg/h have been reported | SE > bolus and high doses Bradycardia | Minimal respiratory depression Decreased clearance (age <1 y) FDA approval: infusion for <24 h |
| Clonidine (α 1 , α 2 ) | 1-5 mcg/kg/dose PO q6h – 8h (max dose: 200 mcg/dose) | Bradycardia | Rebound hypertension IV bolus/continuous sedation under study, few clinical reports. | |
| N -methyl- d -aspartate (NMDA) Receptor Antagonist → Analgesia, Dissociative Sedation | ||||
| Ketamine | 0.5-2 mg/kg IV 3-7 mg/kg IM 5-10 mg/kg PO | 0.5-2 mg/kg/h | ↑HR, stable MAP Lose CV + effects w/ ↓catecholamine | Minimal respiratory depression Active metabolite (weak) Bronchodilation |
Analgesia
Nonopioid analgesics include acetaminophen and nonsteroidal antiinflammatory drugs (NSAIDs). Regimens that combine nonopioid and opioid analgesia are clearly associated with improved pain scores and less total opioid requirement. Nonopioid analgesics should be administered early and scheduled as baseline therapy, using opioids for more severe or breakthrough pain. In the CICU, where we are already challenged with oversedation, the use of these adjuncts cannot be emphasized enough.
Pain, fever, and inflammation are mediated through the cyclooxygenase (COX) pathway via prostaglandin production and action. There are two forms of the COX enzymes (COX-1 and COX-2), of which analgesia is greatest via COX-2 inhibition, whereas COX-1 inhibition may alter platelet aggregation and protection of the gastric lining. Many NSAIDs are nonselective COX inhibitors, whereas acetaminophen is COX-2 specific. Acetaminophen (paracetamol) is the most widely used analgesic and antipyretic worldwide, though with minimal antiinflammatory properties. With the availability of IV formulations of both acetaminophen and ketorolac (an NSAID), the use of nonopioid agents in the ICU is no longer limited by the severity of illness or intolerance of enteral intake. Recent studies have supported the use of ketorolac to provide rapid and effective analgesia among pediatric cardiac patients, with low reported risk of associated side effects such as clinically significant hemorrhage. Changes in routine administration of acetaminophen may increase risk of overdose, requiring early detection because high levels of the metabolite N -acetyl- p -benzoquinone imine (NAPQI) are treated by N -acetylcysteine therapy.
Opioids are potent analgesics that stimulate the µ, κ, and δ opioid receptors in both the peripheral and central nervous systems. Most analgesic properties are mediated via the µ 1 -receptor, with less potent analgesia via the µ 2 -receptor and naturally occurring dynorphin/encephalin-stimulated κ- and δ-receptors. The majority of opioid-associated respiratory depression, bradycardia, and physical dependence occur via the µ 2- receptor activity. Most opioids undergo hepatic metabolism and renal elimination.
Morphine is considered the archetypal opioid. Morphine is highly water soluble, subject to extensive first-pass metabolism, and undergoes glucuronide conjugation to morphine-6 glucuronide, an active and potent metabolite dependent on renal elimination. Cardiac patients with low cardiac output and requiring inotropic support are at risk for decreased metabolism and up to a threefold decrease in clearance among patients with hemodynamic instability. Although normovolemic patients tolerate morphine administration without hemodynamic effects, patients with low cardiac output may have an exaggerated histamine effect, causing impaired compensatory sympathetic reflexes, increased venous capacitance, and decreased perfusion.
Fentanyl is a synthetic opioid that is highly lipid soluble, with a rapid clinical onset, and is converted primarily to inactive metabolites. Fentanyl is routinely used in the perioperative period, providing safe and effective analgesia and sedation to cardiac patients. In the CICU prolonged use of fentanyl infusions for analgesia and sedation may lead to tachyphylaxis and require transition to other opioids. Studies have demonstrated a positive safety profile and clinical benefit of the associated blunted sympathetic stress response in infants and children with LCOS, pulmonary hypertension, and critically balanced systemic/pulmonary circulations.
Hydromorphone (Dilaudid) is a semisynthetic opioid, with greater potency and quicker onset (5 to 10 minutes) compared to morphine. Metabolism occurs via conjugation and forms two nonactive metabolites that undergo renal elimination. The absence of significant associated hemodynamic side effects makes its use an advantageous option for analgesia in the CICU.
Methadone is a long-acting opioid analgesic, most commonly used in the ICU setting for opioid withdrawal. Though the long half-life of methadone prevents its use in critically ill patients with a rapidly changing clinical course, methadone can be as a useful adjunct when a regimen is required for prolonged MV, gradual weaning, or a stable level of analgesia. Potential adverse effects include prolongation of QTc and hypotension.
Remifentanil is a pure µ-receptor agonist with equipotency to fentanyl, an ultrashort onset of action, and a constant context-sensitive half-time regardless of duration of infusion. These properties make remifentanil an important option for intraoperative and procedural anesthesia. The associated bradycardia common with higher doses and hyperalgesic component following prolonged exposure make its usefulness in the CICU for analgesia less apparent.
Tramadol is a synthetic 4-phenyl-piperidine analogue of codeine with a dual mechanism of action; it is a µ-, δ-, and κ-opioid receptor agonist. Though less potent, its lack of respiratory depression and low risk of dependence lead to its role as an adjunct for breakthrough pain and opioid-sparing effects.
Sedation
Benzodiazepines bind to the postsynaptic type A γ-aminobutyric acid (GABA A ) receptor, which increases the affinity of the receptor to GABA. GABA-ergic activity leads to inhibition of the central nervous system (CNS), resulting in sedation, hypnosis, amnesia, anxiolysis, and anticonvulsant effects. In general, metabolism of benzodiazepines is hepatic mediated and elimination is renal mediated.
Midazolam is the most commonly used sedative in the CICU with its rapid onset and ease of titration as a continuous infusion. Midazolam may lead to respiratory depression, especially with coadministration of an opioid. However, it can provide safe sedation/induction for some patients with cardiac disease, usually at lower doses. Although many patients may tolerate benzodiazepine use, administration can result in hypotension, particularly in patients with decreased cardiac function or in conjunction with additional sedative therapies. Midazolam has a short context-sensitive half-time; however, these properties are lost following infusions lasting multiple days, leading to longer duration of action and delayed elimination. The free (active) fraction of midazolam can be increased in critically ill children who are malnourished, have hepatic dysfunction, or are receiving other protein-bound drugs. Midazolam is metabolized via cytochrome P-450 to an equipotent metabolite, of which approximately 80% of conjugated 1-OH-midazolam is renally eliminated; therefore the risk of prolonged sedation should be considered in the setting of renal dysfunction. Due to the reliance on midazolam for continuous sedation, associated tolerance, dependence, and withdrawal is common. The clinical benefit of dose titration occurs at lower doses, with only marginal clinical improvement at higher levels. With the heavy reliance on benzodiazepines in the CICU, including high exposure and prolonged administration, association with delirium, prolonged ICU stay, and MV, and recent concerns regarding neurodevelopment, clinicians must seriously consider the appropriate role of this drug class in ICU sedation regimens.
Lorazepam is a long-acting benzodiazepine used for intermittent sedation and treatment for withdrawal following long-term midazolam infusions. It undergoes metabolism via hepatic-mediated phase II reactions that are not as sensitive to low cardiac output or hepatic dysfunction. Lorazepam infusions are frequently used in the adult ICU but are not used in the pediatric setting due to its long context-sensitive half-time and risk of propylene glycol (an additive in lorazepam) toxicity. Propylene glycol toxicity presents as an osmolar-gap metabolic acidosis and can be fatal if not recognized in a timely manner.
Alpha-2 adrenergic receptor agonists inhibit adenylyl cyclase via the alpha-2-adrenergic receptor located in both the peripheral and central nervous system. The greatest density of receptor activity occurs in the locus coeruleus (pons) and is responsible for mediating the sympathetic responses to stress, reducing brainstem vasomotor center–mediated CNS activation, in addition to treating pain, agitation, and withdrawal. The quality and characteristics of sedation differ based on drug choice. Sedation produced by alpha-2 agonists occurs by decreasing sympathetic neurotransmission, with a clinical picture of a calm-appearing yet easily aroused and attentive patient. (This is in contrast to GABA-ergic agents [benzodiazepines], which suppress arousal, producing an alteration of consciousness and even paradoxical agitation, with negative effects on cognition and behavior. Opioids too have inhibitory effects on the locus coeruleus; therefore when consumption is discontinued, abrupt increases in adrenergic neurotransmission from the locus coeruleus produce significant symptoms of withdrawal). The alpha-2 adrenoceptor activity of both dexmedetomidine (selective alpha-2 adrenergic receptor agonist) and clonidine (nonselective alpha-adrenergic receptor agonist) counteract these withdrawal effects. Due to their broad activity and favorable safety profiles, the roles of the alpha-2 adrenergic receptor agonists in the CICU continues to increase.
Dexmedetomidine is a selective alpha-2 adrenergic receptor agonist that demonstrates both sedative and analgesic properties. Dexmedetomidine is currently approved for procedural sedation of nonintubated patients, though off-label use for prolonged continuous sedation (>24 hours) in the ICU setting among adults, infants, and children is becoming commonplace. To that end, successful use of dexmedetomidine has been demonstrated in multiple studies of critically ill pediatric patients, including those with cardiac disease, for surgical anesthesia, postoperative sedation, and procedural sedation. Similar to ketamine and propofol, dexmedetomidine has opioid- and benzodiazepine-sparing effects when continuous sedation is required. Due to its large volume of distribution, a loading dose is required before starting an infusion for acute sedative/analgesic effects. In infants and younger children the terminal half-life is longer, and therefore a lower maintenance dose is required. Because its properties are favorable for continuous sedation, the longer context-sensitive half-time requires thoughtful titration in advance when weaning is desired. Dexmedetomidine is metabolized via the hepatic-mediated cytochrome P-450 pathway but does not depend on renal elimination and is therefore safe in patients with renal insufficiency. The safety profile of dexmedetomidine has been promising in the critically ill population, though bradycardia and hypotension have been reported regardless of the mode of administration (bolus versus infusion). Uniquely, the electrophysiologic effects of dexmedetomidine may be advantageous in some patients, with fewer episodes of perioperative junctional ectopic tachycardia and supraventricular tachycardia reported. As the routine use of dexmedetomidine increases, the risks of dependence and significance of symptoms, such as agitation and hypertension, will need to be further studied.
Clonidine is a nonselective alpha-adrenergic receptor agonist with greater alpha-2 activity. Historically used for the treatment of hypertension, clonidine is increasingly being tapped for sedation, analgesia, and opioid/benzodiazepine withdrawal. Though the most effective dosing regimen of clonidine in the pediatric intensive care unit (PICU) has not been determined, available evidence supports both infusion and bolus dosing. In a recent report, infusion dosing as high as 3 mcg/kg/h was reported to have excellent analgesic and sedative properties while maintaining stable hemodynamics in patients less than 2 years of age following repair of congenital heart disease. In combination with NSAIDs, local anesthetics, opioids, or ketamine, clonidine been shown to be a successful adjunct for analgesia.
Ketamine is a phencyclidine compound that antagonizes the N -methyl- d -aspartate (NMDA) receptor and has some opioid, nicotinic, and monoamine oxidase action as well. The dissociation created between the thalamocortical and limbic systems leads to a cataleptic-like state of unresponsiveness (sedation), intense analgesia, and amnesia. Despite the significant degree of anesthesia, with appropriate titration both airway reflexes and spontaneous respiration can remain intact. Ketamine is metabolized via N -methylation to form norketamine, a less-potent active metabolite, which undergoes conjugation and renal-mediated elimination. Ketamine undergoes rapid redistribution due to its lipophilicity. Ketamine causes both release of and reuptake inhibition of endogenous catecholamines that increase/maintain vascular tone and increase heart rate and contractility. Patients with severe uncompensated cardiomyopathy often have depleted catecholamine stores; hence the direct myocardial depressant properties of ketamine may be unmasked and lead to hypotension during bolus administration. Psychedelic effects, including emergence delirium, and the risk of tolerance with prolonged use require consideration.
In the perioperative setting, oral ketamine provides excellent sedation for IV access in patients with cardiac disease who cannot tolerate the hemodynamic effects associated with a volatile induction. Intramuscular ketamine provides safe, emergent anesthesia for intubation of uncooperative or critically ill patients who do not have IV access. Ketamine has been successfully used intraoperatively as continuous IV anesthesia for cardiac catheterizations and repair of congenital heart lesions on cardiopulmonary bypass, without significant dose-related elevations in heart rate. With a short context-sensitive half-time, ketamine is ideal for CICU sedation because titration is predictable. Low-dose (<2 mg/kg/h) ketamine infusions provide effective analgesia and decrease the total opioid and benzodiazepine exposure, without clinically significant hypertension, tachycardia, increased intracranial pressure, or hallucinations.
Propofol is an IV-administered sedative-hypnotic commonly used for procedural sedation, intubation, and anesthetic induction and maintenance in pediatric patients. As a lipophilic phenol, propofol potentiates and directly facilitates the action of the GABA A receptor, causing hypnosis, amnesia, and sedation, without analgesia. Propofol has several unique actions, including antiemetic, antipruritic, and anticonvulsant properties. It is metabolized rapidly by the liver with no active metabolites. In the CICU population, propofol has been advantageous as an adjunct for continuous sedation with opioid/benzodiazepine-sparing effects and a short context-sensitive half-time for infusions of short duration (<10 hours) due to its large volume of distribution and increased clearance in pediatric patients. Though low-dose continuous infusions may be tolerated, induction bolus doses (1 to 3 mg/kg) can lead to profound decreases in systemic vascular resistance, myocardial depression, and bradycardia in settings of significant myocardial depression, shunt-dependent pulmonary/systemic flow, or hypovolemia. Propofol has not been associated with withdrawal syndrome and may be advantageous during rapid titration of opioids and benzodiazepines. The use of propofol in the pediatric population must be weighed against the risk of developing propofol infusion syndrome (PRIS). Key risk factors associated with PRIS include high infusion rates (>4 mg/kg/h), infusion duration longer than 48 hours, critical illness, young age, concurrent catecholamine infusion, and steroid use. PRIS results in impaired oxidative phosphorylation, leading to a life-threatening metabolic acidosis, myocardial failure, rhabdomyolysis, and renal failure. The FDA provided recommendations in 2001 against prolonged propofol infusion sedation regimens in critically ill children, and therefore it is usually employed for less than 48 hours or during early extubation in most pediatric ICUs if at all.
Barbiturates are global CNS depressants that provide sedation, amnesia, and anticonvulsant activity. Barbiturates are not routinely used in the CICU because even lower doses can have profound effects on cardiac output and vasomotor tone, leading to poor oxygen delivery. The duration of action is determined by the rate of redistribution from the CNS to other tissue compartments. As these compartments become saturated due to high exposure or prolonged administration, patients are at risk for protracted sedative effects.
Tolerance and Dependence
Prolonged sedation in the ICU setting can lead to iatrogenic withdrawal syndrome (IWS). The most predictive risk factors for IWS are duration and cumulative exposure, with other associated factors, including age, drug class, bolus versus continuous administration, and use of weaning protocols. Prolonged exposure to opioids and sedatives can produce tolerance and dependence. Tolerance results in less clinical effect for a given dose, and therefore higher doses are required to achieve therapeutic effect. Dependence occurs when ongoing drug administration is required to prevent symptoms of abstinence syndrome. Abstinence syndrome can develop within 24 hours following cessation of chronic medications and is characterized by symptoms such as restlessness, insomnia, diaphoresis, tachycardia, hypertension, movement disorders, abdominal cramps, vomiting, diarrhea, delirium, and seizures. Amigoni and colleagues recently reported the extremely high prevalence (>60%) of IWS among survivors of critical illness. Both methadone and lorazepam are commonly required to treat IWS. The hope is that with optimization of targeted sedation and analgesia regimens, IWS will decrease over time.
Implementation of Sedation and Analgesia Regimens
To optimize sedation and analgesia in the CICU, a change in culture is required. It is necessary to routinely monitor pain and sedation scores, consider both pharmacologic and nonpharmacologic interventions, and recognize the physiologic implications of both inadequate treatment and overtreatment. For every patient, every day, three questions should be asked: (1) Where is my patient now? (pain, sedation, and delirium assessments); (2) Where is my patient going? (incorporation of the current physiologic/disease state and establishment of patient goals for the day); and (3) How do I get the patient there? (setting new sedation targets and planning for titration). In a systematic review by Vet and colleagues it is reported that daily sedation targets are achieved only approximately 60% of the time, with oversedation common. In fact, when sedation scores were below the target (oversedation), clinicians seemed to tolerate the more undesirable LOC because weaning of sedating medications rarely occurred. With recent recognition of the dangers of oversedation, clinicians are beginning to tolerate patients’ being more awake to benefit from touch, family present, and early mobilization (EM). In addition, interdisciplinary and multimodal approach, incorporating active waiting, and consideration of intermittent dosing rather than continuous infusions may benefit children in the CICU.
Delirium in the Cardiac Intensive Care Unit
Children are at high risk for developing delirium in the CICU. Delirium is defined as an acute and fluctuating state of neurologic dysfunction, manifested by altered awareness and cognition. It develops as a direct result of the underlying illness or as an unwanted side effect of hospital treatment. This is a medical diagnosis (rather than a psychiatric one) and can be conceptualized as a hospital-acquired complication. Delirium is a reversible process, and duration can be shortened with early detection and appropriate intervention.
Three subtypes of pediatric delirium are described: hyperactive, hypoactive, and mixed. Hyperactive delirium is easily recognized, with refractory agitation impeding caregivers’ attempts to administer necessary treatment. Hypoactive delirium, characterized by age-inappropriate lethargy, apathy, and withdrawal, is often overlooked. In mixed delirium, children vacillate between the motoric subtypes over the course of a 24-hour period. In children, disrupted sleep is a hallmark of all delirium subtypes. It is estimated that without routine screening, most cases of delirium remain undiagnosed.
Etiology of Delirium
The pathophysiology leading to delirium is complex and multifactorial. Hypoxia, inflammation, and altered neurotransmission all play key roles in delirium development. The neuroinflammatory hypothesis suggests that inflammation or oxygen deprivation leads to cytokine release with subsequent modification of the blood-brain barrier (BBB). Alteration in BBB permeability, in addition to dysregulation of the hypothalamic-pituitary-adrenal axis, lead to disruption of second messengers and neurotransmitters. All neurotransmitters are involved to some extent, with acetylcholine playing a prominent role. Neurons may also be directly injured due to oxidative stress, with possible exacerbation by the neurohumoral response, with research suggesting that endogenous glucocorticoids contribute to neuronal injury, particularly in the hippocampus. Delirium is also more likely to occur in those with less physiologic reserve, where serious illness unmasks those brains with the least resilience. Regardless of cause, most researchers agree that delirium represents acute whole brain dysfunction, with interruption of brain network connections and alteration in neurotransmission leading to the cognitive and behavioral changes that we recognize as delirium.
Clinically it is useful to think of pediatric delirium as a consequence of three interrelated factors: the underlying medical illness, unwanted side effects of treatment, and the unnatural and stressful environment in the ICU. Accordingly, children with congenital heart disease are at particular risk. As an example, consider a 4-month-old infant with trisomy 21 and an atrioventricular canal defect. She is admitted with congestive heart failure and undergoes surgical repair. After general anesthesia and bypass surgery the infant is admitted to the CICU. In the postoperative period she is mechanically ventilated, with multiple indwelling lines and tubes, pharmacologically sedated with narcotics and benzodiazepines, and restrained in her ICU bed. There she is immobilized while being exposed to noise and lights 24 hours a day. This child will likely become delirious early in her CICU stay.
Epidemiology of Delirium
The prevalence of pediatric delirium reported in the literature ranges from 13% to 65%, depending on the population studied. A large multinational study (n = 835 subjects) showed an overall frequency of 25%. In this mixed population, factors independently associated with increased delirium risk included need for MV or inotropes, receipt of benzodiazepines and narcotics, and age less than 2 years. For children in the ICU for 6 or more days, delirium frequency increased to 38%. Children with critical cardiac disease have many (if not all) of these risk factors, suggesting that our patients are at particular risk for developing delirium during their hospital stay.
Other single-institution studies have identified an extremely high delirium prevalence in children less than 5 years of age. In two separate studies, delirium rates were 56% and 65% in patients younger than 2 years, and 35% and 40% in patients 2 to 5 years of age. Children with preexisting neurodevelopmental delay are more likely to become delirious when compared to children with typical development. Severity of illness is also independently associated with delirium.
Postcardiotomy delirium is a specific delirium subtype described in adults after cardiac bypass surgery, which can affect up to 70% of patients in a CICU. This high prevalence is attributed to perioperative insults (including hypoperfusion, hypoxia, and cerebral microinfarcts) in an already vulnerable population. Until recently, little was known about delirium in children after bypass surgery.
A single-center prospective observational study followed 194 children after cardiac bypass surgery. These children were screened for delirium daily, and 49% were diagnosed with delirium during their ICU stay. In most children, delirium developed within the first 3 days after surgery, with a median duration of 2 days. Consistent with other pediatric delirium literature, the youngest children (less than 2 years of age) were most likely to develop delirium, as were children with the highest severity of illness, and those with baseline cognitive impairment. Cardiac-specific risk factors that were identified were presence of cyanotic heart disease and duration of bypass. Data suggest that poor nutritional status preoperatively increased the chance for development of postoperative delirium. As a marker of good nutritional status, children with a preoperative albumin level of more than 3 mg/dL had a lower odds of delirium development (odds ratio [OR] for delirium diagnosis 0.2; P = .028). Close attention should be paid to vulnerable children with these risk factors because they are most likely to become delirious during their hospital stay.
Importantly, there are hospital-related factors that contribute to delirium development in at-risk children. These include use of restraints and benzodiazepines. A large multicenter study showed an independent association between the use of physical restraints and delirium (OR 4.0), after adjusting for MV, sedating medication, and other potential confounders. A recent single center study has shown that the use of benzodiazepines confers a fivefold risk of delirium in children, even after controlling for age, developmental delay, severity of illness, MV, and use of other narcotics and sedatives. In adults, dexmedetomidine has been associated with decreased delirium, when compared to benzodiazepine-based sedation. It is likely that sleep deprivation and immobilization both play a role in the evolution of pediatric delirium.
Outcomes Related to Delirium
Not only is delirium highly prevalent, it also is associated with short-term distress and long-term harm. Delirium has been tightly linked to increased time on MV and longer hospitalization. A 2016 study demonstrated a 60% increase in CICU length of stay in children diagnosed with delirium, after controlling for relevant confounders. Even early-onset delirium of only 1 to 2 days’ duration was associated with poor short-term outcomes. Delirium during hospitalization has also been linked to delusional memories and posttraumatic stress symptoms in PICU survivors. From an economic perspective, pediatric delirium is associated with a dramatic increase in hospital costs, estimated at more than $560 million annually in the United States.
Most significantly, delirium has been shown to be an independent predictor of pediatric in-hospital mortality, with a quadrupling of mortality rates beyond that predicted by severity of illness on admission (adjusted OR 4.39; CI 1.96-9.99; P < .001).
Delirium Assessment
Clinical guidelines recommend routine monitoring of all critically ill children for delirium at least once every shift. The gold standard for delirium diagnosis is a psychiatric evaluation using the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) criteria. This is time-consuming and clearly not feasible for twice-daily use in every patient in the CICU. Recognizing the need for delirium diagnosis by nonpsychiatrists, experts have developed appropriate tools for use at the bedside. There are two validated instruments suitable for use in children.
The Pediatric Confusion Assessment Method for the ICU (pCAM-ICU) is an interactive bedside tool for use in children 5 years of age and older ( Fig. 21.1A ). The medical provider assesses for alteration from baseline mental status, then evaluates attention by asking the child to squeeze the provider’s hand in response to the letter A (or as an alternative, using memory pictures if appropriate). If altered LOC is also present, then the pCAM-ICU is positive. Occasionally, when the child is alert and calm, an additional step to assess cognition (yes/no questions and hand gestures) may be required. There is a preschool version, the Preschool Confusion Assessment Method for the ICU (psCAM-ICU) available for use in children between 6 months and 5 years of age (see Fig. 21.1B ). The pCAM-ICU and psCAM-ICU provide point-in-time assessments for delirium and can be completed in a short amount of time.
