Pediatric anesthesia emergence: agitation, PAED scoring, and how to tell it apart from pain
Emergence delirium (ED) — also called emergence agitation — is a dissociative disturbance of consciousness that appears as a child surfaces from general anesthesia: thrashing, kicking, inconsolable crying, disorientation, and non-purposeful movements, with no recognition of the caregiver or surroundings. It was first described by Eckenhoff in 1961 and remains one of the most common early recovery-room complications in pediatric anesthesia. Recognizing who is at risk before the case starts is the first step in prevention.
ED presents as a state in which the child is clearly awake — eyes open, moving — but is not appropriately aware: they thrash, cry inconsolably, do not make eye contact with a parent or nurse, resist comforting, and may not recognize familiar people or the room. Movements are purposeless rather than goal-directed (unlike a child reaching for a painful ear or IV site). The episode typically begins within 5 minutes of emergence, peaks over the next 5–15 minutes, and is usually self-limited, resolving within 15–30 minutes even without treatment; episodes lasting beyond 45 minutes are unusual and should prompt a search for another cause.
ED is distinct from postoperative pain, though the two frequently coexist and are easy to confuse at the bedside — a distinction covered in depth in Stage 4.
Reported incidence varies enormously by definition and technique: roughly 10–15% with propofol total intravenous anesthesia (TIVA) versus 20–80% with sevoflurane in preschool-age children undergoing ENT or ophthalmologic surgery — the highest-risk combination in the literature.
Several risk factors combine additively:
• Age 2–6 years: preschoolers lack the cognitive coping strategies of older children and are most vulnerable to the sensory mismatch of waking up in an unfamiliar place, in pain, without a parent. • High preoperative anxiety: each point increase on the modified Yale Preoperative Anxiety Scale (mYPAS) is independently associated with greater emergence agitation — anxious induction predicts an agitated wake-up. • Sevoflurane maintenance: low blood:gas solubility drives a fast, abrupt return of consciousness (detailed in Stage 2). • Surgical site: ENT procedures (tonsillectomy, adenoidectomy, myringotomy) and ophthalmologic surgery (strabismus repair) carry the highest rates, likely from a mix of airway/facial discomfort, altered sensory input (ear packing, eye patches, blurred vision), and shorter, more stimulating cases with rapid emergence. • Rapid, unmodulated emergence and absence of intraoperative analgesia further raise risk.
The Pediatric Anesthesia Emergence Delirium (PAED) scale (Sikich & Lerman, Anesthesiology 2004) scores five behaviors, each 0 (not at all) to 4 (extremely):
1. The child makes eye contact with the caregiver (reverse-scored) 2. The child's actions are purposeful (reverse-scored) 3. The child is aware of his/her surroundings (reverse-scored) 4. The child is restless 5. The child is inconsolable
Total score ranges 0–20; a score ≥10 defines emergence delirium (the scale was validated with a Cronbach's alpha of 0.89). It should be scored serially — commonly every 5 minutes for the first 30 minutes in the PACU — since agitation evolves quickly during this window.
Sevoflurane became the standard pediatric induction agent because it is non-pungent and well tolerated for mask induction, producing a smooth, hemodynamically stable transition to unconsciousness. The same pharmacokinetic property that makes it a fast, pleasant induction agent — low blood solubility — also drives an abrupt offset at the end of the case, which is mechanistically linked to emergence delirium.
Unlike older volatile agents (halothane, isoflurane), sevoflurane does not irritate the airway, so it can be delivered by face mask without provoking coughing, breath-holding, or laryngospasm — critical for a cooperative mask induction in a frightened child. Induction is typically achieved within 1–2 minutes using an incremental or single-breath vital-capacity technique, with good cardiovascular stability at clinically used concentrations.
A volatile agent's blood:gas partition coefficient describes how readily it dissolves in blood versus staying in alveolar gas. Sevoflurane's coefficient of 0.65 is very low — it equilibrates quickly in both directions, so alveolar (and brain) concentration falls rapidly once the vaporizer is turned off.
Children compound this effect: their alveolar ventilation relative to functional residual capacity is roughly 3–5× higher than in adults, so both uptake and elimination of inhaled agents occur faster. The practical result is that children emerge from sevoflurane anesthesia in a matter of minutes — often before the cortex has had time to reorganize normal sensory integration and orientation, even though motor and brainstem arousal have already returned. This temporal mismatch between "awake" and "oriented" is the leading mechanistic hypothesis for emergence delirium.
Desflurane, with an even lower blood:gas coefficient (0.42) and faster offset than sevoflurane, is associated with similar or higher rates of emergence agitation — supporting rapid emergence itself, not sevoflurane specifically, as the operative mechanism.
Because the abruptness of emergence — not the drug per se — appears to drive ED, several intraoperative strategies reduce its incidence:
• Total intravenous anesthesia (TIVA) with propofol produces a smoother, less abrupt emergence and roughly halves ED incidence compared with sevoflurane. • Where volatile anesthesia is used, a slow, titrated wake-up ("low and slow") rather than an abrupt vaporizer shut-off allows more gradual cortical reorientation. • Adjuncts such as dexmedetomidine and an end-of-case propofol bolus (Stage 5) blunt the transition further, independent of which primary agent was used.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
The child arrives in the post-anesthesia care unit still groggy, and within minutes the picture changes: crying escalates to screaming, limbs thrash against the bed rails, the child pulls at monitors and dressings, and nothing the nursing staff or parent does seems to help. Vital signs swing upward. Recognizing this pattern promptly — and scoring it objectively — guides whether reassurance alone is enough or pharmacologic rescue is needed.
Typical features include: thrashing or kicking that risks falling or self-extubation of a nasal cannula/IV; loud, inconsolable crying unrelieved by parental presence or soothing; disorientation — the child does not track a parent's face or respond appropriately to their voice; non-purposeful movements (flailing rather than reaching for a specific painful area); and dilated pupils with tachycardia and mild hypertension from the associated sympathetic surge. Oxygen saturation is typically preserved, which helps distinguish ED from hypoxia-driven agitation (always confirm this first — see Stage 6).
At the bedside, the five PAED items (eye contact, purposeful action, awareness of surroundings, restlessness, inconsolability — each 0–4) are rated from direct observation over roughly one minute, without disturbing the child to test responses. A running total ≥10 out of 20 meets the validated threshold for emergence delirium; some centers use ≥12 for a more specific (fewer false-positive) diagnosis. Trending the score every 5 minutes shows whether the episode is escalating, plateauing, or resolving on its own.
Not every agitated wake-up needs a drug. Brief, mild agitation that responds to a quiet environment, dim lighting, a parent's voice, and gentle containment can simply be observed through its natural 15–30 minute course. Escalation to pharmacologic rescue (Stage 6) is appropriate when the PAED score is persistently high (≥10–12), the child is at risk of self-injury or dislodging a surgical dressing/airway device, or the episode is prolonged beyond the usual window despite comfort measures and a normal safety check (oxygenation, analgesia, bladder).
A thrashing, crying child in the PACU could be experiencing emergence delirium, undertreated pain, or — commonly — both at once. The distinction matters clinically: sedating a child in true pain leaves the pain untreated and adds unnecessary drowsiness, while giving opioids to a child with pure ED provides no benefit and adds a real risk of oversedation and hypoventilation. A structured bedside assessment resolves the ambiguity in most cases.
Both present as crying, thrashing, and resistance to comfort in a child who cannot yet reliably verbalize what is wrong. Both peak in the same early recovery window. And because sevoflurane and painful surgery (ENT, ophthalmology) are risk factors for both ED and inadequately controlled pain, the two frequently coexist in the same child, muddying the clinical picture further.
A few bedside observations separate the two reasonably well: a child in pain will often localize — reaching toward the ear, throat, or eye — and can frequently be at least partly consoled by being picked up, held, and spoken to softly, especially once analgesia is on board. A child with pure emergence delirium is inconsolable despite comfort measures, does not localize to a specific body part, does not make sustained eye contact or track a parent's face, and often appears not to recognize the caregiver at all. Response to a therapeutic trial of analgesia is itself diagnostic: pain that resolves with an opioid or acetaminophen dose was pain; agitation that persists unchanged was not primarily pain.
When in doubt: score both scales concurrently — PAED for delirium features and FLACC (Face, Legs, Activity, Cry, Consolability, each 0–2, total 0–10) for pain behavior. If FLACC is elevated or the surgical procedure is known to be painful and no analgesia has been given recently, treat pain first with a weight-based opioid or acetaminophen/NSAID and reassess in 10–15 minutes. If agitation persists despite adequate analgesia, normal oxygenation, and an empty bladder, the diagnosis of emergence delirium is supported and dexmedetomidine or propofol rescue (Stage 6) is the more appropriate next step.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
Because ED is unpleasant for the child, distressing for parents, and disruptive to PACU workflow, prevention is preferable to rescue. No single intervention eliminates it, but combining pharmacologic prophylaxis, a gentler anesthetic technique, and family-centered behavioral strategies meaningfully lowers incidence.
Dexmedetomidine is the best-supported preventive agent: an intravenous dose of 0.3–0.5 mcg/kg given at induction or toward the end of the case, or an intranasal premedication dose of 1 mcg/kg roughly 30–45 minutes before induction, reduces ED incidence by approximately half in multiple randomized trials and meta-analyses, with a number-needed-to-treat around 3–4. A single propofol bolus of 1 mg/kg (or a low-dose propofol infusion) given during the last 10 minutes of a sevoflurane case also reduces ED, likely by softening the transition out of anesthesia. Oral midazolam premedication (0.5 mg/kg, max 20 mg, 20–30 minutes before induction) reliably reduces preoperative anxiety and separation distress but has inconsistent direct effect on ED itself, and can prolong recovery-room stay. Clonidine (3–4 mcg/kg), another alpha-2 agonist, is a reasonable alternative to dexmedetomidine where the latter is unavailable. Intraoperative opioid (e.g., fentanyl ~1 mcg/kg) reduces ED when a painful component is anticipated, by pre-treating the pain that would otherwise compound agitation.
Dexmedetomidine remains the single most consistently effective pharmacologic preventive strategy across pediatric anesthesia literature — more so than midazolam premedication, which addresses preoperative anxiety but not emergence agitation itself.
Where feasible, propofol TIVA instead of sevoflurane maintenance lowers baseline ED incidence substantially (Stage 2). If a volatile technique is used, a gradual, titrated emergence rather than an abrupt vaporizer shut-off gives the cortex more time to reorient. Regional or local anesthetic blocks (e.g., peritonsillar infiltration, sub-Tenon block for strabismus surgery) reduce both intraoperative volatile requirements and postoperative pain, indirectly lowering agitation risk from both mechanisms. A quiet, dimly lit PACU bay with minimal early stimulation during the first minutes of emergence is a simple, low-cost adjunct.
Preoperative preparation programs (child life specialists, hospital tours, age-appropriate explanation) and distraction techniques (tablets, videos, bubbles) during induction lower preoperative anxiety, which is itself an independent ED risk factor. Parental presence at induction reduces anxiety at that stage for many children, though its effect on ED specifically is mixed in trials. Parental presence at emergence — having a parent at the bedside as the child wakes, using a familiar voice and a comfort object — is a simple, evidence-supported adjunct that can shorten and soften agitated episodes even when it does not prevent them outright.
When prevention fails and a child is thrashing in the PACU, management follows a deliberate sequence: protect the child from injury, rule out every treatable mimic of ED, and only then reach for sedative rescue medication — reserving opioids for confirmed pain rather than using them as a default sedative.
Before anything else, protect the child from self-injury: pad or raise bed rails, keep a hand gently on the child rather than applying forceful restraint (aggressive physical restraint can escalate agitation and risks injury to both child and staff), and remove or secure anything the child could pull — IV lines, surgical drains, dressings, a nasal cannula. Keep the pulse oximeter on and continuously monitor oxygen saturation and airway patency throughout the episode. Dim the lights, minimize unnecessary stimulation and noise, and bring a parent to the bedside if not already present — a familiar voice is often genuinely calming even before any drug is given.
Agitation in the PACU is not always ED. Systematically exclude, in roughly this order:
• Hypoxia or airway obstruction — check SpO₂ and airway patency first, always, before attributing agitation to ED • Undertreated pain — assess with FLACC and consider a weight-based analgesic trial, especially after ENT/ophthalmologic surgery (Stage 4) • Full bladder — common after longer cases or a caudal/regional block; consider straight catheterization or simply allow more time if voiding is imminent • Hunger or thirst, and hypoglycemia in very young or fasting-fragile children • Residual anesthetic effect itself, which is transient and improves with time alone
Only once these are addressed or excluded should the episode be treated as emergence delirium requiring sedative rescue.
First-line: dexmedetomidine 0.5 mcg/kg IV given over 5–10 minutes — effective with minimal respiratory depression, making it attractive in a still-recovering airway. Alternative: propofol 0.5–1 mg/kg IV in small titrated boluses — fast-acting but carries some resedation risk and may modestly prolong PACU discharge time. Midazolam is generally less favored for rescue, as it can produce paradoxical disinhibition in some young children rather than calming them. Opioids (e.g., fentanyl 0.5–1 mcg/kg IV) should be reserved for cases where pain has been specifically identified — giving an opioid for pure ED adds sedation and respiratory depression risk without treating the underlying process.
Opioids are not a default treatment for postoperative agitation. Give them only when pain has been actively assessed and identified as a contributor — reflexive opioid dosing for undifferentiated agitation risks oversedation and hypoventilation with no benefit if the true process is emergence delirium.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|