The next challenge is what happens when the patient survives the initial injury, sedation begins to lighten, and the bedside suddenly becomes dominated by recurrent episodes of tachycardia, hypertension, tachypnea, diaphoresis, hyperthermia, and extensor or dystonic posturing. These episodes are often called “storming,” but the preferred term is paroxysmal sympathetic hyperactivity, or PSH. The distinction is more than semantic. “Storming” encourages clinicians to label any dramatic autonomic instability after brain injury as a single disease. PSH is a defined syndrome: simultaneous, transient, paroxysmal increases in sympathetic and motor activity occurring in a subset of patients with severe acquired brain injury. Traumatic brain injury is its most common substrate, particularly severe diffuse injury, but PSH also occurs after anoxic injury, large hemorrhage, ischemic stroke, encephalitis, hydrocephalus, and other destructive cerebral processes. The attending-level task is not simply to recognize that the heart rate and temperature are rising. It is to decide whether these changes represent PSH, another immediately dangerous process, or several overlapping processes occurring in the same injured patient.

PSH usually emerges during the first one to two weeks after severe brain injury, although the timing is variable and the syndrome may not become obvious until propofol, opioids, or dexmedetomidine are reduced. That creates a common diagnostic illusion: the patient appears to “develop” PSH during sedation weaning when the medications may previously have been suppressing its manifestations. Episodes are frequently provoked by apparently minor non-noxious stimulation—turning, bathing, suctioning, passive range of motion, urinary retention, bowel distension, or even an examination. They tend to begin abruptly, involve several sympathetic features together, resolve between episodes, and recur in a similar pattern. The trigger-response relationship is one of the most useful bedside clues. A patient who becomes tachycardic, diaphoretic, hypertensive, tachypneic, and rigid every time the left leg is moved has a pattern more consistent with PSH than a patient with persistent sinus tachycardia and fever throughout the day.

The current terminology comes from the 2014 international consensus statement, which replaced more than 30 competing labels, including diencephalic seizures, autonomic dysreflexia, paroxysmal autonomic instability with dystonia, and sympathetic storming. The consensus definition deliberately requires both sympathetic and motor manifestations and emphasizes that they occur simultaneously and paroxysmally. It also produced the Paroxysmal Sympathetic Hyperactivity Assessment Measure, or PSH-AM, which remains the standard structured diagnostic framework despite incomplete external validation. The measure combines a Clinical Feature Scale with a Diagnostic Likelihood Tool. The Clinical Feature Scale grades heart rate, respiratory rate, systolic pressure, temperature, sweating, and posturing from zero to three. The likelihood component asks whether the features occur simultaneously, are episodic, are provoked by normally non-painful stimuli, persist despite treatment of alternatives, recur over several days, and remain otherwise unexplained. The combined score classifies PSH as unlikely below 8, possible from 8 through 16, and probable at 17 or above. The score is not a biomarker and should not overrule contradictory physiology, but it forces clinicians to evaluate the syndrome longitudinally rather than diagnosing it from one hypertensive episode. The original consensus statement remains the foundational definition.

The pattern helps separate these possibilities. Sepsis generally produces sustained rather than highly stereotyped autonomic activation, although rigors can appear paroxysmal. Opioid or sedative withdrawal may produce diaphoresis, tachycardia, hypertension, gastrointestinal activation, mydriasis, and agitation, but usually lacks the reproducible stimulus-induced dystonic or extensor motor component. Pulmonary embolism causes tachycardia and tachypnea but does not usually explain episodic posturing and sweating triggered by repositioning. Intracranial hypertension may cause hypertension and posturing, but classically produces bradycardia rather than tachycardia when Cushing physiology is advanced; more importantly, a new pupil change, sustained ICP elevation, declining compliance, or progressive imaging abnormality must never be attributed to PSH. Seizures can produce autonomic changes and tonic motor activity, but recurrent generalized sympathetic episodes in severe diffuse brain injury are much more often nonepileptic. Continuous EEG is appropriate when the motor or autonomic events remain ambiguous, but the absence of an ictal correlate should redirect the clinician rather than generate endless antiseizure escalation. PSH was historically called “diencephalic seizures,” yet antiseizure medications do not generally treat it unless seizures are independently present.

The physiology is incompletely established, and that uncertainty matters because there is no single lesion, neurotransmitter, or therapeutic target that explains every patient. Older descriptions localized the syndrome to the diencephalon, but modern imaging and network models point toward disconnection among cortical, diencephalic, brainstem, and spinal autonomic systems. Severe diffuse axonal injury may interrupt descending inhibitory pathways that normally constrain responses to sensory input. The excitatory-to-inhibitory ratio model proposes that loss of descending inhibition leaves spinal and brainstem circuits excessively responsive to afferent stimulation. A routine turn then produces an amplified motor and sympathetic response rather than a proportionate one. This explains why episodes are so often triggered by suctioning or passive movement and why reducing afferent input with analgesia or gabapentin may help even when the patient cannot consciously experience or communicate pain.

That mechanism also explains why posturing in PSH should not automatically be interpreted as evidence of new herniation. The motor component may consist of extensor posturing, flexor posturing, dystonia, increased tone, jaw clenching, or a mixture that varies by limb. It reflects abnormal motor network activation occurring with sympathetic discharge. Nevertheless, the diagnosis is longitudinal: a first episode of extensor posturing and hypertension in an acute TBI patient still requires urgent exclusion of increased ICP, seizure, and structural deterioration. Only after a recurrent, stereotyped, stimulus-sensitive pattern emerges and competing causes have been addressed does PSH become the more parsimonious explanation.

Treatment becomes clearer when divided conceptually into aborting individual episodes, reducing their frequency and severity, and limiting the complications produced by the syndrome. These goals require different drugs. A medication that terminates an episode within minutes may be unsuitable as continuous prevention. A preventive medication may need several days of titration and will appear ineffective if judged ten minutes after administration. The common failure is to select one drug, expect it to accomplish every goal, escalate it rapidly, and then conclude that the patient has “refractory PSH” when the pharmacologic strategy was mismatched to the problem.

Before adding medication, remove correctable triggers. Check the bladder and bowel, evaluate painful fractures or wounds, examine lines and tubes, reduce unnecessary suctioning, coordinate care activities, and provide premedication before predictable stimulation. Maintain a neutral thermal environment, replace fluid losses, provide adequate calories and protein, preserve joint range of motion, and treat spasticity and contracture risk. These are not ancillary measures. If every turn or suctioning event provokes a crisis, environmental and procedural modification may reduce the total sympathetic burden as much as another scheduled drug.

For severe acute episodes, rapid analgesic and sympatholytic treatment is usually required. Intravenous opioids, particularly morphine, have the longest clinical history and can terminate episodes quickly by reducing afferent nociceptive input and sympathetic outflow. Small titrated intravenous doses are often used, with the exact dose determined by prior opioid exposure, ventilation, hemodynamics, and renal function. The response can be diagnostically informative but is not specific; pain, ventilator distress, and withdrawal may also improve with opioids. Excessive or repeated opioid use prolongs respiratory suppression, ileus, tolerance, and impaired examination, so successful abortion of episodes should lead to a preventive strategy rather than indefinite rescue dosing.

Dexmedetomidine is particularly useful when rapid titratable sympatholysis is needed in an intubated patient because it reduces central sympathetic outflow while generally preserving respiratory drive. It may facilitate liberation from propofol or benzodiazepines and permit a more interpretable examination. Its limitations are bradycardia and hypotension, especially in patients whose CPP is dependent on systemic pressure. The correct endpoint is not a perfectly flat heart rate if achieving it reduces cerebral perfusion. Clonidine provides a related enteral alpha-2 strategy for longer-term control, but its slower titration, hypotension, bradycardia, and potential rebound sympathetic activation after abrupt discontinuation must be anticipated. Transitioning from dexmedetomidine to clonidine can be useful, but the enteral drug should not be treated as a pharmacologically exact substitute.

Propranolol is frequently chosen because its lipophilicity permits central nervous system penetration and because beta blockade addresses tachycardia, hypertension, and the metabolic consequences of catecholamine excess. Enteral regimens are generally titrated from relatively modest doses, often administered every six to eight hours, while monitoring heart rate, blood pressure, CPP, conduction, bronchospasm risk, and the effect on episode frequency. Propranolol should not be escalated solely to normalize every transient tachycardic value. The patient may still need compensatory cardiac output during fever, infection, hypovolemia, or evolving shock. Labetalol provides mixed alpha- and beta-blockade and may help when hypertension is prominent, but clinical evidence for selecting one beta-blocker over another in PSH is limited.

Gabapentin is often one of the most useful preventive agents when episodes are stimulus-sensitive and the patient has prominent tone, allodynia, or suspected neuropathic afferent amplification. It reduces excitatory neurotransmission through the alpha-2-delta subunit of voltage-gated calcium channels and can dampen the afferent limb that provokes episodes. It generally has less direct hemodynamic effect than propranolol or clonidine, which is valuable when CPP is fragile. The tradeoffs are somnolence, accumulation in renal dysfunction, and the need for enteral administration and titration. Gabapentin should be adjusted for renal clearance, and abrupt cessation after sustained higher-dose therapy may itself contribute to withdrawal symptoms. Evidence consists primarily of case series and observational experience rather than robust trials, a limitation shared by much of PSH pharmacotherapy. A 2021 scoping review found heterogeneous reporting, frequent multidrug therapy, and insufficient evidence to establish a single optimal regimen. That pharmacologic review reinforces that treatment remains mechanism-guided rather than protocol-proven.

Baclofen is most relevant when spasticity and the motor component are dominant. Enteral baclofen may reduce tone but can cause sedation, weakness, hypotonia, and impaired arousal, and its penetration into the central nervous system is variable. Abrupt withdrawal is dangerous and can cause severe hyperthermia, rigidity, autonomic instability, rhabdomyolysis, seizures, and organ failure—a syndrome that can be mistaken for catastrophic recurrent PSH. Intrathecal baclofen has been used for severe refractory spasticity-associated PSH, particularly later in recovery, but it is an invasive escalation with risks of infection, pump or catheter failure, overdose, and life-threatening withdrawal. It is not a routine early ICU intervention.

Bromocriptine is sometimes added when hyperthermia and diaphoresis are particularly resistant, based on dopaminergic effects within hypothalamic thermoregulatory pathways. The evidence is weak, dosing is empiric, and adverse effects include hypotension, nausea, confusion, and rarely vasospastic or fibrotic complications with sustained exposure. Dantrolene may reduce severe peripheral muscle contraction and heat generation but does not treat central sympathetic discharge; hepatotoxicity and generalized weakness limit prolonged use. Benzodiazepines can suppress motor activity and acute agitation, but repeated dosing obscures the examination, promotes delirium, and delays ventilation liberation. They are most defensible when anxiety, withdrawal, or seizures coexist rather than as default long-term PSH therapy.

The practical strategy is therefore phenotype-based combination therapy. A patient whose episodes are brief, stimulus-sensitive, painful-appearing, and hemodynamically tolerated may respond to gabapentin for prevention with limited opioid rescue. A patient dominated by tachycardia and hypertension may benefit from propranolol, provided CPP and cardiac output remain adequate. A ventilated patient with frequent severe episodes may initially need dexmedetomidine, followed by transition to enteral agents. Prominent spasticity may justify baclofen, while refractory hyperthermia may lead to cautious consideration of bromocriptine. Combination therapy is common because different agents target the afferent trigger, central sympathetic output, peripheral adrenergic consequences, pain, and motor activity. The objective is not polypharmacy for its own sake; it is to use complementary mechanisms at lower doses while avoiding the toxicity of forcing one medication to do everything.

Response should be measured with the same discipline used for diagnosis. Record episode frequency, duration, triggers, peak heart rate and pressure, temperature, posturing severity, rescue-medication use, and adverse effects. Repeat the Clinical Feature Scale longitudinally. A lower resting heart rate does not necessarily mean the syndrome is controlled if turning still causes prolonged hyperthermic posturing. Conversely, occasional brief stimulus-associated increases may be acceptable if the patient is otherwise comfortable, mobilizing, and progressing. Treatment should target clinically consequential episodes, not eradicate every autonomic fluctuation from an injured nervous system.

Daily de-escalation is essential. PSH usually diminishes as the brain recovers, but medications accumulate while ownership fragments across the ICU, rehabilitation service, and outpatient setting. Propranolol, clonidine, gabapentin, baclofen, opioids, and bromocriptine may all remain on the list long after the reason for each has become unclear. Taper one medication at a time when possible, beginning with the agent contributing the greatest harm or addressing a feature that has resolved. Avoid abrupt withdrawal of clonidine, baclofen, opioids, benzodiazepines, or sustained high-dose gabapentin. The 2024 Scandinavian Delphi consensus similarly emphasized systematic use of the PSH-AM, nonpharmacologic trigger management, minimization of sedating drugs when rehabilitation participation matters, and regular medication review as the syndrome evolves. That document is expert consensus rather than a randomized treatment guideline, but it provides a practical bridge between neurocritical care and rehabilitation. The Scandinavian consensus is especially valuable because PSH often persists beyond ICU discharge.

References