In cardiac arrest, constricted, light-responsive pupils suggest a favorable prognosis. In contrast, dilated, light-unresponsive pupils are indicative of cerebral anoxia or brain death and constitute an adverse prognostic sign. Notably, the presence of dilated, nonreactive pupils 48 hours post-cardiac arrest is a compelling sign of brain death. However, the absence of pupillary light reflex immediately post-cardiac arrest is more equivocal. Various pharmaceutical agents, including epinephrine—administered to the vast majority of cardiac arrest patients—can induce such pupillary responses. In a comprehensive study by Javaudin et al., involving more than 11,000 cardiac arrest cases, the correlation between the initial pupillary reflex (observed upon emergency department arrival) and the neurological status assessed at 30 days was evaluated. The findings were as follows:

• The absence of a pupillary reflex demonstrated a sensitivity of 72.2% for predicting an unfavorable neurological outcome, indicating that the test failed to identify 27.8% of individuals with poor neurological recovery.
• The specificity of absent pupillary reflex for predicting a poor neurological outcome was 68.8%. This implies that approximately one-third of the patients who ultimately had a favorable neurological recovery exhibited an abnormal pupillary reflex upon initial examination.
• Notably, 31.2% of patients discharged with intact neurological function were observed to lack a pupillary reflex upon their initial presentation.
• An absent pupillary reflex upon presentation was associated with a threefold increased risk of an adverse neurological outcome.

In summary, while the absence of a pupillary reflex upon hospital admission post-cardiac arrest serves as an unfavorable prognostic indicator, its standalone utility in guiding decisions, such as termination of cardiopulmonary resuscitation, is limited.

Pupillary light reflex (PLR)

The pupillary light reflex (PLR) – the constriction of the pupils in response to light – is a fundamental neurological examination finding that reflects brainstem (midbrain) function. In the context of cardiac arrest and the post-resuscitation period, assessment of the PLR has long been considered for its prognostic implications. Clinicians often check pupil size and reactivity as part of initial and ongoing evaluations of patients in cardiac arrest or after return of spontaneous circulation (ROSC). A brisk, reactive pupil generally indicates an intact brainstem and adequate cerebral perfusion, whereas fixed, dilated pupils can signal severe global ischemia or even herniation of the brain stem.

Physiological Basis of the Pupillary Reflex in Cardiac Arrest

The pupillary light reflex is controlled by a simple neural circuit: light signals travel from the retina via the optic nerve (CN II) to the midbrain, where interneurons in the pretectal area activate the Edinger–Westphal nuclei. Parasympathetic fibers (carried by CN III) then prompt sphincter muscles in the iris to constrict the pupil. Under normal conditions, bright light causes rapid bilateral pupillary constriction (a brisk reflex), whereas low light causes dilation. This reflex is highly sensitive to brainstem integrity and cerebral perfusion. During cardiac arrest, circulation to the brain is severely compromised or absent, leading to rapid brainstem hypoxia. As a result, pupils often dilate widely within 30–120 seconds of cardiac arrest and light reactivity is lost when cerebral blood flow ceases. In other words, during prolonged cardiac arrest without effective perfusion, one expects fixed, dilated pupils. Conversely, if some blood flow is maintained (for example, high-quality cardiopulmonary resuscitation or partial ROSC), the midbrain may receive enough oxygen to preserve some pupillary reflex.

Importantly, pupillary size and reactivity are influenced by factors beyond neural activity. Systemic catecholamines released or administered during resuscitation can affect pupil tone – e.g. epinephrine (adrenaline) can cause pupillary dilation via α-adrenergic effects. Likewise, atropine, occassionally used in cardiac arrest, blocks vagal input and can dilate pupils. Environmental factors (bright ambient light or darkness) and core body temperature also alter the reflex; severe hypothermia can reduce or even transiently abolish pupillary reactions despite an intact brainstem. Understanding these factors is crucial for clinicians so that a missing pupillary reflex is interpreted in context, especially in the hectic environment of a cardiac arrest.

Methods for assessing the pupillary light reflex

Standard examination

Traditionally, clinicians assess the pupillary reflex using a penlight or ophthalmoscope. The key observations are pupil size (millimeters of diameter) and the speed/degree of constriction in response to light (brisk, sluggish, or absent). This bedside method is simple and quick, but it has limitations. It is subjective and qualitative – different examiners may disagree on what constitutes a “sluggish” response, and subtle pupillary movements can be missed. Ambient lighting and examiner technique (e.g. light intensity and angle) can also affect the exam. Inter-observer variability in manual pupillary assessments is well documented. Despite these limitations, the penlight exam remains the most widely used method, especially in prehospital and emergency settings due to its immediacy and no requirement for special equipment.

Quantitative Pupillometry

In recent years, automated pupillometry devices have emerged as a novel technology to objectively measure the pupillary light reflex. These handheld infrared devices (e.g. NeurOptics NPi® monitor) measure baseline pupil size, then shine a calibrated light stimulus and record the exact change in pupil diameter, constriction velocity, dilation velocity, and latency. The output often includes the percentage of pupillary constriction (sometimes called qPLR for quantitative PLR) and a composite Neurological Pupil Index (NPi). The NPi is a proprietary score from 0 to 5, where 4-5 is normal, and values <3 indicate abnormal or sluggish reactivity (an NPi of 0 denotes a truly non-reactive pupil). Automated pupillometry thus provides standardized, numeric data that can be tracked over time. It eliminates much of the subjectivity of the manual exam and can detect very small pupillary changes invisible to the naked eye. For example, a pupillometer might detect a 5% constriction that a human examiner would call “non-reactive”. This improved precision has significant implications for prognostication, as discussed later.

Benefits of pupillometry: The technology has shown excellent inter-rater reliability and reproducibility. It also enables clinicians to set specific threshold criteria for prognosis. Indeed, current guidelines recommend using quantitative pupillometers for neuro-prognostication after cardiac arrest, in order to reduce observer variability. Pupillometry is noninvasive and quick (measurements take only seconds). Modern devices store trends, allowing serial monitoring of the pupil reflex over the post-arrest course.

Limitations: Automated pupillometers require dedicated equipment and can be more expensive or not universally available. They also require that the patient’s eyes be accessible (not swollen shut, etc.), and severe ocular trauma or pre-existing ophthalmic conditions (like prior eye surgery) may limit accuracy. Nonetheless, in critical care settings that have pupillometers, they have rapidly become part of standard post-cardiac arrest care.

Novel and experimental approaches

Beyond handheld pupillometers, research is exploring continuous or video pupillography during CPR. Feasibility studies have used video cameras or infrared recording to continuously monitor pupil size during ongoing chest compressions. The goal is to identify real-time changes in pupil size/reactivity that might signal improving perfusion or impending ROSC. One animal study in pigs correlated pupillary dynamics with CPR quality: when coronary perfusion pressure improved during CPR, pupils gradually constricted and began reacting to light, whereas if pupils stayed fixed and dilated through 6 minutes of CPR, resuscitation was uniformly unsuccessful. Early human studies have similarly categorized pupillary changes during out-of-hospital CPR, but this remains an experimental domain. In the future, continuous pupillary monitoring might aid real-time decision-making (for example, to gauge whether CPR is providing adequate cerebral perfusion or to detect ROSC even before a pulse is palpable).

At present, the mainstay methods are the bedside exam and intermittent checks with automated pupillometry.

Pupillary reflex assessment during cardiac arrest

Assessing the pupillary reflex during an ongoing cardiac arrest (during CPR, prior to ROSC) can provide some information, but it must be interpreted with caution. In the hyperacute phase of cardiac arrest, the absence of a light reflex is expected in many cases because of severely reduced brain perfusion. Fixed, dilated pupils during CPR have traditionally been seen as a grave sign of cerebral anoxia. Indeed, if a patient’s pupils are completely unreactive early in an arrest, it indicates that the midbrain is not being perfused well – either due to low blood flow or very severe hypoxia. Clinical dogma has been that “fixed pupils” in arrest suggest a low likelihood of recovery. However, modern evidence shows this is not an absolute indicator of futility. Many cardiac arrest patients have transiently fixed pupils from the arrest itself or medications, yet can survive with good outcomes, especially if high-quality CPR and advanced therapies are delivered.

Crucially, resuscitation guidelines do not recommend using fixed pupils alone as a reason to terminate resuscitative efforts in cardiac arrest. Outcomes are influenced by many factors (no-flow time, low-flow time, cause of arrest, patient comorbidities, etc.), and a single exam finding during CPR is insufficient to make decisions about stopping attempts. The pupillary reflex can reflect the effectiveness of CPR to some extent. If during compressions a patient’s pupils are observed to constrict or show some flicker of reactivity, that is encouraging – it suggests some oxygenated blood is reaching the brain. Thus, the presence of a pupillary reflex during an arrest is a positive sign – it may indicate that ROSC is imminent or that cerebral perfusion is improving.

On the contrary, absence of pupillary reflex during arrest is not a definite finding. Most patients in cardiac arrest will lack pupillary reactivity until circulation is restored; this does not reliably predict poor outcomes. Additionally, certain resuscitation drugs, as noted above, can affect pupils. For example, epinephrine given during CPR often causes pupillary dilation; however, it does not abolish the light reflex once ROSC is achieved. If a patient has regained a pulse, nearly all will have a reactive pupil even with administered epinephrine. One study of in-hospital arrests found 97.5% of patients had reactive pupils within minutes of ROSC despite receiving epinephrine (and many with atropine), indicating that lack of PLR post-ROSC should not simply be blamed on drugs (Achamalla et al). In other words, if pupils remain non-reactive immediately after achieving ROSC, it is likely due to severe neurological injury rather than just medication effects. This emphasizes that while the patient is still in arrest (prior to ROSC), drugs and low flow state can make pupils non-reactive, but as soon as circulation returns, a viable brainstem will usually resume a light reflex, even if epinephrine or atropine were used.

In summary, during an ongoing cardiac arrest, a reactive pupillary reflex (even faintly reactive) is a favorable sign suggesting some preserved brainstem function and a higher probability of ROSC. Fixed, non-reactive pupils during CPR signal that cerebral perfusion is critically low; if they persist despite improved CPR quality, it raises concern for a poor outcome. However, clinicians should not use fixed pupils in isolation to terminate resuscitation – this finding must be weighed with the overall context.

Pupillary reflex in the immediate post-resuscitation period

After ROSC is achieved, the pupillary examination becomes an important component of the neurologic assessment of the patient. In this immediate post-resuscitation phase (hours to the first day after the arrest), pupil size and reactivity are documented repeatedly as part of evaluating brain injury severity. Assessing cerebral injury at this stage is difficult; some patients may remain comatose due to global ischemic injury or due to sedative medications and targeted temperature management. Pupillary responses help stratify risk but are not an absolute predictor at this stage.

Importantly, 48 hours after cardiac arrest, persistently dilated and unreactive pupils strongly suggest brain death, if corroborated by other brainstem areflexia. In brain death, the pupils are typically mid-position or large and do not respond to light at all.

However, in the immediate hours post-ROSC, the absence of a pupillary reflex is less conclusive. Early after resuscitation, resuscitation drugs are still in circulation, the patient may be acidotic or hypotensive, or might be undergoing therapeutic hypothermia – all of which can depress pupillary reactivity. As noted, epinephrine and atropine can dilate pupils but generally do not prevent a light reflex once circulation is restored. Therefore, clinicians are cautious in prognosticating based on a single early pupillary check.

Evidence from clinical studies

Early absence of the pupillary light reflex (PLR) following cardiac arrest is linked to worse neurological outcomes but should not be used in isolation to guide clinical decisions. In a large multicenter study of over 10,000 patients, absence of PLR on admission predicted poor outcome with 72.2% sensitivity and 68.8% specificity. Importantly, 31% of patients who ultimately had good neurological recovery presented initially with non-reactive pupils, highlighting the risk of false pessimism.

Clinical experience supports this finding: patients whose PLR recovers within 24–48 hours post-ROSC are more likely to have favorable outcomes, while those with persistently absent PLR beyond 48 hours are at higher risk of severe hypoxic brain injury. However, accurate prognostication is generally deferred until at least 72 hours after arrest. At that point, if both pupils remain non-reactive and confounders like sedation and hypothermia are excluded, the absence of PLR becomes a highly specific predictor of poor outcome.

Serial examination is essential. Restoration of pupillary reactivity—even if sluggish—is a reassuring sign of recovering brainstem function. Conversely, new asymmetry or fixed dilation, particularly unilateral, may indicate acute neurological deterioration such as herniation and should prompt urgent evaluation.

In the first 24 hours post-arrest, presence of PLR is encouraging, while absence should prompt caution but not immediate conclusions. Prognosis should rely on a combination of neurologic findings, biomarkers, electrophysiology, and imaging—ideally after 72 hours—rather than a single early exam finding.

Pupillary reflex for prognostication in the subacute phase (72 hours and beyond)

At 72 hours post-cardiac arrest, the prognostic value of the pupillary light reflex (PLR) becomes more reliable. By this time, transient effects of sedatives, neuromuscular blockers, and hypothermia have typically resolved, allowing clearer interpretation of brainstem function. Bilateral absence of the PLR at or after 72 hours, confirmed in a normothermic, drug-free patient, is a highly specific predictor of poor neurologic outcome, with multiple studies reporting near 100% specificity for severe disability, coma, or death.

Quantitative pupillometry has strengthened this association. For example, an NPi (Neurological Pupil index) ≤ 2 within the first 72 hours has been shown to correlate with poor 3-month outcomes with absolute specificity in some cohorts. These findings support the inclusion of PLR in multimodal prognostication guidelines, particularly when combined with absent corneal reflexes, non-reactive EEG, or absent somatosensory evoked potentials.

However, while absent PLR at 72 hours strongly predicts poor outcome, its sensitivity is low. Many patients with poor outcomes still exhibit some pupillary reactivity during this period. Therefore, intact PLR at 72 hours does not ensure good neurological recovery, and clinicians should avoid drawing premature conclusions. Instead, further assessment is needed, including neurological examination over time, imaging, and neurophysiological testing.

In summary, absent PLR at 72 hours is a robust marker of irreversible brain injury, justifying discussions about prognosis and potential withdrawal of life-sustaining therapy. Conversely, presence of the reflex requires cautious interpretation and continued observation.

Quantitative Pupillometry in Prognostication

Quantitative pupillometry offers objective, reproducible measurements of pupillary reactivity, improving the precision of neurologic prognostication after cardiac arrest. By replacing subjective interpretation with numerical thresholds—such as qPLR% and the Neurological Pupil index (NPi)—clinicians can better stratify patients.

Evidence supports that at ≥72 hours post-arrest, specific pupillometry thresholds strongly predict poor outcome. An NPi ≤ 2 or qPLR < 4% has shown 100% specificity for unfavorable neurological outcomes, effectively ruling out the possibility of meaningful recovery when these values are confirmed in a drug-free, normothermic state. Importantly, integrating pupillometry with other markers, like neuron-specific enolase, can further increase sensitivity without compromising specificity. These data have led to its inclusion in the 2021 ERC/ESICM guidelines for multimodal prognostication.

Earlier in the course—within 6 to 48 hours—low NPi values may suggest higher risk but are not definitive. Some patients with NPi < 3 early after ROSC still recovered, so early readings are better used for risk stratification than decision-making. However, in specific contexts, such as refractory arrest or ECPR, very low or absent NPi (especially if persistent) may indicate non-survivable injury and support early discussions about futility or even brain death evaluation.

Overall, while early pupillometry offers prognostic insights, the most reliable application remains at or after 72 hours post-arrest, and always in combination with other clinical and diagnostic data.

Pupillary reflex in pediatric cardiac arrest prognostication

In pediatric cardiac arrest, the pupillary light reflex (PLR) is a key component of neurologic assessment, though data are more limited than in adults. As in adults, absent PLR at or beyond 72 hours in a comatose child is highly specific for poor neurologic outcome, although with low sensitivity. Some children may recover despite initially unreactive pupils, highlighting the importance of delayed and cautious prognostication.

Pediatric protocols often recommend waiting longer, sometimes 96 hours or more, before making definitive prognostic judgments. This accounts for potential delays in neurologic recovery and challenges in evaluating reflexes, particularly in neonates and infants with small or difficult-to-assess pupils.

While early absence of PLR can raise concern, it should not be used in isolation. Pediatric guidelines favor a multimodal approach incorporating EEG, neuroimaging, and motor responses. Cases of delayed reflex recovery followed by awakening reinforce the need for serial assessment and patience.

In summary, absent PLR beyond 72–96 hours in children is a strong indicator of poor prognosis, but early absence is not definitive. As in adults, multimodal, time-sensitive evaluation is essential for accurate prognostication in pediatric patients.

Interpreting pupillary reflex findings: Clinical guidance

• During Arrest (No ROSC yet):
  • Pupils reactive: Indicates some brainstem perfusion. This is encouraging; continue aggressive resuscitation. Reactive pupils during CPR suggest that resuscitative efforts are at least partially effective in delivering blood to the brain, and ROSC may be possible. Do not use this alone to predict survival, but it’s a positive sign.
  • Pupils fixed/dilated: Common in arrest due to global ischemia and catecholamines. By itself, this should not prompt stopping CPR, especially if downtime is short and reversible causes are being addressed. Focus on improving CPR quality (high-quality compressions, adequate ventilation) and treat underlying causes. If pupils remain completely unreactive despite optimal resuscitation (e.g., after several minutes of good CPR), it suggests very poor cerebral perfusion – the team should verify other prognostic markers (ETCO₂, ultrasound of heart, etc.) as this could portend low likelihood of ROSC. But the decision to terminate should follow established algorithms, not the pupillary exam alone.
  • One pupil fixed, one reactive: Unusual in pure cardiac arrest. Could indicate a concomitant issue like head trauma or stroke. Manage as a special case – consider intracranial catastrophe if context fits. Usually, bilateral findings are expected in primary arrest.
• Immediately Post-ROSC (0–1 hour):
  • Pupils reactive (even if sluggish): A good sign. Nearly all patients who have just been resuscitated will have at least some pupillary reactivity if brain function is intact. Continue post-arrest care (optimize oxygenation, blood pressure, initiate targeted temperature management if appropriate). Reactive pupils do not guarantee awakening, but they indicate the brainstem is functioning.
  • Pupils not reactive: Concerning, but not definitive in the first minutes to hours. Check for confounders: residual paralytics, heavy sedation, ocular trauma, or high-dose vasopressors. Ensure adequate blood pressure and oxygenation – sometimes pupils that were initially fixed will begin reacting after hemodynamics improve. The Achamallah et al. study noted that by a median of 6 minutes after ROSC, 97.5% of patients had reactive pupils. Thus, if an hour after ROSC a patient still has non-reactive pupils, one should be worried about severe brain injury. Yet, the plan would be to support and reevaluate later, not to make a prognosis on the spot. One exception: if pupils are completely non-reactive and other brainstem reflexes (corneals, spontaneous respirations) are also absent and there’s no sedation on board, one might consider brain death evaluation if the situation fits (e.g., hypoxic brain death can occasionally occur quickly if the insult was massive). But usually it’s too early for that, and further observation is warranted.
• Day 1–2 Post-Arrest:
  • Reflex present: Many patients in this window may still be comatose (due to injury or induced hypothermia) but have intact pupillary responses. This is a favorable sign as it suggests some preservation of neurologic function. Continue comprehensive critical care. If other signs are also positive (e.g. purposeful movements or improving consciousness), it bodes well.
  • Reflex absent: If by 24–48 hours the pupils remain fixed despite normalization of physiology (warmed, normotensive, off sedatives if possible), the outlook is poor. Yet, it is not completely certain – some patients could be slower to recover brainstem function. Neurologists would likely schedule formal prognostication tests for 72 hours or later. It’s too early to call, but such a patient should be monitored closely for any neurological signs. Families can be informed that the exam is very concerning, but that final assessments will be done at 72 hours. Notably, if a decision was made to cool the patient (TTM to 33°C or 36°C for 24 hours), many protocols avoid any prognostic statements until after rewarming and a period of normothermia, because hypothermia can delay neurological recovery and can even slow the pupillary reflex.
• 72 Hours Post-Arrest (and beyond):
  • Reflex absent (both sides): This is a grave prognostic finding. In the absence of confounders, two fixed, non-reactive pupils at 72h have essentially a near-zero false-positive rate for bad outcome. At this point, if the patient is still comatose, most guidelines would consider this (plus at least one other corroborating poor prognostic marker) sufficient evidence to recommend withdrawal of life-sustaining therapy due to likely devastating neurologic injury. Before doing so, careful checks must confirm that drugs (sedatives, paralytics) are truly cleared and that no reversible metabolic factor is present. If confirmed, absent pupils indicate that the patient will not regain consciousness or meaningful brain function.
  • Reflex present: If pupils are reactive at 72h, it rules out the worst prognoses (brain death is off the table), but outcome can still be variable. Some of these patients will awaken and have good recovery; others may remain in a persistent vegetative state or have severe deficits despite intact brainstem reflexes. Additional tests (MRI brain, EEG patterns, evoked potentials, biomarkers) are needed to refine prognosis. The presence of a pupillary reflex suggests that if withdrawal of support is being considered, one should be more cautious and likely wait longer or obtain more data. It is not uncommon that at 72h pupils are reactive but the patient shows no sign of cortical function – some of those patients improve in the ensuing days or weeks, while others do not. Thus, a reactive pupil gives hope but not certainty of a good recovery.
  • Reflex returns late: Occasionally, a patient who initially had absent PLR might show it later. For instance, there are anecdotes of patients with no light reflex for over 24 hours who then show a sluggish reaction on day 3. Such a return of reflex is a positive change and would reset any prior assumptions of irreversibility. Medicine must always allow for such possibilities, which is why serial exams are crucial.

Interpret degrees of response

• Brisk reaction: quick and full constriction to light. Indicates normal brainstem function. In post-arrest patients, a brisk PLR (especially if pupils are mid-size or on the smaller side) is a great sign – often seen in patients who will eventually wake up or have milder injury.
• Sluggish reaction: delayed or reduced amplitude constriction. This could mean partial injury or an ongoing effect of sedatives. Many post-arrest patients have some sluggishness in reflex early on. If sluggish at 72h, it’s treated as still “present” but abnormal; the neurologic prognosis might be guarded, but not hopeless.
• Non-reactive: no noticeable change in pupil size with light. When confirmed in a fully exam-accessible setting (no confounders), this is essentially an on/off indicator of severe brainstem dysfunction. Degrees of non-reactivity can also be quantified by pupillometry – for example, a pupillary constriction of only 1% would appear clinically “absent” and certainly is profoundly abnormal, whereas 3–4% might be visibly subtle but detectable with a device. Pupillometers might grade an abnormal reflex as NPi of 1 or 2, versus a completely absent reflex as NPi = 0.

References

Prognostic performance of early absence of pupillary light reaction after recovery of out of hospital cardiac arrest
F Javaudin 1, B Leclere 2, J Segard 3, Q Le Bastard 4, P Pes 5, Y Penverne 5, P Le Conte 5, J Jenvrin 5, H Hubert 6, J Escutnaire 6, E Batard 4, E Montassier 4, Gr-RéAC 7 Affiliations Expand PMID: 29545138 DOI: 10.1016/j.resuscitation.2018.03.020

Neurological Prognostication in Children after Cardiac Arrest.
Alyssa E Smith a, Stuart H Friess b Author information Article notes Copyright and License information PMCID: PMC7354677 NIHMSID: NIHMS1590975 PMID: 32381279

European Resuscitation Council and European Society of Intensive Care Medicine guidelines 2021: post-resuscitation care