Yes, a person in a coma can often hear you. Coma abolishes wakefulness and awareness, not the ear or the auditory nerve. Most coma patients show brainstem responses to sound, and a substantial minority show cortical responses to speech. Whether that sound becomes a felt experience depends on the depth and cause of the coma and on how it is measured. The honest answer is layered, and two habits change it: close the loop before speaking, and favor familiar voices.
The word "hear" splits into three questions with three different answers. Detection asks whether the ear and brainstem convert sound into nerve signals; this persists in most comas. Discrimination asks whether the cortex distinguishes speech from noise; this is variable and testable. Awareness asks whether the patient experiences what is said; this is the hard question, and science cannot yet answer it for every person. Keeping those layers separate prevents overpromising and underselling at the same time.
The auditory machinery below awareness
Sound travels through the outer and middle ear to hair cells in the cochlea. These cells convert pressure waves into electrical pulses carried by the auditory nerve toward the brainstem. Two features matter for coma. First, the ear is passive hardware — it receives sound automatically, awake or not. Second, the auditory pathway crosses the brainstem at multiple relay stations, so a single lesion rarely silences both ears. This is why the brainstem auditory evoked potential (BAEP) — a scalp recording of the first 10 milliseconds of the response — can be normal for years in people with catastrophic cortical damage.
Above the brainstem, the thalamus gateways signals to the auditory cortex. Wakefulness and awareness depend on separate circuits: cortical neurons in sleep and coma enter a down state, then a burst of activity, about once per second — the slow oscillation of anesthesia and coma. Typically in awake brains, a sound evokes a sharp, millisecond-scale response. In comatose brains, the same sound may evoke slow waves lasting hundreds of milliseconds or nothing recordable at all. The hardware still reaches the cortex; the cortex's reply comes back slower, weaker, or not at all.
A useful analogy: the auditory nerve is a microphone that is always plugged in. The cortex is a mixing board whose power may be off. You can verify the microphone works while the board is dark — and in carefully selected cases, flickers at the board suggest some hidden processing.
What the numbers say about hearing and prognosis
Disorders of consciousness come in grades, and the numbers attached to each grade matter. Coma is defined by absent wakefulness and awareness — eyes do not open to stimulation — persisting at least one hour; it usually does not last beyond a few weeks before resolving to brain death, recovery, or a vegetative state, renamed unresponsive wakefulness syndrome (UWS) in 2010. The minimally conscious state (MCS) was formalized in 2002 as inconsistent but reproducible evidence of awareness, such as visual pursuit or command following. MCS is further split: plus (MCS+) for language-related evidence, minus (MCS-) for non-language evidence.
Quantified comparisons frame the stakes. A 2018 meta-analysis in Annals of Neurology pooled six studies of coma after cardiac arrest and found approximately 30% of patients reached MCS within 72 hours of sedation being stopped. A 2020 British Medical Journal paper reported that about 25% of patients diagnosed as vegetative actually showed covert consciousness on functional MRI — but in a designated scan that required them to follow commands in their heads. A separate International Brain Injury Association working group estimated that roughly 15-20% of patients who appear unresponsive may have covert consciousness. These are minority outcomes, not the rule, and they matter because hearing may be the channel through which some of these patients are reached.
Prognosis hinges on why the brain is offline. In cardiac arrest coma, the 2020 European Resuscitation Council guidelines endorse multimodal assessment: clinical exam delayed to at least 72 hours after rewarming, pupillary and corneal reflexes, somatosensory evoked potentials (N20), electroencephalography, and neuron-specific enolase. No single test rules out recovery. In traumatic coma, outcome depends heavily on age and lesion pattern. In drug-induced coma, recovery is usually complete. Knowing the cause tells you more about likely hearing outcomes than one recording of the auditory pathway can.
Edge cases where the bedside logic breaks
Not every unresponsive patient is in coma, and several states complicate what hearing can mean. In the locked-in syndrome, awareness and hearing are fully intact; only voluntary movement is lost. Its 1966 description by Plum and Posner emphasized that the patient is awake inside. In UWS, the brainstem maintains sleep-wake cycles and eye opening, but cortical function is largely absent; auditory startle and even orienting can occur without awareness. In MCS, command following is the rule. Some patients float between these states across days, so a single exam is a snapshot, not a verdict. Slow recovery can be measured in months: some patients emerge long after early predictions of doom, including one widely reported case described by Fins in 2007 involving a man who recovered conversationally years after being labeled vegetative.
The counterintuitive detail is the temporal one. In a controlled 2015 comparison, hearing provinces differ from what families expect. Some UWS patients show preserved early cortical auditory responses — their brains react to a tone — yet show no evidence of awareness on command-following tasks. Conversely, a small number of behaviorally unresponsive patients show language-evoked activation on fMRI despite absent cortical responses to simple tones. A recording of the auditory pathway cannot diagnose awareness, and a lack of cortical response cannot certify that nothing is heard. The signals and the experience diverge, and clinicians need multiple approaches to each.
What families and clinicians get wrong
The most common misconception is the flat rule that coma patients cannot hear. Evidence supports a spectrum: most retain some level of auditory processing, many process speech in some way, and some are covertly aware. The equal and opposite mistake is to assume that hearing equals comprehension. A patient may hear you without understanding, understand without remembering, or be entirely unable to respond yet later report having heard. Families interpret random limb movements as responses; the unresponsive wakefulness syndrome literature shows that many "responses" are reflexive. A rule of thumb: behaviors that repeat to command, consistently, count; one-off movements do not.
Another misconception concerns the ethics. Some argue that if awareness is uncertain, we should behave as if the patient can hear. The professional caveat is that "act as if" is not the same as "we know they can." Families deserve honest uncertainty, not false comfort. Clinicians should audit their language: "we cannot measure awareness" is defensible; "he can hear you" as a blanket statement is not.
Flag these badly supported bedside claims when you see them:
"Coma patients cannot hear anything" — contradicted by preserved brainstem and cortical responses in many patients. "A normal auditory brainstem response proves awareness is intact" — it proves the ear and lower brainstem work, nothing about the cortex. "A flat EEG means she has no inner life" — some covertly conscious patients produce unrecognizable surface activity; sophisticated tools are needed. "Talking louder will reach him" — volume does not restore cortical processing; familiar, slow speech is the better lever. "He squeezed my hand; he understands" — reflexive grasp is common; only consistent, repeatable command following counts. "The scan says nothing, so stop talking" — absence of evidence on one test is not evidence of absent hearing.
Practical actions when you sit at the bedside
Close the loop before speaking. Say the patient's name, then pause two to five seconds, then speak. This gives slow, metabolically depressed cortex time to complete a cortical response — the same reason anesthesia research uses "oddball" sound sequences with long gaps. Use familiar voices, because speech-evoked responses are larger and more reliable for names and familiar narratives than for neutral tones in the literature on auditory processing. Keep sentences short and concrete: "It's Sarah. I'm here. The dog is fine." Repeat the same content across visits, so recognition is not a fresh problem each time.
Practical thresholds that clinicians and families can act on:
For diagnosis: at least 72 hours after sedation is stopped is the standard window before strong prognostic statements; in cardiac arrest, the same 72 hours after rewarming applies. For command following at the bedside: repeat each command several times; only consistent responses count, and video-record ambiguity for the care team. For families: speak in your normal voice, within your normal range, rather than shouting. For back-referrals: ask whether a formal CRS-R (Coma Recovery Scale-Revised) has been done — the validated bedside measure — and whether the team considered advanced imaging or EEG for covert consciousness. For temperature: keep the room quiet; noise stress does not improve anything.
The best evidence-supported move is not a script but a routine: consistent presence, familiar voice, closed-loop speech. Families should tell the care team what they observe, and clinicians should be candid about what those observations can and cannot establish.
Limits: what no one can currently promise
No test can certify at the bedside that a specific patient hears and understands you. The tools that come closest — repeating a command during fMRI or EEG in the minority of centers that offer it — are expensive, available only at specialized research sites, and results must be interpreted by teams trained in preventing false positives. In 2020, a joint report by several international societies (the ACMUI/DGN/ESICM group) offered a standardized framework, but it does not settle the individual case. The dominant practical truth: in most comas, auditory detection is preserved and meaningful awareness is uncertain. Acting kindly is not the same as making a promise. For now, speak as if it matters, and say plainly that you cannot know for sure.
Turning toward prognosis, the strongest statement a clinician can make is probabilistic. Hearing is common; awareness is the exception; and both can change across days. The way to serve patients and families is to keep the question open — and to keep talking while the science catches up.