A patient who is initially observed to be unconscious can ultimately manifest a variety of clinical states. Some patients will regain full consciousness without intervention, while others will require intensive management and intricate diagnostic testing. Etiologies of persistent unconsciousness can be reversible or permanent.
This article will focus on patients who are in a coma. Coma is defined by Plum and Posner as ‘a state of unresponsiveness in which the patient lies with eyes closed and cannot be aroused to respond appropriately to stimuli even with vigorous stimulation.’ Coma has also been defined objectively as a Glasgow Coma Scale (GCS) <8 or GCS <10.(2)
Coma is generally caused by a temporary or permanent impairment of either the reticular activating system in the brainstem, both cerebral hemispheres, or bilateral thalamus (as described below). The causes of coma can be differentiated into pathology local to the brain(structural), or systemic pathology.
Structural causes include traumatic brain injury, ischemia, hemorrhage, space-occupying lesions, or inflammation. These injuries either destroy an area listed above or exert indirect damage by way of mass effect(compression) or increased intracranial pressure. Increased intracranial pressure impairs global cerebral blood flow and can promote tissue distortion and tissue shifts (herniations).
Systemic causes of decreased consciousness are many, but they generally induce global brain dysfunction by some mechanism. The broad categories can be remembered by the acronym TOMES (Toxins, Organ Failure, Metabolic, Endocrine, Seizures).
The prevalence and relative etiologies of coma vary by institution and patient population. For example, a high-volume trauma center will likely see a significant amount of coma related to traumatic brain injury.
A 2015 systematic review that examined etiologies and prognosis of non-traumatic coma (NTC) highlighted the relative variance between institutions. Overall, the most prevalent etiology of NTC was ischemic or hemorrhage stroke (6-54%), followed by post-anoxia (3-42%), poisoning (<1-39%), and metabolic (1-29%). Despite stroke being the overall most common cause of NTC, the total non-structural causes (37-75%) tended to slightly outnumber the structural causes (28-64%).
The overall mortality was 25-87%. Stroke and post-anoxic coma had the highest mortality at 60-95% and 54-89% respectively. The lowest mortality was found in poisoning and epilepsy which both had rates less than 10%. 
The many causes of coma can be classified as affecting three main areas of the brain :
Bilateral Hemispheric Damage/Effect
Extensive damage to the bilateral cerebral cortex, as can occur with hypoxic-ischemic injury or brain trauma, causes neuronal death and de-innervation of cortical regions. Such patients lose the ability to process and consciously respond to stimuli.
The systemic causes of coma can also be placed in this category, as they produce an abnormal physiologic environment that inhibits neuronal function. This type of pattern is generally reversible if the systemic abnormality can be corrected.
Diencephalic (Thalamic) Injury
The thalamus contains relay nuclei which direct afferent input to the cortex. Thus, lesions to the bilateral thalami can mimic the result of a bilateral cortical injury.
Upper Brainstem Injury
The midbrain and pons have a reticular core that contains regions that are responsible for promoting a wakeful state. Lesions to this area, the so-called reticular activating system, can inhibit consciousness and result in a comatose state.
History regarding an unconscious patient is based on supplementary data. Questioning a person who has good knowledge of the recent history of the patient is preferable. 
Abrupt onset of altered mental status points towards drug poisoning or acute structural lesions such as trauma or stroke. Most metabolic disorders and compressive structural injuries have a relatively gradual onset.
Neurological abnormalities or a headache preceding coma point toward a structural lesion. Specifically, cranial nerve abnormalities can suggest brainstem involvement. Metabolic disturbances usually cause diffuse forebrain dysfunction manifesting as confusion, delirium, or encephalopathy before the coma.
Knowing the patient’s medical history can provide important clues to the diagnosis. A history of chronic cardiopulmonary, hepatic, or renal disease may be contributing. Use of or access to sedative or psychoactive drugs supports intoxication. 
It is essential to have an idea of the patient’s state of health before their presentation. Understanding that prognosis in many comatose patients is poor, a pre-morbid condition will inform decisions regarding the escalation of care and appropriateness of cardiopulmonary resuscitation. 
The initial step in the evaluation of an unconscious patient is to evaluate for the basic signs of life. The American Heart Association recommends examining for a pulse, followed by assessing for airway patency and breathing pattern. If the patient does not have a pulse or does not have a regular breathing pattern, BLS/ACLS is indicated.
For patients with a pulse, who are breathing adequately, the evaluation shifts to a neurological examination. The ideal neurologic examination would serve to determine the site and nature of the neurological lesion and to determine prognosis. The examination is considered most useful if the patient is well perfused, normothermic, normoglycemic, and without the effects of neurologically active toxins or medications. 
The initial step is to evaluate for reactivity, using objective measures. Address the patient verbally, and then progress to light shaking, then progress to more intense mechanical stimulation. Sufficient stimulus to the supraorbital ridge, nail beds, or temporomandibular joint can be painful without risk of tissue injury. Response to these painful stimuli should be graded bilaterally, in case of a focal spinal cord lesion. If these measures do not produce a response, vigorously pressing the knuckles up and down the sternum should arouse any patient who is not deeply comatose. 
The best practice for reporting level of responsiveness is to document specifically how the patient reacted to the external stimulus that was provided for testing. Coma scales such as GCS(Glasgow Coma Scale) and FOUR(Full Outline of Unresponsiveness) exist to aid in objectifying results for provider communication, trending, and prognosis. 
Initial eye position and movements should be noted. Nystagmus is uncommonly seen in Unconscious patients but may indicate an irritating brain lesion or even occult seizure activity. Size of the pupils (using a pupilometer) and reaction to light should be documented. A strong light should be used, as pupillary responses may be sluggish in unconscious patients. 
If available, a funduscopic exam may reveal essential findings such as papilledema or subhyaloid hemorrhage.
Examination of eye movement with oculocephalic and oculovestibular testing can provide information regarding cranial nerves III, IV, VI, and VIII. In general terms, patients with intact brainstem functioning should have a normal response to these maneuvers. Oculovestibular testing via caloric stimulation can be performed in patients who do not have a cleared C-spine.
The corneal reflex can be established by stimulating the cornea and observing for blinking. Stimulation should elicit both a direct and consensual response and suggests normal function of CN V and VII. It is recommended to stimulate the cornea with drops of normal saline to prevent corneal damage.
The gag and cough reflexes suggest normal function of cranial nerves IX and X. 
Finally, motor function is tested by examining muscular tone, movement patterns (if any), and peripheral tendon reflexes.
The physical exam should be repeated at least daily, in a sequential fashion and documented systematically.
Walker and O’Brien  published a recommended short neurologic exam summary:
Neuroimaging is extremely valuable in evaluating the unconscious patient, as an intracranial insult is common and important to define as soon as possible in the management. Computed Tomography (CT) of the brain is commonly the first study performed and can reveal several key findings including hemorrhage, infarction, space-occupying lesion, herniation, edema, and hydrocephalus. MRI/MRA is a worthwhile study in cases that are still unclear after initial neuroimaging. 
Serum testing usually includes, at least: Complete blood counts, serum electrolytes including calcium and glucose, renal/hepatic functions, coagulation panel, and toxin studies. 
Lumbar puncture should be considered in all patients, especially if the diagnosis remains unclear after the above testing – or if a CNS infection is suspected. If meningitis is suspected, empiric antibiotics and antivirals should not be delayed by the lumbar puncture procedure. The risk of herniation in patients with supratentorial mass effect is controversial , but CT should generally still be performed before lumbar puncture. CSF analysis should include opening pressure, cell count, gram stain, glucose, protein, culture, and viral PCR testing. 
Because etiology of unconsciousness is often initially unclear, initial treatment paradigms occur before full evaluation or diagnostics.
Plum and Posner  offer the principles of initial management of comatose patients:
If the above goals can be completed, management can then be tailored based on the results of diagnostic testing. Patients who are persistently comatose despite initial interventions often require a high level of care such as an intensive care unit, if that is consistent with the patient’s established wishes.
The differential diagnosis for the causes of coma is broad. Some reasons are straight forward such as anoxic brain injury, cerebrovascular accidents, seizures, and poisonings but some categories have a wide differential in and of themselves. For example, metabolic causes of coma may include hepatic encephalopathy, uremia, electrolyte abnormalities, and endocrine disorders. Infections, shock, disorders of temperature regulation, respiratory failure, and trauma can also cause coma. Additionally, when examining an unconscious patient, it is crucial to differentiate organic causes of coma from disorders that cause psychiatric unresponsiveness such as catatonia, severe depression, conversion disorder, and malingering.  Psychogenic unresponsiveness is more likely to include active lid closing, reactive pupils, nystagmus, variable motor tone, eupnea or hyperventilation, the absence of pathologic reflexes, and a normal EEG. 
The prognosis of patients who are acutely comatose is variable and highly dependent on the etiology of coma, the severity of brain injury and individual patient factors. The Glasgow Outcome Scale  is used to evaluate outcomes for research purposes.
Traumatic Brain Injury (TBI) has been the most studied etiology of coma, likely due to its high prevalence and highly variable outcomes. Estimated mortality in patients in a coma from TBI ranges from 40 to 50% . Predictors that are linked to outcome include patient age, motor findings, neuro-ophthalmologic signs, secondary injuries, neuroimaging findings, and duration of coma. 
Patients with non-traumatic etiologies of coma that do not recover quickly have an even worse prognosis than the TBI cohort. Patients who have non-traumatic loss of consciousness lasting more than six hours had a one-month mortality rate of 76%, in one study . NTC outcomes are quite variable by etiology. NTC due to depressant overdose, demyelinating disease (ADEM), or auto-immune encephalitis (Bickerstaff Brainstem Encephalitis) carries excellent prognosis with adequate supportive care. Vascular causes such as large CVA or Sub Arachnoid Hemorrhage carry a grave prognosis, as does hypoxic-ischemic injury. Again, neuro-ophthalmologic signs and motor function were consistent predictors of long term outcomes. 
Coma is caused by a temporary or permanent impairment of the bilateral forebrain (thalamus/cerebrum) or reticular activating system in the brainstem. Because of the diverse reasons for a coma, the condition is best managed by an interprofessional team that consists of a neurologist, internist, intensivist, primary care provider, neurosurgeon, and an ophthalmologist. A systematic evaluation of the unconscious patient is recommended. Because many cases of coma are reversible, management of unconscious patients necessitates thorough history-taking, patient evaluation, stabilizing treatment, and diagnostic testing occurring contemporaneously. The prognosis of persistently comatose patients is variable but generally poor. The most consistent prognostic exam findings are neuro-ophthalmologic responses and motor function.  (Level V)
|||Plum F,Posner JB, The diagnosis of stupor and coma. Contemporary neurology series. 1972; [PubMed PMID: 4664014]|
|||Horsting MW,Franken MD,Meulenbelt J,van Klei WA,de Lange DW, The etiology and outcome of non-traumatic coma in critical care: a systematic review. BMC anesthesiology. 2015 Apr 29; [PubMed PMID: 25924678]|
|||Cooksley T,Rose S,Holland M, A systematic approach to the unconscious patient. Clinical medicine (London, England). 2018 Feb; [PubMed PMID: 29436445]|
|||Walker MC,O'Brien MD, Neurological examination of the unconscious patient. Journal of the Royal Society of Medicine. 1999 Jul; [PubMed PMID: 10615273]|
|||Weiss N,Regard L,Vidal C,Luque Y,Taldir G,Vallet H,Diehl JL,Fagon JY,Guerot E, Causes of coma and their evolution in the medical intensive care unit. Journal of neurology. 2012 Jul [PubMed PMID: 22231871]|
|||Huffman JC,Stern TA, Assessment of the Awake but Unresponsive Patient. Primary care companion to the Journal of clinical psychiatry. 2003 Oct [PubMed PMID: 15213789]|
|||Jennett B,Bond M, Assessment of outcome after severe brain damage. Lancet (London, England). 1975 Mar 1; [PubMed PMID: 46957]|
|||Consensus conference. Rehabilitation of persons with traumatic brain injury. NIH Consensus Development Panel on Rehabilitation of Persons With Traumatic Brain Injury. JAMA. 1999 Sep 8; [PubMed PMID: 10485684]|
|||Levy DE,Bates D,Caronna JJ,Cartlidge NE,Knill-Jones RP,Lapinski RH,Singer BH,Shaw DA,Plum F, Prognosis in nontraumatic coma. Annals of internal medicine. 1981 Mar; [PubMed PMID: 7224376]|
|||Nguyen KPL,Pai V,Rashid S,Treece J,Moulton M,Baumrucker SJ, Prognostication in Anoxic Brain Injury. The American journal of hospice & palliative care. 2018 Nov [PubMed PMID: 29614864]|