American Academy of Ophthalmology Web Site:
Original URL:

Morning Rounds

The Case of the Unconscious Car Enthusiast
By Thomas J. Mehelas, MD, and Ray A. Brinker, MD
Edited By Thomas A. Oetting, MD

It was 2 a.m. on a cold February night and Keith Hackett* had not been seen since lunchtime. This wasn’t unusual. The 32-year-old was a lifelong racing fan who spent much of his free time tinkering with his sports car, often past midnight. But when his wife checked on him before going to bed, she found him unconscious and unresponsive on the garage floor, as was his dog. The car engine was off, but she noticed that a nearby space heater was running and the doors and windows were all closed. She called 911.

Six Weeks of Acute Care

The EMT service transported Mr. Hackett to a small community hospital where he was treated with 100 percent oxygen by mask. His oxygen partial pressure (PO2) was 338 mmHg; his blood car- boxyhemoglobin (COHb) was not tested.

Later that morning, he was transported to a large regional hospital. In that emergency department he was intubated, sedated and paralyzed. A spectrophotometer was used to test his COHb, which was 4.4 parts per million (the normal range is from 0.0 to 1.5 ppm) while on a ventilator with a fraction of inspired oxygen (FIO2) of 100 percent. Thereafter, he was treated with hyperbaric oxygen at 2.5 atmospheres (atm) for 2.5 hours per session, with two sessions daily for one week.

However, he remained comatose. He underwent a tracheostomy and gastric tube placement. The G-tube leaked and caused peritonitis that resolved after laparoscopic revision and treatment with broad-spectrum antimicrobial drugs. Mr. Hackett had one grand mal seizure that responded to Depakote (divalproex sodium). An initial CT scan of the brain was normal. An MRI of the brain that included diffusion weighted imaging revealed marked supratentorial and periventricular increased white matter signal that included the corpus callosum (but, surprisingly, spared the basal ganglia) and was consistent with leukomalacia (Fig. 1).

What's Your Diagnosis?
Figure 1: What's Your Diagnosis? An MRI revealed marked supratentorial and periventricular increased white matter signal that included the corpus callosum but, surprisingly, spared the basal ganglia.

After six weeks of acute care he had only “vegetative reflexes” to painful stimuli, occasional spontaneous eyelid opening, normal pupillary exam and no purposeful eye movements. He had no formal eye examination while he was hospitalized.

His wife was devastated when she was told that he had a dismal prognosis. He was sent to a rehabilitation facility with a diagnosis of anoxic encephalopathy.

Rehabilitation Goes Well

In the rehabilitation facility, he did improve. Three months later he was seen by his optometrist who noted best- corrected vision of 20/70 in the right eye and 20/50 in the left eye, absent color plate identification, normal pupillary exam, incomplete, conjugate jerky eye movements, normal slit-lamp and funduscopic examination, normal tonometry and left field loss on confrontation visual fields.

Mr. Hackett was referred for a neuro-ophthalmic examination one month later because of his blurred vision (poorly described), which made it hard for him to watch TV. He was in a wheelchair due to difficulty with balance and leg strength, but was sitting upright and alert. He no longer had a tracheostomy tube. His wife said his short-term memory was poor and, because he was adopted, his family eye history was not known. His best-corrected visual acuity was around 20/30 in both the right eye and left eye with a normal Amsler grid examination in each eye. He identified nine of 15 color plates in each eye. His pupillary, ocular motility, external/ocular adnexa, tonometry, biomicroscopic and funduscopic exams were normal. Automated perimetry (Humphrey central 24-2 threshold testing) revealed a dense left inferior quadrantanopsia (Fig. 2). The exam was unchanged three months later, and thereafter his eye exams were performed by his optometrist.

Morning Rounds Figure 2
Figure 2: He Undergoes Automated Perimetry.  Mr. Hackett was found to have a dense left inferior quadrantanopsia.


Mr. Hackett’s visual field loss was caused by leukoencephalomalacia affecting his cerebral afferent pathway. His anoxic encephalopathy was secondary to carbon monoxide poisoning.

Carbon monoxide is a common cause of lethal poisoning, with the winter being the peak season for accidental death by carbon monoxide in the northern United States. It often results from exposure to gasoline-powered engines or generators in poorly ventilated areas. Intentional exposure to this lethal gas is at least five times more likely to result in death than accidental exposure.

Carbon monoxide is found in tobacco smoke, and a smoker’s blood carboxyhemoglobin may be 10 to 15 percent (vs. 1 to 3 percent in nonsmokers).

Diagnostic tests of carbon monoxide poisoning include an arterial hemoglobin oxygen saturation and carbon monoxide level. When the latter is greater than 30 percent, headache, confusion, fatigue, coma and death can occur. The first line of treatment is to remove the affected patient from the source of carbon monoxide. Next, high-flow oxygen is used. Inspired 100 percent oxygen lowers the elimination half-life from three hours to 90 minutes, and hyperbaric oxygen lowers the elimination half-life to 30 minutes. At 2 atm there is a 14-fold increase in the blood’s dissolved oxygen. Specifically, the beneficial effects of hyperbaric oxygen include diminished carboxyhemoglobin half-life, improved tissue clearance of carbon monoxide, reversed inhibition of cytochrome oxidase, reduced cerebral edema and the prevention of CNS lipid peroxidation.1 Bed rest, sedation and treatment of fever are also important to reduce the amount of hypoxic cellular damage. For patients who have attempted suicide, a psychiatric examination is mandatory.

Brain infarction can result from lowered arterial perfusion or decreased oxygenation. If oxygenation is the problem (when perfusion is maintained), the visual cortex, thalamic nuclei, basal ganglia (particularly the globus pallidus), hippocampus and deep cerebellar nuclei are preferentially affected as they are metabolically very active. Myelin basic protein (MBP) is altered and attacked by lymphocytes.2 Demyelination leading to neuronal necrosis has been reported with carbon monoxide poisoning as has oligodendrocyte apoptosis and brain edema.3 Following exposure, permanent brain damage can occur between four and 21 days.

An MRI scan confirmed Mr. Hackett’s cortical visual loss, but there are reported cases of carbon monoxide poisoning– related encephalopathy in which neuroimaging (CT and MRI) is normal. In such cases SPECT (single proton emission tomography) and PET (position emission tomography) can confirm the encephalopathy.4

In order to make the diagnosis of carbon monoxide poisoning, a high level of suspicion must be maintained. Patients with carbon monoxide poisoning may present with “cherry red” skin (anoxia without cyanosis), headaches, reduced cognition, syncope, seizures, angina, arrhythmias or coma. They can develop a polyneuropathy, tremor, ataxia and encephalopathy. Reported visual sequelae from carbon monoxide poisoning include retinal hemorrhages, ocular dipping5 (a slow, downward cyclic dipping of the eyes followed by rapid upward gaze that, unlike ocular bobbing, may be associated with spontaneous horizontal gaze), optic neuropathy,6 cerebral dyschromatopsia, visual agnosias and cortical blindness.

Prevention Is Key

The most important lesson of this case is to prevent carbon monoxide poisoning. And though it’s uncommon for ophthalmologists to be involved acutely in such near lethal carbon monoxide poisonings, they may play an important later role in documenting visual disability.

* Patient name is fictitious.

1 Dean, B. S. et al. Am J Emerg Med 1993;11:616–618.
2 Hampton, T. JAMA 2004;292:1542.
3 Muratra, T. et al. J Neurol Neurosurg Psychiatry 2001;71:250–253.
4 Moster, M. L. et al. Surv Ophthalmol 1996;40:395–399.
5 Ropper, A. H. Arch Neurol 1981;38:297–299.
6 Simmons, I. G. and P. A. Good. Eye 1998;12:809–814.

Dr. Mehelas is in private practice at Reed Vision in Toledo, Ohio and was formally associate professor of surgery and neurology at the Medical College of Ohio in Toledo where Dr. Brinker, now retired, was professor of radiology.