Early prognosis in anoxic coma: an analysis of the major clinical criteria.

• Author: DeGiorgio, Christopher M.

The Clarence Herbert case(1) has had substantial impact on the care of comatose patients in the state of California and in the nation as a whole, yet the medical details demonstrate the dangers implicit in early coma prognosis. In this article, the major coma prognosis criteria are applied to this and to another illustrative case involving an early prognosis of "hopelessness," but with a very different outcome. The criteria are also analyzed from medical and statistical perspective and recommendations regarding their use in treatment decisions are made.

If withdrawal of nutrition and hydration or other basic levels of care is dependent on prognosis, then prognostication that is less than reasonably certain will result in inappropriate deaths of patients. Because of medicolegal, economic, and other pressures, physicians are making prognoses of irreversibility very early in the course of coma, the most famous case being that of Clarence Herbert. This was the first legal precedent in the state of California for the withdrawal of nutrition and hydration in a comatose patient. It became a landmark case, cited by the California Medical Association,(2) the Los Angeles County Medical and Bar Associations,(3) the American Dietetic Association,(4) the Hastings Center,(5) the State of California Department of Health Services,(6) and various authors for their guidelines and positions regarding the withdrawal of nutrition and hydration in comatose patients.(7) Furthermore, it has been interpreted by many as a precedent for the proxy exercise of a patient's right to forego medical treatment on a constitutional basis in the state of California.(8) The medical aspects of the case, however, which were crucial to the judicial opinion, have not been critically reviewed in the literature.(9)

Such a case should be analyzed thoroughly from a medical perspective if it is to have such impact on the way medicine is and will be practiced. Most importantly, the Herbert case and others like it raise critical questions about the validity of early coma prognosis. Are early prognostic signs reliable enough to be useful in the management of comatose patients? Which, if any, prognostic criteria are sufficiently reliable? Can life support be justifiably withdrawn based on currently available early prognostic criteria? What is the minimum duration of coma necessary before a diagnosis of irreversibility can be reliably made? What degree of certainty (or false positive error) is acceptable for determinations of prognosis? Lastly, do criteria for irreversibility have the requisite degree of certainty to fulfill the ethical standards of the American Medical Association?(10)

The case history of Clarence Herbert is presented below, with emphasis on the neurologic signs documented in the medical record. Another patient, DV, who was also given a poor prognosis early in coma but whose outcome was markedly different, is presented to illustrate the unreliability of early prognosis in individual cases. The medical records of both Clarence Herbert and DV were reviewed by both authors. Because Clarence Herbert's case is a matter of public record and is widely known, his real name is used. For purposes of confidentiality, the other patient, who is alive, is referred to by her initials.

The purpose is to analyze currently available coma prognostic criteria using Clarence Herbert and DV as models for medical decisionmaking. The authors assume the best of intentions on the part of the patients' physicians.

Case Report: Clarence Herbert

Clarence Herbert was a fifty-five year old white male, admitted to the hospital on August 24, 1981, for a routine ileostomy closure. On August 26 surgery was performed without incident. The patient was transferred to the recovery room where he was successfully extubated and was noted to respond to voice. At approximately 1:18 p.m. he was found cyanotic and bradycardic; an airway was inserted and ventilation initiated. He was soon noted to be pulseless, and cardiopulmonary resuscitation (CPR) was begun. Atropine, epinephrine, sodium bicarbonate, and lidocaine were administered sometime between 1:16 and 1:25. At 1:25 systolic blood pressure was 130 mm Hg. The chart is inconsistent concerning the timing of these events, and the exact duration of hypoxia, hypotension, bradycardia, and asystole cannot be determined. The resuscitation was relatively easy, as the surgeon charted later that day: "[P]atient had cardio respiratory arrest in recovery room. Rapidly intubated and heart massage started. Rhythm returned promptly."

After transfer to the intensive care unit (ICU), he was noted by the nursing staff to be unresponsive to verbal or painful stimuli. Pupils were 7 mm and unreactive. The respiratory flow sheet indicated that from twenty minutes after the arrest until early the next morning, the patient had some spontaneous respirations, in that his respiratory rate was intermittently greater than the ventilator setting. Two hours post-arrest, the pupils were minimally reactive to light. Jerking of the extremities upon suctioning was noted at 4:00 p.m. and increased thereafter. At 7:00 p.m. the cardiologist in charge of the ICU documented intact oculocephalic reflexes, superior deviation of the eyes, and no response to pain. Intravenous diazepam was administered for myoclonus at a rate of 10 mg per hour for a total of 90 mg by 5:00 a.m. on August 27. During the early morning, the nursing staff noted repeatedly that the pupils had become more reactive to light. Also, consistent with the respiratory therapist's observations, the nurse noted around midnight that the patient assisted the ventilator on occasion; however, after 5:00 a.m., both the respiratory flow sheet and the nursing notes document absence of further spontaneous respirations.

At 12:00 noon on August 27, twenty-three hours post-arrest, the cardiologist in charge of the ICU spoke to the family about Clarence Herbert's prognosis. The chart does not indicate what was communicated. At 12:20, he charted no change in status, no spontaneous respirations, and the fact that the patient responded to painful stimuli with increased myoclonus. The cardiologist also noted constricted pupils, disconjugate gaze, and absent doll's eye reflex. He wrote: "Question brainstem infarct anteceding respiratory arrest. Plan: Supportive treatment--wife expresses wish of no heroics. Neuro consult this p.m."

At 4:30 p.m. on August 27, one day post-arrest, the first formal neurologic examination was charted by the neurologic consultant. Mental status was described as "comatose, unresponsive to verbal or noxious stimuli." Generalized, fine myoclonic activity was noted, which increased with stimulation. Sternal compression resulted in transient eye opening. Regular, rapid, and spontaneous respirations were noted after removal of the ventilator for one minute. Pupils were 4 mm and reactive to light. Corneal and oculocephalic reflexes were intact. Gag reflex was intact. No purposeful movements were noted. Upon stimulation, tone was normal, and deep tendon reflexes were normally reactive in the upper extremities and hyperactive in the lower extremities. Slight withdrawal to plantar stimulation was noted, with flexor plantar responses. The neurologist's assessment was "severe diffuse anoxic cerebral damage with myoclonic activity," without evidence of a brainstem stroke. Recommendations were made to discontinue diazepam, continue nonheroic supportive care, and obtain an electroencephalogram (EEG).

On August 28, at 7:30 a.m., less than two days post-arrest, the cardiologist noted some respiratory effort when the patient was taken off the ventilator for a brief time. He charted: "Will discuss termination of respiratory support with wife." The following order was written at the same time: "Supportive care: Do not treat cardiac arrhythmias; do not treat hyper- or hypo-tension." One hour later the internist charted: "Spoke with wife. She wishes patient not to be kept on the respirator. Also consents to an autopsy." At 8:45 a.m. a nurse initialed the following untimed order from the internist: "Remove patient from respirator after family has arrived." Forty-five minutes later, a telephone order was issued by the internist to cancel the EEG scheduled the previous day by the neurologist. Around two hours later, the internist rescinded the cancellation, noting that he had spoken with the neurologist, who wanted the EEG and more time to evaluate the patient.

An EEG was performed around forty-eight hours post-arrest and was interpreted as showing "generalized, polymorphic delta, predominately in the 1-2 Hz range, with some activity into the 3-4 Hz range, mostly of low voltage, with some medium voltage activity, especially in the referential montages." The presence or absence of reactivity was not mentioned in the EEG report. The same afternoon, two days post-arrest, the neurologist noted that the patient no longer had myoclonus, but was comatose with decerebrate posturing, which increased slightly with noxious stimuli. Pupils were 5 mm and reactive, eyes were deviated upwards, and oculocephalic reflexes were intact. The neurologist's assessment was: "Severe anoxic cerebral injury. A poor prognosis is suggested by failure to improve in 48 hours. Recommendations--non-heroic supportive care." The neurologist related his prognosis to the family.

On August 29, three days post-arrest, the nurses continued to note brisk pupillary reactivity; however, the internist charted that the patient had "dilated and fixed pupils" and "no spontaneous respirations." He wrote: "Have again explained the patient's hopeless condition to the family, and they insist we take Mr. Herbert off the respirator." At 9:15 a.m. the same day, the patient was rapidly removed from the ventilator. After a period of shallow respirations, his breathing stabilized.

On August 31, five days post-arrest, the neurologist's examination revealed that...