Clinical Question: Are there any clinical signs that are useful in predicting the prognosis of survivors of cardiac arrest who are comatose?
Setting: Various (meta-analysis)
Study Design: Meta-analysis (other)
Synopsis: Information obtained during physical examination that is accurate in predicting the prognosis of survivors of cardiac arrest who are comatose would be useful for physicians and families in making treatment decisions. Two authors independently searched MEDLINE, bibliographies of retrieved articles, and physical examination textbooks for English-language articles that assessed the usefulness of clinical examination in predicting the prognosis of comatose survivors of cardiac arrest. Eleven studies that evaluated the outcomes of 1,914 patients met the authors' inclusion criteria. Disagreement regarding study inclusion and methodologic quality was resolved by consensus.
Five clinical signs were found to be useful in the prediction of death or poor neurologic outcome: (1) absent corneal reflex at 24 hours (positive likelihood ratio [LR+] = 12.9); (2) absent pupillary response at 24 hours (LR+ = 10.2); (3) absent withdrawal response to pain at 24 hours (LR+ = 4.7); (4) no motor response at 24 hours (LR+ = 4.9); and (5) no motor response at 72 hours (LR+ = 9.2). The positive likelihood ratio is a measure of how well a positive finding predicts a poor outcome.
After pooling the data, the pretest probability of poor outcome was 77 percent (i.e., before applying the test, 77 percent of the 1,914 patients were identified as likely to have a poor outcome). Using the test with the highest possible LR (i.e., corneal reflex) increased the odds of identifying a poor outcome to 97 percent. No clinical findings were useful in predicting a good neurologic outcome. Data evaluating the usefulness of more than one test together were scant and unreliable.
Bottom Line: Five clinical examination maneuvers are useful in predicting a poor prognosis in comatose survivors of cardiac arrest: (1) absent corneal reflex at 24 hours; (2) absent pupillary response at 24 hours; (3) absent withdrawal response to pain at 24 hours; (4) no motor response at 24 hours; and (5) no motor response at 72 hours. Prediction of the prognosis within 24 hours should not be made on the clinical examination alone. (Level of Evidence: 2a)