|
Correction
to
D'Onofrio et al.,
|
| |||||||||||||||||||||||||||||||||||
My first shock (of many) was that the patients were randomly assigned to receive either lorazepam or a saline placebo, even though it was recognized that without treatment a much higher rate of seizures was expected. I was anticipating a dose–response study or a comparison with another antiepileptic medication.
In explaining the need for their study, the authors stated: "Prevention of recurrent seizures related to alcohol use is important." So why was it acceptable to withhold treatment from some patients? Furthermore, two of the studies cited by the authors2,3 document the efficacy of lorazepam for this indication. Indeed, D'Onofrio et al. stated in the Methods section that lorazepam is an accepted treatment. Why was it necessary to perform this study at all?
My next shock came when I read that "consent was initially waived for any patients who were unable to give informed consent at presentation because of intoxication or postictal phenomena." If patients are unable to give true informed consent for entry into an important study, it should be obtained from a relative or guardian; otherwise, such patients should not be enrolled under any circumstances, even for a short period. It is important to prevent the mistreatment of patients who cannot protect themselves.
It is hard for me to believe that any patient who had a seizure would consent to participate in a study in which effective treatment might be withheld if the patient understood that this might happen and also understood that he or she was likely to have more seizures. The patients could not have been aware of the dubious value of this study. Therefore, it seems unlikely that true informed consent was obtained from any of the participants.
If lorazepam or another effective antiepileptic medication had been administered to the patients in the control group instead of saline, approximately 18 patients would not have suffered from recurrent seizures within six hours, 11 would not have needed to be hospitalized, and 6 would not have needed to return to the emergency department.1 This was the cost in human suffering that was needed to demonstrate that lorazepam is superior to saline for the prevention of recurrent alcohol-related seizures. It was a high price to pay for a trivial result.
Mitchel B. Sosis, M.D., Ph.D.
5804 Parade Field Way
Lansdale, PA 19446
References
Since most clinicians would consider a withdrawal seizure a complication of withdrawal and a risk factor for progression to the more serious withdrawal syndrome of delirium tremens (which still carries a considerable mortality rate), how did the authors determine that these patients could or should be released with or without medication rather than be admitted for observation and controlled detoxification?
Robert Matz, M.D.
Mount Sinai Medical Center
New York, NY 10029-6574
To the Editor: We refer Dr. Sosis to our article, which clearly states that at the time of the study, the standard of care at both institutions for patients who presented with a recurrent seizure related to alcohol was similar to that at most urban emergency departments in the United States and included supportive care only.1 Therefore, at no time did we withhold an accepted treatment.
For years, we as well as other researchers have sought to improve the care of patients with recurrent seizures related to alcohol. In the 1980s and early 1990s, we exclusively used the anticonvulsant phenytoin without success.2 At the time of this study, it was not recognized that treatment with lorazepam would reduce the rate of recurrent seizures related to alcohol, since no previous randomized, controlled trial had addressed the issue. The study cited by Dr. Sosis included patients who were admitted to an inpatient rehabilitation center and observed for five days for primary prevention of any signs of alcohol withdrawal.3
We believe that Dr. Sosis is confusing two very separate issues: the care of the patient presenting to an emergency department with a recurrent seizure related to alcohol and the primary prevention of withdrawal symptoms, which include seizures, in a controlled inpatient setting. In the first situation, the patient is more likely to refuse a referral to a detoxification program and resume alcohol consumption on discharge. The risks and benefits of the administration of lorazepam in this setting had not been identified. The second scenario involves the monitoring of patients in an inpatient setting for signs of withdrawal. The evidence in this situation suggests therapy with benzodiazepines.4
In response to Dr. Matz: Patients who had signs and symptoms of moderate to severe alcohol withdrawal during the study period were treated with benzodiazepines, as we described.1 Patients were not given oral medications on discharge from the emergency department unless they were admitted to the hospital or a detoxification unit. The criteria for admission to the hospital included the presence of recurrent seizure, the progression of withdrawal symptoms to moderate or severe intensity, and the presence of coexisting illness requiring admission.
A 48-hour interval for follow-up was chosen because seizures related to alcohol withdrawal typically occur within this period after the cessation of drinking. Delirium tremens is often a later complication of withdrawal. Extended follow-up could have been confounded by repeated cycles of drinking and abstinence.
We agree that seizures are a serious complication of alcohol withdrawal. Therefore, we offered treatment for substance abuse to every patient. Unfortunately, most of the patients refused these services.5
Finally, in the abstract and in the Results section of our article we incorrectly reported the odds ratio for hospital admission in the placebo group as compared with the lorazepam group as 2.1 (95 percent confidence interval, 1.1 to 4.0; P=0.02). The correct odds ratio is 1.71 (95 percent confidence interval, 0.92 to 3.2; P=0.09). We regret the error.
Gail D'Onofrio, M.D.
Yale University School of Medicine
New Haven, CT 06519-1315
Niels K. Rathlev, M.D.
Andrew S. Ulrich, M.D.
Boston University School of Medicine
Boston, MA 02118
References
| |||||||||||||||||||||||||||||||||||
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
