June 2009
Reviewer: Jeff C. Huffman, MD
Pharmacological management of delirium in hospitalized adults: a systematic evidence review
Campbell N, Boustani MA, Ayub A, et al
J Gen Intern Med 2009 Jul; 24(7):848-53; Epub 2009 May 8
J Gen Intern Med 2009 Jul; 24(7):848-53; Epub 2009 May 8
Background: Despite the frequency and significance of delirium in hospitalized patients, the literature on the pharmacological prevention and treatment of delirium is scattered and sparse, and there is no FDA-approved treatment for this condition.Methods: The authors systematically searched several major indexes (e..g., Pubmed) for articles between 1966 and October 2008 that described randomized trials of medication for the prevention/treatment of delirium in adults; they excluded studies involving children/adolescents, patients with axis I psychiatric disorders, and delirium related to alcohol use/withdrawal. Each study was appraised by three independent reviewers and assessed for quality on a 1-5 JADAD scale (5=highest quality).Results: The authors found 4 studies of delirium management and 9 of delirium prevention.
The only study of delirium management of JADAD quality higher than 2 was a study of AIDS patients admitted with acute medical illness; that study found haloperidol to be superior to both chlorpromazine and lorazepam in terms of both delirium severity and rates of adverse events (lorazepam in particular was associated with confusion and sedation). The other studies of delirium treatment were of low quality and compared antipsychotics with each other, finding no differences in outcome measures in all cases.With respect to delirium prevention, two highly rated studies of antipsychotics (haloperidol, risperidone) versus placebo found the antipsychotics to be superior to placebo in terms of either delirium incidence or length/severity (this is also consistent with the APM Research Award-winning presentation by Larsen, et al, regarding olanzapine prevention of delirium in surgical patients). A pair of placebo-controlled studies evaluating donepezil and a study evaluating a precursor of acetylcholine (citicholine) all found that these agents were ineffective in preventing delirium. Single studies of other agents (demedetomidine, nitrous oxide, benzodiazepines) were of mixed quality and results but did not overwhelmingly favor these agents. A single study of gabapentin found that it reduced delirium at higher rates than placebo.Discussion: This interesting review does not tell us anything earth-shattering: it suggests that antipsychotics are the treatments of choice for both the management of active delirium and prevention of delirium, and benzodiazepines as monotherapy appear to be a bad idea for non-alcohol related delirium. Indeed, these are essentially the recommendations of the APA treatment guidelines for delirium management from 1999. The acteylcholinesterase inhibitors don’t seem to be effective in this setting, and no one has found a new agent that appears to be highly effective (aside from the somewhat surprising study of gabapentin). Of interest, the studies of antipsychotics used pretty low doses—for example, no study of haloperidol used a daily dose higher than 6.5 mg/day, much lower than the daily doses that are sometimes used clinically for patients with agitated delirium.More than anything, this review underscores the incredible dearth of data in this area of great clinical importance in psychosomatic medicine (and in medicine in general). The fact that there has been minimal study of delirium treatment, no well-designed study comparing antipsychotics, and that the quality of current data in all of these studies (with small sample sizes and a wide variety of patient populations, dosing, and duration of treatment) essentially lead us back to using our clinical experience in managing our patients with delirium rather than evidence-based medicine. This is, of course, a very difficult population to study for a variety of reasons, but hopefully the coming years will bring greater clarity on the optimal agent, dose, and duration of treatment.At this point, it seems that haloperidol probably remains the treatment of choice for delirium given the many years of clinical experience with this agent and the lack of any evidence to suggest that newer agents are superior, but judicious use of the second generation antipsychotics also seems reasonable. It is of some interest that there are now three small positive studies of delirium prevention using antipsychotics, and the use of prophylactic antipsychotics is an intriguing idea for patients who may be at very high risk for developing delirium (e.g, patients undergoing surgery who have a history of post-op delirium).
The only study of delirium management of JADAD quality higher than 2 was a study of AIDS patients admitted with acute medical illness; that study found haloperidol to be superior to both chlorpromazine and lorazepam in terms of both delirium severity and rates of adverse events (lorazepam in particular was associated with confusion and sedation). The other studies of delirium treatment were of low quality and compared antipsychotics with each other, finding no differences in outcome measures in all cases.With respect to delirium prevention, two highly rated studies of antipsychotics (haloperidol, risperidone) versus placebo found the antipsychotics to be superior to placebo in terms of either delirium incidence or length/severity (this is also consistent with the APM Research Award-winning presentation by Larsen, et al, regarding olanzapine prevention of delirium in surgical patients). A pair of placebo-controlled studies evaluating donepezil and a study evaluating a precursor of acetylcholine (citicholine) all found that these agents were ineffective in preventing delirium. Single studies of other agents (demedetomidine, nitrous oxide, benzodiazepines) were of mixed quality and results but did not overwhelmingly favor these agents. A single study of gabapentin found that it reduced delirium at higher rates than placebo.Discussion: This interesting review does not tell us anything earth-shattering: it suggests that antipsychotics are the treatments of choice for both the management of active delirium and prevention of delirium, and benzodiazepines as monotherapy appear to be a bad idea for non-alcohol related delirium. Indeed, these are essentially the recommendations of the APA treatment guidelines for delirium management from 1999. The acteylcholinesterase inhibitors don’t seem to be effective in this setting, and no one has found a new agent that appears to be highly effective (aside from the somewhat surprising study of gabapentin). Of interest, the studies of antipsychotics used pretty low doses—for example, no study of haloperidol used a daily dose higher than 6.5 mg/day, much lower than the daily doses that are sometimes used clinically for patients with agitated delirium.More than anything, this review underscores the incredible dearth of data in this area of great clinical importance in psychosomatic medicine (and in medicine in general). The fact that there has been minimal study of delirium treatment, no well-designed study comparing antipsychotics, and that the quality of current data in all of these studies (with small sample sizes and a wide variety of patient populations, dosing, and duration of treatment) essentially lead us back to using our clinical experience in managing our patients with delirium rather than evidence-based medicine. This is, of course, a very difficult population to study for a variety of reasons, but hopefully the coming years will bring greater clarity on the optimal agent, dose, and duration of treatment.At this point, it seems that haloperidol probably remains the treatment of choice for delirium given the many years of clinical experience with this agent and the lack of any evidence to suggest that newer agents are superior, but judicious use of the second generation antipsychotics also seems reasonable. It is of some interest that there are now three small positive studies of delirium prevention using antipsychotics, and the use of prophylactic antipsychotics is an intriguing idea for patients who may be at very high risk for developing delirium (e.g, patients undergoing surgery who have a history of post-op delirium).