Reviewer: Jeff C. Huffman, MD
Administration of olanzapine to prevent postoperative delirium in elderly joint-replacement patients: a randomized controlled trial
Larsen KL, Kelly SE, Stern TA, et al
Psychosomatics 2010; 51(5):409-418
Background: Postoperative delirium is common, costly, and associated with substantial negative medical outcomes. However, there has been minimal prospective study of delirium prevention among patients undergoing surgery.
Methods: The authors performed a double-blind placebo-controlled randomized trial of perioperative olanzapine for patients undergoing joint-replacement surgery (knee/hip) at an urban orthopedic hospital. Olanzapine was selected because antipsychotics are the treatment of choice for delirium and because olanzapine is available in an easily dissolved form that can be given pre- and post-procedure. Eligible subjects included all patients 65 and older, along with any patient with a history of postoperative delirium, who were scheduled for joint-replacement surgery at the hospital; patients with dementia, significant alcohol use, and current use of antipsychotics were excluded.Enrolled patients received 5mg olanzapine or placebo immediately prior to surgery and 5mg olanzapine or placebo shortly post-procedure. Patients were evaluated for study outcome measures (MMSE, Delirium Rating Scale-Revised-98 [DRS-R-98], Confusion Assessment Method [CAM]) by research assistants/nurses on a daily basis for 8 days or until discharge, whichever came first. A psychologist investigator reviewed the data on a daily basis to make a determination about whether patients met DSM-III-R criteria for delirium (this was the primary outcome measure). Study staff also recorded time to delirium, perceived pain, narcotic usage, length of admission, and disposition, among other clinical metrics, as secondary outcome measures.
Results: In total, 495 subjects were enrolled; not all potential subjects could be screened by study staff due to manpower/volume issues. All but 3 enrollees were 65 or older. Of these 495 subjects, 95 dropped out prior to receiving medication or undergoing surgery, leaving 400 completers. On the primary outcome measures, the incidence of delirium was significantly lower in the olanzapine group (14.3% vs. 40.2%; p<0.001), and time to delirium significantly longer in the olanzapine group; delirium did last longer in patients receiving olanzapine (2.2 days vs. 1.6 days; p=0.02), and was more severe (mean max score 16.44 on DRS-R-98 vs. 14.5; p=0.02). Significantly fewer subjects required post-admission rehab stays (59% vs. 70%; p=0.02) in the olanzapine group; other secondary outcomes were not different between groups. On multivariable logistic regression that included multiple demographic and medical factors associated with delirium in prior studies, olanzapine treatment remained independently associated with prevention of delirium.
Commentary: This is an imperfect but important study in the field of psychosomatic medicine. The study was limited by an inability to recruit subjects consecutively (essentially a convenience sample), a high dropout prior to surgery, and lack of preoperative cognitive assessments; assessment of the impact of olanzapine on delirium duration and severity may also been confounded by the fact that patients did not receive standardized delirium treatment. Furthermore, this study was essentially only in older adults and in a single hospital. Nearly all patients were > 65 and those < 65 had history of preoperative delirium.All that said, this was a well-designed study that enrolled a large number of subjects, followed them daily in a structured manner, did account for complexity of surgery/anesthesia class and other delirium risk factors in its multivariable assessments, and utilized an approach that could be replicated in clinical practice. The findings are impressive—olanzapine was not associated with adverse effects/events, and was linked with a substantial reduction in risk of delirium in this high-risk population. These findings also occur in the context of a few other studies of antipsychotic prophylaxis of delirium that similarly found antipsychotics to reduce incidence or duration of delirium.Further studies to replicate this result and to assess longer-term follow-up will be beneficial; we are getting closer to a time when periprocedure use of antipsychotics may become part of routine practice for patients undergoing joint-replacement and other surgeries, especially if future studies find reductions of cost, length of stay, readmission rates, or mortality. It may be useful to have such reports enter the surgical literature as well, since these are the physicians who will be managing these patients.