Current issues of ACP Journal Club are published in Annals of Internal Medicine


Meperidine and benzodiazepines were associated with postoperative delirium

ACP J Club. 1995 May-June;122:80. doi:10.7326/ACPJC-1995-122-3-080

Source Citation

Marcantonio ER, Juarez G, Goldman L, et al. The relationship of postoperative delirium with psychoactive medications. JAMA. 1994 Nov 16; 272:1518-22.



To determine whether an association exists between medications with known psychoactive properties and the development of delirium after surgery.


Nested case-control study within a cohort analytic study.


General, orthopedic, and gynecologic surgery services at Brigham and Women's Hospital, Boston, Massachusetts.


91 case patients with delirium after surgery and 154 control patients (mean age 73 y, 50% women) were identified from among 1341 patients who were admitted to the hospital for major noncardiac procedures and who were enrolled in the cohort study. The case patients developed delirium during days 2 through 5 after surgery. Each case patient was matched to up to 2 control patients who had the same risk for delirium before surgery (based on 7 risk factors identified and validated in the cohort study: age; poor cognitive function; Specific Activity Scale class; self-reported alcohol abuse; markedly abnormal serum sodium, potassium, or glucose levels before surgery; aortic aneurysm surgery; and noncardiac thoracic surgery) and who were still hospitalized on the day after surgery when the case patient became delirious.

Assessment of risk factors

Data on medications were obtained from the nurses' medication administration records by a reviewer blinded to the hypothesis of the study. For case patients, exposures to narcotics, benzodiazepines, and anticholinergics were recorded for the 24 hours before delirium developed. For the control patients, medication exposures were recorded for the same 24-hour period after surgery as that of the matched case patient.

Main outcome measure

Beginning on the second day after surgery, delirium was diagnosed using either the Confusion Assessment Method criteria or documentation of altered mental status in both the medical record and the nursing intensity index.

Main results

Among individual narcotic agents, only meperidine was associated with delirium (odds ratio [OR] 2.7, 95% CI 1.3 to 5.5). Epidural analgesia was associated with delirium (OR 2.3, CI 1.2 to 4.4) whereas patient-controlled intravenous administration was not (OR 1.1, CI 0.5 to 2.2). Benzodiazepines were associated with delirium (OR 3.0, CI 1.3 to 6.8); long-acting benzodiazepines had a stronger association than short-acting agents (OR 5.4, CI 1.0 to 29.2 vs OR 2.6, CI 1.1 to 6.5), and high-dose benzodiazepines had a stronger association than low-dose benzodiazepines (OR 3.3, CI 1.0 to 11.0 vs OR 2.6, CI 0.8 to 9.1). Anticholinergic agents were not associated with delirium (OR 1.5, CI 0.6 to 3.4).


Exposure to meperidine, epidural analgesia, and benzodiazepines after surgery was associated with delirium.

Source of funding: In part, Agency for Health Care Policy and Research.

For article reprint: Dr. T.H. Lee, Section for Clinical Epidemiology, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA. FAX 617-278-1048.


By matching for the risk for delirium before surgery and by systematically ascertaining actual medication use, this carefully designed study avoids many of the pitfalls of previous investigations. Unfortunately, many of the drugs of interest were taken by only a few patients, limiting the statistical power of the study and reducing the likelihood of finding a significant association between delirium and individual narcotics, benzodiazepine subgroups, and anticholinergics.

That no narcotic other than meperidine was associated with delirium could have been due to the difference in the amount of drug given (equianalgesic doses of narcotics vary considerably [1]), differences in the route of administration, or a unique property of meperidine. It is known that in the presence of renal insufficiency or after high, recurrent doses, normeperidine can accumulate and result in neurotoxicity (2). The Acute Pain Management Guideline Panel (1) has suggested that meperidine be used only for brief courses of treatment in otherwise healthy patients who cannot tolerate other narcotics.

The possibility of confusing the causal direction is a concern with both benzodiazepines and meperidine. For example, the association between benzodiazepine use and delirium could be caused by the effect of the benzodiazepine or by prescription of the drug for sleeplessness and agitation, 2 early signs of delirium. The study, however, offers yet another reason to limit the use of benzodiazepines. For meperidine, it is unclear whether delirium resulted from the pain for which meperidine was prescribed (3), or from the meperidine itself.

Epidural analgesia was associated with delirium but, as the authors state, this association may have been confounded by the predominant use of meperidine for epidural analgesia.

Rosanne M. Leipzig, MD, PhD
New York Medical CollegeNew York, New York, USA


1. Acute Pain Management Guideline Panel. Acute pain management: operative or medical procedures and trauma. Clinical Practice Guideline. AHCPR; Feb 1992.

2. Kaiko RF, Foley KM, Grabinski PY, et al. Central nervous system excitatory effects of meperidine in cancer patients. Ann Neurol. 1983;13:180-5.

3. Lynch EP, Lazor MA, Gellils JE, et al. The impact of postoperative pain on the development of postoperative delirium. Anesth Analg. 1998;86:781-5.