Pharmacologic Calvinism: Why Drugs Should Be Used for Indications, Not Side Effects

Abstract & Commentary

Synopsis: Diphenhydramine was associated with significantly increased risk of inattention, disorganized speech, altered consciousness, urinary retention, and increased length of stay in hospitalized patients older than 70 years.

Source: Agostini JV, et al. Arch Intern Med. 2001;161:2091-2097.

Agostini and colleagues, who were funded by the National Institute on Aging, undertook a prospective study of adverse effects of diphenhydramine (Benadryl) in hospitalized patients older than the age of 70 years. They hypothesized that this anticholinergic medication would increase symptoms of delirium in this vulnerable group. Four hundred twenty-six patients were prospectively enrolled. Of the 426 patients, 114 (27%) of them received diphenhydramine. This medication was prescribed as a hypnotic 68% of the time. For 21% of the time, it was administered for transfusion prophylaxis, and 3% of the time, it was given for allergies or for pruritis. The 114 patients who received diphenhydramine and the 312 who did not were similar in all the important variables, including age, Mini Mental Status Exam Score (MMSE), baseline sleeping difficulty, number or medications prior to admission, gender, and impairment in activities of daily living. Those who got diphenhydramine had statistically significant increases in delirium symptoms (relative risk [RR], 1.7; confidence interval [CI], 1.3-2.3), MMSE decline > 3 points (RR, 1.8; CI, 1.0-3.2), altered consciousness (RR, 3.2; CI, 1.6-6.1), abnormal psychomotor activity (RR, 2.3; CI, 1.2-3.3), behavioral disturbance (5.6, 1.0-29.9), and inattention (RR, 3.0; CI, 1.5-5.9) compared with baseline. Cognitive assessments were carried out by investigators blinded both to the study hypothesis and to the patients’ diphenhydramine use. Patients who received diphenhydramine also had increased length of stay (7 vs 6 days) and a greater likelihood of urinary catheter placement (RR, 2.5; CI, 1-6). The risk of adverse effects increased in a dose-dependent manner.

Agostini et al also found that about a quarter of diphenhydramine doses were administered inappropriately (eg, to those with known urinary retention or with no known history of transfusion reaction).

Comment by Barbara A. Phillips, MD, MSPH

A large majority of patients who received diphenhydramine in this study got it as a sleeping aid, prescribed while they were in the hospital. The issue of sleep disturbance and sleeping pills in the elderly is like the elephant in the room that nobody wants to talk about. In this study and others,1 about half of the older patients had baseline sleeping difficulty. Sleep, like most everything else, deteriorates with aging. The response of many clinicians is what Dr. Wallace Mendelson has referred to as "Pharmacologic Calvinism."2 Instead of treating sleeping difficulty with any of the available effective hypnotics, many clinicians use antihistamines, antipsychotics, or antidepressants for their side effect of somnolence. I think there are 2 problems with this: 1) antihistamines, antipsychotics, and antidepressants are not as effective in inducing sleep as are hypnotics (especially the newer agents); and 2) in general, antihistamines, antipsychotics, and antidepressant drugs have more side effects than the newer hypnotics.

In the standard text of sleep medicine, Roehrs and Roth say, "Although studies have shown that H 1 antihistamines do increase sleepiness in healthy normal individuals, no studies have clearly established the dose range over which hypnotic effects in people with insomnia might be found. Low-dose antidepressants have also been used as hypnotics. It is the sedating side effect of the drug that is being sought. However, the antidepressants have cardiotoxic side effects and anticholinergic side effects that make this class of drugs a poor choice as a hypnotic in the absence of clinical depression."3

Walsh and Schweitzer recently reported that: 1) pharmacologic treatment of insomnia fell dramatically from 1987 to 1996; and 2) the use of antidepressants to treat insomnia has grown substantially.4 Trazodone is a "prime offender" in this category. They present evidence that the use of antidepressants in insomnia has grown because of concern about dependence rather than because of recognition and treatment of depression in those reporting insomnia. Trazodone is associated with significant side effects, including daytime somnolence, orthostatic dizziness/hypotension, and priapism.5,6 Further, trazodone appears to improve sleep in the non-depressed patient only in the short term and is less effective than newer hypnotic agents, such as zolpidem.7

We used to undertreat pain, and now make aggressive attempts to identify and relieve it appropriately. I suspect that in the not too distant future, we will look back on our currently poor job of addressing very treatable sleep disturbances in the same way.


1. Phillips BA, Ancoli-Israel S. Sleep Medicine. 2001;2: 99-114.

2. Wallace Mendelson, MD, personal communication, many occasions.

3. Roehrs T, Roth T. Hypnotics: Efficacy and adverse effects. In: Kryger MH, Roth T, Dement WC, eds. Principals and Practice of Sleep Medicine. 3rd ed. Philadelphia, Pa: WB Saunders; 2000:417.

4. Walsh JK, Schweitzer PK. Sleep. 1999;22:371-375.

5. Nierenberg AA, Peck PE. J Clin Psychopharmacol. 1989;9:42-45.

6. Thompson JW, et al. J Clin Psychiatry. 1990;51:430-443.

7. Walsh JK, et al. Human Psychopharmacology. 1998; 3:191-198.

Dr. Phillips, Professor of Medicine, University of Kentucky; Director, Sleep Disorders Center, Samaritan Hospital, Lexington, KY, is Associate Editor of Internal Medicine Alert.

Readers are Invited

Readers are invited to submit questions or comments on material seen in or relevant to Internal Medicine Alert. Send your questions to: Neill Larmore—Reader Questions, Internal Medicine Alert, c/o American Health Consultants, P.O. Box 740059, Atlanta, GA 30374. For subscription information, you can reach the editors and customer service personnel for Internal Medicine Alert via the Internet by sending e-mail to We look forward to hearing from you.