Benzodiazepines are among the most commonly prescribed drugs in elderly patients (1). They are used for the pharmacological management of symptoms of anxiety and insomnia in the elderly, and depressed elderly patients often receive benzodiazepines instead of antidepressants (2). Prescription of these benzodiazepines for elderly persons raises concerns about long-term use, dependence, cognitive impairment, loss of efficacy, and safety (3). Regular use of benzodiazepines is one of the risk factors in postoperative confusion (4). However, the relationship between postoperative confusion and psychiatric symptoms—such as depression, anxiety, and cognitive dysfunction—in elderly patients treated by benzodiazepines remains unclear. [It is possible that the preoperative psychiatric state and benzodiazepine use are associated with postoperative confusion, because the preoperative depression, anxiety, and cognitive dysfunction are associated with developing postoperative confusion (5).]
Many elderly patients are treated with benzodiazepines for prolonged periods (6). Long-term use of benzodiazepines can accelerate cognitive decline in these patients (7). Benzodiazepine withdrawal symptoms in the elderly include confusion (8). Withdrawal symptoms have been associated with long-term use, and when benzodiazepines are abruptly discontinued, the severity of the withdrawal symptoms depends on the length of time of consumption (9). There is no described relationship between postoperative confusion and duration of benzodiazepine exposure. In this study, we investigated the relationship between postoperative confusion and preoperative anxiety, depressive state, cognitive function, or duration of benzodiazepine exposure in elderly patients treated with benzodiazepines.
The study was approved by the medical ethics committee of our institution. Informed consent was obtained from all patients and their families. We studied 385 patients ranging in age from 65 to 80 yr. Patients with a history of congestive heart disease, transient ischemic attack, stroke, liver cirrhosis, and liver dysfunction (defined by a serum total bilirubin concentration >2.0 mg/dL); mild or heavy alcohol drinkers who consumed ≥25 g of alcohol a day for several years; or those with regular use of antipsychotic, antidepressant, or anticholinergic drugs were excluded. Three hundred twenty-eight patients (85%) participated in this study. The patients were divided into two groups: patients who were preoperatively treated with benzodiazepines for management of anxiety, insomnia, and depression and those who were not. All patients in the study underwent elective orthopedic surgery—total knee arthroplasty for 246 patients and total hip arthroplasty for 82 patients—at the Hakodate Watanabe Hospital and the Hirosaki National Hospital.
Anesthesia was induced with IV propofol 1.5 mg/kg and fentanyl 2 μg/kg, and tracheal intubation was facilitated by vecuronium 0.1 mg/kg IV. Anesthesia was maintained with an infusion of propofol at a rate of 5–6 mg · kg−1 · h−1 until skin closure. Fentanyl was given according to response to vital signs such as systolic blood pressure and heart rate, which were controlled within 20% of preoperative values. Acetated Ringer’s solution was infused at a rate of 5–7 mL · kg−1 · h−1 to all patients. The lungs were mechanically ventilated with 30% oxygen in air to maintain the end-tidal pressure of CO2 at 35–40 mm Hg. No patient received packed red blood cell transfusions. Nasopharyngeal temperature was monitored continuously with an electric thermistor and was maintained at 36.0°C–37.0°C by using a warming blanket and by controlling the temperature in the operating room. After surgery, all patients were treated with nonsteroidal analgesics (diclofenac sodium 50-mg suppository) every 8 h for incisional pain. If they complained of pain, they were treated with pentazocine 15 mg. Nasal oxygen (6 L/min) was administered for 24 h after the end of anesthesia. The patients were monitored with a pulse oximeter. On the first postoperative day, the patients who demonstrated an Spo2 <95% were excluded from this study. Benzodiazepine treatment in benzodiazepine users was continued until the day of operation and was again administered 72 h after cessation of benzodiazepines. All patients were premedicated with 0.1 mg/kg of oral diazepam 60 min before the start of anesthesia.
Postoperative confusion was assessed by using the confusion-assessment method, which is a sensitive and reliable method (10). It is composed of four key features: 1) acute change in mental status with a fluctuating course, 2) inattention, 3) disorganized thinking, and 4) altered level of consciousness. Diagnosis of confusion requires both of the first two features. The patients were examined at least once daily with the confusion-assessment method until the seventh day after operation, and the assessment of confusion was performed by the nursing staff whenever a patient was found to be mentally changed. Pain scores were evaluated by nurses every 24 h after that time. Pain was estimated by using a 100-mm visual analog scale (VAS) (with 0 mm representing no pain and 100 mm representing the worst imaginable pain). The nurses were blinded to the purpose of this study.
Psychometric testing was performed in the afternoon before surgery for investigating whether preoperative depression, anxiety, and cognitive dysfunction are associated with developing postoperative confusion. The Japanese version of the State-Trait Anxiety Inventory (STAI) was used to assess the anxiety state of patients (11). State anxiety (transient anxiety that varies according to the situation) and trait anxiety (stable personality disposition reflecting the general level of fearfulness) were evaluated. A score in state anxiety >42 and a score in trait anxiety >44 are indicative of an anxious state. Cognitive function was assessed by using the Mini-Mental State (MMS) test (12). The MMS examination is a general cognitive screening test that investigates orientation to time and place, calculation ability, level of attention, short-term recall, language ability, and the ability to copy a geometric design. The maximal score is 30 points, and a score <23 is indicative of cognitive impairment. Depression was assessed according to the depression scale by Koenig et al. (13). The maximal score is 11 points, and a score ≥4 is indicative of clinical depression.
Data are expressed as mean ± sd. Comparisons between groups in mean duration of anesthesia and surgery, mean volume of blood loss, and fentanyl and pentazocine consumption were analyzed with Student’s t-test. Differences in STAI, MMS, depression scores, and VAS pain scores between groups were analyzed by repeated-measures analysis of variance followed by Bonferroni’s correction. The incidence of postoperative confusion was analyzed by χ2 testing. P values <0.05 were considered significant.
Among 328 patients, 57 (17%) reported taking benzodiazepines. The number of long-term benzodiazepine users (daily use for more than 1 yr) was 37 (11%) (mean, 4.8 ± 1.1 yr). No patient was taking a dose larger than the maximal dose recommended for elderly patients. The benzodiazepine users were treated by diazepam (1 patient), ethyl loflazepate (15 patients), chlordiazepoxide (2 patients), clotiazepam (1 patient), bromazepam (1 patient), alprazolam (7 patients), and triazolam (30 patients). Ten patients were treated for anxiety, 31 patients were treated for insomnia, 9 patients were treated for anxiety and insomnia, and 7 patients were treated for anxiety and depression.
There were no significant differences in age, average weight, mean duration of anesthesia and surgery, mean volume of blood loss, and total fentanyl and pentazocine consumption between benzodiazepine users and nonusers (Table 1). There were no preoperative hemodynamic differences in systolic and diastolic blood pressure and heart rate between benzodiazepine users and nonusers. There were no significant differences in educational level or the rate of smoking between benzodiazepine users and nonusers. We could not find significant differences in MMS, STAI (state anxiety and trait anxiety), or depression scores between benzodiazepine users and the nonusers (Table 2).
Postoperative confusion during the first 72 h after the end of operation occurred in 15 (26%) of 57 benzodiazepine users and in 34 (13%) of 271 benzodiazepine nonusers. The incidence of postoperative confusion was significantly more frequent (P < 0.01) in benzodiazepine users than in benzodiazepine nonusers. Eight (25%) of 31 patients treated for insomnia and 7 (27%) of 26 patients treated for anxiety and/or depression developed postoperative confusion. The patients who developed a score <23 on the MMS were 5 (9%) of 57 benzodiazepine users and 8 (3%) of 271 benzodiazepine nonusers (P < 0.05). All patients who developed a score <23 on the MMS among benzodiazepine users, and 7 of 8 patients who developed a score <23 on the MMS among nonusers, developed postoperative confusion.
We found no significant differences in age between patients with and without postoperative confusion in benzodiazepine users and nonusers. There was no significant relationship between postoperative pain scores or total pentazocine consumption and postoperative confusion in benzodiazepine users and nonusers. Postoperative confusion 72 h after the end of operation occurred in 1 (2%) of 57 benzodiazepine users and in 1 (0.3%) of 271 benzodiazepine nonusers.
Postoperative confusion occurred in 13 (35%) of 37 long-term benzodiazepine users (daily use for more than 1 yr) and in 2 (10%) of 20 short-term benzodiazepine users (daily use for <1 yr). The incidence of postoperative confusion was significantly more frequent in long-term benzodiazepine users than in short-term users and in nonusers. The patients who developed a score <23 on the MMS were 5 (14%) of 37 long-term benzodiazepine users and 0 (0%) of 20 short-term users. All long-term benzodiazepine users who had a score <23 on the MMS developed postoperative confusion. The incidence of patients with impaired cognitive function was more frequent in the long-term benzodiazepine users than nonusers (P < 0.001). The STAI (state anxiety) in long-term benzodiazepine users was 42.9 ± 5.3, which was significantly more than the 37.0 ± 5.5 in nonusers and 37.3 ± 5.8 in short-term users (Table 3).
This study showed that postoperative confusion was significantly more frequent in patients with regular use of benzodiazepines. Impaired cognitive function was more frequent in the long-term benzodiazepine users. Thus, impaired cognitive function in the long-term benzodiazepine users appeared to result partly in the increased occurrence of postoperative confusion in benzodiazepine users. Fastbom et al. (14) suggested that cognitive function impaired by benzodiazepines was associated with protection by increased γ-aminobutyric acid transmission against glutamate neurotoxicity. According to the recommendations of the United States Food and Drug Administration, benzodiazepine use should not exceed four months (15). However, many elderly patients are treated for prolonged periods (6). The number of French people aged 60 years and older who are long-term benzodiazepine users is estimated at more than 1,000,000 (7). Although we have no information on the number of long-term benzodiazepine users in Japanese people, in this study, 11% of elderly patients were long-term benzodiazepine users. Therefore, for elderly people with cognitive dysfunction, prolonged use of benzodiazepines cannot be considered negligible.
The increased postoperative confusion in benzodiazepine users may result from withdrawal symptoms, which may include anxiety, restlessness, sleep disturbance, nausea, and psychosis (8). Benzodiazepine withdrawal symptoms in the elderly include confusion, differing from what is seen in young patients. In this study, 65% of patients with regular use of benzodiazepines were long-term benzodiazepine users, and the incidence of postoperative confusion in long-term benzodiazepine users was more frequent. More severe withdrawal symptoms occur after discontinuation of long-term use of benzodiazepines (16). Moss and Lanctot (17) demonstrated that benzodiazepine withdrawal plays a role in the occurrence of confusion in older patients. Thus, for elderly patients who have been long-term benzodiazepine users, abrupt discontinuation of benzodiazepines might be a significant problem.
In this study, anxiety scores were significantly higher in long-term benzodiazepine users than in short-term users and nonusers. There was no significant difference in the incidence of postoperative confusion between patients treated for insomnia and patients treated for anxiety and/or depression. The long-term benzodiazepine users appeared to feel more intense preoperative anxiety, independent of the preoperative psychometric condition. It is unclear whether a preoperative anxiolytic state is associated with postoperative confusion. Simpson and Kellett (18) demonstrated no relationship between a preoperative anxious state and postoperative confusion. However, Aakerlund and Rosenberg (19) reported that preoperative anxiety is associated with postoperative confusion.
In this study, no patient demonstrated an Spo2 <95%, and there were no significant differences in age or VAS pain scores between groups. All patients underwent total knee arthroplasty and total hip arthroplasty. In addition, all operations were performed by two surgeons and one anesthesiologist. Therefore, the effects of the surgical procedure on postoperative confusion appear to have been minimal. Thus, the postoperative confusion was not caused by hypoxia, postoperative pain, or surgical procedure in this study. The MMS is a widely used instrument for assessing cognitive function. The sensitivity and specificity were 80% and 98% (20). To assess anxiety, the most commonly used instrument is the Spielberger STAI. The STAI questionnaire is a global test and measures trait and state anxiety. Depression was assessed with the brief depression scale made by Koenig et al. (13), because the questionnaire is simple and easy for elderly patients to use. The sensitivity was 83% and specificity was 77% (13). The brief depression scale is a sensitive and reliable method for assessment of depression. We compared the incidence of postoperative confusion between 37 long-term benzodiazepine users and 20 short-term users. The number of patients was relatively small. Further study may be needed. Although the nurses did not know the results of preoperative depression, anxiety, or cognitive conditions or whether the patients took benzodiazepines, they collected the data of postoperative pain scores and the occurrence of postoperative confusion. Therefore, the data that concern postoperative pain scores and postoperative confusion may not be blinded.
In conclusion, long-term benzodiazepine use, but not short-term use of benzodiazepines, is a risk factor for developing postoperative confusion. The frequent incidence of postoperative confusion is partly associated with cognitive dysfunction.
1. Mort JR, Aparasu RR. Prescribing potentially inappropriate psychotropic medications to the ambulatory elderly. Arch Intern Med 2000;160:2825–31.
2. Heston LL, Garrad J, Makris L, et al. Inadequate treatment of depressed nursing home elderly. J Am Geriatr Soc 1992;40:1117–22.
3. Wang PS, Bohn RL, Glynn RJ, et al. Hazardous benzodiazepine regimens in the elderly: effects of half-life, dosage, and duration on risk of hip fracture. Am J Psychiatry 2001;158:892–8.
4. Marcantonio ER, Juarez G, Goldman L, et al. The relationship of post-operative delirium with psychoactive medications. JAMA 1994;272:1518–22.
5. Aakerlund LP, Rosenberg J. Postoperative delirium: treatment with supplementary oxygen. Br J Anaesth 1994;72:286–90.
6. Kirby M, Deniham A, Bruce I, et al. Benzodiazepine use among the elderly in the community. Int J Geriatr Psychiatry 1999;14:280–4.
7. Paterniti S, Dufouil C, Alperovitch A. Long-term benzodiazepine use and cognitive decline in the elderly: the epidemiology of vascular aging study. J Clin Psychopharmacol 2002;22:285–93.
8. Rickels K, Schweizer E, Case WG, et al. Long-term therapeutic use of benzodiazepines. I. Effects of abrupt discontinuation. Arch Gen Psychiatry 1990;47:899–907.
9. Wolf B, Grohmann R, Biber D, et al. Benzodiazepine abuse and dependence in psychiatric inpatients. Pharmacopsychiatry 1989;22:54–60.
10. Inouye SK, Dyck CHV, Alessi CA, et al. Clarifying confusion: the confusion assessment method. Ann Intern Med 1990;113:941-8.
11. Iwata N, Mishima N, Shimizu T, et al. Positive and negative affect in the factor structure of the State-Trait Anxiety Inventory for Japanese workers. Psychol Rep 1998;82:651–6.
12. Folstein MF, Folstein SE, McHugh PR. "Mini-Mental State": a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98.
13. Koenig HG, Cohen HJ, Blazer DG, et al. A brief depression scale for use in the medically ill. Int J Psychiatry Med 1992;22:183–95.
14. Fastbom J, Forsell Y, Winblad B. Benzodiazepines may have protective effects against Alzheimer disease. Alzheimer Dis Assoc Disord 1998;12:14–7.
15. Food and Drug Administration. Prescribing of minor tranquilizers. FDA Drug Bull 1980;10:2–3.
16. Busto U, Sellers EM, Naranjo CA, et al. Withdrawal reaction after long-term therapeutic use of benzodiazepines. N Engl J Med 1986;315:854–9.
17. Moss JH, Lanctot KL. Iatrogenic benzodiazepine withdrawal delirium in hospitalized older patients. J Am Geriatr Soc 1998;46:1020–2.
18. Simpson CJ, Kellett JM. The relationship between preoperative anxiety and post-operative delirium. J Psychosom Res 1987;4:491–7.
19. Aakerlund LP, Rosenberg J. Postoperative delirium: treatment with supplementary oxygen. Br J Anaesth 1994;72:286–90.
© 2004 International Anesthesia Research Society
20. MacKenzie DM, Copp P, Shaw RJ, et al. Brief cognitive screening of the elderly: a comparison of the Mini-Mental State Examination (MMSE), Abbreviated Mental Test (AMT) and Mental Status Questionnaire (MSQ). Psychol Med 1996;26:427–30.