The Long-Term Prescription of Benzodiazepines, Psychotropic Agents, to the Elderly at a University Hospital in Japan

Tohoku J. Exp. Med., 2007,The 212Long-Term, 239-246 Benzodiazepine Prescription in the Elderly 239

KYOKO NOMURA, MUTSUHIRO NAKAO, MIKIYA SATO and EIJI YANO

Department of Hygiene and Public Health, School of Medicine, Teikyo University, Tokyo, Japan

NOMURA, K., NAKAO, M., SATO, M. and YANO, E. The Long-Term Prescription of Benzodiazepines, Psychotropic Agents, to the Elderly at a University Hospital in Japan. Tohoku J. Exp. Med., 2007, 212 (3), 239-246 ── Benzodiazepines are useful and effective psychotropic agents used worldwide. However, the long term use of the drugs can lead to serious adverse health effects such as psychomotor and cognitive impairment, especially in the elderly. In Japan, there are very few reports concerning long-term use of benzodiazepines, and no countermeasures have been instituted. Thus, this study assessed the characteristics of long-term prescription of benzodiazepines at a university hospital in Japan. A cross-sectional study using the database of a computer ordering system examined 4,239 adult outpatients who were prescribed benzodiazepines at a university teaching hospital between July 2002 and June 2003. The patients were divided into two groups: those with long-term (≧ 3 months) and short-term (≦ 2 months) prescriptions. A logistic regression model was used to analyze the effect of patient age on long-term benzodiazepine prescription. Adjusting for patient gender, pharmacological half-life of the drug, and department group, a logistic regression model showed that long-term benzodiazepine prescription occurred more frequently in older patients (p < 0.0001 in trend tests) and varied according to the physician’s specialty (p < 0.0001). Benzodiazepines were more frequently prescribed for long term in the elderly by internal medicine group (p = 0.003). Of the patients older than 71 years (n = 1,105), 86% were assigned to the long-term group and were more likely to have been prescribed benzodiazepines by an internist than a surgeon (p < 0.0001). The appropriate prescription of benzodiazepines in the elderly should be included in the educational programs at teaching hospitals, and rational prescribing needs to be monitored carefully. ──── benzodiazepine; computer ordering system; elderly; long-term prescription; university hospital © 2007 Tohoku University Medical Press

Benzodiazepines are useful and effective are less toxic than barbiturates when they are used psychotropic agents used worldwide. Their thera- in overdose and have thus become one of the most peutic actions as anxiolytics, sedative hypnotics, prescribed classes of drugs. A literature (Woods anticonvulsants, and muscle relaxants have led to et al. 1992) reported that about 2% of the adult their use as first-line treatments. In addition, they populations in the United States (4 million peo- Received March 7, 2007; revision accepted for publication May 12, 2007. Correspondence: Kyoko Nomura, M.D., M.P.H., Ph.D., Department of Hygiene and Public Health, School of Medicine, Teikyo University, 2-11-1, Kaga, Itabashi, Tokyo 173-8605, Japan. e-mail: [email protected] 239 240 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 241 ple) and United Kingdom (1 million people) were is unclear if previous risk profiles are being con- prescribed benzodiazepines as hypnotics or tran- sidered in light of the long-term prescription of quilizers. In Japan, benzodiazepines are quite these drugs to this age group. Therefore, we popular in clinical practice, although a population- investigated long-term benzodiazepine prescrip- based survey needs to be conducted to determine tions especially to the elderly at a university the prevalence of benzodiazepine use in the gen- teaching hospital in Japan. eral population (Yamawaki 1999). In 1999, the Teikyo University School of Despite their usefulness, benzodiazepines Medicine installed a computer ordering system have serious adverse effects, including sedation, (COS), which has been used to monitor drug pre- psychomotor and cognitive impairment, and scriptions in every clinical department at the hos- memory loss, and they induce withdrawal reac- pital. Using data available through the COS, the tions and physical dependence (Nelson and university has launched several research projects Chouinard 1999). Hence, it has been recom- to study evidence-based health care services at the mended that their use be limited to short interven- hospital (Nakao et al. 2007). This study was part tions (Committee on Safety of Medicine 1988; of a project to assist physicians in developing Royal College of Psychiatrists 1988; Fick et al. more rational protocols for appropriate drug pre- 2003; Ashton 2005). Nevertheless, many epide- scription. miological studies (Magrini et al. 1996; Simon et al. 1996; Egan et al. 2000a, b; Gray et al. 2003; MATERIALS AND METHODS Fialova et al. 2005; Voyer et al. 2006) in Europe, Data set the United States, and Canada, have reported that We created a new data set of patient information benzodiazepines are prescribed for the long term, using Access 2000 (Microsoft Japan, Tokyo) based on especially to the elderly, women with psychologi- the original COS database accumulated at Teikyo cal problems and chronic physical diseases, and University Itabashi Hospital, Tokyo, Japan (1,154 hospi- users of long-acting type benzodiazepines. Of tal beds). This university teaching hospital has 19 these three types of users, the elderly constitute department clinics, with approximately 60,000 outpa- the most vulnerable population because the long- tients visiting annually. The variables retrieved from the term use of benzodiazepines increases the risks of COS database were patient gender and age, prescribing falls and fractures (Pierfitte et al. 2001; Wagner et department, benzodiazepine pharmacological half-life, al. 2004), and cognitive and memory changes and the date of benzodiazepine prescription. (Paterniti et al. 2002; Barker et al. 2004) that could dramatically decrease their quality of life. Patients Therefore, although the high prevalence of use Fig. 1 is a flow chart of our study subjects. The among the elderly has been reported repeatedly, it subjects consisted of 5,959 adult outpatients who had

Fig. 1. A flow chart of the study subjects. 240 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 241

visited the hospital and were prescribed benzodiazepines room, plastic surgery, and cardiovascular surgery. The by physicians between July 2002 and June 2003. After internal medicine group consisted of the general internal excluding 1,720 patients who were prescribed benzodi- medicine and neurology departments. The psychiatry azepines at two or more different clinics, 4,239 patients group was composed of the psychiatry and mind/body (59% female, mean age 57 years) receiving benzodiaze- medicine departments, and the “other” group included pines were selected for the analyses. The patients were the pediatrics, dermatology, radiology, anesthesiology, divided into two groups depending on the prescription and rehabilitation departments and the hemodialysis length based on previous studies (WHO Collaborating center. Centre for Drugs 1996; Zandstra et al. 2002) ; the long- Although the diagnostic information was prelimi- term group consisted of those with prescription periods nary during the study period (the annual report of the of 3 months or more (mean ± S.D.: 8.0 ± 2.7 months) and hospital, unpublished), the major ICD-10 diagnoses in the short-term group had prescription periods of 2 internal medicine were angina pectoris, followed by months or less (1.4 ± 0.5 months). In Japan, according to chronic renal failure, primary lung cancer, old myocardi- the Health Insurance Law, physicians can write a pre- al infarction, and hepatocellular carcinoma, while those scription for hypnotics for up to 14 days; hence, patients in surgery were inguinal hernia, followed by gastric can- who take the drug continuously have to see a physician cer, breast cancer, pancreatic cancer, and hepatocellular at least twice a month. The data set used in this study carcinoma. The total number of new outpatients in the contained the date of prescription, and we defined the psychiatry group was 1,549, of which 19.5% had depres- prescription time period based on this date. sive disorders. During the study period, the numbers of physicians employed in the aforementioned groups were Benzodiazepines 190 in the internal medicine group, 354 in the surgery Referring to the Anatomical Therapeutic Chemical group, 60 in the psychiatry group and 84 in the other Classification System (WHO Collaborating Centre for department group. Drugs 1996) index groups, including N05BA (N = ner- vous system, N05B = anxiolytics, 01 = diazepam, 02 = Statistical analysis medazepam, 04 = oxazolam, 06 = lorazepam, 08 = bro- Chi-square tests were used to compare patient mazepam, 12 = alprazolam, 18 = ethyl loflazepate, 19 = gender, department group, and pharmacological half-life etizolam, 21 = clotiazepam, 22 = cloxazolam, 23 = tofi- between the long- and short-term groups, and the t-test sopam), N05CD (N05C = hypnotics and sedatives, 01 = was applied to compare age between the two groups. To flurazepam, 02 = nitrazepam, 03 = flunitrazepam, 05 = clarify the patient characteristics in each age category, triazolam, 06 = lormetazepam, 09 = brotiazepam, 10 = patient age was transformed into four groups (15–40, quazepam), and N5CM (others, rilmazafone hydrochlo- 41–59, 60–70, and 71 years or older) relying on the dis- ride). The following benzodiazepines were categorized tribution quartiles, and a logistic regression model was by pharmacological half-life. The short half-life (< 10 applied to assess whether the categorized age was associ- hrs) group comprised brotiazepam, clotiazepam, etizolam, ated with long-term benzodiazepine prescription after lormetazepam, tofisopam, and triazolam; the intermedi- adjusting for patient gender, department group, and phar- ate half-life (10–30 hrs) group included alprazolam, bro- macological half-life (Model I). In the same logistic mazepam, estazolam, flunitrazepam, lorazepam, nitraze- regression model, the statistical interaction effect of the pam, oxazolam, and rilmazafone hydrochloride; and the oldest age group with the department group was also long half-life (> 30 hrs) group consisted of cloxazolam, assessed. In order to evaluate the independent effects of diazepam, ethyl loflazepate, flurazepam, medazepam, patient gender, department group, and pharmacological and quazepam. half-life on long-term prescription in the elderly, the same logistic model was also calculated using these Department clinics stratified age groups (Model II). The department clinics were categorized into four Univariate and adjusted odds ratios (OR) for long- groups: surgery, internal medicine, psychiatry, and term benzodiazepine prescriptions were computed along “other”. The surgery group included the departments of with the 95% confidence intervals (CI). The Cochrane– general surgery, orthopedic surgery, urology, ophthalmol- Armitage trend test was applied to assess the age group ogy, otorhinolaryngology, neurosurgery, the emergency trends, and the Wald chi-square test was used to deter- 242 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 243 mine the significance of the department group. P < 0.05 proportion of long-term prescriptions also was considered significant, and SAS ver. 8.12 for increased. The relationship between long-term Windows (SAS Institute, Cary, NC, USA) was used for benzodiazepine prescription and age remained the analyses. significant after controlling for patient gender, department group, and pharmacological half-life RESULTS (p < 0.0001 in trend tests). The department group Table 1 shows the differences in patient age was also significantly associated with long-term and gender, department group, and pharmacologi- benzodiazepine prescription in the multivariate cal half-life between the long- and short-term pre- model (data not included in Table 2). scription groups. The mean age was 9 years older Benzodiazepines were more frequently prescribed in the long-term group than in the short-term by internal medicine group (adjusted OR, 1.93; group. The patients given prescriptions in the 95% CI, 1.58–2.36) and psychiatry group (adjust- internal medicine or psychiatry department groups ed OR, 2.14; 95% CI, 1.67–2.73) than by surgery were more likely to be assigned to the long-term group. Gender and pharmacological half-life prescription group than those given prescriptions were not significantly associated with long-term in the surgery and “other” department groups. prescription. Statistical interactions between the Table 2 shows the results of Model I, esti- elderly (≧ 71 years old) and long-term prescrip- mating the OR for long-term benzodiazepine pre- tion were observed with internal medicine (p = scription in association with the age categories. 0.003); benzodiazepines were more frequently Patient age was significantly associated with long- prescribed for long term in the elderly by internal term prescriptions; as patient age increased, the medicine group.

TABLE 1. Patient age and gender, and clinical departments between long- and short-term prescription groups (n = 4,239). Long-term group (n = 3,240) Short-term group (n = 999) prescribed 3 months or longer prescribed 2 months or shorter Age (y/o) Means S.D. Means S.D. p valuea 59 17 50 19 < 0.000 n % n % Gender 0.107 Male (n = 1,748) 1,358 78 390 22 Female (n = 2,491) 1,882 76 609 24 Department group < 0.0001 Surgery (n = 988) 680 69 308 31 Internal medicine (n = 1,714) 1,398 82 316 18 Psychiatry (n = 1,419) 1,082 76 337 24 Others (n = 118) 80 68 38 32 Pharmacological half-lifeb 0.059 Long (n = 718) 512 71 206 29 Intermediate (n = 950) 726 76 224 24 Short (n = 1,515) 1,130 75 385 25 a Based on chi-square test for gender, department group, and pharmacological half-life, and t-test for age. b 1,056 patients who were prescribed more than one different pharmacological half-life benzodiazepines were excluded. 242 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 243

Table 3 shows the results of Model II, the than surgeons, whereas no such tendency was OR of the long-term prescription for benzodiaze- observed in psychiatrists and physicians in the pines in the elderly (≧ 71 years old). Internists “other” group. were more likely to prescribe benzodiazepines

TABLE 2. Unadjusted and adjusted odds ratios (OR) of the age effect on long-term prescription of benzodiazepines (n = 4,239; Model I)a. Long-term Short-term Univariate analyses Multivariate analysesb group group (n = 3,183) Variables (n = 3,240) (n = 999) 95% CI 95% CI n % n % OR Upper Lower p valuec OR Upper Lower p valuec Age (y/o) < 0.0001 < 0.0001 15-40 (n = 985) 598 61 387 39 ― ― ― ― ― ― 41-59 (n = 1,157) 877 76 280 24 2.03 1.68 2.44 2.67 2.12 3.35 60-70 (n = 992) 810 82 182 18 2.88 2.35 3.54 4.23 3.28 5.45 71- (n = 1,105) 955 86 150 14 4.12 3.33 5.11 5.82 4.49 7.56 a Long-term = 1 vs short-term = 0. b Adjusting for patient gender, department group and pharmacological half-life; 1,056 patients who were prescribed more than one different pharmacological half-life benzodiazepines were excluded. c The Cochrane-Amitage trend test was used for age.

TABLE 3. Unadjusted and adjusted odds ratios (OR) of long-term prescription of benzodiazepines in the elderly aged 71 and older (n = 1,105; Model II)a.

Long-term Short-term Univariate analyses Multivariate analyses group group (n = 922)b Variables (n = 955, 86%) (n = 150) 95% CI 95% CI n % n % OR Upper Lower p valuec OR Upper Lower p valuec Gender 0.025 0.551 Male (n = 427) 369 86 58 14 ― ― ― ― ― ― Female (n = 678) 586 86 92 14 1.00 0.70 1.43 0.89 0.61 1.31 Department group < 0.0001 < 0.0001 Surgery (n = 139) 222 78 63 22 ― ― ― ― ― ― Internal medicine (n = 663) 600 91 63 10 2.70 1.85 3.96 2.91 1.95 4.34 Psychiatry (n = 285) 120 86 19 14 1.79 1.03 3.14 1.50 0.73 3.08 Others (n = 18) 13 72 5 28 0.74 0.25 2.15 0.62 0.21 1.90 Pharmacological half-lifed 0.059 0.548 Long (n = 162) 140 86 22 14 ― ― ― ― ― ― Intermediate (n = 277) 243 88 34 12 1.12 0.63 2.00 1.00 0.55 1.79 Short (n = 483) 403 83 80 17 0.79 0.48 1.32 0.68 0.41 1.15 a Long-term = 1 vs short-term = 0. b Adjusting for patient gender, department group and pharmacological half-life. c P values based on the Wald chi-squared test. d 183 patients who were prescribed more than one different pharmacological half-life benzodiazepines were excluded. 244 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 245

long-term prescription, regardless of whether it DISCUSSION was benzodiazepine, compared with the internal In contrast to their use and safety in younger medicine group. populations, benzodiazepines should be limited in Nevertheless, the characteristics of patients the elderly because long-term use can have seri- who were prescribed benzodiazepines over the ous health consequences, such as a fall, hip frac- long-term were similar in relation to patients in ture, cognitive impairment, sedation, and impaired the internal medicine and surgery groups in terms activity (Pierfitte et al. 2001; de Rekeneire et al. of patient age (66 ± 14 years in the internal medi- 2003; Fick et al. 2003; Wagner et al. 2004). cine group and 63 ± 14 years in the surgery Despite the recommendation that intervention group). The patients’ underlying health condi- with benzodiazepines be short based on several tions in both groups differed from those in the guidelines (Committee on Safety of Medicine psychiatry group. However, our additional analy- 1988; Royal College of Psychiatrists 1988; Fick ses, which included all prescribed patients, et al. 2003; Ashton 2005), previous studies repeat- regardless of benzodiazepine prescription, showed edly showed that the long-term use of these drugs that the prescription period in the internal medi- was very common among the elderly (Magrini et cine group was longer than that in the surgery al. 1996; Simon et al. 1996; Egan et al. 2000a, b; group, which led us to speculate that patients in Gray et al. 2003; Fialova et al. 2005; Voyer et al. the internal medicine group were more likely to 2006). The majority of these previous studies have chronic illnesses than those in the surgery were conducted overseas, and there are very few group. Nevertheless, a chronic health condition reports on elderly users in Japan (Miyamoto et al. does not justify the long-term use of benzodiaze- 2002). This study demonstrated that, in Japan, pine. Moreover, our result showed that internists benzodiazepines were more likely to be pre- were more likely to prescribe benzodiazepines for scribed for a longer term as patient age increased. the long-term, especially in the oldest age group. We discuss the results of this study below. Since the long-term use of benzodiazepine leads In this study, among patient gender, depart- to serious adverse effect in the elderly, any physi- ment group, and pharmacological half-life, cian who might prescribe benzodiazepines department group was found to be a significant requires special attention. At our university hos- determinant of long-term prescriptions; internists pital, hundreds of physicians prescribe benzodiaz- were more likely to write long-term prescriptions epines, and the finding concerning physician spe- as patient age increased (Model I), and internists cialty might better reflect the educational impact were more likely to write the long-term prescrip- of conferences and the advice of supervisors, tions in the elderly compared to surgeons (Model rather than individual prescribing preferences. II). However, the results require careful interpre- Therefore, we suggest that increased education be tation. Referring to the annual report of the hos- provided to those working in department clinics pital, it was found that the underlying disease of concerning the rational prescription of these drugs patients differed, according to the department. to the elderly. Thus, the prescription period might differ accord- Regarding to pharmacological half-life of ing to the characteristics of the patients’ underly- benzodiazepines, a guideline (Nelson and ing condition in each department clinic. Taking Chouinard 1999) denoted that short- and interme- this into account, we performed sub-analyses that diate-beta half-life compounds carry a greater risk included all of the prescribed patients and con- of rebound and withdrawal reactions, and drug firmed that approximately 80% of those visiting dependence than long-acting agents, which the surgery group received drug prescriptions resulted in long-term use of benzodiazepines. written for 2 months or less (short-term in this However, we did not find an association between study). Hence, it was suggested that patients in the pharmacological half-life of benzodiazepines the surgery group were less likely to receive a and long-term prescription, and the result was not 244 K. Nomura et al. The Long-Term Benzodiazepine Prescription in the Elderly 245

altered even in the analyses that included patients in the use of benzodiazepines among older people who had received long-term prescription in more do not necessarily lead to decreased incidence of than two different clinics, and after adjusting for hip fracture. Therefore, the adverse outcome of gender and age. Hence, our findings do not sup- benzodiazepine use remains controversial and port the idea that the difference in the pharmaco- further evidence must be accumulated. logical half-life determines the long-term use of With the limited evidence described above, benzodiazepines. our findings indicate that long-term benzodiaze- There are a few study limitations that need to pine prescriptions increase with patient age, and be discussed. First, because this study is a cross- suggest that educational intervention at depart- sectional study, our result does not immediately ment clinics is useful to promote the rational pre- indicate a causal relationship between long-term scription of benzodiazepines. A discussion of the prescription and patients’ age. Second, the dataset rational prescribing of benzodiazepines in the excluded 1,720 patients who were prescribed ben- elderly should be mandated in the education pro- zodiazepines at two or more different clinical grams for physicians in university teaching hospi- departments. If the characteristics of the excluded tals, and the actual prescribing needs to be moni- population differed from those of our analyzed tored carefully and continuously. sample, our results would be biased. To address this issue, we also analyzed the entire dataset and Acknowledgments still found that long-term prescription increased We thank Mr. Hidefumi Tomisawa, Teikyo with patient age after adjusting for patient gender, University Hospital, for his assistance with the data age, and pharmacological half life of benzodiaze- collection. pines. Third, although short-term intervention is recommended with benzodiazepine use, the long- References term use of benzodiazepines is sometimes Ashton, H. (2005) The diagnosis and management of benzodiazepine dependence. Curr. Opin. Psychiatry, 18, 249-255. required in patients who need benzodiazepines for Barker, M.J., Greenwood, K.M., Jackson, M. & Crowe, S.F. treatment reasons (e.g., augmentation therapy (2004) Cognitive effects of long-term benzodiazepine use: a meta-analysis. CNS Drugs, 18, 37-48. with other drugs such as lithium in mania Committee on the Safety of Medicines (1988) Benzodiaze- [Gouliaev et al. 1996], antipsychotics in psychotic pines, dependence and withdrawal symptoms. Curr. Probl., agitation, and selective serotonin reuptake inhibi- 21, 1-2. de Rekeneire, N., Visser, M., Peila, R., Nevitt, M.C., Cauley, tors in panic disorder). Given that psychiatrists J.A., Tylavsky, F.A., Simonsick, E.M. & Harris, T.B. (2003) usually treat such patients, our additional analy- Is a fall just a fall: correlates of falling in healthy older per- ses, which excluded patients who received benzo- sons. The Health, Aging and Body Composition Study. J. Am. Geriatr. Soc., 51, 841-846. diazepines in the psychiatry group, showed that Egan, M., Moride, Y., Wolfson, C. & Monette, J. (2000a) Long- the long-term benzodiazepine prescription still term continuous use of benzodiazepines by older adults in Quebec: prevalence, incidence and risk factors. J. Am. increased with age. Fourth, rational prescription Geriatr. Soc., 48, 811-816. on demand such as PRN (pro re nata, “as need- Egan, M., Wolfson, C., Moride, Y. & Monette, J. (2000b) Do ed”) was tolerated with long-term use of benzodi- patient factors alter the relationship between physician characteristics and use of long-acting benzodiazepines? J. azepines, even in the elderly (Perry and Wu Clin. Epidemiol., 53, 1181-1187. 1984). However, as our dataset did not include Fialova, D., Topinkova, E., Gambassi, G., Finne-Soveri, H., information about the dose of benzodiazepines Jonsson, P.V., Carpenter, I., Schroll, M., Onder, G., Sorbye, L.W., Wagner, C., Reissigova, J. & Bernabei, R. (2005) prescribed, the result of our study requires careful Potentially inappropriate medication use among elderly interpretation. Also of importance, the relation- home care patients in Europe. JAMA, 293, 1348-1358. Fick, D.M., Cooper, J.W., Wade, W.E., Waller, J.L., MacLean, ship between benzodiazepine use and the inci- J.R. & Beers, M.H. (2003) Updating the Beers criteria for dence of hip fracture among older people has potentially inappropriate medication use in older adults: been the focus of many studies, with conflicting results of a US consensus panel of experts. Arch. Intern. Med., 163, 2716-2724. results. For example, Wagner et al. (2007) report- Gouliaev, G., Licht, R.W., Vestergaard, P., Merinder, L., Lund, H. ed that policies that lead to substantial reductions & Bjerre, L. (1996) Treatment of manic episodes: zuclo- 246 K. Nomura et al.

penthixol and clonazepam versus lithium and clonazepam. (2001) Benzodiazepines and hip fractures in elderly peo- Acta Psychiatr. Scand., 93, 119-124. ple: case-control study. BMJ, 322, 704-708. Gray, S.L., Eggen, A.E., Blough, D., Buchner, D. & LaCroix, Royal College of Psychiatrists (1988) Bulletin of the Royal A.Z. (2003) Benzodiazepine use in older adults enrolled in College of Psychiatrists, 12, 107-108. a health maintenance organization. Am. J. Geriatr. Psychi- Simon, G.E., VonKorff, M., Barlow, W., Pabiniak, C. & Wagner, atry, 11, 568-576. E. (1996) Predictors of chronic benzodiazepine use in a Magrini, N., Vaccheri, A., Parma, E., D’Alessandro, R., Bottoni, health maintenance organization sample. J. Clin. Epidemiol., A., Occhionero, M. & Montanaro, N. (1996) Use of ben- 49, 1067-1073. zodiazepines in the Italian general population: prevalence, Voyer, P., Verreault, R., Mengue, P.N. & Morin, C.M. (2006) pattern of use and risk factors for use. Eur. J. Clin. Phar- Prevalence of insomnia and its associated factors in elderly macol., 50, 19-25. long-term care residents. Arch. Gerontol. Geriatr., 42, Miyamoto, M., Hirata, K., Miyamoto, T., Iwase, T. & Koshikawa, 1-20. C. (2002) Hypnotic prescriptions in a university hospital: Wagner, A.K., Zhang, F., Soumerai, S.B., Walker, A.M., Analysis of data from the computer-ordering system. Gurwitz, J.H., Glynn, R.J. & Ross-Degnan, D. (2004) Psychiatry Clin. Neurosci., 56, 305-306. Benzodiazepine use and hip fractures in the elderly: who is Nakao, M., Takeuchi, T. & Yano, E. (2007) Prescription of ben- at greatest risk? Arch. Intern. Med., 26, 1567-1572. zodiazepine in comparison with SSRI and SNRI for outpa- Wagner, A.K., Ross-Degnan, D., Gurwitz, J.H., Zhang, F., tients attending a Japanese university hospital. Int. J. Clin. Gilden, D.B., Cosler, L. & Soumerai, S.B. (2007) Effect of Pharm. Therap., 45, 30-35. New York State regulatory action on benzodiazepine pre- Nelson, J. & Chouinard, G. (1999) Guidelines for the clinical scribing and hip fracture rates. Ann. Intern. Med., 146, use of benzodiazepines: pharmacokinetics, dependency, 96-103. rebound and withdrawal. Can. J. Clin. Pharmacol., 6, WHO Collaborating Centre for Drugs (1996) Statistics Meth- 69-83. odology Guidelines ATC Classification and DDD Assign- Paterniti, S., Dufouil, C. & Alperovitch, A. (2002) Long-term ment. Oslo, Norway, WHO/NCM. benzodiazepine use and cognitive decline in the elderly: the Woods, J.H., Katz, J.L. & Winger, G. (1992) Benzodiazepines: Epidemiology of Vascular Aging Study. J. Clin. Psycho- use, abuse, and consequences. Pharmacol. Rev., 44, pharmacol., 22, 285-293. 151-347. Perry, S.W. & Wu, A. (1984) Rationale for the use of hypnotic Yamawaki, S. (1999) The use and development of anxiolytics agents in a general hospital. Ann. Intern. Med., 100, in Japan. Eur. Neuropsychopharmacol., 9, S413-419. 441-446. Zandstra, S.M., Furer, J.W., van de Lisdonk, E.H., Bor, J.H., Pierfitte, C., Macouillard, G., Thicoipe, M., Chaslerie, A., Zitman, F.G. & van Weel, C. (2002) Differences in health Pehourcq, F., Aissou, M., Martinez, B., Lagnaoui, R., status between long-term and short-term benzodiazepine Fourrier, A., Begaud, B., Dangoumau, J. & Moore, N. users. Br. J. Gen. Pract., 52, 805-808.