Psychotropic Medication and Insomnia Complaints in Two ...

2 downloads 0 Views 77KB Size Report
Sep 5, 1996 - habits of study populations in France and Montreal, current consumption ... 1Director, Centre de recherche Philippe Pinel de Montréal, Montreal,. Quebec. ..... Institut National de Statistiques et Evaluation Economique (INSEE).
Psychotropic Medication and Insomnia Complaints in Two Epidemiological Studies Maurice M Ohayon, MD, DSc1, Malijaï Caulet, MD 2

Objective: This study compared prescribed psychotropic medication patterns for reported sleep disorders in French and Quebec samples. Method: The first study was undertaken in France (N = 5622) and the second in the metropolitan area of Montreal (N = 1722). Lay interviewers used a specialized knowledge-based system for the purpose of evaluating sleep disorders by telephone. Results: Results showed similar prevalence of insomnia complaints in both samples (20.1% and 17.8%, respectively). A higher level of psychotropic consumption was found in France (11.7% [95% confidence interval (CI), 10.9 to 12.5]) compared with Quebec, however, where consumption was less than half the French rate (5.5% [95% CI, 4.4 to 6.6]). Both studies identified females and the elderly as the primary consumers of these drugs. For approximately two-thirds of both samples, sleep-promoting medications were prescribed for a year or longer, revealing a chronicity of the consumption. Approximately 4 out of 5 prescriptions for sleeping medications were ordered by general practitioners in both samples. Conclusion: These findings clearly show a higher prevalence of psychotropic drug use in the French compared with the Quebec population. The patterns of consumption and prescription, however, are quite similar in both studies. (Can J Psychiatry 1996;41:457–464) Key Words: anxiolytic, epidemiology, hypnotic, insomnia complaints

T

Besides these methodological differences, there are national variations in health care services as well as cultural differences within both the population and the medical profession (for example, public awareness and medical training in the treatment of sleep disorders). In the course of 2 epidemiological studies of the sleep habits of study populations in France and Montreal, current consumption (that is, consumption ongoing at the time of the interview) of sleep-promoting and anxiety-reducing drugs was investigated using the same methodology. This article compares psychotropic drug consumption in these 2 general populations, which present relatively similar rates of sleep disorders.

he use of psychotropic drugs in the treatment of sleep disorders is extremely widespread. Relevant epidemiological studies are numerous and demonstrate that psychotropic drug consumption varies between 2% and 10%, depending on the country considered (1–6). This variability is attributable in part to the study methodologies employed. Indeed, there are considerable disparities in this regard, such as in the time intervals considered (for example, current consumption over the past month or over the past year) and the specific focus of the study (for example, particular attention paid to a specific class of medication, such as hypnotics).

This paper was presented at the XIXth Collegium Internationale Neuro-Psychopharmacologicum (CINP) Congress, Washington, DC, 27 June–1 July, 1994. Manuscript received November 1995, revised April 1996. 1 Director, Centre de recherche Philippe Pinel de Montréal, Montreal, Quebec. 2 Neurologist, Scientist, Centre de recherche Philippe Pinel de Montréal, Montreal, Quebec. Address reprint requests to: Dr Maurice M Ohayon, Centre de recherche Philippe Pinel de Montréal, 10905, boulevard Henri-Bourassa Est, Montréal, QC H1C 1H1 Can J Psychiatry, Vol 41, September 1996 457

Method Subjects In 1993, two epidemiological studies of sleep habits, sleep-related symptoms, and sleep disorders were conducted by telephone. The first study was performed in France (7). The target population comprised noninstitutionalized residents of 15 years of age and older. A representative sample was obtained

458

The Canadian Journal of Psychiatry

using a stratified probabilistic approach. Sample selection, with respect to distribution among the 9 geographic areas of approximately equal population (Nielsen areas) of France, was based upon the results of the 1990 France census survey (8) and, with respect to age and gender, was based upon the Kish selection method (9). The Kish survey sampling technique uses 8 selection tables to determine the subject in a given household to be interviewed in order to maintain a representative sampling. A selection table is randomly assigned to a household before the number is called. Based on the number of subjects in the household, their gender, and their age, the table indicates which household member should be interviewed. The second survey was conducted in the greater metropolitan area of Montreal (approximately 1 762 121 Francophones aged 15 years or older), where approximately 40% of all French-speaking Quebecers reside. The target population consisted of all noninstitutionalized French-speaking residents of 15 years of age or older. A representative sample was drawn using the same procedure as in the French sample. Geographic distribution was established based on the 1991 Canadian census survey (10). In France, 6966 subjects were approached and 5622 interviews were completed (80.8% of the potential sample). In the metropolitan area of Montreal, the participation of 2117 people was solicited, and 1722 interviews were completed (81.3% of the potential sample). Excluded from the studies were non-French speakers, those who suffered from a hearing or speech impairment, and those who had an illness precluding the possibility of an interview. Any selected phone number with no answer was dialled again at least 6 times at different evening hours and on different days, including weekdays and weekends, before being replaced due to lack of response. Individuals who declined to participate in the study upon first contact were solicited twice more before being eliminated. The participation rates were calculated on the basis of the number of eligible persons contacted. Rejected telephone numbers (for example, a business number, a number no longer in service, a number for which there was no answer after a minimum of 6 attempts, or a number meeting exclusion criteria) were not included in the calculation of the participation rates. Instrument Telephone interviews were conducted by lay interviewers using Sleep-Eval, a knowledge-based system (11,12), a previously validated (13–15), nonmonotonic, level-2 expert system with a causal reasoning mode. General Functioning. The system was used to conduct the epidemiological study and administer the questionnaire through the use of 3 primary “modules.” The first is a selection module designed to select the appropriate interviewee. Data information entries included the number of the selection

Vol 41, No 7

table, the number of persons in the household 15 years of age or older, and the sex and the age of each. The software then indicated which person to interview. The second module, a logical reasoning component (inference engine), prompted the interviewer to ask specific questions based on the interviewee’s previous responses. Diagnostic hypotheses of sleep and mental disorders were formulated and verified after responses to all questions were provided. Questions not relevant to a specific individual were preemptively eliminated by the system. Consequently, subjects with no particular sleep difficulties would undergo only a brief interview (10 to 15 minutes), while subjects with documented sleep disturbances (with or without mental disorders) would necessarily respond to longer interviews of up to 90 minutes in some cases. It was the system that selected and phrased the questions, with the interviewer merely asking them as they appeared on the monitor screen. The third component of the system is a file management module, which is designed to classify subjects in the study. The classification categories were a) subjects with sleep problems, b) subjects without sleep problems, c) subjects who refused to be interviewed, d) telephone numbers called more than 6 times unsuccessfully, and e) telephone numbers to be renewed. Once a file was definitively classified in any but the last category (e), it was impossible for the interviewer to access it. Variables Explored by the System. The aspects of medical and social history that were systematically investigated in the present study included 1) sociodemographics; 2) medical consultations in the past 6 months and illnesses treated; 3) quantity and quality of sleep (bedtime, sleep latency, wake-up time, duration of nighttime sleep, presence and frequency of nocturnal awakenings, daytime sleepiness, number of daily naps, and satisfaction with sleep); 4) DSMIV psychiatric diagnoses of mood and anxiety disorders; and 5) current or past consumption of medication to promote sleep or alleviate anxiety (name of medication, dosage, number of pills, name of physician who prescribed the medication [if applicable], and duration of consumption). In the case of past treatment, subjects were asked when treatment started and ceased. The name of each medication was later verified and categorized according to classification proposed by the French national dictionary of prescriptions (Vidal) and the Canadian Compendium of Pharmaceuticals and Specialties (CPS). The populations were subsequently divided into 2 distinct groups: subjects satisfied with their quality of sleep and not taking psychotropic medication of the anxiolytic or hypnotic type (SQS group) and subjects dissatisfied with the quantity and quality of their sleep and/or taking sleep-promoting medication (DQS group).

September 1996

Psychotropic Medication and Insomnia

459

Table I Types of Sleep-Promoting Medication Medication Hypnotics

France % (n = 549) 37.9

Anxiolytics Other psychotropics Nonpsychotropics Unknown

Quebec % (n = 63) 22.2a

50.0 4.6 6.6 1.0

54.0 7.9 a 11.1 6.3a

a

Intercountry differences at P < 0.01.

Figure 1. Prevalence of DQS subjects in France and Quebec by gender and age group. Intercountry differences *P < 0.005; **P < 0.01;***P < 0.05.

Subjects in Quebec ranged between 15 and 100 years of age. Under the same criteria, 17.8% (95% CI, 16 to 19.6) of these were DQS. The DQS proportions for men and women, however, were relatively similar (male: 15.9% [95% CI, 13.4 to 18.4]; female: 19.5% [95% CI, 16.9 to 22.1]). Rates were also relatively stable across age groups, except for women aged 55 years or over, where the proportion of subjects in the DQS category was significantly higher compared with the

Analyses Table II a Drugs Most Frequently Used as Sleep-Promoting Medication

To compensate for potential biases (from such factors as an uneven response rate in different demographic groups or the absence of a telephone in a fraction of households), a weighting procedure was used to adjust for sample design in both studies. Weights were derived from the following equation: weight (Wi ) = nexpected / nobserved, where the nexpected represents the number of subjects expected in the sample for a given area, age group, and gender, and the nobserved represents the number of subjects in the sample obtained for this cell. Consequently, after weighting, the French sample was split 52.1% female to 47.9% male (56.1% and 47.9% before the weighting procedure). In Quebec, the gender breakdown of the weighted sample was 52% women and 48% men (55.4% and 44.6% before the weighting procedure). The results presented include weighted rates and unweighted n values. Intercountry comparisons were made using the z statistic. Ninety-five percent confidence intervals (95% CI) are also provided. The minimum statistical significance was P < 0.05.

Generic Name

A significantly higher DQS proportion was observed in women than in their male counterparts (24.4% [95% CI, 22.8 to 26.0] versus 15.6% [95% CI, 14.3 to 16.9]). DQS proportions were significantly higher for women over the age of 54 years compared with women aged 45 years or under. Among men a significant difference was observed between the group aged 65 years or older and the other age groups (Figure 1). The median duration of insomnia complaints was 5 years.

Lorazepam (Ax) Bromazepam (Ax) Zopiclone (H) Oxazepam (Ax)

24.9 14.6 6.2 5.7

Flunitrazepam (H) Zolpidem (H) Loprazolam (H) Noctranb (H) Clorazepate (Ax) Meprobamate (Ax) Othersc

4.6 4.6 3.2 3.0 2.8 1.9 27.8 Quebec (n = 63)

Results In France, the subjects were between the ages of 15 and 96 years. Those in the DQS subset represented 20.1% (95% CI, 19.1 to 21.1) of the French sample.

% France (n = 549)

a

Lorazepam (Ax)

32.7

Bromazepam (Ax) Clonazepam Oxazepam (Ax) Flurazepam (H) Temazepam (H) Amitriptyline (Ad) Trimipramine (Ad) Othersc

5.7 5.7 5.2 5.2 (2)d (2)d (2)d 36.8

Drug classification according to the Vidal (France) and CPS (Canada). Clorazepate acepromazine acetopromazine. Compounds cited less than 10 times. d n value. Intercountry differences not significant. Ad = antidepressant. Ax = anxiolytic. H = hypnotic. b c

460

The Canadian Journal of Psychiatry

Vol 41, No 7

Table III Duration of Sleep-Promoting Drug Consumption Duration ≤ 6 months 6−12 months

France % (n = 549) 19.6 8.6

1−5 years ≥ 5 years

40.4 31.5

Quebec % (n = 63) 24.4 10.1 35.9 29.6

Intercountry differences not significant.

Figure 2. Prevalence of sleep-promoting drug consumption in France and Quebec by gender and age group. Intercountry differences *P < 0.005; **P < 0.01; ***P < 0.05.

younger age groups (see Figure 1). As in the French sample, the median duration of insomnia complaints was 5 years. Sex and age differences in the distribution of DQS between the 2 populations are attributable to the fact that a higher proportion of young men (< 45 years of age) and a lower proportion of older men (≥ 55 years of age) were classified as being DQS in Quebec than in France (P < 0.05). In females, the proportion of subjects classified as DQS among those aged 45 years or older was lower in Quebec than in France (P < 0.05). Sleep-Promoting Medication To the question, “Are you presently taking medication to help you sleep?” 9.9% (95% CI, 9.1 to 10.7) of French subjects responded in the affirmative. Consumption was higher among women (12.7% [95% CI, 11.5 to 13.9]) than among men (6.8% [95% CI, 5.8 to 7.8]; P < 0.0001) and increased with age. While relatively uncommon among young subjects, consumption increased considerably beginning at the age of 45 years, and affected one-quarter of the elderly (≥ 65 years of age) (P < 0.0001). Current use of sleep-promoting medication was reported by 3.8% (95% CI, 2.9 to 4.7) of the Quebec sample. The consumption rate for men (2.8% [95% CI, 1.7 to 3.9]) was relatively similar to that for women (4.7% [95% CI, 3.3 to 6.1]). Here, too, consumption increased with age, especially among subjects aged 65 years or over. Consumption was twice as high in France as in Quebec among subjects aged 45years or over (P < 0.01) (Figure 2). The names of the sleep-promoting medications were provided by subjects during the interviews and later verified, resulting in the identification of 46 different sleep-promoting

psychotropic drugs in France and 15 in Quebec. For more than half of the subjects, the medication used to facilitate sleep was, in fact, an anxiolytic (Table I). Hypnotics ranked second. Consumption of both these classes of drugs was significantly higher in France than Quebec. Approximately half of the medications in the category “other psychotropics” were antidepressants. Lorazepam and bromazepam were the most commonly used compounds (Table II). The majority of subjects using such medications had been doing so for at least one year. In France, 71.9% of consumers belonged to this category compared with 65.5% in Quebec (Table III). Most consumers (83.3% in France and 77.4% in Quebec) obtained their prescription from a general practitioner. Psychiatrists were a distant second (6.5% and 11.3%, respectively). Anxiety-Reducing Medication To the question, “Are you presently taking medication to alleviate anxiety?” 6.7% (95% CI, 6.0 to 7.4) of French subjects responded affirmatively. Such drugs were consumed by a higher proportion of women (8.7% [95% CI, 7.7 to 9.7]) than men (4.5% [95% CI, 4.7 to 5.3]; P < 0.0001). Consumption increased with age, but less dramatically than the increase seen with the consumption of sleep-promoting medication. The use of anxiety-reducing drugs was reported by approximately 4.5% of subjects under the age of 45 years, and increased slightly across the other age groups (Figure 3). In Quebec, 3% (95% CI, 2.2 to 3.8) of the population took anxiety-reducing medication. Proportionally speaking, more

Table IV Types of Anxiety-Reducing Medication Medication Anxiolytics Antidepressants Hypnotics Other psychotropics Nonpsychotropics Unknown a

France % (n = 342) 52.9 12.5 2.4 3.7 23.7 4.9

Intercountry differences at P < 0.05.

Quebec % (n = 53) 53.0 13.5 6.7a 5.6 16.5 4.8

September 1996

Psychotropic Medication and Insomnia

461

Overall Psychotropic Drug Consumption When the responses to the 2 questions mentioned previously were combined—with each subject counted only once and those subjects not taking any psychotropic medication excluded—psychotropic drug consumption totalled 11.7% (95% CI, 10.9 to 12.5) within the French population and 5.5% (95% CI, 4.4 to 6.6) within the Quebec population (P < 0.0001). In both countries, the percentage of women (France: 14.7%; Quebec: 7.2%) taking such medication was twice as high as the percentage of men (France: 8.4%; Quebec: 3.6%). As shown in Figure 4, overall consumption was nearly twice as high in France as in Quebec. This difference was significant for all groups over the age of 35 years. In France, 53.6% of the DQS subjects used psychotropic medication compared with only 26.3% of the DQS subjects in Quebec (P < 0.0001) (Table VII). While the difference between the women of both populations was indeed significant, the difference between the men of France and Quebec Figure 3. Prevalence of anxiety-reducing drug consumption in France and Quebec by gender and age group. Intercountry differences *P < 0.005; **P < 0.01; ***P < 0.05. Table V a Drugs Most Frequently Used as Anxiety-Reducing Medication

women than men (4.3% [95% CI, 3 to 5.6] versus 1.6% [95% CI, 0.7 to 2.5]; P < 0.0001) used this type of drug. Consumption increased with age.

Generic Name Bromazepam (Ax) Prazepam (AX) Lorazepam (Ax) Clorazepate (Ax) Fluoxetine (Ad) Alprazolam (Ax) Clomipramine (Ad) Nordazepam (Ax)

Overall, the consumption of anxiety-reducing drugs was twice as high in France as in Quebec (P < 0.001). This spread was apparent particularly between the groups of subjects under the age of 35 years, women aged 35 to 44 years, women aged 55 to 64 years, and men aged 65 years or older (see Figure 3).

For approximately half the cases in both studies, the use of medication for the purpose of alleviating anxiety had been ongoing for over one year (Table VI). With these drugs, too, general practitioners were the most common prescribers (74.6% and 75.1% of French and Quebec consumers, respectively). Psychiatrists were second, with 10.4% of French consumers and 20.5% of Quebec consumers having obtained their prescription from these specialists (P < 0.01).

16.9 7.5 6.8 5.0 4.3 2.7 2.4 2.3

b

Alpidem (Ax) Amitriptyline (Ad) Othersc

The names of the medications consumed were provided by subjects during the interviews and later verified, resulting in the identification of 47 different anxiety-reducing psychotropic drugs in France and 16 in Quebec. For more than half of the cases in both populations, the drugs used to alleviate anxiety were anxiolytics. Antidepressants ranked second, followed by hypnotics. A significantly higher consumption of hypnotics was observed in Quebec (P < 0.05) (Table IV). Regarding the compounds used, topping the lists in France and Quebec were bromazepam and lorazepam, respectively. The latter drug was much less commonly used in France (P < 0.001) (Table V).

% France (n = 342)

1.9 1.5 50.7 Quebec (n = 53)

a

Lorazepam (Ax) Bromazepam (Ax) Clonazepam Alprazolam (Ax) Lithium Flurazepam (H)

27.1d 9.6 7.3 7.3 (2)e (2)e

Oxazepam (Ax) Amitriptyline (Ad)

(2) (2)e

Others

24.7

e

Drug classification according to the Vidal (France) and CPS (Canada). b Deleted in 1993. c Compounds cited less than 10 times and included 16.6% of “natural compounds” in France. d Intercountry differences at P < 0.001. e n value. Ad = antidepressant. Ax = anxiolytic. H = hypnotic.

462

The Canadian Journal of Psychiatry

Vol 41, No 7

Table VI Duration of Anxiety-Reducing Drug Consumption Duration ≤ 6 months 6−12 months 1−5 years ≥ 5 years

France % (n = 342) 37.6 10.4 27.1

Quebec % (n = 53) 31.2 20.5a 28.1

24.9

20.2

a

Intercountry differences at P < 0.01.

Table VII Prevalence of Psychotropic Drug Consumption in DQS Subjects Subgroups

Figure 4. Prevalence of psychotropic drug consumption in France and Quebec by age group. Intercountry differences *P < 0.005; **P < 0.01; ***P < 0.05.

was even more appreciable. Where age is concerned, the proportion of psychotropic drug consumers aged 55 years or over was, for all intents and purposes, the same in both populations (74.7% in France versus 64.7% in Quebec; difference deemed nonsignificant). Finally, a sizeable portion of the consumers in both populations suffered from mental disorders. Anxiety disorders were the most frequently encountered, affecting 65.7% of French consumers and 61.6% of Quebec consumers. Also, major depression was found in 19.7% and 18.1% of French and Quebec consumers, respectively. These base rates for both anxiety and depression are not statistically different. Conclusion One limit of this study is that it was a comparison between the whole country of France and only the metropolitan area of Montreal. Three arguments in response to this limitation can be given, however: 1) the comparison between the metropolitan area of Paris and the whole French sample did not reveal significant differences in the rates of psychotropic drug consumption, 2) the metropolitan area of Montreal represents nearly half (40%) of the French-speaking Quebecers, and 3) the direction of the differences were the same when Paris and the whole French sample were compared with the metropolitan area of Montreal; that is, rates of psychotropic drug consumption remained twice as high. The rural versus urban area factor, therefore, did not account for the intercountry rate differences. The present study empirically demonstrates that just over twice as many French citizens use prescribed psychotropic

Total Women Men < 55 years old ≥ 55 years old a

% 53.6 56.4 49.0 37.1 74.7

France 95% CI 50.6 to 56.6a 52.6 to 60.2a 44.1 to 53.9a 32.7 to 41.5a 71.2 to 78.2

% 26.3 31.9 18.7 14.1 64.7

Quebec 95% CI 21.4 to 31.2 25.0 to 38.8 12.0 to 25.4 9.6 to 18.6 53.9 to 75.5

Intercountry differences at P < 0.0001.

medication as Montreal-area Quebecers. In both populations, the overwhelming majority of these patients (nearly 90%) were prescribed benzodiazepines. Females and the elderly constituted the majority of consumers. Such treatment was typically chronic in that it had been ongoing for more than one year. Finally, general practitioners were the usual prescribers of such medication. These observed rates are highly consistent with others reported in the literature. Laurier and others (3) recently found that 5.6% of Quebecers used prescribed medication, including tranquillizers, sedatives, and sleeping pills. Quera-Salva and others (5) reported that 10% of the French population employed drugs to facilitate sleep, although they did not specify what percentages used what types of medication. Accounting for these differences is important because if the true base rates of sleep disturbance are similar in the 2 populations, either one population is overmedicated, or the other population is denied treatment. One possible explanation for the discrepancy lies in the medical and/or patient cultures in France and Quebec. Medical training may direct more French physicians to prescribe for minor geriatric sleep problems or Quebec physicians to discount geriatric complaints of sleep disturbance. The 2 approaches to medical training may differentially emphasize or down-play possible side effects of such medications in geriatric patients. Additionally, French geriatric patients come to expect relief from insomnia from their physicians, while Quebecers may be more likely to construe such events as a benign consequence of aging or to seek other remedies. Such speculation awaits further investigation.

September 1996

Psychotropic Medication and Insomnia

463

A more prosaic explanation for the higher French medication rate lies in the different payment schedules of the respective national health care systems existing in the 2 areas. In Quebec, only welfare recipients and those aged 65 or older are reimbursed for medication expenses, while in France all such expenses are covered. Additionally, available private health care schemes are out of reach for a considerable segment of the Quebec working class. This consideration may explain a significant component of the discrepancy, since in the over 65 DSQ group, where drug coverage is quite similar, consumption rates of the 2 samples do not differ significantly.

side effects, be pursued in combatting this public health problem that affects approximately 20% of both populations sampled in this study.

Despite the dissimilarity in consumption prevalence in the 2 samples, similarities were observed with reference to duration of consumption and type of prescribing physician— usually general practitioners. More focused analyses of the effectiveness of different drug therapies for sleep facilitation (16) yielded inconclusive results. Over 50% of subjects demonstrated no improvement in sleep quality when compared with a control DSQ group taking no medication.

• Study conducted in a noninstitutionalized population. • Insomnia complaint was based on the subjective perception of

One difference in prescription patterns that did emerge was that hypnotics represented 38.9% of psychotropic drug consumption in France versus only 22.2% in Quebec. Anxiolytics represent the most frequently prescribed class of medication for insomnia in both localities. This is understandable if most patients complain of anxiety in addition to, or as an integral component of, their sleep disturbances. In this regard, 29% of Quebec insomniacs who were prescribed an anxiolytic had a concurrent diagnosis of anxiety disorder, whereas in France this figure jumped to 50%. Furthermore, it is well known that some benzodiazepines have both hypnotic and anxiolytic properties. When using a drug, however, physicians should ensure that pharmaceutical trials have been performed to prove its efficacy and safety for the purpose for which it is prescribed. Pharmaceutical dictionaries, such as the CPS in Canada and Vidal in France, always provide the main purpose of the drug and the main indications for its use. Individuals whose insomnia was treated with antidepressant medication tended predominantly (over 80% in France and 70% in Quebec) to suffer concomitantly from a spectrum of depressive disorders, including major depressive disorder, dysthymic disorder, or depressed mood. In both venues, however, unexpectedly high proportions of people suffering from various mental disorders received no medication, including 40% suffering from major depression in both samples and 66% and 46% suffering from anxiety disorders in the Quebec and France populations, respectively. Unfortunately, there was an increased tendency on the part of physicians in France to prescribe hypnotics with other psychotropic agents (17.4% in France versus 4% in Quebec). We suggest that alternate treatment approaches, including educating patients regarding methods of sleep hygiene and prescribing more specific hypnotic compounds with fewer

Clinical Implications • Insomnia complaints are a common symptom in the general population.

• General practitioners widely use anxiolytics in the treatment of sleep complaints.

• Women and the elderly are the primary chronic psychotropic drug consumers ($ one year).

Limitations

the patient.

• Drugs and indications were based on self-report.

Acknowledgements This study was supported by the “Fond de la Recherche en Santé du Québec (FRSQ)” #931597 and Synthelabo, CNS France.

References 1. Allgulander C, Nasman P. Regular hypnotic drug treatment in a sample of 32,679 Swedes: associations with somatic and mental health, inpatient psychiatric diagnoses and suicide, derived with automated record-linkage. Psychosom Med 1991;53:101–8. 2. Dunbar GC, Perera MH, Jenner FA. Patterns of benzodiazepine use in Great Britain as measured by a general population survey. Br J Psychiatry 1989;155:836–41. 3. Laurier C, Dumas J, Grégoire JP, Duval L. L’utilisation des tranquilisants, sédatifs et somnifères: analyse des données de 1’enquête Santé Québec. In: Direction des Communications Ministère de la Santé et des Services sociaux. Québec: Les cahiers de recherche; 1990. 4. Mellinger GD, Baiter MD, Uhlenhuth EH. Insomnia and its treatment. Arch Gen Psychiatry 1985;42:225–32. 5. Quera-Salva MA, Orluc A, Goldenberg F, Guilleminault C. Insomnia and use of hypnotics: study of a French population. Sleep 1991;14:386–91. 6. Vazquez-Barquero JL, Diez-Manrique JF, Pena C, Arenal-Gonzalez A, Cuesta MJ, Artal JA. Patterns of psychotropic drug use in a Spanish rural community. Br J Psychiatry 1989;155:633–41. 7. Ohayon M. Epidemiological study on insomnia in the general population. Sleep 1996;19(3 Suppl)7S–15S. 8. Institut National de Statistiques et Evaluation Economique (INSEE). Recensement de la population de 1990. Paris: INSEE; 1991. 9. Kish L. Survey sampling. New York: Wiley; 1965. 10. Statistics Canada. Profiles. Ottawa: Statistics Canada; 1993. 11. Ohayon M. Sleep disorders questionnaire and decision trees of the Eval system. Quebec: Bibliothèque nationale du Québec; 1994. 12. Ohayon M. Use of an expert system (Eval) in mental health epidemiological surveys. In: Barahona P, Veloso M, Bryant J, editors. Proceedings of the 12th International Congress of the European Federation for Medical Informatics, 1994 May 22–26; Lisbon: Medical Informatics in Europe (MIE 94). p 174–9.

464

The Canadian Journal of Psychiatry

13. Ohayon M, Caulet M. Adinfer: experience of an expert system in

Federation for Medical Informatics, 1994 May 22–26; Lisbon: Medical Informatics in Europe (MIE 94). p 90–5.

psychiatry. In: Lun KC and others, editors. Med Info (MEDINFO 92). Amsterdam: Elsevier Science Publishers; 1992. p 615–9. 14. Ohayon M. Validation of a knowledge based system (Adinfer) versus human experts. In: Barahona P, Veloso M, Bryant J, editors. Proceedings of the 12th International Congress of the European

Vol 41, No 7

15. Ohayon M. Validation of expert systems: examples and considerations. In: Greenes RA and others, editors. Med Info (MEDINFO 95). Amsterdam: Elsevier Science Publishers; 1995. p 1071–5. 16. Ohayon MM, Caulet M. Insomnia and psychotropic drug consumption. Prog Neuropsychopharmacol Biol Psychiatry 1995;19:421–31.

Résumé Objectif : L’étude compare la prescription de psychotropes pour traiter les troubles du sommeil en France et au Québec. Méthode : La première étude a eu lieu en France (N=5622), la seconde dans la région métropolitaine de Montréal (N=1722). Des intervieweurs non spécialisés se sont servis d’un système expert conçu pour évaluer par téléphone les troubles du sommeil. Résultat : Les résultats montrent une prévalence semblable de l’insomnie dans les 2 échantillons (20,1 % et 17,8 %, respectivement). La prescription de psychotropes est plus fréquente en France (11,7 % [intervalle de confiance (IC) de 95 %; 10,9 % à 12,5 %]) qu’au Québec où l’utilisation de ces molécules est 2 fois moindre (5,5 % [IC de 95 %; 4,4 % à 6,6 %]). Les 2 études indiquent que les femmes et les personnes âgées ont recours plus fréquemment à ces médicaments. Environ les 2 tiers des sujets des 2 échantillons ont obtenu une ordonnance d’un an ou plus, signe d’une consommation chronique. Dans les 2 cas échantillons, environ 4 ordonnances sur 5 pour des somnifères proviennent d’un omnipraticien. Conclusion : Ces résultats révèlent clairement une plus grande utilisation des psychotropes en France qu’au Québec. Néanmoins, les habitudes de consommation et de prescription restent semblables dans les 2 populations.