Vanna Schiralli Marion McIntosh

Benzodiazepines: Are We Overprescribing?SUMMARYThe authors made a survey of benzodiazepineuse in the Family Practice Units at TorontoGeneral Hospital and report the findings.They have examined, among other factors,drugs used, reasons for use, and perceivedwithdrawal symptoms. The results indicatedthat 24.3% of respondents had takenbenzodiazepines in the previous year, and12.2% in the previous two weeks. There wasno difference in the percentage of use ofbenzodiazepines by males and females. Thisstudy confirms that diazepam was the mostcommon drug used in all age ranges. Finally,6.1% of benzodiazepine users stated that theyhad attempted an overdose. (Can FamPhysician 1987; 33:927-934.)

RESUMELes auteurs ont effectue une enquete concernantl'utilisation des benzodiazepines dans les unites depratique familiale reliees au Toronto GeneralHospital et font etat de leurs constatations. Ils ontexamine, entre autres facteurs, les medicamentsuteliss, les raisons invoqudes et les sympt6mes desevrage perVus. Les resultats indiquent que 24.3%des repondants ont pris des benzodiazepines aucours de l'annee precedente et que 12.2% de ceux-cien ont utflise au cours des deux dernieres semaines.On n'a pas constate de difference dans lepourcentage d'utilisation des benzodiazepimes entreles hommes et les femmes. Cette etude confirmeque, dans tous les groupes d'age, le diazepam fut lemedicament le plus frequemment prescrit.Finalement, 6.1% des utilisateurs debenzodiazepines ont mentionn6 etre alles jusqu'al'intoxication.

Key words: benzodiazepine prescribing, drug use, drug abuse

Dr. Schiralli, a certificant of theCollege, is a member of theDepartment of Family &Community Medicine in TorontoGeneral Hospital and a lecturer inthe Department of Family andCommunity Medicine at theUniversity of Toronto. Dr.McIntosh, who is also a certificantof the College, is an assistantprofessor and Co-ordinator of Drugand Alcohol Training in theDepartment of Family &Community Medicine at theUniversity of Toronto and Directorof Ambulatory Medical Services ofthe Addiction ResearchFoundation. Reprint requests to:

Dr. Vanna Schiralli, FamilyPractice Service, Team 2, Toronto

CAN. FAM. PHYSICIAN Vol. 33: APRIL 1987

General Hospital, 200 ElizabethStreet, Toronto, Ont. M5G 2C4

B ENZODIAZEPINES, it has beenclaimed, are the most frequently

prescribed drugs in the world.' The

Annual Ontario Prescription Survey(1979) shows that from 1970 to 1978,prescriptions for minor tranquillizers,as a percentage of all prescriptions forpsychotropic medications, had in-creased from 31% to 44%. The in-

crease was largely accounted for byan increase in prescriptions for benzo-diazepines.2 In 1983, the authors of a

provincial government review of theuse of benzodiazepines reported thatfrom 1977 to 1982, there was a de-crease in the use of benzodiazepine

tranquillizers and an increase in theuse of benzodiazepine hypnotics.Overall, the total use of benzodiaze-pines had not changed.3 With suchextensive use of the benzodiazepines,the potential for misuse and abuse be-comes high. Several studies have ad-dressed this problem.4-6The family physician is in a unique

position as a primary care giver tocontrol the use of these drugs and toexplore with his or her patients alter-native means to cope with their anxi-ety. Many self-report surveys havebeen done to estimate the prevalenceand patterns of use of drugs and alco-hol. The validity and reliability ofself-reported drug-use surveys havebeen studied, and these studies indi-cate that self-reported drug use repre-

927

M-- 1-

sents an underestimate of true use.7-9Self-reported alcohol use was studiedby Pernanen, who reported similarfindings.10

Self-report surveys have many spe-

cific limitations in that they are sub-ject to interviewer bias, to problemsof recall, and, often, to difficulties inproving the validity of the data. Theydo, however, provide correlational re-

lationships between personal habitsand health, though they do not con-

tribute to our understanding of thereasons for particular associations. I

The present study is a survey offamily-practice patients and as suchomits institutionalized people, such as

those in prisons, hospitals or militaryestablishments. In addition, thosewho visit their family physicians more

often have a higher probability ofbeing included in the sample. Sincethe use of benzodiazepines can causea short-term memory loss, the prob-lem of recall in our study will be even

greater than usual.12-13The purpose of our study was to

determine the prevalence and patternsof use of benzodiazepines in the Fam-ily Practice Units at the Toronto Gen-eral Hospital. Specifically, questionswere asked about types of benzodia-zepines used, the sources from whichthe drugs were obtained, reasons foruse, tolerance, perceived withdrawalsymptoms, whether users felt thattheir use was a problem, and overdoseattempts.A comparison was made of demo-

graphic characteristics of non-users

and users. Those who had taken a tran-quillizer at least once in the past 12months (not including pre-operativepatients) were classified as users.

MethodSix hundred questionnaires were

distributed in the Family PracticeUnits of the Toronto General Hospi-tal. Two hundred questionnaires eachwere given to Team 1, Team 2, bothlocated within the hospital, and to theSt. George Health Centre, locatedseveral blocks from the hospital.The questionnaires were distributed

to each patient who had not pre-viously filled one out as part of a reg-ular visit to the doctor. No age re-

strictions were involved. In theintroductory paragraph of the ques-tionnaire, the participants were askednot to identify themselves in any wayand to place completed questionnaires

CAN. FAM. PHYSICIAN Vol. 33: APRIL 1987

in the box provided at the front desk.The first section of the questionnaireincluded general demographic ques-

tions. Those who stated that they hadtaken a tranquillizer or sleeping pill inthe preceding 12 months (not includ-ing any administered pre-operatively)were asked to complete the secondsection of the questionnaire, which in-cluded specific questions about theirdrug use.Of the 600 questionnaires handed

out, 510 were returned. The responserate was 85%. Of the 510 respon-dents, 135 stated that they had taken a

tranquillizer or sleeping pill with inthe last 12 months. Those who con-sidered medications other than benzo-diazepines to be tranquillizers (i.e.,antihistamines, neuroleptics or antide-pressants) were omitted. There were11 of these patients. The data were

analysed by a difference-of-propor-tions test (2-tailed Z value) to mea-

sure significance.The patients surveyed during the

study are typical of those seen in a

family practice unit. The majority(61.6%) were in the 21 -40 agerange, 22.0% in the 41-60 age

range, and 12.2% in the 61-80range. A few patients remained out-side the 20-80 age range: 3. 1% were

under 20 years of age, and 1.0% were

over 80 years. The ratio of females tomales was 70:30. The majority of pa-tients (56.3%) considered themselvesprofessionals by occupation; 29.7%held clerk/secretarial positions; 7.2%were the skilled tradespersons, and6.7% were labourers. Of our respon-dents, 11.9% were unemployed,62. 1% were working full time, 13.0%worked part time, 10.5% were house-wives, and 2.5% were retired.

ResultsIn this study, 24.3% of respondents

claimed to have used tranquillizers or

sleeping pills at least once in the pre-vious 12 months. Of these respon-dents 50% reported use within theprevious two weeks. The sex of re-

spondents was not significant. Theproportion of male to female respon-dents who used benzodiazepines(30.0:70.0) was the same as the pro-portion that did not use them(30.3:69.7). In addition, in no agerange, was there any statistical dif-ference in the number of males andfemales who reported use. Whenquestioned about the benzodiazepine

used, 55.4% of users reported havingused diazepam; 44.6% gave the tradename "Valium". Of all tranquillizerusers, 52.2% of those aged 21-40,66.6% of those aged 41-60, and60.9% of those aged 61-80 wereusing diazepam (Table 1). Lorazepamwas the second most commonly usedbenzodiazepine (19.0%) followed byflurazepam (10.7%), triazolam(9.9%), and oxazepam (8.3%). Only3.3% of respondents used chlordiaze-poxide (N= 121, Table 2). The re-spondents who could not recall thename of their tranquillizer or sleepingpill totalled 11.6%. It is possible that7.5% of this group would be identi-fied as users of antihistamines, neuro-leptics or antidepressants if the namesof all drugs were available. This pos-sibility was disregarded in our analy-sis because of the small number of re-spondents (N= 0.65) involved.The benzodiazepines were obtained

by prescription by 84.4% of users,and for 80.6% of users, they wereprescribed by the family physician.Although it is difficult to substantiatethe response in a teaching setting in-volving changing residents, 76.4% ofrespondents stated that they alwaysobtained their benzodiazepines fromthe same physician. Prescription re-newals were common: close to 32%of respondents had had more than fourrenewals, although more than 40% hadhad none (Table 3).The reasons listed by patients at the

time of the study for their use ofbenzodiazepines included insomnia(55.3%), stresses of daily living(26.3%), irritability or nervousness(22.8%), acute anxiety (21.1%), de-pression (18.4%), backache or othermuscle spasm (10.5%), headache(7.9%), night before an event causingunusual stress (7.9%), heart problem(3.6%), jet lag (2.6%), and fun

Table 1Percentage of Benzodiazepine UsersWho Reported Using Diazepam byAge

OtherBenzo-

Age Diazepam diazepine21-40 52.2 47.8 N = 4641-60 66.6 33.4 N = 3361-80 60.9 39.1 N = 23

Total N = 112

Note: Age groups < 20 and > 80 wereomitted because N = 1 in each group.

929

(0.9%) (N= 114, Table 4). However,many (60.7%) of the respondentsusing benzodiazepines had discussedtheir current reasons for the use withtheir doctor. A significant proportionof all users (40%) were taking thebenzodiazepines for a reason otherthan that for which their doctor hadprescribed the drug.

Patients described a variety ofsymptoms experienced when theirmedication was discontinued. Theseare illustrated on Table 5; they in-clude insomnia (68.0%), depression(32.0%), and nervousness or anxiety(32.0%).We found that 11.5% of benzodia-

zepine users considered that their usewas a problem, and 6.1% attemptedto overdose on the drug.

DiscussionAlthough many of the findings in

our study are consistent with those ofother studies, we noted a few dif-ferences. Smart and Adlaf, in ahousehold survey of Ontario adultsfound that 15.4% of the sample re-ported tranquillizer or sleeping-pilluse in the previous year.14 A cross-national comparison study made byBalter found that prevalence of use inthe preceding year varied from 17.6%in Belgium to 7.4% in the Nether-lands. The reported use in the UnitedStates was 12.9%. 15

In our study 24.3% of respondentsclaimed to have used tranquillizers orsleeping pills in the previous year,

Table 2Types of Benzodiazepines Used

Name PercentageDiazepam (Valium 55.4Lorazepam (Ativan) 19.0Flurazepam (Dalmane) 10.7Triazolam (Halcion) 9.9Oxazepam (Serax) 8.3Chlordiazepoxide 3.3Do not recall 11.6

N = 121 (Respondents could identifymore than one drug if applicable.)

Table 3Number of Prescription Renewals inBenzodiazepine Users

Number Percentage> 4 31.83-4 5.61-2 20.6

0 41.1

N = 121

and of these users, 50% reported usewithin the previous two weeks. The1981 Canada Health Survey, whichhad a base of 40,000 people, foundthat 10% reported tranquillizer use inthe previous year, and 4.8% reporteduse in the previous two weeks.16

Although both in our study and inthe Canada Health Survey, approxi-mately 50% of those who had takentranquillizers in the previous year hadalso taken them in the previous twoweeks, the number of respondentswho reported use in the previous yearwas higher in our study than in othersurveys. The difference in these sta-tistics may derive from the sub-groupof the population participating in ourstudy. Because we studied patientsseen in the family-practice depart-ment, those who visited their familyphysician more frequently had ahigher probability of being includedin the study sample.

In our survey we found no statisti-cal difference in any age range be-tween the males and females whoreported using tranquillizers andsleeping pills. This finding contradictsthe findings of several studies whichshow that adult females at eachage level receive more prescriptionsfor psychotropic drugs than domales.'7' 18 Both the 1981 CanadaHealth Survey and Smart and Adlaffound that the sex ratio was 2:1 in fa-vour of females for use of tranquil-lizers and sleeping pills.'16 14 An On-tario study by Bass and another byAnderson showed that men andwomen with identical symptoms at-tending a family-practice clinic re-ceived a different number of prescrip-tions for benzodiazepines. The

Table 4Reasons for Benzodiazepine Use

Reasons PercentageInsomnia 55.3Stresses of daily living 26.3Irritability (Nervousness) 22.8Acute anxiety 21.1Depressi on 18.4Backache or othermuscle spasm 10.5

Headache 7.9Night before anunusual stress 7.9

Heart problem 3.6Jet lag 2.6Fun 0.9

N = 114 (Respondents could identifymore than one reason.)

women received a greater number, al-though there was no difference in theother therapies that they received.'9, 20The most commonly prescribed

benzodiazepine was diazepam (Va-lium); 55.4% of respondents reportedits use. In fact, a higher proportion ofthose who used diazepam were in the41-60 and over-60 age groups, com-pared to other benzodiazepine users.Diazepam has a long half-life. It canbe found in the plasma for a meanperiod of 30 hours, and an actualrange of 20 to 170 hours.21 Elderlypersons metabolize most drugs moreslowly than do younger persons; fur-thermore, there is an increased vol-ume of distribution which increases adrug's half-life.22 23 It has also beenshown that adverse drug reactions in ahospitalized population occurred morefrequently in the elderly age group.24All these factors may warrant either adecrease in dosage or a change to ashorter-acting benzodiazepine.

Divisions of opinion exist about thedose and time necessary to producedependence on benzodiazepines, al-though there is no longer any dis-agreement that dependence of a physi-cal or psychological nature candevelop.25 The major indication ofdependence is the development ofwithdrawal symptoms which havebeen clearly established at therapeuticdose levels.26-28Our study suggests that a large per-

centage of patients report withdrawalsymptoms (21.7%), while Khan in ageneral practice survey of withdrawalsymptoms, reported a 17% inci-dence.29 The most common symptom

Table 5Perceived Withdrawal SymptomsSymptomsInsomniaDepressionNervousness or anxietySweatingMuscular pain and

stiffnessPalpitationsIntolerance of loud noises

or bright lightsTremorNumbness or tinglingBad DreamsHeadacheFaintingDizziness

Percentage68.032.032.024.0

24.016.0

16.012.04.04.04.04.04.0

N = 25 (Respondents could identifymore than one symptom.)

CAN. FAM. PHYSICIAN Vol. 33: APRIL 1987

-

931

of withdrawal was insomnia. Some ofthe withdrawal symptoms can be attri-buted to the underlying anxiety thatprecipitated the use of the drug ini-tially, though such anxiety is not thecause in all instances because in somecases withdrawal symptoms do not re-semble those of anxiety (e.g., abruptweight loss).30 Mellor describes thesymptom groups in diazepam with-drawal. He states that diazepam with-drawal can be distinguished from anx-iety by the symptom groups, thepresence of tremor and myoclonus,and the relief of symptoms by a testdose; a non-benzodiazepine tranquil-lizer such as hydroxyzine, on theother hand, has little effect.31 In ourstudy, it is not possible to determinewhether the symptoms reported indi-cate a physical dependence. However,the symptoms reported are those pre-ceived by the patients on withdrawaland, thus, maintain benzodiazepineuse.

In our study, 6.1% of drug usersclaimed to have overdosed on benzo-diazepines. Many overdoses are im-pulsive acts, and the drugs used reflecttheir availability.32' 33 In a study ofoverdoses made in Toronto emergencydepartments, benzodiazepines ac-counted for 34% of overdoses, andhalf of these were taken in combina-tion with other drugs.34 Thus, al-though there are few reported cases ofdeaths with benzodiazepine overdosealone, its combination with otherdrugs may be lethal.More important may be the psycho-

motor, learning and memory impair-ment that benzodiazepines cancause. 12, 13 One of the first large-scalestudies of motor impairment under"natural" conditions was made in Bri-tain. The prescribed drugs used bypeople in the three months prior to in-jury or death while driving a car, mo-torcycle or bicycle were comparedwith those used by a large number ofcontrols. There was a highly signifi-cant association between the use ofminor tranquillizers and the risk of aserious road accident.35 Binnie con-firmed this finding.36

Depression was the reason given forthe use of benzodiazepines by 18.4%of the participants in our study. It issurprising that they should be used forthis reason, as benzodiazepines canproduce widespread depression of thecentral nervous system.37 An aggrava-tion of depressive symptoms has been

shown in some patients not treatedconcurrently with antidepressants.38Self-destructive ideation leading tosuccessful suicide in a few cases hasbeen attributed to diazepam therapy byseveral authors.39'40

In a study of prisoners at MillhavenPenitentiary, using prisoners as theirown controls, it was found that violentor aggressive incidents were most fre-quent in those using benzodiaze-pines.41 Although there is some evi-dence that benzodiazepines may be ofsome use for depressive disorders inwhich anxiety is a major component,the adverse depressive effects shouldbe kept in mind, and caution should beused in prescribing these drugs in thesecircumstances.25, 42, 43

Several authors have addressed theuse of benzodiazepines in family med-icine, but all have used either a chartreview or a review of prescriptionswritten, and not a survey of the pa-tients themselves. '9, 44-49

Bass reviewed 233 patients in hisdepartment of Family Medicine whopresented with specific emotionalcomplaints. He found that the pre-scribing of minor tranquillizers signifi-cantly related to the frequency of visitsto a physician, previous tranquillizeruse, and female gender. He found too,that 35% of his patients after sixmonths had received at least one pre-scription for minor tranquillizers. 9

In a review of recorded prescrip-tions, Rosser found that prescribing ofbenzodiazepines per capita increasedwith increasing age. Diazepam wasused approximately four times as fre-quently as chlordiazepoxide.45 In asubsequent study of written prescrip-tions Rosser found a significant shift toshorter-acting benzodiazepines: overfour years the use of the shorter-actingdrugs rose to 80% from 16%. He alsofound a 15% decline in prescriptionsfor benzodiazepines between 1978 and

.461982.4As a result of what we found in our

study, we believe that benzodiazepinesare being used too liberally in our fam-ily-practice unit. Benzodiazepineshave a place in relieving acute anxietyin the short term, but because of theirpotential adverse effects, they shouldbe used cautiously in the long term.This caution is especially important inthe elderly, for whom shorter-actingbenzodiazepines are preferable.We must take the responsibility of

offering our patients alternative ways

of dealing with their anxieties. Anxi-ety is, after all, a normal part of life.®

References1. Greenblatt DJ, Shader RI. Benzodiaze-pines (Parts 1 and 2). N Engl J Med 1974;291:1011-5, 1239-43.2. Annual Ontario Prescription Survey.Distribution of psychotropic prescriptionsin Ontario 1970- 1978. Toronto: OntarioMinistry of Health. Data on file at the Ad-diction Research Foundation, Toronto,1979.3. Saskatchewan Joint Commission onDrug Utilization. Report #7. 3475 AlbertStreet, Regina, 1983.4. Tennant FS, Day CM, Ungerleider JT.Screening for drug and alcohol abuse in ageneral medical population. JAMA 1979;242(6):533-5.5. Swanson DW, Weddige RL, MorseRM. Abuse of prescription drugs. MayoClin Proc 1973; 48:359 -67.6. Woody GE, O'Brien CP, Greenstein R.Misuse and abuse of diazepam: an increas-ingly common medical problem. Int J Ad-dict 1975; 10(5):843-8.7. Smart RG, Jarvis GK. Do self-reportstudies of drug use really give dependableresults? Can J Criminol Corrections 1981;23( 1):83 -92.8. Smart RG. Recent studies on the validityand reliability of self-reported drug use1970-74. Can J Criminol Corrections1975; 17:326-33.9. Whitehead PC, Smart RG. Validity andreliability of self-reported drug use. Can JCriminol Corrections 1972; 14:1-7.10. Pemanen K. Validity of survey data onalcohol use. Res Adv in Alcohol and DrugProblems 1974; New York: Wiley.11. Cooperstock R, Parnell P. AddictionResearch Foundation: Report on the preva-lence and patterns of use of psychotropicsubstances. Geneva: World Health Organi-zation, 1980; 15-20.12. Wittenborn JR. Effects of benzodiaze-pines on psychomotor performance. Br JClinic Phacrmatcol, Supplement 1, 1979;7:61-7.13. Kleinknecht RA, Donaldson DA. Re-view of the effects of diazepam on cogni-tive and psychomotor performance, J NervMentDis 1975; 161(6):399-411.14. Smart RG, Adlaf EM. Trends in alco-hol and drug use among Ontario adults: re-port of a household survey 1982. Sub-study No. 1234. Toronto: AddictionResearch Foundation, 1982.

15. Baiter MB, Manheimer DT, MellingerCD, et al. A cross-national comparison ofanti-anxiety/sedative drug use. Curr MedRes Opin 1984; 8 suppl. 4:5-20.16. Health and Welfare Canada, StatisticsCanada. The heallth of Calnadialns: reportof the Calnalda Heallth Survey, catalogue82-538E. Ottawa: Minister of Supply andServices, 1981.

CAN. FAM. PHYSICIAN Vol. 33: APRIL 1987 933

ProdiemIndications:The relief of acute and obstinate constipa-tion, the early stages of more severe simpleconstipation and constipation of the elderlycaused by loss of muscle tone and neuro-musculature reflex. For use in patients beingweaned off harsh laxatives or where astraight bulk forming preparation has failedto provide regular bowel evacuation.Contraindications:The presence of nausea, vomiting, abdomi-nal pain or symptoms of an acute abdomenor fecal impaction.Precautions:For patients with a history of esophageal dis-orders. If mild cramping occurs, dosageshould be reduced.Dosage:Adults -1 to 2 level teaspoonfuls (5 g to 10 g)should be placed in the mouth and swal-lowed unchewed with at least 240 ml of anycool beverage including juice, milk or water.This dose may be taken before breakfastand/or in the evening depending on the con-dition being treated, its severity and indivi-dual responsiveness.Children -Forchildren aged 6tol2years,thedosage is half the adult dose.Supplied:Each 5g (1 level teaspoonful) of dark brownminty-tasting granules contains 2.71g ofpsyllium hydrophilic mucilloid and .62 g ofsenna pod, 1.5mg of sodium, 30 mg of potas-sium and fumishes 3.5 calories.Available in lOOg and 250g canisters.

Prodiem PlainIndications:The relief of simple, chronic andspastic constipation and forconstipation as-sociated with pregnancy, convalescenceand advanced age. For use in special dietslacking in residue fibre and in the manage-ment of constipation associated with irritablebowel syndrome, diverticulitis, hemorrhoidsand anal fissures.Contraindications:The presence of nausea, vomiting, abdomi-nal pain or symptoms of an acute abdomenor fecal impaction.Precautions:For patients with a history of esophagealdisorders.Dosage:Adults -1 to 2 level teaspoonfuls (5g to 10g)should be placed in the mouth and swal-lowed unchewed with at least 240 ml of anycool beverage including juice, milk or water.This dose may be taken before breakfastand/or in the evening, depending on thecon-dition being treated, its severity and indivi-dual responsiveness.Children - For children aged 6 to 12 years,the dosage is half the adult dose.Supplied:Each 5 g (1 level teaspoonful) of light brownminty-tasting granules contains 3.25g ofpsyllium hydrophilic mucilloid, 1.5 mg of so-dium, 30mg of potassium and furnishes 3.5calories.Available in 10Og and 250g canisters.

_IRorer Canada Inc._ Bramalea, Ontario L6T 1C3

17. Chaiton A, Spitzer WO, Roberts RS, etal. Patterns of medical drug use: a commu-nity focus. Can Med Assoc J 1 976;1 14(1):33-7.18. Archer A, Benner M. Women's use ofpsychotropic medication: a community sur-vey, Can Fam Physician 1980; 26:867-71.19. Bass MJ, Baskerville JC. Prescribingof minor tranquillizers for emotional prob-lems in a family practice. Can Med Assoc J1981; 125:1225-6.20. Anderson JE. Prescribing of tranquil-lizers to women and men. Can Med AssocJ 1981; 125:1229-32.21. Sellers EM. Clinical pharmacology andtherapeutics of benzodiazepines, Can MedAssoc J 1978; 1 18:1533-8.22. MacLeod SM, Giles HG, Bengert B, etal. Age and gender-related differences indiazepam pharmacokinetics. J Clin Phar-macol 1979; 19:15-9.23. Cooper AJ. Benzodiazepine update: aguide to rational prescribing. Mod MedCanada 1983; 38(2):209-18.24. Hurwitz N. Predisposing factors in ad-verse reactions to drugs. Br Med J 1969;1(5643):536-9.25. Cooperstock R, Hill J. The effects oftranquillization: benzodiazepine use inCanada. Ottawa: Health and Welfare Can-ada, 1982.26. Pevnick JS, Jasinski DR, HaertzenCA. Abrupt withdrawal from therapeuti-cally administered diazepam. Arch GenPsych 1978; 35:995-8.27. Petursson H, Lader MH. Benzodiaze-pine dependence. Br J Addict 1981;76:133-45.28. Owen RT, Tyrer P. Benzodiazepinedependence: a review of the evidence.,Drugs 1983; 25:385-98.29. Khan A, Homblow AR, Walshe JWB.Benzodiazepine dependence: a generalpractice survey. N Z Med J 1981; 94:19-21.30. Benzodiazepine withdrawal, Lancet1979; 1(8109):196.31. Mellor CS, Jain VK. Diazepam with-drawal syndrome: its prolonged and chang-ing nature, Can Med Assoc J 1982;127:1093 -6.

32. Proudfoot AT, Park J. Changing pat-tern of drugs used for self-poisoning. BrMedJ 1978; 1:90-3.

33. Busto U, Kaplan HL, Sellers EM. Ben-zodiazepine associated emergencies inToronto. Am J Psych 1980; 137(2):224-7.

34. Sellers EM, Marshman JA, KaplanHL, et al. Acute and chronic drug abuseemergencies in Metropolitan Toronto. Int JAddict 1981; 16(2):283-303.35. Skegg DCC, Richards SM, Doll R.Minor tranquillizers and road accidents. BrMedJ 1979; 1(6168):917-9.36. Binnie GAC. Psychotropic drugs andaccidents in a general practice. Br Med J1983; 287:1349-50.37. Goodman Gilman A, Goodman LS,Gilman A. The pharmacological basis oftherapeutics. 6th ed. New York: MacMil-lan Publishing 1980.38. Bulletin of the World Health Organiza-tion, 1983; 61(4):551-62.39. Ryan HF, Merrill FB, Scott GE. In-crease in suicidal thoughts and tendenciesassociated with diazepam therapy. JAMA1968; 203:1137-9.40. Hall RCW, Joffe JR. Aberrant re-sponse to diazepam: a new syndrome. Am JPsych 1972; 129:738 -42.41. Workman DG, Cunningham DG. Ef-fect of psychotropic drugs on aggression ina prison setting. Can Fam Physician 1975;21(1 1):63-6.42. Schuster CR, Bergman J, Hartel CR.Assessing the impact on public health, in-dividual deficit and organ system damageassociated with psychoactive substanceuse. Rockville, MD: National Institute onDrug Abuse, International Activities Pro-gram, 1980; p. 76.43. Castano GB, Conti L. Some consider-ations on the role of benzodiazepines in thetreatment of depression. Br J Clin Pharma-col 1981; 11:23s-9s.44. Rosser WW, Simms JG, Patten DW, etal. Improving benzodiazepine prescri-bing in family practice through review andeducation. Can Med Assoc J 1981; 124:147- 54.45. Rosser WW. Patterns of benzodiaze-pine usage in a family medicine centre.Can Faim Physicians 1980; 25:718- 20.

46. Rosser WW. Benzodiazepines: part oflifestyle in the 1980's. Can Fam Physician1984; 30: 193- 8.47. Hasday JD, Karch FE. Benzodiazepineprescribing in a family medicine centre.JAMA 1981; 246(12):1321-5.48. Sandifer MG, Stolldorf D. Psycho-tropic drug prescribing in a family medi-cine residency program. J Fam Pract 1980;11(7):1077- 80.49. Seifert RD, Clarens RD, SorensenMD, et al. Benzodiazepine utilization .in afamliy medicine residency program. J FamPract 1982; 15(3):497-500.

CAN. FAM. PHYSICIAN Vol. 33: APRIL 1987