Psychoactive Medication Use in
Intermediate-Care Facility Residents
Mark Beers, MD; Jerry Avorn, MD; Stephen B. Soumerai, ScD; Daniel E. Everitt, MD;
David S. Sherman, RPh; Susanne Salem, MHSA

Despite the large number of elderly patients in nursing homes and the intensity of medication use there, few current data are available on patterns of medication use in this setting. We studied all medication use among 850 residents of 12 representative intermediate-care facilities in Massachusetts. Data on all prescriptions and patterns of actual use were recorded for all patients during one month. On average, residents were prescribed 8.1 medications during the month (interquartile range, 7.4 to 8.8) and actually received 4.7 (range, 4.2 to 5.4) medications during this period. More than half of all residents were receiving a psychoactive medication, with 26% receiving antipsychotic medication. Twenty-eight percent of patients were receiving sedative/hypnotics during the study month, primarily on a scheduled rather than an as-needed basis. Of patients receiving a sedative/hypnotic, 26% (range, 14% to 41%) were taking diphenhydramine hydrochloride, a strongly anticholinergic hypnotic. Of those receiving one of the benzodiazepines, 30% were receiving long-acting drugs, generally not recommended for elderly patients. The typical benzodiazepine dose was equivalent to 7.3 mg per patient per day of diazepam. The most commonly used antidepressant was amitriptyline hydrochloride, the most sedating and anticholinergic antidepressant in common use. These data indicate that despite growing evidence of the risks of psychoactive drug use in elderly patients, the nursing home population studied was exposed to high levels of sedative/hypnotic and antipsychotic drug use. Suboptimal choice of medication within a given class was common, and use of standing vs as-needed orders was often not in keeping with current concepts in geriatric psychopharmacology. Additional research is needed to assess the impact of such drug therapy on cognitive and physical functioning, as well as to determine how best to improve patterns of medication use in this vulnerable population.
(JAMA 1988;260:3016-3020)

From the Program for the Analysis of Clinical Strategies (Drs Beers, Avorn, Soumerai, Everitt, Mr Sherman, and Ms Salem), the Division on Aging (Drs Beers, Avorn, Soumerai, and Everitt and Mr Sherman), and the Department of Social Medicine and Health Policy (Drs
Avorn and Soumerai), Harvard Medical School, Boston;
the Gerontology Division, Department of Medicine, Beth Israel Hospital, Boston (Drs Avorn and Everitt);
and the Geriatric Research, Education and Clinical
Center of Brockton/West Roxbury Veterans Administration Medical Center, Boston (Dr Everitt). Dr Beers is now at the Center for Health Sciences, Department of Medicine, UCLA, Los Angeles.
Reprint requests to the Program for the Analysis of Clinical Strategies, Harvard Medical School, 643 Huntington Ave, Boston, MA 02115 (Dr Avorn).

ON ANY given day, there are more people in nursing home beds than in acute-care hospital beds in the United States. Although nursing home resi-
For editorial comment see p 3054.
dents are among the most intensely
medicated patients,1 little is known
about the details of drug prescription and use in this large population of insti-

tutionalized elderly persons. This arti¬ cle describes a large-scale study of all
medication use in 850 residents of 12
representative Massachusetts nursing
homes. It represents one of the few de¬
tailed surveys of medication use per¬
formed in the last decade with a large cohort of long-term-care residents.
In 1976, the Office of Long-Term Care ofthe Department of Health, Edu¬ cation, and Welfare published one of the first comprehensive studies of medica¬ tion use in such a population, based on a sample of 3458 nursing home patients covered by the Medicaid and Medicare programs.2 The report suggested seri¬ ous problems of overuse and misuse of drugs in this population. Patients were found to be taking an average of 6.1 drugs, and nearly 50% of all residents were prescribed tranquilizers (a catego¬ ry that included antipsychotic medica¬ tions and minor tranquilizers). Approxi¬ mately 12% of all residents were prescribed two or more psychoactive drugs, with antipsychotic medications prescribed much more often than minor tranquilizers. About one third of all patients were prescribed sedative/hyp¬ notics, and 9% were taking antidepressants. Of those receiving antidepressants, 38% were prescribed amitriptyline hydrochloride.
In a more recent study reviewing Medicaid patients residing in nursing homes in Tennessee, Ray et al3 found that 42% ofthose patients received anti¬ psychotic medication. The authors con¬ cluded that their findings provided "epidemiologic evidence suggesting misuse of antipsychotic drugs in nursing
homes." Further accounts of the over¬
use of psychoactive drugs in nursing homes were presented during widely

•

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/9122/ by a Harvard University User on 02/21/2017

reported congressional hearings in
1980.4
Data collected regarding drug use in nursing home patients have relied on computerized databases that focused on Medicaid and Medicare patients. Al¬ though these sampling methods have the advantage of including large num¬ bers of patients, they exclude half of long-term-care patients who pay for nursing home care through their own resources and, thus, may not be repre¬ sentative of the nursing home popula¬ tion in general. In addition, most stud¬
ies have not collected data on actual use
of as-needed (pro re nata [PRN]) or¬ ders, and so have not been able to measure the intensity of use of med¬ ication administered through this com¬
mon nurse-mediated and very variable mechanism. Smaller studies have exam¬
ined drug use in a single nursing home5"7 and may not be generalizable.
During the past decade, considerable
evidence has accumulated and been
published concerning the dangers of polypharmacy in elderly patients.1,8 Rec¬
ommendations have included restrained
use of medications, particularly those affecting the central nervous system,
and the substitution of safer alterna¬
tives for many older drugs. Amitriptyline, the first available tricyclic anti¬ depressant, is highly sedating and anticholinergic, two properties that of¬ ten make it undesirable for elderly pa¬ tients. Newer and safer antidepressants are now available and preferable. Diphenhydramine hydrochloride, an antihistamine with strong sedating properties, has been commonly used to induce sleep in the nursing home set¬ ting, but it too is strongly anti¬ cholinergic and, thus, is not a drug of choice for elderly patients.9 Flurazepam hydrochloride, diazepam, and chlordiazepoxide hydrochloride, the older benzodiazepines with very long half-lives in elderly patients,10"12 have been shown to present the risk of increased toxicity in aged patients; newer, shorter-acting benzodiazepines are generally prefera¬ ble when benzodiazepines are needed.11
Increasing numbers of studies of tox¬ ic reactions to psychoactive drugs in el¬ derly patients have demonstrated asso¬ ciated morbidity, including a high incidence of acute confusional states,13 hip fractures,14 and oversedation. The clinical literature on antipsychotic med¬ ications has failed to show specific bene¬ fit for most nonpsychotic elderly pa¬
tients,15"17 although their toxicity has
been well documented.18 There is good evidence that these drugs have been seriously overused in nursing home patients."
Although the knowledge base of geri-

atric pharmacology has improved and
such information has become more
available, there have been remarkably few large-scale data published on pat¬ terns of medication use, and particular¬ ly on psychoactive drug use, in nursing
homes in recent years; nor has there
been any systematic evaluation of the impact of developments in geriatric re¬ search and teaching on actual practice in long-term-care facilities. This study was designed to explore this important but poorly documented aspect of the
health care system.
METHODS
In developing a targeted educational intervention aimed at improving drug use in nursing homes, we studied base¬
line patterns of medication use among residents of 12 representative interme¬ diate-care facilities (ICFs) in Massachu¬
setts. A complete list of all ICFs in Mas¬
sachusetts was obtained from the
Massachusetts Department of Public Health. Because the eventual goal was to provide an educational intervention,
homes with fewer than 40 ICF-level res¬ idents were excluded. We identified 339 homes in Massachusetts that met this
criterion; those in eastern Massachu¬ setts were considered for further study. Tb ensure that the homes represented typical geriatric facilities rather than those caring for deinstitutionalized mental hospital patients, homes were eliminated if they had greater than 20% of their residents admitted from inpatient psychiatric hospitals. Homes with active nurse practitioner prescribing were likewise excluded, as were facili¬ ties with preexisting relationships be¬ tween the authors and consulting phar¬ macists or physicians in the homes. The remaining 26 homes were given infor¬ mation about the drug-use interven¬ tion. Ten homes declined to participate in a randomized trial, primarily because of changes in ownership or administra¬ tion or labor disputes. Of the remaining 16 homes, 12 were enrolled in the study
and four were retained as alternates on
the basis of geographic distribution. A pilot survey of drug use over a twoweek period revealed no difference in average use of antipsychotic and seda¬ tive drugs between the study homes and
the alternate homes.
Data Collection and Analysis
An interactive data collection pro¬
gram was written to guide research as¬ sistants in capturing medication-use data on a portable computer. In each facility, all medication data were re¬
corded for all residents for an entire
month. For all drugs prescribed, the program prompted for data regarding

medication name (trade or generic), current regimen, dosage changes, start and stop orders, route of administra¬ tion, and other changes in regimen dur¬ ing the month. A PRN use module made it possible to document the doses of PRN medications actually administered each day and sum them over the month. The software combined data on a given drug recorded as either its generic or trade name, as well as summing sched¬ uled doses and PRN doses actually ad¬ ministered. Additional programming made it possible to convert regimens into equivalent milligrams of a proto¬ type compound for each group (eg, chlorpromazine hydrochloride or diazepam), as well as analyze use by broad categories of drug (eg, all antipsychotic medication). In addition, the program also recorded the name of each patient's physician, the patient's age, sex, date of admission, payment source, current di¬
agnoses, and the date and reason for
hospitalizations that occurred during the study period.
Data were collected regarding all pa¬ tients residing in the 12 study homes
(n = 850). More than half (n = 441) of the
study subjects were Medicaid patients. Because nursing homes rather than pa¬ tients were sampled, analyses were performed with the nursing home as the unit of analysis. In addition, to control for the potential influence of outlier fa¬
cilities and to make our results more
generalizable to other ICFs, we mea¬ sured drug utilization as the median and interquartile range of the measures of drug use. This method is more conser¬ vative when there are significant interhome differences in specific usage cate¬ gories. To avoid considering drugs written for PRN use but not given, medications are described as having been "used" only if they were adminis¬ tered to the patient on at least five days in the study month.
Nursing Homes Sampled
Tb ensure that the homes studied
were representative of ICFs through¬ out the state, we compared all available
characteristics of each of the 339 Massa¬ chusetts ICFs that had 40 or more beds
with those of the study homes, as indi¬ cated in Table 1. The study homes were found to be quite similar to the universe of homes regarding patterns of owner¬ ship and management; as noted in Table 1, about 70% of Massachusetts homes
with intermediate-care beds are owned
and operated by for-profit corporations. Additional data for nonstudy homes
were obtained from the Massachusetts
Department of Public Health. Table 2
shows that the homes studied were com¬
parable with all ICFs of similar size

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/9122/ by a Harvard University User on 02/21/2017

within 96 km (60 miles) of Boston with
respect to the proportion of residents on Medicaid, under age 65 years, and over
age 85 years.
RESULTS
Demographics and Overall Drug Use The total sample consisted of 850 pa¬
tients in 12 facilities; the median home had 70 patients, with an interquartile range of 60 to 80 patients. Unless other¬ wise stated, all results reported herein are the median of the 12 nursing home proportions or means; the interquartile ranges are indicated in parentheses.
As in most nursing home populations, the majority of residents, or 84% (range, 76% to 87%), were female; their mean age was 85 years (range, 84 to 96 years). Sixteen percent of patients were
male, with a mean age of 82 years
(range, 76 to 84 years). Of all drug or¬
ders written, 48% (n = 7168 orders;
range, 37% to 52%) were written for PRN use.
The average number of different
medications ordered per patient during the month was 8.1 (range, 7.4 to 8.8). One fifth of all prescriptions were for over-the-counter analgesics and cathar¬ tics. Analysis of drugs actually given at least once each month and those given on at least five days each month demon¬ strated that patients received 5.5 (range, 4.7 to 6.0) or 4.7 (range, 4.2 to
5.4) different medications on average,
respectively. Since these estimates in¬ cluded occasional short-term therapies (eg, antibiotics), the average number of concurrent drugs per day would be slightly lower. Only 35 patients (4%)
had no medication orders. There was no
significant difference in overall drug use
between men and women.
Psychoactive Drug Use Nearly two thirds of the residents
(65%; range, 58% to 71%) had orders written for one or more psychoactive medications, and 53% (range, 48% to
59%) actually used psychoactive drugs on five or more days during the study
month (Table 3). These included seda¬
tive/hypnotics (including benzodiazepine tranquilizers), antipsychotics, and antidepressants. Barbiturates, antiseizure medications, antiparkinsonian drugs, lithium carbonate, and antihypertensives with known psychoactive properties were not included. Fortythree percent of orders (range, 35% to 48%) written for psychoactive drugs were written for PRN use, but only 20%
of those PRN orders were used. There was no difference between Medicaid
and non-Medicaid patients in overall psychoactive drug use or in any subcategoryofdruguse.

Table 1.—Characteristics of Study Homes and All Massachusetts Nursing Homes With 40 or More Intermediate-Care Facility Beds

Characteristics
Mean No. of intermediate-care facility beds
Mean No. of total beds
Ownership, % Individual or partnership Nonprofit For-profit corporation

Study Homes (n = 12)
75 101
25
67

All Massachusetts
Homes (N=339)
66 108
12 72

Table 2.—Patient Characteristics (Means) in Study Homes and All Massachusetts Nursing Homes With 60 or More Intermediate-Care Facility Beds*

Residents
On Medicaid
From psychiatric institutions <Age 65 y >Age 85 y

Study Homes, % (n = 12)
62 (SD = 26)
4 (SD = 6) 49 (SD = 10)

All Massachusetts
Homes, %(n = 109) 73 (SD = 23) 3 (SD = 9) 8 (SD = 12) 43 (SD = 14)

•Study homes are compared with all Massachusetts nursing homes with 60 or more intermediate-care facility beds within 96 km (60 miles) of Boston; based on data provided by the Massachusetts Department of Public Health.

Table 3.—Patterns of Prescription and Actual Use of Psychoactive Medications* in 12 Nursing Homes (850 Patients)

No. of
Psychoactive Drugs
£3

Medications Ordered,
Median %
(Interquartile Range)
34.9 (29.1-41.5)
38.8 (32.4-41.9)
20.6 (19.0-26.3)
6.6 (1.3-8.7)

Medications Used.f
Median %
(Interquartile Range)
47.1 (40.9-52.4)
32.9 (29.9-39.3)
15.2 (12.8-20.1)
2.6 (0.0-4.3)

•Includes antipsychotics, sedative/hypnotics, and antidepressants only. tPatient received drug at least five days in study month.

Table 4.—Psychoactive Drug Use by Residents With Specific Diagnoses

No. (%) of Patients

Diagnosis

Receiving Any Receiving Receiving

Receiving

Psychoactive Antipsychotic Antidepressant Sedative/Hypnotic

AM Patients Medication Medication Medication

Medication

Organic brain syndrome

150

94 (63)

58 (39)

8 (5)

57 (38)

Alzheimer's disease

228

164 (72)

99 (43)

33 (14)

94 (41)

Schizophrenia/psychosis

42 (82)

36 (71)

7 (14)

15 (29)

Depression

116

96 (83)

46 (40)

47 (40)

56 (48)

About one fifth of patients (21%) had orders for two psychoactive medica¬ tions during the study month; 7% of pa¬
tients had orders for three or more psy¬ choactive medications. Actual use was
somewhat less (Table 3).
Antipsychotic Drug Use One third of patients (270) had orders
written for antipsychotic medications. During the study month, 26% (range, 20.9% to 27.8%) of all residents actually used a neuroleptic on five or more days. Most of these patients had a combina¬ tion of scheduled neuroleptic prescrip¬
tions with additional PRN orders. Of all
orders for antipsychotic medication written, 42% (range, 26.6% to 45.5%) were for PRN use, and 58% for regular-

ly scheduled regimens. However, only 11% (range, 0% to 17%) of PRN orders were activated during the study month. Of those patients taking antipsychotics, the median of the average daily dose by home was 65 mg per patient per day (range, 42 to 94 mg per patient per day) in milligram equivalents of chlorproma-
zine20,21; for individual patients, the high¬
est dose observed was 2000 mg per day. Only 36 patients receiving antipsychotic medications carried a diagnosis of schizophrenia or other psychosis; Alz¬ heimer's disease, organic brain syn¬ drome, and depression were each more common diagnoses in these patients (Table 4).
Haloperidol was the most commonly prescribed antipsychotic medication

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/9122/ by a Harvard University User on 02/21/2017

and was given to 43% (range, 26% to 52%) of those receiving antipsychotics and to 10% ofthe total sample. Thioridazine hydrochloride was the second most commonly prescribed antipsychotic, given to 33% (range, 25% to 50%) of those receiving antipsychotics and to 7% (range, 6% to 14%) of the total population.
Sedative/Hypnotic Drug Use Forty percent (range, 34% to 51%) of
patients were prescribed sedative/hyp¬ notic drugs, and 28% (range, 23% to 41%) actually used these drugs for at least five days during the study month. Of these, 82% were receiving such a l medication on a regularly scheduled ba¬ sis. Most orders for minor tranquilizers and hypnotics were for regularly sched¬ uled use; while many patients had both scheduled and PRN orders, just 13% of patients prescribed sedative/hypnotics had only PRN orders. Twenty-three percent (range, 19% to 31%) of PRN orders for drugs in this class were used.
About one fourth (26%; range, 14% to
41%) of all patients receiving a seda¬ tive/hypnotic were prescribed diphenhydramine. Twenty percent (range, 11% to 25%) of all patients and 63% (range, 48% to 74%) of those receiving a sedative received a benzodiazepine. As with sedative/hypnotics in general, the majority ofbenzodiazepine orders were prescribed as a standing order (87%; range, 64% to 100%). Of those receiving benzodiazepines, 51 patients (30%; range, 15% to 39%) were taking the long-acting drugs flurazepam, diazepam, or chlordiazepoxide; for 67% of these patients, the regimen was a stand¬ ing order to be administered at least daily. The average consumption of ben¬ zodiazepines was 7.3 mg per patient per day (range, 6.7 to 9.5 mg per patient per day) in milligram equivalents of diazepam.20'21'22
Antidepressant Drug Use Fourteen percent (range, 10% to
17%) of patients used antidepressants.
These orders were written almost ex¬
clusively as regularly scheduled regi¬ mens, though two patients received the drug as a PRN bedtime hypnotic. There
was no difference between men and
women in overall use of antidepres¬ sants. While 47 (39%) ofthe 119 patients receiving antidepressants had a diagno¬ sis of depression recorded, the majority
(61%) did not (Table 4).
Of those receiving antidepressants, the largest group (26%; range, 18% to 35%) received amitriptyline. The other antidepressants used were desipramine hydrochloride, doxepin hydrochloride, trazodone hydrochloride, imipramine

hydrochloride, nortriptyline hydrochlo¬ ride, and maprotiline hydrochloride.
COMMENT
These data suggest that despite a growing literature on the risks of polypharmacy in elderly patients, total drug use has remained high in this sample of nursing homes. The Department of Health, Education, and Welfare Office of Long-Term Care study of 1976 found that nursing home residents had pre¬ scriptions for an average of 6.1 concur¬ rent medications per patient on the day of the survey. On average, patients in our study homes were prescribed 8.1 different medications during a month. Of these orders, an average of 4.7 drugs were actually used on five or more days by these residents. Though drugrelated adverse effects are well publi¬ cized and the seriousness ofsuch compli¬ cations are well known,6,22"25 those chronically institutionalized elderly pa¬
tients who are least able to cope with
drug side effects appear to be at contin¬ ued risk for experiencing them.
The nursing homes from which these findings come are reasonably represen¬
tative of homes in Massachusetts and
the Northeast in general. However, there are potential limits to the generalizability of these findings. We studied only intermediate-care patients. These are patients in need of less care than
those admitted to skilled-nursing facili¬
ties, a selection criterion that would, if anything, tend toward lower medica¬ tion use. Facilities with high propor¬ tions of deinstitutionalized psychiatric patients were not included in the study. On the other hand, the inclusion of all short-term drugs given during the study month (eg, antibiotics) somewhat
increases the estimated number of
drugs prescribed in comparison with the Department of Health, Education, and Welfare study, which was based on a single-day survey.
The use ofPRN medications presents
special problems in nursing homes. Such orders place the responsibility of deciding to implement drug therapy on the nursing staff. It has been previously
shown that extensive use of PRN medi¬
cation may adversely affect patient care26,27 if monitoring is poor and if the nursing staff is not adequately trained in the use and side effects of such drugs. Some authors have also suggested an
increased cost to the institution when PRN orders are overused.27 The data
presented herein document a very high level of PRN ordering of psychoactive medications, of concern in a level of care (ICF) with fewer requirements for the presence of skilled nursing surveillance. Nearly half of all orders for psychoac-

tive drugs were written for PRN use. In practice, nurses initiated use of only
about one fifth of these PRN orders.
Ironically, in those situations in which PRN use is generally preferred over regularly scheduled use (eg, as seda¬ tive/hypnotics), the data indicate that PRN use accounts for only a small frac¬ tion of drug administration.
In this study, over half of all patients were receiving psychoactive drugs; about one in four patients was taking two or more psychoactive medicines. Reliance on scheduled regimens of psychoactive medicines indicates that these drugs are not used transiently for periods of special need. Instead, this pattern of use is compatible with the
concept of sedation as "chemical
restraint."28
In many cases, the best drug within a therapeutic class was underutilized. This is exemplified by the high use of diphenhydramine as a hypnotic, a drug
with strong anticholinergic properties9
that can lead to confusion and worsening of dementia, constipation, and urinary retention.18,29 Likewise, nearly one third of those who received benzodiazepines were prescribed older, long-acting drugs that are known to accumulate in elderly patients12,30,31 and are associated with a high frequency of toxic reactions
in this age group.32 These findings differ from that of James,33 who found a level
of use of hypnotics much lower than this study did, and lower than that reported
in the rest of the literature. He attrib¬
uted this difference to more vigorous drug utilization review in the five facili¬ ties that agreed to participate in his
survey.
Amitriptyline, the antidepressant with the highest anticholinergic and se¬ dating properties,10,34 and least favored for geriatric patients,35 was given to about one fourth ofthose taking an anti¬ depressant, despite the availability of medications less toxic for elderly pa¬ tients, such as the secondary amines. (This percentage of use is comparable with that reported in 1976, although the 1976 analysis ofthe Office of Long-Term Care considered perphenazine as an antidepressant, somewhat lowering the percentage of amitriptyline use.)
The usefulness of antipsychotic medi¬ cations in nonpsychotic, elderly pa¬ tients has been questioned,14"16 although these drugs represent a common ap¬ proach to the very difficult problem of managing disruptive behavior in de¬ mented patients who may be a risk to themselves or others. The high frequen¬ cy of toxic reactions to these drugs is well documented, with many older pa¬ tients who take them experiencing orthostatic hypotension, Parkinson's syn-

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/9122/ by a Harvard University User on 02/21/2017

drome, tardive dyskinesia, akathisia, worsened confusion, dry mouth, consti¬ pation, oversedation, and urinary in¬ continence.1832,3637 Although the total use of these drugs in this study is slightly lower than that in earlier reports, previ¬ ous studies included many chronically institutionalized psychotic patients, es¬
timated to make up at least 8% of all
nursing home residents3839; these pa¬ tients were in large part excluded from our sample.
These findings indicate that there is

need for considerably more effort to im¬ prove the use of psychoactive drugs in institutionalized elderly patients. Psy¬
choactive medications continue to be
used with high frequency and selection of drug is often suboptimal. Further analysis is required of the factors that influence drug use in nursing homes. Many of these may represent problems of inadequate staffing as well as infor¬ mation deficits among prescribers. There is a pressing need to study the clinical impact of such high levels of psy-

choactive drug use in this population to determine whether more judicious use
of medications would result in reduced
levels of disability in these already im¬ paired elderly patients.
This work was funded by a grant from the John
A. Hartford Foundation.
The authors are grateful to Susanne Bellavance
for administrative assistance in the coordination of
this research, Christina Casteris, MS, for analysis of the data, Sharon Hawley for data collection, Gail Gregg for consultation on software development, and Steven Rappaport, MD, for his assistance in the early phase of this project.

References
1. Rowe JW, Besdine RW (eds): Health and Disease in Old Age. Boston, Little Brown & Co Inc, 1982, p 39. 2. Physician's Drug Prescribing Patterns in Skilled Nursing Facilities, publication (OS)76\x=req-\ 50050. Office of Long-Term Care, US Dept of Health, Education, and Welfare, 1976.
3. Ray WA, Federspiel CF, Schaffner W: A study ofantipsychotic drug use in nursing homes: Epidemiologic evidence suggesting misuse. Am J Public Health 1980;70:485-491. 4. Hearings Before the Select Committee on Aging, House of Representatives, 96th Cong, 2nd sess
(1980).
5. Ingman S, Lawson I, Pierpaoli PG, et al: A survey of the prescribing and administration of drugs in a long-term care institution for the elderly.
J Am Geriatr Soc 1975;23:309-316. 6. Kalchthler T, Coccaro E, Lichtiger S: Incidence of polypharmacy in a long-term care facility. J Am Geriatr Soc1977;25:308-313.
7. Bergman HD: Prescription of drugs in a nursing home. Drug Intell Clin Pharm 1975;9:365-368. 8. Avorn J, Lamy P, Vestal R: Prescribing for the elderly\p=m-\safely.Patient Care 1982;16:14-62. 9. Regesteinn QR: Insomnia and sleep disturbances in the aged. J Geriatr Psychiatry 1980;
13:153-171.
10. Lippman S: Drug therapy for depression in the elderly. Postgrad Med 1983;73:159-173. 11. Allen RM: Tranquilizers and sedative/hypnotics: Appropriate use in the elderly. Geriatrics
1986;41:75-88. 12. Meyer BR: Benzodiazepines in the elderly. Med Clin North Am 1982;66:1017-1034. 13. Smith L: Aging and drug response: Research priorities and future directions, in Jarvik L, Greenblatt DJ, Harman D, et al (eds): Clinical

Pharmacology and the Aged Patient. New York, Raven Press, 1981, pp 207-215. 14. Ray WA, Griffin MR, Schaffner W, et al: Psychotropic drug use and the risk of hip fracture. N Engl J Med 1987;316:363-369. 15. Helms P: Efficacy of antipsychotics in the treatment of the behavioral complications of de-
mentia: A review of the literature. J Am Geriatr
Soc 1985;33:206-209. 16. Barnes R: Efficacy of antipsychotic medication in behaviorally disturbed dementia patients. Am J Psychiatry 1982;139:1170-1174. 17. Raskin D: Antipsychotic medications and the
elderly. J Clin Psychiatry 1985;46:136-140. 18. Jenike M: Psychosis, rage and violence, in Handbook of Geriatric Psychopharmacology. Littleton, Mass, PSG Publishing Co Inc, 1985, pp 22\x=req-\
29.
19. Butler R: The overuse of tranquilizers in older patients. Int J Aging Hum Dev 1976;7:185-187. 20. AMA Drug Evaluations, ed 5. Chicago, American Medical Association, 1986, p 112. 21. Goodman A, Gilman L, Gilman A (eds): The Pharmacological Basis of Therapeutics. New York, MacMillan Publishing Co Inc, 1980, p 350. 22. Everitt DE, Avorn J: Drug prescribing for the elderly. Arch Intern Med 1986;146:2393-2396. 23. Lang L, Kabat H: Drug interactions in nursing home patient prescriptions. JAMA 1970;10:674\x=req-\
677.
24. Cooper J, Wellins I, Fish KH Jr, et al: A seven\x=req-\ nursing home study: Frequency of potential drug\x=req-\ drug interactions. J Am Pharm Assn 1975;15:24\x=req-\ 27,31. 25. Lavanturier MD, Talley RB, Hefner DL, et al: Drug utilization and potential drug-drug interactions. J Am Pharm Assn 1976;16:77-81. 26. Howard J, Strong K Sr, Strong KE Jr: Medica-

tion procedures in a nursing home: Abuse of PRN orders. J Am Geriatr Soc 1977;25:83-84. 27. Brown C, Kirk K: PRN drug utilization in Indiana long-term care facilities. Hosp Pharm 1984;19:692-696.
28. Covert A: The use of mechanical and chemical
restraints in nursing homes. J Am Geriatr Soc 1977;25:85-89. 29. Resnick N: Urinary incontinence in the elderly. Medical Grand Rounds 1984;3:281-290. 30. Lader M, Petursson H: Rational use of anxiolytic/sedative drugs. Drugs 1983;25:514-528. 31. Greenblatt D, Sellers E, Shader R: Drug therapy: Drug disposition in old age. N Engl J Med 1982;306:1081-1088. 32. Salzman C: Key concepts in geriatric psychopharmacology: Altered pharmacokinetics and polypharmacy. Psychiatr Clin North Am 1982;5:181\x=req-\
190.
33. James DS: Survey of hypnotic drug use in nursing homes. J Am Geriatr Soc 1985;33:436-439. 34. Glassman A, Carino J: Adverse effects of tricyclic anti-depressants: Focus on the elderly. Adv Biochem Psychopharmacol 1984;39:391-398. 35. Georgotos A, Cooper T: The treatment of affective disorders in the elderly. Psychopharmacol Bull 1983;19:226-237. 36. Baldessarini R: Antipsychotic agents, in Chemotherapy in Psychiatry. Cambridge, Mass, Harvard University Press, 1977, p 44. 37. Vestal R: Drug use in the elderly: A review of the problems and special considerations. Drugs 1978;16:358-382. 38. Rovner B, Rabins P: Mental illness among nursing home patients. Hosp Community Psychiatry 1985;36:119-120. 39. Carling PJ: Nursing homes and chronic mental patients. Schizophr Bull 1981;7:574-579.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/9122/ by a Harvard University User on 02/21/2017