The American Journal of Geriatric Pharmacotherapy E.R. Hajar et al.

Adverse Drug Reaction Risk Factors in Older Outpatients

Emily R. Hajjar, PharmD, 1,2Joseph T. Hanlon, PharmD, MS,1-3 Margaret 6. At-k, PhD,’
Catherine I. Lindblad, PharmD,‘,* Carl F. Pieper, DrPH,4s5Richard J. Sloane, MPH,4
Christine M. Ruby, PharmD?r6-* and Kenneth E. Schmader, MD4r6s7

‘Institute for the Study of Geriatric Pharmacotherapy and the Department of Experimental and Clinical Pharmacology,
College of Pharmacy, University of Minnesota, %eterans Affairs Medical Centel; 3Division of Health Services Research and
Policy, School of Public Health, University of Minnesota, Minneapolis, Minnesota, 4Center for the Study of Aging and Human
Development, and 5Department of Biostatistics and Bioinformatics, Duke University Medical Center, “Veterans Affairs
Medical Center, Durham, and 7Department of Medicine, Division of Geriatrics, and 8School of Pharmacy, University of North
Carolina, Chapel Hill, North Carolina

ABSTRACT received, means and 95% CIs were calculated. Con-

Background: Adverse drug reactions (ADRs) are
common in older (age 265 years) outpatients (preva-
lence, 5%-35%), but there is no consensus on factors
that put these patients at high risk for ADRs. Identifying
a uniform set of risk factors would be helpful to develop
risk models for ADRs for older outpatients and to imple-
ment targeted interventions for those patients at high
risk for ADRs.
Objective: The aim of this study was to identify
potential risk factors for ADRs in older outpatients
through a survey of geriatric experts and to determine
their prevalence.
Methods: A comprehensive literature search was
conducted to find published articles on ADRs in older
patients. Forty-four potential risk factors were identified
through the literature search and 6 additional factors
were suggested by the expert panel. Through a modi-
fied 2-round survey, based on the Delphi consensus
method, of an expert panel of 5 physicians and 5 phar-
macists, the probability that each of these 50 potential
factors could contribute independently to placing an
older outpatient at high risk for an ADR was rated on a
5-point Likert scale. After the survey responses were
sensus was defined as a lower 95% confidence limit 24.0.
Potential risk factors that reached consensus were then
applied to a sample of older outpatients to determine
their prevalence.
Results: After 2 rounds, the expert panel reached con-
sensus on 21 factors, including 12 medication-related
factors and 9 patient characteristics. The most prevalent
medication-related risk factors were opioid analgesics;
warfarin; non-acetylsalicylic acid, non-cyclooxygenase-2
nonsteroidal anti-inflammatory drugs; anticholinergics;
and benzodiazepines. The most prevalent patient charac-
teristics included polypharmacy, multiple chronic medical
problems, prior ADR, and dementia.
Conclusions: An expert panel was able to reach a
consensus on potential risk factors that increase the risk
for ADRs in older outpatients. Many risk factors were
common in a sample of older outpatients. Future
research is needed to determine the predictive validity of
these risk factors for ADRs in older outpatients. (Am J
Geriaty Plmrmacother. 2003;1:82-89) Copyright 0 2003
Excerpta Medica, Inc.
Key words: aged, drug utilization, adverse drug reac-
tions, quality assurance, risk.

INTRODUCTION ADRs are a major threat to the health-related quality of

Older (age 265 years) outpatients often consume many life of older outpatients and account for the expenditure
different medications to treat acute and chronic condi- of billions of health care dollars each year in the United
tions. Avid drug use places older outpatients at risk for States.2Jj>7
adverse drug reactions (ADRs). The World Health A few studies3-5 have analyzed risk factors for ADRs in
Organization defines an ADR as “a reaction that is nox- older outpatients. However, the measured risk factors
ious and unintended and that occurs at doses normally were not consistent across these studies. Identifying a
used in humans for prophylaxis, diagnosis, or therapy.“’
ADRs occur commonly in older outpatients, with a Accepted for publication November 24, 2003.
reported prevalence ranging between 5% and 35%.2-s Copyright 0 2003 Excerpta Medica, Inc. 1543.5946/03/$19.00
J

82 December 2003 Volume 1 l Number 2

 E.R. Hajjar et al. The American Journal of Geriatric Pharmacotherapy

uniform set of risk factors would be helpful to develop sensus method27 of providing feedback information on
risk models for ADRs for older outpatients and to the responses of the group was conducted.8,27 The list of
implement targeted interventions for those patients at potential risk factors was sent to a panel of 5 physicians
high risk for ADRs. To meet this goal, one approach is and 5 pharmacists (see Acknowledgments) who were
to survey an expert panel to identify potential risk factors selected because of their work in the field of geriatrics and
for ADRs, measure the prevalence of the identified risk ADRs in older outpatients. During round 1, the panel was
factors in a population, and then determine the predic- asked to rate the probability that each of the 44 potential
tive validity of prevalent factors for ADRs. Fouts et al* factors could contribute independently to placing an older
conducted a survey of a 15-member expert panel to outpatient at high risk for an ADR on a 5point Likert
establish risk factors for drug-related problems (includ- scale (scale: 1 = definitely not a risk, 2 = doubtful as a risk, 3 =
ing ADRs) in older residents of long-term care facilities equivocal-may or may not be a risk, 4 = probably a risk,
and to determine their prevalence among a sample of and 5 = definitely a risk). Panelists also were asked to sug-
those residents. Koecheler et al9 also used a survey gest any additional risk factors not already included. The
method to elicit the opinions of an S-member expert distributional properties for round 1 of the survey data
panel of ambulatory care pharmacists to determine risk were summarized using basic descriptive statistics (eg,
factors for drug-related adverse outcomes and their means and 95% CIs) for the ratings of each factor. Risk
prevalence in an ambulatory population. These studies factors were classified into 3 categories: accepted, rejected,
were not restricted to older outpatients. or equivocal8 Based on previously published work8 by
The purpose of this study was to identify potential risk our research group, accepted risk factors were those with a
factors for ADRs in older outpatients through a survey lower 95% confidence limit 24.0. Rejected risk factors
of a panel of geriatric experts and to determine the were those with an upper 95% confidence limit ~3.0. All
prevalence of these risk factors in a sample of older out- other risk factors were categorized as equivocal,
patients. In round 2, the survey again was sent to each of the
10 panelists. It included risk factors that were deter-
METHODS mined to be equivocal, and provided each panelist with
We modified methods previously used by Fouts et al8 that the mean group response and his or her own response
identified risk factors for drug-related problems in elder- from round 1. Additional risk factors were added to the
ly patients in long-term care facilities. With the help of a survey in round 2 if 22 panelists had suggested them in
medical librarian, we conducted an extensive literature round 1. Accepted risk factors for round 2 were those
review focusing on ADRs in older outpatients using a with a lower 95% confidence limit 24.0 after the survey
search of the MEDLINE and International Pharma- process was completed; all other factors were rejected.
ceutical Abstracts databases for English-language articles The prevalence of the consensus risk factors was mea-
(key terms: adverse drtid reactions, aged, and epidemiolo- sured in a random sample of older outpatients from the
8% years: 19662002). We also conducted a manual Geriatric Evaluation and Management (GEM) Drug
search of the reference lists from identified articles and Study.28l29 Details of the GEM Drug Study design and
the authors’ article files, book chapters, and reviews to methodology are presented elsewhere. Briefly, this study
identify additional articles. We identified 4 articles3-5310 was conducted at 11 Veterans Affairs (VA) Medical Centers
that focused on ADRs in older outpatients and 14 arti- and examined patients at hospital discharge. Patients
cles11-24 that reported information on ADRs in elderly were eligible if they were aged r65 years, had been hos-
patients that led to hospitalization. We also considered pitalized on a medical or surgical ward for ~48 hours,
articles that assessedADRs in adult outpatients of all ages and were considered by the study investigators, based on
and other survey studies conducted to determine risk fac- the inciusion criteria, to be frail. The sample for this
tors for ADRs in various populations.8,9,25>26 Two mem- study included 808 outpatients at discharge from the
bers of the clinical investigative team (E.R.H., J.T.H.) hospital. The primary data source was medical records.
reviewed the identified articles and compiled a prelimi- Medications were organized according to the VA
nary list of potential risk factors. A final list of 44 poten- Medication Classification System.30
tial risk factors was decided on in a meeting with the rest
of the clinical investigative team (Table I). Statistical Analysis
To reach consensus on the potential risk factors, a 2- Descriptive statistics (eg, percentage and mean [SD])
round modified Delphi survey based on the Delphi con- were used to summarize the risk factors. SAS software

I 83
The American Journal of Geriatric Pbarmacotherapy E.R. Hajar et al.

version 8.2 (SAS Institute Inc., Cary, North
was used for the statistical analysis.
RESULTS
Round 1
The response rate for the 10 panelists was 100% in
round 1 of the survey. Two potential patient risk factors
were eliminated based on low ratings-married and white
race, with mean values (95% CI) of 1.8 (1.4-2.2) and 1.9
(1.4-2.4), respectively (Table I). Sixteen risk factors
were accepted by consensus (medication-related factors
[medication class]: anticholinergics; antipsychotics; ben-
zodiazepines; corticosteroids; non-acetylsalicylic acid, non-
cyclooxygenase-2 nonsteroidal anti-inflammatory drugs
[non-ASA, non-COX-2 NSAIDs]; opioid analgesics; and
Carolina) sedative/hypnotics; medication-related factors [specific
medication]: chlorpropamide, lithium salts, theophylline
salts, and warfarin salts; patient risk factors: demen-
tia, multiple chronic medical problems, polypharmacy,
prior ADR, and renal insufficiency [creatinine clearance
~50 mL/min]).
Three additional patient risk factors were suggested in
round 1 by 22 panelists and also subsequently were
included in round 2 (multiple pharmacies used, use of
self-medication with nonprescription medication, and
poor self-reported health). Two factors included in
round 1 were thought to be difficult to interpret by pan-
elists (antidepressants and multiple provider visits).
These factors were further clarified and subclassified into
more discrete terms for round 2 (selective serotonin

Table I. Potential adverse drug reaction (AD!?) risk factors included in round 1 of the Delphi survey.

Medication-related factors Specific medication (cont’d)

Class of medication Glyburide
ACE inhibitors Insulin (all types)
Antiarrhythmics Lithium salts’
Anticholinergics” Potassium chloride
Anticonvulsants Theophylline salts*
Antidepressants+ Warfarin salts*
Antiparkinsonians
Patient characteristics
(nonanticholinergics)
Advanced age (285 y)
Antipsychotics’
Dementia*
Benzodiazepines”
Depression
Beta-blockers
Female sex
Calcium channel blockers
Hearing loss
Corticosteroids’
inability to read
Diuretics
Low body weight or BMI ~22 kg/m2
Hz-receptor antagonists
Married*
New prescription for an antibiotic
Multiple chronic medical problems*
Nitrates
Multiple doses of medication per day
Non-ASA, non-COX-2 NSAIDs’
Multiple provider visits*
Opioid analgesics*
Polypharmacy”
Sedative/hypnotics*
Prior ADR’
Specific medication Recent hospitalization
ASA Regular use of alcohol (>I fl oz/d)
Chlorpropamide” Renal insufficiency (CrCI ~50 mUmin)*
Digoxin White race*

ACE = angiotensin-converting enzyme; H, = hrstamrne 2; non-ASA, non-COX-2 NSAlDs = non-acetylsalicylic acid, non-cyclooxygenase-2
nonsteroidal anti-inflammatory drugs; BMI = body mass index; CrCl = creatinine clearance.
‘Accepted after round 1 and so was omitted from round 2.
‘Changed to “tricyclic antidepressants,” “selective serotonin reuptake Inhibitors:’ and “other antidepressants” in round 2.
*Eliminated after round 1.
OChanged to “multiple outpatient visits” and “multiple prescribers” in round 2.
2

a4
I
 E.R. Hajjar et al. The Amcvican Jwmal of Geriatric Pharmacotherapy

reuptake inhibitors [SSRIs], tricyclic antidepressants

[TCAs], and other antidepressants; and multiple outpa- Table II. Potential adverse drug reaction risk factors
tient visits and multiple prescribers, respectively). included in round 2 of the Delphi survey.

Medication-related factors
Round 2
Class of medication
The response rate was also 100% for round 2 of the
ACE inhibitors*
survey. Eight new factors were included in round 2 due
Antiarrhythmics’
to suggestion/renaming as described previously, togeth
Anticonvulsants*
er with the 24 factors determined to be equivocal in
Antiparkinsonians (nonanticholinergics)”
round 1 (Table II). Round 2 further eliminated 27
Beta-blockers*
potential risk factors. Five more risk factors were accept-
Calcium channel blockers*
ed by the panel (TCAs, advanced age [285 years], multi-
Diuretics*
ple prescribers, recent hospitalization, and regular use of
H,-receptor antagonists*
alcohol). Table III shows all of the factors that were elim-
inated after rounds 1 and 2. The panel reached consen- New prescription for an antibiotic”
sus on a total of 21 of 50 potential risk factors (Table IV). Nitrates*
SSRls”+
TCAs+*
Prevalence of Final List of Risk Factors
The final list of risk factors for ADRs was applied to a Other antidepressants*+
sample of 808 older outpatients from the GEM Drug Specific medication
Study. 28,29 Two of the factors were not assessed: recent ASA’
hospitalization (because all patients had a hospitalization Digoxin”
within the previous 6 months) and regular use of Glyburide’
alcohol (because those data were not collected in the Insulin (all types)”
original study). The prevalences of the risk factors Potassium chloride*
in the final list are summarized in Table V.28>31The most
Patient characteristics
prevalent medication-related risk factors, in descending
Advanced age (285 y)*
order, included opioid analgesics (197 patients
Depression*
[24.4%]); warfarin (118 patients [ 14.6%]); non-ASA,
Female sex*
non-COX-2 NSAIDs (114 patients [ 14.1%]);
Hearing loss”
anticholinergics ( 106 patients [ 13.1%]); and benzodi-
Inability to read*
azepines (87 patients [lO.S%]). The most prevalent
Low body weight or BMI ~22 kg/m2*
patient characteristics included polypharmacy (674
Multiple doses of medication per day*
patients [83.4%]), multiple chronic medical problems
Multiple outpatient visits’§
(214 patients [26.5%]), prior ADR (153 patients
Multiple pharmacies used”11
[lS.9%]), dementia (83 patients [10.3%]), and renal
Multiple prescribers’*
insufficiency (42 patients [5.2%]). Two medication-
Poor self-reported health”11
related risk factors, chlorpropamide and lithium, had
Recent hospitalization*
prevalences of zero.
Regular use of alcohol (>I fl oz/d)*
Use of self-medication with nonprescription
DISCUSSION
medication”11
A modified Delphi consensus method identified 8 medica-
tion classes, 4 specific medications, and 9 patient ACE = angiotensin-converting enzyme; H, = histamine 2; SSRls =
characteristics as risk factors for ADRs for older outpa- selective serotonin reuptake inhibitors; TCAs = tricyclic antidepres-
tients. Polypharmacy was the only patient characteristic sants; ASA = acetylsalicylic acid; BMI = body mass Index.
to reach complete consensus. A number of previously *Eliminated after round 2.
+Changed from “antidepressants” after round 1.
published studies4@J0J1 also have shown that poly-
*Accepted after round 2.
pharmacy is a risk factor for ADRs. Five of the 8 ‘Changed from “multiple provider visits” after round 1.
medication classes were psychotropics. Psychotropic ISuggested by a panel member in round 1 and subsequently was
medications have been commonly reported to cause included in round 2.

I 85
The American Journal of Geriatric Phavmacotherapy E.R. Hajar et al.

may be attributable to the fact that anticholinergics are

Table III. Potential adverse drug reaction risk fac- known to cause confusion and delirium in old-
tors eliminated after rounds 1 and 2 of er patients.35 Three of the 4 specific medications includ-
the Delphi survey. ed as risk factors have known narrow therapeutic ranges
(lithium salts, theophylline salts, and warfarin salts), and
Medication-related factors
the fourth (chlorpropamide) is a hypoglycemic agent with
Class of medication
an extended half-life. Digoxin, a drug with a narrow
ACE inhibitors
therapeutic range, and glyburide did not reach consensus
Antiarrhythmics
as risk factors. In a study by Shorr et al,36 glyburide was
Anticonvulsants
found to have the same risk of causing hypoglycemia as
Antiparkinsonians (nonanticholinergics)
chlorpropamide.
Beta-blockers
Calcium channel blockers
Diuretics
Table IV. Adverse drug reaction (ADR) risk fac-
HZ-receptor antagonists
tors in rank order on a 5-point Likert
New prescription for an antibiotic
scale* from the Delphi survey.
Nitrates
Other antidepressants Mean
SSRIS Scale
Specific medication Risk Factor Score 95% Cl
AS A
Digoxin Medication-related factors
Glyburide Class of medication
Anticholinergics 4.9 4.7-5.0
Insulin (all types)
Benzodiazepines 4.9 4.7-5.0
Potassium chloride
Antipsychotics 4.7 4.4-5.0
Patient characteristics Sedative/hypnotics 4.7 4.4-5.0
Depression Non-ASA, non-COX-2 NSAlDs 4.5 4.2-4.8
Female sex TCAs 4.5 4.2-4.8
Hearing loss Opioid analgesics 4.5 4.2-4.8
Inability to read Corticosteroids 4.4 4.1-4.7
Low body weight or body mass index Specific medication
Married Chlorpropamide 4.7 4.4-5.0
Multiple doses of medication per day Theophylline salts 4.6 4.349
Multiple outpatient visits Warfarin salts 4.6 4.2-5.0
Multiple pharmacies used Lithium salts 4.4 4.0-4.8
Poor self-reported health Patient characteristics
Use of self-medication with nonprescription Polypharmacy 5.0 5.0-5.0
medication Dementia 4.7 4.4-5.0
White race Multiple chronic medical problems 4.7 4.4-5.0
Renal insufficiency (CrCI ~50 mL/min) 4.7 4.3-5.0
ACE = angiotensrn-converting enzyme; H, = histamine 2; SSRls = Recent hospitalization 4.5 4.2-4.8
selective serotonin reuptake inhibitors; ASA = acetylsalicylic acid. Advanced age (285 y) 4.5 4.2-4.8
Multiple prescribers 4.5 4.2-4.8
Regular use of alcohol (~1 fl oz/d) 4.4 4.1-4.7
ADRs.3~4~10-12~14~17~18~23~26One particular psychotropic Prior ADR 4.4 4.0-4.8
drug class, benzodiazepines, received high consensus from
the group (mean [95% CI], 4.9 [4.7-5.0]), probably non-ASA, non-COX-2 NSAlDs = non-acetylsalicylic acid, non-
cyclooxygenase-2 nonsteroidal anti-inflammatory drugs; TCAs = trt-
because benzodiazepines, regardless of half-life, have been
cyclic antidepressants; CrCl = creatinine clearance.
associated with impaired cognition, hip fractures, and *Scale: 1 = definitely not a risk, 2 = doubtful as a risk, 3 = equivocal-
falls.32-34 Anticholinergics also received high consensus may or may not be a risk, 4 = probably a risk, and 5 = definitely a
from the panel (mean [95% CI], 4.9 [4.7-5.0]), which risk.

86 I
 E.R. Hajjar et al. The American Journal qf Geriatric Phavmacotberapy

reported range was 2.7 to 4.2 for both prescription and

TableV. Prevalence of potential adverse drug nonprescription medication use. The higher mean num-
reaction (ADR) risk factors in frail older ber of medications may be due to the frailty and multi-
(age 265 years) patients at discharge from ple comorbidities of the outpatients in the GEM Drug
hospital (N = 808).*** Study. 28,29 The prevalence of opioid analgesic use
(24.4%) was similar to the data reported in another study
No. (%) of
that assessed ADRs of older outpatients in a Medicare
Risk Factor Patients+
population (21.9%).3 The high prevalence may be attrib-
Medication-related factor uted to the fact that both scheduled and as-needed opi-
Class of medication oid use were included and that all patients were recently
Opioid analgesics 197 (24.4) discharged from the hospital, with many of them possi-
Non-ASA, non-COX-2 NSAlDs 114 (14.1) bly needing opioid analgesics for pain control. The find-
Anticholinergics 106 (13.1) ing that chlorpropamide and lithium had prevalences of
Benzodiazepines 87 (10.8) zero may be due to the fact that health care profession-
Corticosteroids 76 (9.4) als may have an increased awareness of the potential for
TCAs 52 (6.4) these agents to cause ADRs, with the result that these
Antipsychotics 18 (2.2) drugs are not being commonly prescribed.
Sedative/hypnotics 16 (2.0)
There are several limitations to this study. The preva-
Specific medication lence data came from a sample of frail, older outpatients
Warfarin salts 118 (14.6) who were veterans and therefore may not be generaliz-
Theophylline salts 33 (4.1) able to the entire population of elderly outpatients.
Chlorpropamide 0 (0.0) There might have been an underascertainment of the
Lithium salts 0 (0.0)
prevalence of some of the risk factors because some of
Patient characteristics the data were not collected for the original GEM study
Polypharmacy (25 medications)* 674 (83.4) (amount of alcohol consumed) or may have been omit-
Multiple (~2) chronic medical problems* 214 (26.5)
ted from patients’ medical records. It is also likely that
Prior ADR 153 (18.9)
some of the patient characteristics are not independent
Dementia 83 (10.3)
Renal insufficiency (CrCI ~50 mL/min) 42 (5.2) of one another (eg, advanced age and renal insufficien-
Advanced age (285 y) 37 (4.6) cy, or dementia; polypharmacy, and multiple chronic
Multiple prescribers 32 (4.0) medical problems or multiple prescribers).

non-ASA, non-COX-2 NSAlDs = non-acetylsalicylic acid, non- CONCLUSIONS

cyclooxygenase-2 nonsteroidal anti-inflammatory drugs; TCAs = trt- An expert panel was able to reach a consensus on poten-
cyclic antidepressants; CrCl = creatinine clearance.
tial risk factors that increase the risk for ADRs in older
‘Two of the factors were not assessed: recent hospitalization
(because all patients had a hospitalization within the previous 6 outpatients. Many risk factors were common in a sample
months) and regular use of alcohol (because those data were not of older outpatients. Future research is needed to deter-
collected in the original study). mine the predictive validity of these risk factors for
Some patients had >l risk factor. ADRs in older outpatients.
*Mean (SD) comorbtdities defined using the Charlson index,31 2.5 (1.9).
§Mean (SD) concomitant medications, 8.7 (4.2).
ACKNOWLEDGMENTS
Financial support was provided by grants ROl-AG-
14158 from the National Institute on Aging (Bethesda,
It was interesting to note how common certain risk Maryland), K24-AI-51324-01 from the National
factors were in our sample of older outpatients. The Institute of Allergy and Infectious Diseases (Bethesda,
number of medications defining polypharmacy has var- Maryland), and from the VFW Endowed Chair in
ied between studies and our group used a definition of Pharmacotherapy for the Elderly, College of Pharmacy,
25 medications, making polypharmacy the most preva- University of Minnesota (Minneapolis, Minnesota).
lent patient characteristic. The mean number of medica- We thank Kathy Robbins for her help with the litera-
tions was higher than that reported by other community ture searches, and the following expert panelists who
based studies of older individuals,37 in which the generously participated in the survey: Michael J.

I a7
The American Journal of Geriatric Phavmncotherapy E.R. Hajjar
Anderson, PharmD, Ovations Pharmacy Solutions,
Minneapolis, Minnesota; Arlene S. Bierman, MD, MS,
Center for Outcomes and Effectiveness Research,
Agency for Healthcare Research and Quality, Rockville,
Maryland; Barry J. Cusack, MD, Idaho Veterans Affairs
Medical Center, Boise, Idaho; Howard A. Fink, MD,
MPH, Geriatric Research, Education, & Clinical Center,
Veterans Affairs Medical Center, Minneapolis,
Minnesota; Shelly L. Gray, PharmD, MS, School of
Pharmacy, University of Washington, Seattle,
Washington; Timothy J. Ives, PharmD, MPH, School of
Pharmacy, The University of North Carolina, Chapel
Hill, North Carolina; Michael J. Koronkowski, PharmD,
College of Pharmacy, University of Illinois, Chicago,
Illinois; Teresa C. McCarthy, MD, MS, University of
Minnesota, Minneapolis, Minnesota; Michael D.
Murray, PharmD, MPH, Purdue University School of
Pharmacy, and Regenstrief Health Center, Indianapolis,
Indiana; and Ronald I. Short-, MD, MS, University of
Tennessee College of Medicine, Memphis, Tennessee.
REFERENCES
1. Venulet J, ten Ham M. Methods for monitoring and docu-
menting adverse drug reactions. Int J Clin Pharmacol
Ther. 1996;34:112-129.
2. Hanlon JT, Gray SL, Schmader KE. Adverse drug reac-
tions. In: Delafuente JC, Stewart RB, eds. Therapeutics in
the Elderly. 3rd ed. Cincinnati, Ohio: Harvey Whitney
Books; 2001:289-314.
3. Gurwitz JH, Field TS, Harrdld LR et al. Incidence and
preventability of adverse drug events among older persons
in the ambulatory setting. JAMA. 2003;289:1107-1116.
4. Hanlon JT, Schmader KE, Koronkowski MJ, et al. Adverse
drug events in high risk older outpatients. J Am Geriatr
SOL 1997;45:945-948.
5. Chrischilles EA, Segar ET, Wallace RB. Self-reported
adverse drug reactions and related resource use. A study of
community-dwelling persons 65 years of age and older.
Ann Intern Med. 1992;117:634640.
6. Kohn LT, Corrigan JM, Donaldson MS, eds, for the
Committee on Quality of Health Care in America, Institute
of Medicine. In: To Ew Is Human: Building a Safer Health
System. Washington, DC: National Academy Press; 2000.
7. Ernst FR Grizzle AJ. Drug-related morbidity and mortal-
ity: Updating the cost-of-illness model. JAm Pharm Assoc
(Wash). 2001;41:192-199.
8. Fouts MM, Hanlon JT, Pieper CF, et al. Identification
elderly nursing facility residents at high risk for drug-
related problems. Consdt Pharm. 1997;12:1103-1111.
88I
et al.
9. Koecheler JA, Abramowitz PW, Swim SE, Daniels CE.
Indicators for the selection of ambulatory patients who warrant
pharmacist monitoring. Am J Hosp Pbarm. 1989;46:729-732.
10. Gray SL, Mahoney JE, Blough DK Adverse drug events in
elderly patients receiving home health services following hos-
pital discharge. Ann Pharmacother. 1999;33:1147-1153.
11. Grymonpre RE, Mitenko PA, Sitar DS, et al. Drug-associated
hospital admissions in older medical patients. JAm Geriatr
Sot. 1988;36:1092-1098.
12. Gray SL, Sager M, Lestico MR Jalaluddin M. Adverse
drug events in hospitalized elderly. J Gerontol A Biol Sci
Med Sci. 1998;53:M59-M63.
13. Mannesse CK, Derkx FH, de Ridder MA, et al.
Contribution of adverse drug reactions to hospital admis-
sion of older patients. Age Ageind. 2000;29:35-39.
14. Chan M, Nicklason F, Vial JH. Adverse drug events as a
cause of hospital admission in the elderly. Intern Med J.
2001;31:199-205.
15. Bero LA, Lipton HL, Bird JA. Characterization of geriamc drug-
related hospital readmissions. Med Care. 1991;29:989-1003.
16. Co1 N, Fanale JE, Kronholm l? The role of medication
noncompliance and adverse drug reactions in hospitaliza-
tions of the elderly. Arch Intern Med. 1990;150:841-845.
17. Hallas J, Worm J, Beck-Nielsen J, et al. Drug related events
and drug utilization in patients admitted to a geriatric hos-
pital department. Dan Med Bull. 1991;38:417420.
18. Williamson J, Chopin JM. Adverse reactions to prescribed
drugs in the elderly: A multicentre investigation. Age
&einB. 1980;9:73-80.
19. Smucker WD, Kontak JR Adverse drug reactions causing
hospital admission in an elderly population: Experience with a
decision algorithm. JAm Board Fam &act. 1990;3:105-109.
20. Marcantonio ER McKean S, Goldfinger M, et al. Factors
associated with unplanned hospital readmission among
patients 65 years of age and older in a Medicare managed
care plan. Am JMed. 1999;107:13-17.
2 1. Hohl CM, Dankoff J, Colacone A, A&lo M. Polypharmacy,
adverse drug-related events, and potential adverse drug inter-
actions in elderly patients presenting to an emergency
department. Ann Emerg Med. 2001;38:666-671.
22. van Kraaij DJ, Haagsma CJ, Go IH, Gribnau FW. Drug use
and adverse drug reactions in 105 elderly patients admitted
to a general medical ward. Netb J Med. 1994;44:166-173.
23. Cunningham G, Dodd TR Grant DJ, et al. Drug-related
problems in elderly patients admitted to Tayside hospitals,
methods for prevention and subsequent reassessment. A&e
Ageing. 1997;26:375-382.
of 24. Courtman B J, Stallings SB. Characterization of drug-related
problems in elderly patients on admission to a medical
ward. Can J Hosp Pharm. 1995;48:161-166.
 E.R. Hajar et al. The Anzerican Journal of (Geriatric Pharmacotberapy

25. Gandhi TK, Burstin HR, Cook EF, et al. Drug complica- dinal studies: Development and validation. J Chronic Dis.
tions in outpatients. JGen Intern Med. 2000;15:149-154. 1987;40:373-383.
26. Hutchinson TA, Flegel KM, Kramer MS, et al. Frequency, 32. Shorr RI, Bauwens SF, Landefeld CS. Failure to limit
severity and risk factors for adverse drug reactions in adult quantities of benzodiazepine hypnotic drugs for outpa-
out-patients: A prospective study. J Chronic Dis. 1986;39: tients: Placing the elderly at risk. Am J Med. 1990;89:
533-542. 725-732.
27. Jones J, Hunter D. Consensus methods for medical and 33. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls
health services research. BMJ. 1995;311:376-380. among elderly persons living in the community. N En& J
28. Schmader KE, Hanlon JT, Pieper CF, et al. Effectiveness Med. 1988;319:1701-1707.
of geriatric evaluation and management on adverse drug 34. Ray WA, Griflin MR, Schaffher W, et al. Psychotropic drug
reactions and suboptimal prescribing in the frail elderly. use and the risk of hip fracture. N En& J Med. 1987;316:
Am J Med. In press. 363-369.
29. Cohen HJ, Feussner JR, Weinberger M, et al. A controlled 35. Tune LE. Anticholinergic effects of medication in elderly
trial of inpatient and outpatient geriatric evaluation and patients. J Clin Psychiatry. 2001;62(Suppl21):11-14.
management. N En& J Med. 2002;346:905-912. 36. Shorr RI, Ray WA, Daugherty JR, Griffin MR. Individual
30. Veterans Affairs Medication Classification System sulfonylureas and serious hypoglycemia in older people.
[Appendix IV]. In: US Pharmacopeia Dispensing J Am Geriatr Sot. 1996;44:751-755.
Information. 20th ed. Rockville, Md; 2000:3220-3237. 37. Stewart RB. Drug use in the elderly. In: Delafuente JC,
31. Charlson ME, Pompei I’, Ales KL, MacKenzie CR. A new Stewart RB, eds. Therapeutics in the Elderly. 3rd ed.
method of classifying prognostic comorbidity in longitt- Cincinnati, Ohio: Harvey Whitney Books; 2001:235-256.

Address correspondence to: Emily R. Hajjar, PharmD, College of Pharmacy, University of Minnesota, 7-115
Weaver-Densford Hall, 308 Harvard Street SE, Minneapolis, MN 55455. E-mail: hajj0003@umn.edu

1 89