JAMDA xxx (2017) 1e19

JAMDA
journal homepage: www.jamda.com

Review Article

Comprehensive Literature Review of Factors Influencing Medication

Safety in Nursing Homes: Using a Systems Model
Ali Azeez Al-Jumaili BS Pharm, MS *, William R. Doucette PhD
The University of Iowa College of Pharmacy, Pharmacy Practice and Science Department, Iowa City, IA

a b s t r a c t

Keywords: Objectives: The objectives of this review were to identify the work system factors influencing
Medication safety medication safety measures [adverse drug events (ADEs), adverse drug reactions, or medication
adverse drug events errors (MEs)], to determine the incidence of ADEs, and describe the most common ADEs in nursing
nursing homes
homes (NHs).
SEIPS model
Methods: A comprehensive literature review was conducted using PubMed and CINAHL to identify
studies investigating factors that influence ADEs, adverse drug reactions, and MEs in NHs
and skilled nursing facilities. An initial search identified related studies over 3 decades (1985-
2016). Studies were classified according to Systems Engineering Initiative for Patient Safety model
factors.
Results: Sixty studies were included in this review, which identifies 5 categories of work system factors
affecting medication safety in NHs: persons (resident and staff), organization, tools and technology,
tasks, and environment. The personal characteristics of NH residents included age, number and types
of scheduled medications, and number and types of comorbidities. In addition, inadequate nursing
staff medication knowledge and training are usually associated with administration MEs. Organiza-
tional factors include interprofessional collaboration, physician and pharmacist accessibility, and staff/
resident ratio. A high staff number plays an essential role in preventing MEs and fracture incidents. The
technology (barcode medication system) and tools (ME-reporting systems, ADE trigger tool, and
potentially inappropriate medication criteria) can enhance the detection of MEs and ADEs. Workload
and time pressure negatively impact NH staff task performance. Use of an ADE trigger tool by
healthcare providers enhanced the ability to identify ADEs more than 50-fold over 6 months. Several
environmental characteristics such as staff distraction and interruption negatively influence medica-
tion safety in NHs. The incidence rates of ADEs in NHs ranged from 1.89 to 10.8 per 100 resident-
months. The most common ADEs were bleeding, thromboembolic events, hypoglycemia, falls, and
constipation.
Conclusions: The Systems Engineering Initiative for Patient Safety model is a useful framework for
investigating the factors contributing to ADEs. Multiple work-system factors affect the medication safety
of NH residents. A comprehensive study is needed to quantify the influence of various work-system
factors on ADEs in NHs.
Ó 2016 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

According to the U.S. Nursing Home Data Compendium, in 2015
nursing home (NH) facilities were home of more than 1.4 million
residents, 85% aged
65 years old and 41.3% aged
85 years old.1,2
Women were 66.8% of the residents and non-Hispanic white
people accounted for more than three-quarters (76.1%) of all
The authors declare no conflicts of interest.
* Address correspondence to Ali Azeez Al-Jumaili, BS Pharm, MS, PhD Candidate,
The University of Iowa College of Pharmacy, Iowa City, IA 52242.
E-mail address: aliazeezali-aljumaili@uiowa.edu (A.A. Al-Jumaili).
residents.2 NHs provide both custodial and medical care to their
residents. Skilled nursing facilities (SNFs) provide short-term
skilled nursing and rehabilitation services to patients after
discharge from hospitals, while NHs provide long-term and in-
termediate care for mostly older residents.3 According to the
National Study of Long-Term Care Providers (2013-2014), there
are 15,600 NHs in the United States.2 Ninety percent of facilities
are dually certified as a SNF and a NH.3,4
Medications are commonly administered in NHs, which creates
some risk for residents, related to unsafe medication use. One type

http://dx.doi.org/10.1016/j.jamda.2016.12.069
1525-8610/Ó 2016 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
2 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
of problem, medication error (ME), is defined as “a failure in
treatment process that leads to, or has the potential to lead to
harm to the patient.”5 Another concern is adverse drug events
(ADEs), which are any unintended medication-induced injuries
that require monitoring, intervention, or hospitalization, or result
in death.4,6,7 ADEs can be either preventable MEs or adverse drug
reactions (ADRs) (Figure 1).9 ADRs are usually unintended conse-
quences, which often occur at normal therapeutic doses, such as
first dose hypotension induced by angiotensin converting enzyme
inhibitors (ACE-Is).5,10 Nevertheless, some ADRs are preventable
because they are due to MEs such as when an overdose of hy-
dralazine is administered causing a lupus-like syndrome.8
Although researchers have studied these medication safety is-
sues, that work typically has not been reported using a systematic
comprehensive approach.
The Systems Engineering Initiative for Patient Safety (SEIPS)
model was developed by Carayon et al11 in 2006 to comprehen-
sively explain the effects of a work system and process on health
outcomes. A work system is composed of 5 components: Persons,
organization, technologies and tools, tasks, and environment.12
The SEIPS model has been used in patient safety studies in
various healthcare settings. In this model, an individual is the
center of the work system and can be a person (patient and/or
provider) or group of people. The organization consists of struc-
tures external to a person within which work is performed. Tools
and technologies are items or devices that are used to conduct
tasks, and they have characteristics such as usability, accessibility,
familiarity, and portability. Tasks are the specific actions within
the larger work process and can be difficult, complex, varied, and
ambiguous. The environment includes physical and safety envi-
ronment factors. The process is a series of tasks that conducted by
the 5 work system components. For instance, the process for
identifying and managing ADEs is a series of tasks conducted by a
healthcare practitioner or team of practitioners using certain tools
and technologies and working under specific environment and
organizational conditions.11,13
Fig. 1. Relationship between MEs, ADRs, and ADEs (Adapted from Aronson and Ferner, 2005 with additions).8 LTCP, long-term care pharmacy.
This review provides background information about long-term
facility practices and contributes to enhancing the quality of
care and medication safety in NHs. The objectives of this
review were to use SEIPS model to identify the work system fac-
tors [persons (resident and staff), organization, tools and
technology, tasks, and environment] influencing medication
safety measures (ADEs, ADRs, or MEs) among NH residents, to
determine the incidence of ADEs, and describe the most common
ADEs in NHs.
Methods
An initial search identified related studies over 3 decades (1985-
2016). The search had 2 components: a systematic review of the
literature (2000-2016) and a search of references and related articles
(1985-2016). The systematic review included studies available in
PubMed and CINAHL and published between January 1, 2000, and May
9, 2016. January 1, 2000, was chosen as the start date because previous
literature reviews have examined the epidemiology of ADEs in NHs
between 1985 and 2005.7,14,15
We included studies with 1 or more of the following outcome
measures: ADEs, potential ADEs, ADRs, or MEs. We then examined
citations and related articles to identify studies that included
SEIPS factors. This led to the inclusion of some earlier studies
(1985-2000). All studies were original research and written in
English.
We excluded studies conducted in assisted living facilities, age-
care facilities, and residential-care homes, personal opinion arti-
cles, and review articles (Figure 1). Several medication safety and
NH keywords were used to cover as many related studies as
possible (Table 1). The authors over years have not used a single
term to describe medication safety consistency, and the terms
ADEs and ADRs were at times used interchangeably.16 After
excluding articles that were irrelevant to our objectives, we were
left with a total of 60 studies covering a 3-decade period (1985-
2016) (Figure 2).
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
Table 1
Literature Review Keywords
PubMed keywords (“drug-related side effects and adverse reactions”
[MeSH] OR Adverse Drug Event* [text word] OR
Adverse Drug Reaction* [text word] OR Adverse
Reaction* [text word] OR Drug Side Effects [text
word] OR medication safety [text word] OR
medication errors [text word] OR medication-
related problems [text word] OR drug-related
problems [text word]) AND
(“Nursing Homes” [Mesh] OR “Intermediate Care
Facility” [Mesh] OR Nursing Home* [text word]
OR “Skilled Nursing Facilities” [Mesh] OR
Extended Care Facilities [text word] OR Extended
Care Facility [text word] OR Skilled Nursing
Facility [text word] OR Skilled Nursing Facilities
[text word] OR long-term facility [text word] OR
long-term facilities [text word] OR long-term care
facility [text word] OR long-term care facilities
[text word]).
CINAHL headings (MH “Adverse Drug Eventþ”) OR “adverse drug
reaction” AND (MH “Skilled Nursing Facilities”)
OR (MH “Nursing Homesþ”).
Results
The results have been organized by SEIPS model factors to
address objective 1, whereas objective 3 was described within the
first component (persons) of the SEIPS model. The most common
unit of measurement for ADE incidence was event number per 100
resident-months. The ADE incidence rate in NHs ranged from 1.89 to
10.8 per 100 resident-months.9,17,18 The 60 reviewed studies were
Fig. 2. Flowchart of literature study selection and excluded articles.
3
classified according to the 5 main SEIPS concepts. Fifty-one studies
investigated one factor each [person (n ¼ 23), technology/tool
(n ¼ 14), organization (n ¼ 6), task (n ¼ 7), environment (n ¼ 1)], 7
studies had 2 factors each, and 2studies19,20 had 3 factors. The
studies are from 9 different countries: USA (n ¼ 41), United
Kingdom (n ¼ 5), Canada (n ¼ 4), Belgium (n ¼ 4), Norway21,22
(n ¼ 2), The Netherlands23 (n ¼ 1), Italy24 (n ¼ 1), Ireland25
(n ¼ 1), and Taiwan26 (n ¼ 1). The designs of the included studies
are as follows: retrospective cohort (n ¼ 19), cross-sectional
(n ¼ 14), mixed method (n ¼ 7), prospective cohort (n ¼ 6), pre-
and postintervention (n ¼ 7), case-control (n ¼ 3), cluster ran-
domized trial (n ¼ 2), case crossover (n ¼ 1), and randomized
controlled trial (n ¼ 1) (Table 2).
Persons
Thirty studies investigated the influence of the personal charac-
teristics of NH residents or NH staff on the incidence of ADEs, ADRs, or
MEs. Twenty-six of these studies investigated the relationships be-
tween resident characteristics, such as age, cognitive disability,
number of comorbidities, number of scheduled medications, type of
medications, and ADEs. For this review, medications are treated as a
person characteristic. Four studies examined the effect of the clinical
and medication knowledge, training, and credentials of NH practi-
tioners: physicians, nurses, certified medical assistants (CMAs), and
certified nursing assistants (CNAs).20,41,48,61 The outcome measures
varied from the total actual ADEs, potential ADEs, ADRs, MEs, or
adverse drug withdrawal events to the specific medication class-
associated ADEs or specific ADEs (bleeding, hypoglycemia, falls, and
fractures).
4 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
Two studies found that older NH residents (
75 years and

85 years) had a higher risk of adverse anticoagulant events and
fractures, respectively.40,63 Another study using North Carolina
Medication Error Quality Initiative data revealed that residents aged

75 years had significantly higher anticoagulant MEs.63 One study
found that 6% of NH residents experienced fractures, and most of
those were age
85 years and taking anticonvulsants, antidepres-
sants, or thiazides.40 Four studies found that a large number of
comorbidities had positive relationship with the incidence of
ADEs.17,26,30,57 Seven studies reported that NH residents with several
scheduled medications (
5) were at higher risk for different
ADEs.17,18,24,30,33,36
Fifteen ADE-related studies reported 1 or more of the following 6
medications classes to be most frequently associated with ADEs:
psychotropic drugs (antipsychotics, antidepressants, anxiolytics, and
hypnotics),9,18,22,24,28e30,33,38e40,43,55,58,72,75 cardiovascular agents (di-
uretics, angiotensin-converting enzyme inhibitors [ACEIs], and
digoxin),17,18,22,24,28,29,38,40,43,55,70,72 opioid analgesics,17,18,30,38,43,60
anticoagulants (warfarin),9,17,18,40,43,55,63 antibiotics,9,18,33,38,70,74 and
antidiabetics (insulin).
In addition to detecting high-risk medications, the ADEs asso-
ciated with these medications have been identified. A study
involving 6577 residents at 136 Veterans Health Administration
facilities found that Veterans Affairs residents with antipsychotic,
antidepressant, and anti-anxiety medications had a higher risk of
fall incidents [odds ratio (OR) ¼ 1.15, 1.13, and 1.39 respectively].39
Oral anticoagulants such as warfarin are associated with bleeding
and thromboembolism, and 12.8% of these were found to be pre-
ventable.57 A cohort study in Connecticut found a higher incidence
of adverse warfarin events (18.8 per 100 resident-months) among
warfarin users.44 These studies determined that warfarin-induced
bleeding mostly occurred because of inadequate monitoring or
delayed response to abnormal laboratory values.
Antibiotics, opioids, analgesics, and insulin also are frequently
involved in ADEs. A Canadian cohort study of 607 NHs showed resi-
dents with a high use of antibiotics had a higher incidence rate (13.3%)
of antibiotic-associated adverse events such as Clostridium difficile
diarrhea and gastroenteritis. The most common antibiotics associated
with adverse events were penicillins followed by fluoroquinolones,
sulfonamides, and cephalosporins.74 A prospective cohort study at 6
Dutch NHs revealed that 96% of residents taking opioids and 74% of
residents with Parkinson disease used a laxative because they suffered
from constipation.23 In Pennsylvania, a recent retrospective cohort
study reported a 12.8% incidence of drug-induced hypoglycemia, and
insulin was associated with the majority (98.7%) of the cases in 4
NHs.72
Acute kidney injury (AKI) and falls are also frequent ADEs in
NHs. A study examining data available for 4 NHs found the med-
ications associated with AKI alerts were diuretics, ACEIs/angio-
tensin II receptor blockers, and antibiotics.70 An Italian study found
a fall incidence of 1.38 per 100 resident-months, and 48.1% of these
falls resulted in injuries. Those authors also observed that NH
residents taking
4 concurrent medications, and psychotropic
agent(s) (antipsychotics, benzodiazepines and antidepressants)
were at high risk of falls.24 Similarly, a more recent study which
recruited 594 residents from 2 NHs in Massachusetts found
benzodiazepine initiation led to a high risk of fall incidents (OR 3.8,
95% CI, 0.6-10.0).75 To summarize, bleeding, antibiotic-induced
diarrhea, hypoglycemia, AKI, and falls are common ADEs among
NH residents.
Four studies examined the relationship between the personal
characteristics of the NH staff and safe medication management. A
Belgian study surveying 516 NH staff (nurses and CNAs) revealed
that the staff did not have enough knowledge of drug-food
interactions, drug side effects, therapeutic effects, and correct use
of inhalation devices.20 Similarly, in another Belgian study,
educational sessions for NH nurses and CNAs provided by a phar-
macist about drug-interactions, proper use of inhalation devices,
and special medication administration warnings significantly
reduced medication preparation and administration errors.48 On
the other hand, an observational study found no difference in
administering error rates according to the credentials of the
administering staff (registered nurses [RNs], licensed practical
nurses [LPNs], or CMAs).41
Organization
Eleven ADE-related studies have evaluated the relationship be-
tween NH organizational factors, such as organizational culture,
staffing (RN, LPN, CNA) number and ratio, interdisciplinary commu-
nication and collaboration, and ME-incidence report-
ing.19,20,30,32,35,37,40,42,47,59,76 A culture of blame still governs many NH
organizations where the facility administration blames NH staff for
MEs rather than encouraging staff to report them. Two studies found
that some NH staff (20% of nurses and CNAs) reported feeling
vulnerable to punishment and that was preventing them from
reporting MEs.35,37 In addition to barriers to reporting MEs, lack of
accessible ME reporting systems and feedback to the reporter are also
problems in NHs.42
A Canadian study confirmed that staff shortage can be associated
with medication safety problems in long-term facilities.77 Using 6-
year secondary data from the online Survey and Certification and
Reporting System and Nursing Home Quality Initiative for SNFs
from 8 southeastern U.S. states, Walsh et al65 found that an increase
in the number of CMAs decreased the number of deficiency cita-
tions from the state for unnecessary medication use or high ME
rates (
5%) and also lowered pharmacy citations. Likewise, Spector
et al40 concluded that a high CNA/resident ratio (
49/100) is
inversely associated with fracture incidents. A lack of interdisci-
plinary collaboration is also a significant factor influencing medi-
cation safety in institutional healthcare settings. Four mixed
method studies in 4 different countries (United Kingdom, USA,
Belgium, and Canada) have shown that suboptimal interdisciplinary
collaboration is a barrier to medication safety within NHs.19,20,47,59
Other common organizational barriers are limited resident infor-
mation available for physicians when they prescribe from outside
NHs, and low accessibility to physicians and consultant
pharmacists.19,20,59
Tools and Technology
Tools and technologies may include vital signs and laboratory test
equipment, electronic health records, ME-reporting systems, ADE
trigger tool, potentially inappropriate medication (PIM) criteria (eg,
Beers and Screening Tool of Older Person’s Prescriptions), and bar-
code medication administration systems. Sixteen studies aimed to
develop, evaluate, validate, and/or compare new tools or technolo-
gies used by NH practitioners to identify ADEs or PIMs, report MEs,
and enhance medication prescribing, managing, and mon-
itoring.21,25,33,39,42,43,45,46,49,52,54,62,66e68,72 Two of these studies
described and tested the use of the ADE trigger tool, which examines
laboratory values, blood drug levels, use of antidotes, and resident
assessment protocols.45,67 This is a “low tech” tool that has been
found to increase ADE identification rate 50-fold over the conven-
tional chart extraction method. It takes an average time of 8.8 (5.7)
minutes to assess 1 NH resident chart.67,78 A British study found a
nurse-led mental medication monitoring helped to identify and
address 80 psychotropic agent-related problems.69
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
Two studies in the United States developed fall assessment tools
to specify fall risk factors and help to reduce fall incidence rates.39,52
A study involving secondary data of 136 Veterans Affairs NHs
developed fall resident assessment protocol and found that one of
the risk factors was taking antipsychotic, antianxiety, or antidepres-
sant medications.39 A randomized cluster trial which evaluated a
clinical informatics tool, Geriatric Risk Assessment MedGuide
(GRAM) software found a lower incidence of falls, delirium, and
death and an increase in the number of hospitalizations.52 An Irish
study using Screening Tool of Older Person’s Prescriptions and Beers
criteria to identify PIM use in NH residents found the criteria have
the potential to reduce emergency department visits.79 Culley et al72
found a computerized clinical surveillance system accurately detects
drug-associated hypoglycemia with incidence of 9.5 per 1000 resi-
dent-days.
The technologies and tools used for medication prescription,
distribution, and administration, and ME reporting may affect
rates of ADEs and MEs. A mobile device with drug reference
software used by 236 American NH physicians prevented potential
ADEs in NHs.62 An English study, which assessed a pharmacy-led
barcode medication system in 13 long-term facilities, found that
nursing staff experienced less stress and pressure, fewer in-
terruptions, and more awareness of MEs compared with tradi-
tional paper-based medication administration record.53 A
Canadian study, which examined medication distribution tech-
nology, found that using medication distribution technology
increased the detection of MEs.68 All in all, technologies and tools
can prevent MEs, increase ME identification, and reduce stress
and ADEs.
Tasks
Medication management tasks include prescribing, dispensing,
administering, documenting, and monitoring. Workload and time
pressure may negatively impact medication safety. Suboptimal task
performance at each of the 5 stages of medication management can
lead to MEs and ADEs (Figure 1). Examples of MEs are wrong medi-
cation, wrong dose, wrong dose frequency, wrong duration, wrong
strength, wrong route of administration, wrong patient, unnecessary
medication, omission dose, inappropriate documentation, allergic or
contra-indicated medication, drug-drug interaction, drug-food inter-
action, and no medication follow-up.5 Wrong dose is the most com-
mon type of ME in NHs with a rate of 19%-65.1% of all administered
medication doses.50,63,64
Ten ME- and ADE-related studies classified the phases that are
most commonly associated with MEs or ADEs. Four of 6 studies using
the North Carolina ME reporting system found MEs occurred most
commonly during the administration phase with a rate of 10.2% to
69%.43,50,55,63,64,71 In 2 mixed method studies, nursing staff indicated
they do not have enough time to conduct their administering and
monitoring tasks carefully.19,20
Environment
Environmental factors may include the facility’s physical layout,
temperature, air quality, light, noise, location, available space, number
of beds/residents, work distractions, and interruptions.12 Four studies
reported that NH staff considered distraction and interruption during
medication administration to be barriers to safe medication man-
agement and associated with MEs.19,20,41,59 Barker et al found NH size
(
100 beds vs <100 beds) had no influence on ME rate.31 A Canadian
study conducted in 4 long-term facilities identified several physical
environment influences MEs including chart storage space, nursing
5
station location, noise, lighting, and privacy in the nursing station and
administration area.59
Discussion
Nineteen studies19,31,32,38,41e43,48e50,53,55,59,60,63e65,68,71 in this
review measured the factors associated with MEs rather than
ADEs or ADRs, and about half of them focused on resident-related
factors, which usually are nonmodifiable. Nine of the 19 studies
investigating MEs were conducted in North Carolina, USA, which
implemented a mandatory online Medication Error Quality
Initiative system in 2004.38,49,50,55,60,63e65,71 Eight other studies
used an “ADRs” term to describe unintentional medication-
induced injuries.27e29,36,45,66,69,73 Several of the included studies
qualitatively measured the relationship between medication
safety outcome and organization factors.
More than 2 dozen previous studies have found that Person factors,
such as the personal characteristics of NH residents, such as age, and
number and types of scheduled medications, as well as the number
and types of comorbidities influence the incidence rate of ADEs, ADRs,
and MEs. In addition, NH staff characteristics including medication
knowledge, education, and training affect the rate of administering
MEs.48 Many researchers agree that there is a high risk associated with
specific medication classes, including psychotropic agents, anticoag-
ulants, antidiabetics, opioids, and antihypertensive agents, which can
cause common ADEs such as bleeding, thromboembolic events, hy-
poglycemia, falls, constipation, and acute kidney injury. Two Belgian
studies found weakness in NH staff medication knowledge and sug-
gested that educational/training sessions would be helpful to improve
medication safety and reduce MEs.41,61 Thus, minimizing the use of
these high-risk medications could reduce the incidence of ADEs and
ADRs.
As has been noted, adequate NH staff numbers help to limit
preventable adverse events such as falls and fractures, which may
be induced by medications such as psychotropic agents. Overall,
organizational factors, such as interprofessional collaboration,
physician and pharmacist accessibility, and staff-resident ratio,
play essential roles in preventing MEs and fracture incidents.
Hence, increasing interdisciplinary collaboration, staff/resident
ratio, and visitor professionals’ accessibility should enhance
medication safety in NHs.
For tools and technology, several studies have identified different
types of high-tech (Geriatric Risk Assessment MedGuide software)
and low-tech (ADE trigger tool) tools and technologies that can be
helpful in minimizing and identifying ADEs. According to the litera-
ture, conventional chart reviews and incident reports are unreliable
ways to measure ADEs because they are time consuming.80 However,
out of 10 studies that employed medical chart extraction to identify
ADEs,9,17,30,36,44,57,67,78,81,82 8 used retrospective conventional chart
review, and only 2 used an ADE trigger tool, and this was for validation
purpose.67,78 Though this tool is relatively simple, accurate, flexible,
consistent, and reliable,80,83 with several versions available,45,67,78,80,83
many NH consultant pharmacists typically do not use it during
monitoring tasks.
The ADE trigger tool was developed and validated in the United
States, but other countries have their own versions. For instance
Pharmanurse, an instrument to facilitate nurse-driven ADR screening
that was developed and evaluated in 6 Belgian NHs, received an 83%
satisfaction rate from physicians.66 ME reporting systems have been
found to be easy and helpful for reporting and classifying MEs in NHs.
They can specify the severity, harmfulness, and preventability of MEs
in addition to identifying the most common MEs and most frequent
medications associated with these errors.43,49 The absence of an ME
6 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2
Summary of Literature Review Studies (N ¼ 60)

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

1. Soon 198527 USA 1) Develop 2 years Pre and post 826 patients in 10 NHs ADRs After implementing the
(person and tool/ pharmacist-led ADR (Sept 1978- intervention program, the incidence of
tech) monitoring and Aug 1980) design. moderate to severe ADRs
reporting program 2 phases study was decreased from 27% to
in NHs. 20% and a significant
2) Determine the reduction in the number of
effect of sex, age, and ADR-related acute care
number of drugs on hospitalizations.
the incidence of The most common groups of
ADRs. drugs causing ADRs were
digitalis glycosides,
antipsychotics, sedatives
and hypnotics, diuretics,
and anti-inflammatory
agents.
2. Gerety et al 199317 1) Determine the 1.5 years (Feb Retrospective 175 residents at an ADEs and ADWEs 95 residents experienced 201
Texas, USA (person) incidence and 1988-Sept chart review academic VA NH ADEs, 12 needed
severity of ADEs and 1989) hospitalization and 1
ADWEs in a NH resident died whereas 60
population, and (2) residents had 94 ADWEs
measure the without death. Common
association between risk factors for both ADEs
demographic and and ADWEs: number of
clinical diagnoses, number of
characteristics, and medications, and
ADE/ADWEs. hospitalization during the
NH stay. The drug classes
accounted for most of ADEs
and ADWEs were
cardiovascular, CNS, anti-
inflammatory, analgesic,
and gastrointestinal drugs.
3.Cooper, 199628 To assess ADRs in a 4 years Prospective cohort 332 residents at 2 NHs ADRs 217 out of 332 residents had a
Georgia, USA geriatric NH study mean 1.9  1.3 ADRs over
(person) residents. 30 days. The drugs most
commonly causing ADRs
were diuretics,
antipsychotics, anxiolytics,
potassium supplements,
digoxin, NSAIDs, insulin,
theophylline, H2-
antagonists, anti-infections,
anticonvulsants, and
thyroid agents.
4. van Dijk et al 199823 To measure the 2 years Prospective cohort 2355 residents at 6 Constipation Calcium channel blockers,
The relationship study NHs (as ADE) calcium salts and ferrous
Netherlands(person) between laxative use salts, and drugs with
(presence of anticholinergic side-effects
constipation) and had the highest relative risk
other drug use for constipation.
96% of the residents who had
opiates received a laxative
drug, and 74% of residents
with Parkinson disease used
a laxative.
5. Cooper 199929 To assess ADR- related 4 years Prospective 332 residents present ADR- related 64 ADR-associated
Georgia, USA hospitalizations observational for 30 or more days hospitalizations hospitalizations in 52 of the
(person) from nursing facility study at 2 rural SNFs 332 residents (15.7%). The
residents. most common ADRs were
for NSAIDs (30),
psychotropic-related fall
with fracture (14), digoxin
toxicity (5), and insulin
hypoglycemia (4).
6. Gurwitz et al 20009 To assess the incidence 1 year Retrospective 18 community-based ADEs and The incidence was 1.89 per
Massachusetts, USA and preventability of cohort study NHs potential ADEs 100 resident-months for
(person) ADEs and potential ADEs and 0.65 per 100
ADEs in NHs. resident-months for
potential ADEs. Around 51%
of ADEs were preventable.
Most common drugs
associated with ADEs were
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 7

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

psychoactive drugs
(antipsychotics,
antidepressants, and
sedatives/hypnotics),
antibiotics, and
anticoagulants.
7. Field et al 200130 To determine 1 year (Spring Case-control study 410 cases vs 410 ADEs and risk 410 ADEs (4.17 per 100
Massachusetts, USA incidence and risk 1997) nested within a control at 18 NHs factors for ADEs resident-months) and 226
(person) factors of ADEs prospective study of them were preventable
ADEs. Risk factors taking
anti-infective,
antipsychotic,
antidepressants, and
opioids, higher than 5 scores
in Charlson comorbidity
index, number of scheduled
medications (5-9) was also
risk factors (OR 3). ADEs
mostly occurred during
monitoring stage.
8. Barker et al 200231 To identify prevalence A Prospective 36 SNFs and hospitals Medication MEs in doses were 19%:
USA (environment) of medication cohort study administration wrong time ¼ 43%,
(administration) errors omission ¼ 30%, wrong
errors, MEs. dose ¼ 17%, unauthorized
drug ¼ 4%.
7% of MEs had potential ADEs.
No different in ME rate by size
or type of facility.
9. Weinberg et al To determine whether 6 months Prospective Total of 31 residents ME, falls, median Both medians of CNA (4 vs 5;
200232 Georgia, USA admissions to SNF (Jan- analytical admitted skilled staffing (CNAs P  .001) and nurse (3 vs 4;
(organization) received equivalent July 2000). study nursing facility and nurses) P  .001) staffing levels
care on weekdays as level on duty were significantly lower on
opposed to on weekend than on weekday
weekends with day shifts, but there was no
regard to CNA and significant difference in MEs
nurse staffing levels. detected.
10. Ruths et al 200322 Type and frequency of 1997 Cross-sectional 1354 residents at 23 MRPs, and ADRs Potential MRPs were
Norway (person) potential medication NHs identified in 76% of
problems (MRPs) residents: ADRs (26%),
and types of causing inappropriate drug choice
medications for indication (20%), and
underuse of beneficial
treatment (13%). Most
common MRPs were
excessive sedation,
cognitive deterioration,
extrapyramidal, and
anticholinergic side effects.
Psychoactive (antipsychotic)
drugs contributed for 38%
and ACE inhibitors for 12.5%
of all MRPs.
11. Lapane and Hughes To modify and test a Retrospective All residents in 30 Proportion of The modified high-risk
200433 North screening tool for descriptive SNFs residents who screening tool proved to be
Carolina, USA (tool/ identification of NH study. trigger high-risk practical and clinically
tech) residents at high risk screen for relevant in preventing ADEs
for preventable preventable in NH residents in the
ADEs. ADEs Fleetwood Phase III
evaluation. The median
proportion of residents
triggered was one-third. The
risk factor for ADEs was
using 7 or more medications
with 1 medication being an
antidepressant.
12. Perri et al 200534 To identify the 3 months Cohort design Review 1117 resident Prevalence of A total of 519 (46.5%) patients
Georgia, USA prevalence of (Mar medical records in inappropriate received at least 1
(person) inappropriate 1- May 31, 15 Georgia NHs with medication use inappropriate medication
medication use 2002) a high risk of and adverse and 143 (12.8%) patients
among NH residents polypharmacy health outcomes experienced at least 1
using the Beers adverse health outcome.
criteria and identify Inappropriate medication use
the relationship increased the probability of
(continued on next page)
8 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

between experiencing at least 1

inappropriate drug adverse health outcome
use and the more than 2-fold.
probability of an
adverse health
outcome.
13. Gurwitz et al Incidence and risk 9 months Case-control 1247 residents of long- ADEs 815 ADEs (9.8%), 42% of ADEs
200518 Connecticut factors of ADEs (2000- study term care settings were preventable (N ¼ 338),
and Ontario, Canada 2001) (8336.4 residents- 23% of ADEs were serious
(person) month) and 4 % life threatening.
Antipsychotics,
antidepressants,
anticoagulants (warfarin),
loop diuretics, anti-
infective, and opioids were
the most common drugs
associated with ADEs.
Risk factors were high
number of scheduled
medications from different
categories.
14. Scott-Cawiezell To explore the current Mixed-method Staff members were Medication safety The staff shared their
et al 200635 culture of blame and with a case surveyed in 5 concerns about the lack of
Missouri, USA organizational study and Midwestern NHs feedback and the lack of
(organization) elements in NHs staff member resolution of medication
survey system concerns. The staff
stated the realities of the
underreporting of MEs and
“getting into trouble” if an
error is found.
15. Nguyen et al To determine the 12 months Retrospective 335 residents aged ADRs 207 ADRs were identified.
200636 California, relationship (Oct 1998- cohort study
65 years at a SNF A positive correlation
USA (person) between the use of 9 Sept 1999) between the use of
9
or more different different scheduled
scheduled medications and ADRs was
medications and the found among these NH
occurrence of ADRs residents.
in NH residents
16. Hughes and Lapane To evaluate whether Summer and Cross-sectional 367 nurses and 636 34 items on Approximately 40% of nursing
200637 Ohio, USA perceptions of fall of study. nursing assistants at different staff found it difficult to
(organization) patient safety in NHs 2003 26 NHs in Ohio aspects of make changes to improve
vary by length of resident safety things most or all of the
employment, type of time.
employee, and shift Similar proportions indicated
worked. that management seriously
considered staff suggestions
to improve resident safety;
50% of staff reported
management discussions
with staff to prevent
recurrence of mistakes.
1/5 of the staff reported
feeling punished and 2/5
reported that reporting of
errors was seen as a
“personal attack.”
17. Hansen et al Incidence and types of 9 months Case-control Using mandatory MEs, preventable 9272 MEs (24 per home/
200638 North MEs and number of (Jan-Sept study Web-based recording MEs and 9 months), 66 (1.1%) of MEs
Carolina, USA (task) error-induced harms 2004) (secondary system (Med (MARx) ME-induced caused harm (ADEs) and 1.8
data analyses) of 384 state-owned harms preventable MEs per 100
NC NHs resident-month.
Pharmacists involved in 7%
and physician in 8% of MEs.
CNS agents (eg, lorazepam)
(16%), analgesics (eg,
hydrocodone) (11%), and
anti-infective agents (3%).
18. French et al 200739 To assessment fall risk 1 year (fiscal Retrospective, 6577 Fall risk Risk factors for falls were the
USA (person and factors beyond the year 2005) clustered NH residents at 136 characteristics use of assistive devices such
tool/tech) current fall RAP (secondary National VHA long- from the MDS as canes, walkers, crutches,
triggers for NH data analysis) term care NH or the use of wheelchairs,
residents using the foot problems,
Alzheimer, or other dementias
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 9

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

Minimum Data Set and use of antipsychotic

(MDS). medications (OR 1.15,
P  .0039), antianxiety
medications (OR 1.13,
P  .0323), or
antidepressant medications
(OR 1.39, P  .0001)
19. Spector et al To determine whether 1 year (Jan 1, Panel study Residents aged 65 and Fractures In 1996, 6% of NH residents at
200740 USA (person resident and facility 1996) with 1-year older of a nationally incidence the beginning of the year
and organization) characteristics and follow-up representative experienced a fracture
prescription (prospective sample of NHs from during their NH stay(s).
medications design). the MEPS Resident risk factors
influence the included aged 85 and older,
incidence of admitted from the
fractures in NHs. community, exhibited
agitated behaviors, and used
both wheelchair and cane or
walker. Use of
anticonvulsants,
antidepressants, opioid
analgesics, iron,
bisphosphonates, thiazides,
and laxatives were
associated with fractures. A
high certified nurse aide
ratio was negatively
associated with fractures.
20. Scott-Cawiezell To determine the 90 days Descriptive and 39 medication Medication Level of staff credential had no
et al 200741 impact of various exploratory study administrator staff at administration influence on ME rates.
Missouri, USA levels of (cross-sectional). 5 Midwestern NHs discrepancies, RNs had more interruptions
(person and credentialing among (Observed and number of during their medication
environment) NH staff who deliver medication distractions and administration, and these
medications (RN, administration, interruptions more interruptions were
LPN, or CMA) on ME. distractions and associated with more ME
resident’s rates
medical chart)
21. Handler et al To identify Mar-April Mixed methods 28 healthcare Modifiable 14 (70%) survey items had
200742 Philadelphia, organizational-level 2005 (Nominal group professionals barriers to scores that categorized
USA (organization) and individual-level sessions, (physicians, report MEs them as immediate action
modifiable barriers followed by pharmacists, factors, 9 (64%) of which
to medication error cross-sectional advanced were organizational
reporting in NHs. mailed survey). practitioners, and barriers.
nurses) and104 The most 3 modifiable
responded to the organizational factors were
survey. from 4 NHs (1) lack of a readily available
ME reporting system, (2)
lack of instructions on how
to report a ME, and (3) lack
of comments to the reporter
or other staff members on
MEs that have been
reported.
22. Pierson et al To describe the 1 year Cross-sectional 25 NHs in the state of Number and 2731 discrete ME instances in
200743 North implementation and (evaluation North Carolina specific 23 (92%) NHs. 51 (8%) MEs
Carolina, USA (tool/ evaluation of a web- survey) characteristics were having a serious
tech) based medication of MEs reported. patient impact requiring
error reporting monitoring/intervention or
system (assess worse. The most common
usability and the MEs were dose omission
potential for the (32%), overdose (14%),
system to prevent underdose (7%), wrong
errors). patient (6%), wrong product
(6%), and wrong strength
(6%).
Seven drugs were implicated
in one-third (28%) of all
errors: lorazepam,
oxycodone, warfarin,
furosemide, hydrocodone,
insulin and fentanyl; 20
sites (86% of respondents)
found the system easy to
use and it would increase
(continued on next page)
10 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

accuracy of reporting and

improve patient safety.
23. Gurwitz et al Assess warfarin- 1 year (April Retrospective 490 residents taking Incidence of 720 ADEs and 253 potential
200744 Connecticut, related ADEs 1, cohort Study warfarin at 25 NHs warfarin-related ADEs. Incidence of adverse
USA (task) 2003-Mar (medical ADEs warfarin-related events
31, records’ review) (bleeding) was 18.8 per 100
2004) resident-months among
warfarin users. 57% of sever,
11% of serious, 1% of life-
threatening, and 1% of fatal
ADEs were preventable.
Preventable errors most
commonly occur at
prescribing and monitoring
stages.
24. Handler et al To develop a Cross-sectional: Panel of 13 physicians, Signals’ list for Panelists reached consensus
200845 consensus list of Literature search 10 pharmacists, and potential ADRs agreement on 40 signals: 15
Pennsylvania, USA laboratory, followed by a 13 advanced laboratory and medication
(tool/tech) pharmacy, and 2 -rounds of practitioners. combinations, 12
Minimum Data Set modified medication concentrations,
(MDS) signals that Delphi survey 10 antidotes, and 3 Resident
can be used by Assessment Protocols
computer system in (RAPs). Highest consensus
the NH to detect scores (4.6) were for
potential ADRs. naloxone for opioid
analgesics; phytonadione
for warfarin, dextrose,
glucagon, or liquid glucose
for hypoglycemic agents,
and INR for warfarin.
25. Gurwitz, et al To evaluate the Cluster-randomized 1118 residents at 2- The incidence, There were 10.8 ADEs per 100
200846 USA (tool/ efficacy of controlled trial. large long-term care preventability resident-months and 4.0
tech) computerized facilities. and severity preventable events per 100
provider order entry of ADEs resident-months on
(CPOE) with clinical intervention units. CPOE
decision support for with decision support did
preventing ADEs. not reduce the ADE rate or
preventable ADE rate in the
long-term care setting. The
most common medication
categories associated with
preventable ADEs were
antipsychotic agents,
anticoagulants, diuretics,
antiplatelet agents,
cardiovascular drugs,
hypoglycemic agents, and
antidepressants.
26. Barber et al 2009.19 To determine the Mixed method: 256 residents Prevalence and 178 (69.5%) of residents had
UK (organization, prevalence, causes Medication recruited in 55 potential harm one or more MEs. The mean
environment, and and potential harm regimen reviews homes of MEs potential harm from
person) of prescribing, and interviews prescribing, monitoring,
monitoring, with home staff, administration, and
dispensing and doctors and dispensing errors was 2.6,
administration pharmacists 3.7, 2.1, and 2.0 (0 ¼ no
errors in UK care harm, 10 ¼ death),
homes respectively.
Contributing factors from the
89 interviews included
doctors who were not
accessible, did not know the
residents and lacked
information in homes when
prescribing; home staff’s
high workload, lack of
medicines training and drug
round interruptions; lack of
teamwork between home,
and pharmacy.
Miscommunication was
within and between the
home GP practice and
pharmacy. The great
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 11

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

majority of monitoring
errors (90.6%) resulted from
a failure to request
monitoring.
27. Tjia et al 200947 To describe nurses’ Mixed-method: 375 nurses completed Nurses’ Nurses identified several
Connecticut, USA perceptions about A survey and the questionnaire perceptions of barriers to effective nurse-
(organization) barriers to nurse- telephone and 21 nurses nurse-physician physician communication
physician interviews of completed communication include: lack of physician
communication in licensed nurses qualitative openness to
NHs interviews at 26 NHs communication, lack of
professionalism, and
language barriers.
Most frequent barriers were
feeling hurried by the
physician (28%), finding a
quiet place to call (25%) and
difficulty reaching the
physician (21%).
28. Baranzini et al Role of polypharmacy 3 years (July Retrospective 1198 NH residents in a Incidence of falls 695 falls (in 293 residents),
200924 Italy (person) and well known 2004-Dec observational NH 141 residents (48.1%) had
medications on 2007) study fall-induced injuries: 95
incidence of falls (67.4%) minor and 46
(32.6%) major cases.
Risk factors for falls were
antipsychotics,
antihypertensive,
benzodiazepines and
antidepressants,
4 more
medications, and

1 of psychotropic
medications.
29. Verrue et al 201048 To investigate the Mar 2007- Before-after 18 nurses and 28 Rate of Interactive training sessions
Belgium (person) impact of an June interventional nursing aides in 2 administration covered crushing and
educational session 2007 study (direct Belgian NHs. errors splitting of medications,
on the medication observation of drug-drug and drug-food
administration error medication interactions, correct use of
rate. administration inhalation medication,
errors) generic drug names, and
special warnings concerning
the administration of
warfarin and alendronate. In
both NHs, the overall ME
rate (preparation errors and
administration errors)
decreased significantly after
the intervention.
30. Halvorsen et al To describe an Spring 2006 Descriptive 142 NH residents at 3 Drug-related The multidisciplinary
201021 Norway innovative team interventional NHs problems (DRPs) meetings increased
(tool/tech) intervention to study knowledge on drug
identify and resolve (Pharmacist-led treatment and improve the
DRPs in Norwegian medication communication between
NHs. reviews and case nurse, pharmacist and
conferences) physician members, and the
quality of the patients’
treatment. Pharmacists
identified 719 DRPs in 140
residents, of which 504
were confirmed by the
physician and nurses, and
476 interventions were
completed. “Unnecessary
drug” and “monitoring
required” were the most
frequently identified DRPs.
Drugs for treating the
nervous system, the
gastrointestinal tract and
endocrine were most
commonly problematic.
31. Greene et al 201049 To describes the first Oct 1, 2006- Cross-sectional 203 NHs Number and Data available from ME
North Carolina (tool/ year of use of the Sept 30, design characteristics reporting system can be
tech) medication error 2007 of MEs used to reduce MEs. A total
reporting system of 5823 ME reports were
(continued on next page)
12 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

and evaluate how it submitted by 203 sites (52%)

provides rich using the new system
information to target during the reporting year,
patient safety. and a median of 18 ME
reports per site. 612 (10.5%)
MEs were categorized as
serious which included
drugs given to the wrong
patient (RR 4.39), lab-work
error (RR 2.40), wrong
product given (RR 2.22) and
medication overdoses (RR
1.49) and more likely to
occur on second shift.
32. Crespin et al To measure the 3 years (fiscal Cross-sectional 294 NHs. MEs, repeated 15,037 MEs induced 123 (0.8)
201050 North association between years 2006- study Self-reported MEs by MEs and harms. 5615 repeated MEs
Carolina, USA (task) repeat MEs and 2008) MEQI individual ME-induced induced 68 (1.2%) harm
patient harm error system harms (ADEs). Wrong dose errors
were the most common
type (56.5%), and MEs
mostly occurred during the
administration phase
(49.1%) or the
documentation phase
(38.8%).
33.Milligan et al Evaluate ADEs in 5 years (Jan1, Retrospective NHs, residential ADE incidents 684 incidents with insulin and
201151 UK (task) people with DM in 2005-Dec design homes and hospices according to 84 incidents related to oral
care home settings 31, (secondary (National Reporting 5-point scale hypoglycemic agents (OHA).
2009) data analyses) and Learning Service (No harm, 69% of Insulin incidents
Data) dd, death) were due to administration
problem and 24% of OHA
errors were due to wrong/
unclear dose. 94% of insulin
incidents were no/low harm
and 1 death. 92% of OHA
were with no or low harm
34. Dilles et al 201120 To identify and Nov 2008- Mixed-method 12 nurses represented Nurses barriers Expert meetings found 4
Belgium compare the Mar study (started 6 NHs in an expert to safe barrier categories: nurses,
(environment, relevance of barriers 2009 with an expert meeting and 246 medication interdisciplinary
organization and, that nurses in NHs meeting and nurses and 270 CNAs management cooperation, organizational
person) experience in followed by from 20 NHs culture or structure, and
medication a survey) completed the patient or family. The survey
management survey found the most common
barriers were being
interrupted during
preparation, high work
pressure, lack of time for
double-checking
medication during
administration, not
knowing enough on
interactions, not enough
information of the
physicians during
monitoring, barriers in
interdisciplinary
cooperation, not knowing
enough on side-effects, lack
of time to perform the task
with care, limited
accessibility of physicians
and not knowing enough on
therapeutic effects. In
general, barriers in
medication monitoring
were the highest.
35. Lapane et al 201152 To determine whether 2003 and Randomized 491 residents at 25 Incidence of GRAM-triggered monitoring
USA (tool) using a clinical 2004 cluster trial of NHs serviced by 2 potential plans for 491 residents.
informatics tool that GRAM reports long-term care delirium, falls, Residents in the intervention
implements and automated pharmacies. ADEs-induced homes experienced fewer
prospective monitoring plans hospitalizations, falls, less potential delirium,
monitoring plans for falls and and mortality and death, but more
reduces the delirium hospitalizations than in the
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 13

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

incidence of comparison (usual care)

potential delirium, homes.
falls, ADEs-induced
hospitalizations, and
mortality.
36. Wild et al 201153 To evaluate the effects 12 weeks Pre- and A total of 49 staff (35 Staff awareness When NH RNs and residential
England, UK (tool/ of a pharmacy-led training postintervention care staff and 14 of occurrence home care staff use the
tech) barcode medication (Jan 2008- design. RNs) from 13 homes of ‘near PBMS, they suffer less stress
system (PBMS) in Dec 2010) completed the pre- misses’(potential and pressure, are more
care homes PBMS survey. After MEs) aware of MEs, and have
training, 43 fewer interruptions than
interviews and 5 when they use the P-MAR
focus groups system.
conducted.
37. Siddiqi et al 201154 To evaluate the 10 months Mixed methods 6 care homes Number of Stop delirium was
UK (tool/tech) feasibility of an before delirium successfully implemented in
intervention, ‘Stop and after and falls the study homes and
Delirium!’, to intervention reduced the number of falls
prevent delirium in study. and prescribed medications.
care homes for older There was evidence
people supporting positive changes
in staff attitudes and
practice after the
intervention.
38. Desai et al 201155 Evaluate the transition 3 years (fiscal Cross-sectional 394 NHs (MEQI MEs 27,759 MEs of them 2919
North Carolina, USA to NH errors and years 2007- study reporting system) (11%) errors during
(task) their harms on 2009) (Secondary transition to a NH, and these
patients data analyses) transition MEs were more
likely to be harmful. Most
frequent medications were
warfarin, insulin,
lorazepam, hydrocodone,
oxycodone, and furosemide.
39. Field et al 201156 To evaluate warfarin 1 year (2007- Randomized 435 residents at 26 Possible and There were 782 possible
Connecticut, USA management 2008) controlled NHs (13 intervention preventable warfarin-related ADEs, 83 of
(organization) communication trial using and 13 control) warfarin-related them were preventable and
protocol on facilitated ADEs. 183 potential ADEs. The rate
warfarin-safety and telephone of serious preventable
identified discussion events was 0.39 in
preventable ADEs between intervention homes and
resulted from nurses and 0.62 in control homes. The
warfarin physicians. INR measurements were
significantly more
controlled in intervention
homes compared with
control homes.
40. Tjia et al 201257 To test whether 1year (Oct 1, Prospective 435 residents taking Rate of AWEs: Preventable AWEs were 12.8
Connecticut, USA residents with 2007- cohort warfarin (218 of INR
4.5, % for residents with
(person) dementia Dec31, embedded in them with bleeding, low dementia and 9.99 % for
experience higher 2008) a clinical trial dementia) from 26 hematocrit, residents without dementia
rate of adverse- NHs orders for out of those taking warfarin.
warfarin-events vitamin K, and Residents with dementia had
(AWDs) thrombo-embolic higher risk of AWEs
events compared with residents
without dementia.
41. Huybrechts et al To assess risks of 2001-2005 Retrospective 75, 445 new users of Antipsychotic Compared with risperidone,
201258 USA (person) mortality associated cohort study antipsychotic drugs drugs-induced users of haloperidol had an
with use of with secondary mortality increased 180-day risk of
individual data mortality (HR 2.07) and
antipsychotic drugs users of quetiapine a
in elderly residents decreased risk (HR 0.81).
in NHs. The data suggest that the risk
of mortality with these
drugs is generally increased
with higher doses and
seems to be highest for
haloperidol and least for
quetiapine.
42. Mahmood et al To examine the A mixed method Cross-sectional survey Medication Physical design, such as
201259 Canada physical (focus groups, comprised 54 nurses preparation and medication room layout is a
(environment and environments observing working at 4 long- administration major cause of potential
organization) influencing MEs in medication term facilities errors MEs during the medication
preparation and preparation phase. Social
(continued on next page)
14 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

long-term care administration, environment, interruptions,

facilities and a nursing noise, and staff shortages
staff survey) were major sources of errors
during medication
administration.
Staff and organizational
factors that played an
important role included:
insufficient nursing staff
(94.3 %); overwork, stress or
fatigue (86.8 %); poor
training (80.4 %); and
miscommunicating among
practitioners (76.9 %).
Lacking computerized
medical records was
important issue leading to
errors (45.3 %).
43. Desai et al 201360 To characterize 2 years (2010 cross-sectional 396 North Carolina MEs and A total of 32,176 individual
North Carolina, USA analgesic MEs and to and 2011) analysis NHs (MEQI data) analgesic-induced ME incidents were reported
(person) evaluate their MEs over a 2-year period, 12.3%
association with (n ¼ 3949) of which were
patient harm. analgesic MEs. Of these
analgesic MEs, opioid
involved in 3105 and
nonopioid analgesics in 844
errors. Opioid errors were
more likely to be wrong
drug errors, wrong dose
errors, and administration
errors compared with
nonopioid errors
44. Lemay et al 201361 To describe NH June 2011 Cross-sectional 138 NH leaders and knowledge of Only 24% of NH leaders
Connecticut, USA leadership and design. Survey 779 direct care staff antipsychotic identified at least 1 severe
(person) direct care staff of leadership adverse events adverse effect of
members’ and direct care for residents antipsychotics; 13% of LPNs
knowledge of staff of nursing with dementia and 12% of RNs listed at least
antipsychotic ADEs homes 1 severe adverse effect.
and perceptions Leaders were satisfied with
about the the training that staff
effectiveness of received to manage
antipsychotics residents with challenging
behaviors (62%).
45. Handler et al To quantify the use 2010 Cross-sectional 558 valid surveys from Use of mobile 236 of participating NH
201362 USA and perceived study 800 NH physicians devices, and physicians used mobile
(technology) benefits of mobile who attended the beliefs about devices to assist with
devices in 2010 American the effectiveness prescribing in the NH.
preventing ADEs in Medical Directors of drug reference Physicians with 15 or fewer
the NH setting Association Annual software in years of clinical experience
through surveying Symposium preventing ADEs. were 67% more likely to be
nationally mobile device users. Almost
representative all (98%) user physicians
sample of NH reported performing an
physicians. average of 1 or more drug
lookups a day to assist with
prescribing, and 1 to 2
lookups a day for potential
drug-drug interactions, and
most (88%) believed that
drug reference software had
prevent at least 1 potential
ADE in the last month.
46. Desai et al 201363 To characterize 2 years (2010 Cross-sectional Residents of 396 NHs MEs and There were 32,176 ME
North Carolina anticoagulant MEs and 2011) study. taking warfarin, anticoagulant incidents. Of these errors,
(person and task) and to evaluate their heparin, or MEs 1623 (5%) were
association with enoxaparin (MEQI anticoagulant MEs. 291
patient harm reporting system) (0.95%) were harmful MEs
and 29 (1.79%) harmful
anticoagulant errors.
Anticoagulant errors were
significantly more likely to
cause patient harm
compared with all other
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 15

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

MEs. Anticoagulant errors

were more likely due to
wrong follow-up (33%), and
shift change and more
common among residents
with
75 year-old.
47. Desai et al 201364 To evaluate the 2 years (2010 Cross-sectional All the residents at 398 MEs 32,176 individual ME
North Carolina (task) relationship and 2011) study NHs (MEQI and 2004 incidents. Harm (ADEs)
between drug class National Nursing ranges from 0.3 % to 2.8 %.
utilization rates and Home Survey CNS agents accounted for
their involvement in (NNHS)) 33.17%, cardiovascular
MEs agents for 9.74%, metabolic
agents for 8.75%, anti-
infective agents for 6.48%
and anticoagulants for 6.47%
of the total MEs. Most MEs
were due to wrong dose
(43.8%-56.9%) errors and
wrong follow-up (15.7%-
38.1%).
48. Walsh et al 201465 To evaluate the effect 2004-2010 Retrospective NHs in 8 southeastern Nurse numbers, Use of certified medication
NC, USA (person) of medication aide (Staffing levels U.S. states number and aide had no statistically
use on other nurse and inspection severity of MEs significant reduction in RN
staffing, deficiencies, deficiencies and number of or LPN use, and had no effect
and Nursing Home from the Online citations on deficiencies for
Quality Initiative Survey and (penalties) significant or harmful MEs.
(NHQI) health Certification received. Increased medication aide
outcome measures and Reporting use was associated with
System and fewer pharmacy citations
NHQI data) and decreased the
probability that a facility
received a deficiency
citation for unnecessary
drug use or having a
medication error rate
greater than or equal to 5%.
49. Dilles et al 201366 To develop and assess Nov 2010 Intervention study 418 residents at 8 NHs Number of Healthcare professionals gave
Belgium (tool/tech) the effect of a with a pre- and ADRs detected the Pharmanurse
software post- test design by nurses, and intervention a score of 7 of
(Pharmanurse) to (interdisciplinary ADRs confirmed 10 for the potential to
facilitate nurse- medication by general improve pharmacotherapy,
driven ADR review) practitioners and 83% of the physicians
screening in NHs. were satisfied about nurses’
screening for ADRs. Nurses
observed 1527 potential
ADRs in 81% of the 418
residents (mean per
resident 3.7). Physicians
confirmed 821 ADRs in 60%
of the residents (mean per
resident 2.0)
50. Marcum et al To determine the 1 month Sept Retrospective chart 321 veteran residents Potential ADE The overall (PPV of the ADE
201367 USA (tool/ utility of an ADE 29, 2010- review using in 3 VA nursing trigger tool was 40.1% (65/
tech) trigger tool in VA Oct 27 ADE triggers facilities 162), and the average time
nursing facilities and 29, 2010. - 14 lab results to complete resident
to describe the most and 13 drug assessments was
common types of concentrations 8.8 minutes. 50.5% (n ¼ 162)
potential ADEs of veterans had at least 1
detected with the abnormal laboratory value
trigger tool. contained in the trigger tool
over a month.
The most common potential
ADEs were AKI,
hypokalemia,
hypoglycemia, and
hyperkalemia.
51. Lin et al 201326 To measure the 1 year (2010) Retrospective 184 elderly patients at PIMs and After investigating the
Taiwan (person) relationship chart review 4 long-term care ADE-induced presence or absence of PIMs
between presence of institutions hospitalization for patients transferred to
PIMs and ADE- the hospital with ADEs and
induced unexpected illnesses, the
hospitalizations results showed no
statistically significant
(continued on next page)
16 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

differences between the 2

groups of patients (with or
without PIMs).
52. Baril et al 201468 To study a medication Nov 2008- Pre- and 800 patients at 6 NHs MEs MDT detected MEs faster
Quebec Province, distribution Mar 2012 postinterventional (data collection used resulting in less severe
Canada (tool/tech) technology’s (MDT) study ( an voluntary reporting consequences for patients.
impact on MEs in automated process by nursing MDT might detect a
public NHs pharmacy staff) medication not
packaging device administered to a patient
combined immediately at the next
with mobile administration and that can
dispensing significantly reduce MEs in
carts) NHs.
53. Grace et al 201425 To determine the 1 year (2011) Retrospective Acutely ill 165 PIMs and 157 patients (95.2%) were
Ireland (tool/tech) prevalence of PIM cohort study residents from 22 PIM-related prescribed at least 1 PIM; 50
use in NH residents (medication NHs seeking medical ED admission. patients (30.3%) were
aged
65 years review using assistance at the ED. considered to have a
presenting to the ED, Beers and correlation between their
and to identify the STOPP criteria) ED admission; and the PIM
potential role of prescribed. The use of
PIMs in ED criteria such as Beers and
attendances. STOPP may be a useful guide
for physicians coordinating
the long-term care of NH
residents and may have the
potential to reduce
attendances at ED. The most
common cause for attending
the ED was a fall (27.3%).
54. Jordan et al 201469 To explore feasibility Nov 2012- Before-and-after 6 nurses and 11 ADRs 141 new and 56 follow-up
UK (task) and clinical impact Feb 2013 intervention service users at 3 problems (ADRs) potentially
of nurse-led study (record care homes in Wales. associated with
medication review using antipsychotic,
monitoring using the WWADR profile antidepressant, or
West Wales ADR listing 80 antiepileptic medicines
Profile for mental questions) have been identified by the
health medicines profile. The nurses needed
20-25 minutes to
administer the profile. The
service users benefited from
the nurse-led medication
monitoring.
55. Handler et al To determine the 1 year (Feb Retrospective All residents at 4 NHs Drug-induced 249 residents had 668 drug-
201470 incidence of drug- 2012-Feb design Clinical AKI using induced AKI alerts with rate
Pennsylvania, USA associated acute 2013) Surveillance RIFLE criteria of 0.41 cases per 100
(person) kidney injury (AKI) Software System resident days. Most
using the RIFLE to monitor lab common associated
(SCr) and medications were diuretics,
medication data) ACEIs/ARBs, and antibiotics.
A person with a length of
stay of 100 days had a 30%
probability of having at least
one drug-associated AKI
event.
56. Lane et al 201471 To evaluate structure- 1 year (Oct Retrospective 138 SNFs that reported MEs and 581 transition MEs and of
North Carolina, USA process-related 2006- Sept design (MEQI MEs ME-induced them 73 (12.6%) caused
(task) factors that 2007) data). harms harm.
contribute to MEs 35.6 % of documenting errors
and harm during and 30.1% of administering
transition period at a errors caused harms.
SNF Prescribing MEs were much
less frequent than
administration MEs, but
were much more likely to
induce harm.
57. Culley et al 201572 To determine whether 6 months Retrospective 1101 residents at 4 Hypoglycemic Hypoglycemia can be
Pennsylvania, USA a clinical (Oct cohort. NHs alerts (FSBG detected using a clinical
(person and tool/ surveillance system 2012- April The computer 70 mg/dL) surveillance system; 772
tech) can be used to detect 2013) generated alerts alerts involving 141
drug-induced prospectively residents were detected.
hypoglycemia during the The incidence of drug-
events and measure routine care induced hypoglycemia
(TheraDoc) events was 9.5 per 1000
(continued on next page)
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19 17

Table 2 (continued )

Author, Year, Country, Main Objective(s) Period of Design No. of Participants/ Outcome Measure Selected Findings
SEIPS domain Study settings

their incidence in NH resident-days (0.95 per 100

residents. resident-days). 90 (63.8%)
residents had a glucose level
of 55 mg/dL or less. Insulin
orders were associated with
762 (98.7%) alerts while the
rest of alerts caused by oral
hypoglycemic agents
(sulfonylurea and
meglitinide derivative) and
fluoroquinolone
58. Dilles et al 201573 To assess the value of Dec 2009 Cross-sectional 68 residents at 2 NHs Potential ADRs The correspondence between
Belgium (person) NH residents’ ADR design (residents nurse and resident reports
reports. were structurally ranged from 43% (dry
interviewed and mouth) to 88 %
nurse surveyed) (arrhythmia). Over 90% of
the residents used
medications causing tired/
sedates, arrhythmias,
abdominal pain,
constipation, nausea/
vomiting, confusion,
drowsiness, muscle pain/
weakness. The medications
included nervous system,
cardiovascular, GIT, and
anticoagulant medications.
59. Daneman et al To determine whether 2 years (Jan Retrospective 110, 656 older NH Antibiotic-related The high-use NH was
201574 Ontario, high antibiotic use is 2010-Dec longitudinal residents from 607 ADEs correlated with an increased
Canada (person) associated with a 2011) open-cohort NHs in Ontario risk of antibiotic-related
greater risk of study (using adverse events (OR 1.24;
antibiotic-induced population-based P ¼ .003). Antibiotic-related
adverse events administrative harms included Clostridium
among NH residents databases) difficile, diarrhea or
gastroenteritis, antibiotic-
resistant organisms, allergic
reaction.
Penicillins were the most
commonly prescribed
agents, followed by
fluoroquinolones,
Sulfonamides, first-
generation cephalosporins,
nitrofurantoin, and
macrolides.
60. Berry et al 201675 To determine whether 2.5 years Case-crossover 594 NH residents from Fall incidents The risk of falls was higher in
Massachusetts, USA there is an acute (Sept design 2 facilities who fell at the 24 hours following
(person) increased risk of falls 2010-May least once within benzodiazepine initiation
in the days following 2013) 2.5 years compared with other times
a change to an (OR 3.79). There was no
antipsychotic or significant difference in risk
benzodiazepine following antipsychotic
initiation (OR 2.42).
Discontinuing a
benzodiazepine was
associated with a
significantly reduced risk of
fall (OR 0.26).
Benzodiazepines probably
cause fall due to their
influence on memory and
balance.

ADWE, adverse drug withdrawal events; CNS, central nervous system; CPOE, computerized provider order entry; ED, emergency department; FSBG, finger-stick blood
glucose; GIT, gastrointestinal tract; GRAM, Geriatric Risk Assessment MedGuide; HR, hazard ratio; INR, international normalized ratio; MDS, Minimum Data Set; MDT,
medication distribution technology; MEPS, Medical Expenditure Panel Survey; MEQI, Medication Error Quality Initiative; NHQI, Nursing Home Quality Initiative; PBMS,
pharmacy-led barcode medication system; P-MAR, paper-based medication administration record; PPV, positive predictive value; RAP, resident assessment protocol; RIFLE,
risk, injury, failure, loss of kidney function, or end-stage kidney disease; RR, risk ratio; STOPP, Screening Tool of Older Person’s Prescriptions; VA, Veterans Affairs; VHA,
Veterans Health Administration.
18 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
reporting system is a barrier to safe medication management within
NHs.42 In addition, the data available in ME reporting systems can be
used by scholars, healthcare practitioners, and policy maker to
enhance NH resident medication safety.
Of the 4 studies that found poor monitoring to be commonly
associated with MEs or ADEs, 3 studies examined warfarin that re-
quires close and frequent monitoring to avoid bleeding.21,44,63,64
Monitoring errors were commonly associated with warfarin treat-
ment because NH practitioners failed to measure the
international normalized ratio frequently or to respond to out of range
international normalized ratio values.44 Documentation errors were
also commonly associated with MEs in 3 North Carolina studies.43,55,71
These documentation errors may be due to heavy workload, time
pressure, or underestimating of consequences. Overall, this review
shows that MEs or ADEs are most often due to administering, moni-
toring, and/or documenting errors, with much lower rates in the
phases of prescribing and dispensing errors.
Several facility environmental characteristics such as layout, noise,
interruption, and distraction can affect medication safety in NHs. Four
out of 5 of the environment-related studies were mixed method
design19,20,41,59 and 2 of them either observed medication manage-
ment processes41 or interviewed staff19 whereas the other 2 studies
surveyed the nursing staff about the effect of facility environment
such as distractions and interruptions on medication safety.20,59
Limitations
In this review, 1 author (A.A.) systematically evaluated studies
found using PubMed and CINAHL only. However, these are the most
common database for health-related articles. Sixteen studies were
added through review of references since they were related to the 5
work system factors.
Conclusions
The incidence rates of ADEs in NHs ranged from 1.89 to 10.8 per
100 resident-months. Personal characteristics of NH residents such as
age, number and types of scheduled medications, and number and
types of comorbidities influence incidence rates of ADEs, ADRs, or
harmful MEs. NH staff medication knowledge and training affect the
incidence of administering MEs. The most common ADE-related
medications were psychotropic agents, anticoagulants, antidiabetics,
opioids, and antihypertensive agents. Common ADEs were bleeding,
hypoglycemia, falls, fractures, and constipation. Organizational factors
such as interprofessional collaboration, physician and pharmacist
accessibility, and staff number play essential roles in preventing MEs.
Several environmental characteristics such as staff distraction and
interruption influence medication safety in NHs. Different types of
tools and technologies can be helpful in minimizing and identifying
ADEs or PIM. Workload and time pressure can negatively impact NH
staff. ADEs or MEs are most often due to administering, monitoring,
and documenting errors. The SEIPS model is a useful framework for
identifying factors associated with medication safety in NHs.
References
1. Center for Medicare and Medicaid Services (CMS). Nursing Home Data Com-
pendium 2015 Edition. Available at: https://www.cms.gov/Medicare/Provider-
Enrollment-and-Certification/CertificationandComplianc/Downloads/nursinghome
datacompendium_508-2015.pdf. Accessed June 17, 2016.
2. Harris-Kojetin L, Sengupta M, Park-Lee E, et al. Long-term care providers and
services users in the United States: Data from the National Study of Long-Term
Care Providers, 2013e2014. National Center for Health Statistics. Vital Health
Stat 2016;3(38). Available at: https://www.cdc.gov/nchs/data/series/sr_03/
sr03_038.pdf. Accessed June 17, 2016.
3. Lee Adler JM, Federico FA. IHI skilled nursing facility trigger tool for measuring
adverse events. Available at: http://www.ihi.org/resources/Pages/Tools/
SkilledNursingFacilityTriggerTool.aspx?utm_campaign¼tw&utm_source¼hs_
email&utm_medium¼email&utm_content¼29848197&_hsenc¼p2ANqtz-_pkhk0v
DijmBp7enatjRqHG_yU1DJxVjuKt0TDxFQbWMRyZ8B4LMNN7bFK9UQDGsFv5kres
0yuzgaGuBVRaY6hfOed5A&_hsmi¼29848197. Accessed May 27, 2016.
4. Levinson DR. Adverse events in skilled nursing facilities: national incidence
among medicare beneficiaries. Department of Health and Human Services,
Office of Inspector General. Available at: http://oig.hhs.gov/oei/reports/oei-06-
11-00370.pdf. Accessed June 2, 2016.
5. Ferner RE, Aronson JK. Clarification of terminology in medication errors:
definitions and classification. Drug Saf 2006;29:1011e1022.
6. Center for Clinical Standards and Quality SCG. Adverse Drug Event Trigger Tool.
In: Services DohH, ed. Baltimore, MD: Center for Medicare and Medicaid
Services, 2015. Available at: https://www.cms.gov/Medicare/Provider-Enroll
ment-and-Certification/SurveyCertificationGenInfo/Downloads/Survey-and-Cert-
Letter-15-47.pdf. Accessed November 12, 2015.
7. Handler SM, Wright RM, Ruby CM, Hanlon JT. Epidemiology of medication-
related adverse events in nursing homes. Am J Geriatr Pharmacother 2006;4:
264e272.
8. Aronson JK, Ferner RE. Clarification of terminology in drug safety. Drug Saf
2005;28:851e870.
9. Gurwitz JH, Field TS, Avorn J, et al. Incidence and preventability of adverse drug
events in nursing homes. Am J Med 2000;109:87e94.
10. Bain KT. Adverse drug reactions and current state of drug regimen review in
nursing facilities: Need for a change? Consult Pharm 2007;22:586e592.
11. Carayon P, Schoofs Hundt A, Karsh BT, et al. Work system design for patient
safety: The SEIPS model. Qual Saf Health Care 2006;15:i50ei58.
12. Holden RJ, Carayon P, Gurses AP, et al. SEIPS 2.0: A human factors framework
for studying and improving the work of healthcare professionals and patients.
Ergonomics 2013;56:1669e1686.
13. Carayon P, Wood KE. Patient safety: The role of human factors and systems
rngineering. Studies Health Technol Inform 2010;153:23e46.
14. Scott-Cawiezell J, Vogelsmeier A. Nursing home safety: A review of the liter-
ature. Ann Rev Nurs Res 2006;24:179e215.137.
15. Pepper GA, Towsley GL. Medication errors in nursing homes: Incidence
and reduction strategies. J Pharmaceut Finance Economics Policy 2007;16:
5e133.
16. Schatz SNaW, Robert J. Adverse drug reactions. PSAP, CNS/Pharmacy Practice,
2015; Available at: https://www.accp.com/docs/bookstore/psap/2015B2.
SampleChapter.pdf. Accessed June 2, 2016.
17. Gerety MB, Cornell JE, Plichta DT, Eimer M. Adverse events related to drugs and
drug withdrawal in nursing home residents. J Am Geriatr Soc 1993;41:
1326e1332.
18. Gurwitz JH, Field TS, Judge J, et al. The incidence of adverse drug events in two
large academic long-term care facilities. Am J Med 2005;118:251e258.
19. Barber ND, Alldred DP, Raynor DK, et al. Care homes’ use of medicines study:
Prevalence, causes and potential harm of medication errors in care homes for
older people. Qual Safety Health Care 2009;18:341e346.
20. Dilles T, Elseviers MM, Van Rompaey B, et al. Barriers for nurses to safe
medication management in nursing homes. J Nurs Scholarsh 2011;43:
171e180.
21. Halvorsen KH, Ruths S, Granas AG, Viktil KK. Multidisciplinary intervention to
identify and resolve drug-related problems in Norwegian nursing homes.
Scand J Prim Health Care 2010;28:82e88.
22. Ruths S, Straand J, Nygaard HA. Multidisciplinary medication review in nursing
home residents: What are the most significant drug-related problems? The
Bergen District Nursing Home (BEDNURS) study. Qual Saf Health Care 2003;12:
176e180.
23. van Dijk KN, de Vries CS, van den Berg PB, et al. Constipation as an adverse
effect of drug use in nursing home patients: An overestimated risk. Br J Clin
Pharmacol 1998;46:255e261.
24. Baranzini F, Diurni M, Ceccon F, et al. Fall-related injuries in a nursing home
setting: Is polypharmacy a risk factor? BMC Health Serv Res 2009;9:228.
25. Grace AR, Briggs R, Kieran RE, et al. A comparison of beers and STOPP criteria
in assessing potentially inappropriate medications in nursing home
residents attending the emergency department. J Am Med Dir Assoc 2014;15:
830e834.
26. Lin JJ, Yeh TY, Lau HL. Elderly patients with inappropriate medication corre-
lations with adverse drug events or unexpected illnesses in long-term care
institutions. Aging Male 2013;16:173e176.
27. Soon JA. Assessment of an adverse drug reaction monitoring program in
nursing homes. Can J Hosp Pharm 1985;38:120e125.
28. Cooper JW. Probable adverse drug reactions in a rural geriatric nursing home
population: A four-year study. J Am Geriatr Soc 1996;44:194e197.
29. Cooper JW. Adverse drug reaction-related hospitalizations of nursing facility
patients: A 4-year study. South Med J 1999;92:485e490.
30. Field TS, Gurwitz JH, Avorn J, et al. Risk factors for adverse drug events among
nursing home residents. Arch Intern Med 2001;161:1629e1634.
31. Barker KN, Flynn EA, Pepper GA, et al. Medication errors observed in 36 health
care facilities. Arch Intern Med 2002;162:1897e1903.
32. Weinberg AD, Lesene AJ, Richards CL, Pals JK. Quality care indicators and
staffing levels in a nursing facility subacute unit. J Am Med Dir Assoc 2002;3:
1e4.
33. Lapane KL, Hughes CM. Identifying nursing home residents at high risk for
preventable adverse drug events: Modifying a tool for use in the Fleetwood
Phase III Study. Consult Pharm 2004;19:533e537.
 A.A. Al-Jumaili, W.R. Doucette / JAMDA xxx (2017) 1e19
34. Perri M III, Menon AM, Deshpande AD, et al. Adverse outcomes associated with
inappropriate drug use in nursing homes. Ann Pharmacother 2005;39:
405e411.
35. Scott-Cawiezell J, Vogelsmeier A, McKenney C, et al. Moving from a culture of
blame to a culture of safety in the nursing home setting. Nurs Forum 2006;41:
133e140.
36. Nguyen JK, Fouts MM, Kotabe SE, Lo E. Polypharmacy as a risk factor for
adverse drug reactions in geriatric nursing home residents. Am J Geriatr
Pharmacother 2006;4:36e41.
37. Hughes CM, Lapane KL. Nurses’ and nursing assistants’ perceptions of pa-
tient safety culture in nursing homes. Int J Qual Health Care 2006;18:
281e286.
38. Hansen RA, Greene SB, Williams CE, et al. Types of medication errors in North
Carolina nursing homes: A target for quality improvement. Am J Geriatr
Pharmacother 2006;4:52e61.
39. French DD, Werner DC, Campbell RR, et al. A multivariate fall risk assessment
model for VHA nursing homes using the minimum data set. J Am Med Dir
Assoc 2007;8:115e122.
40. Spector W, Shaffer T, Potter DE, et al. Risk factors associated with the occur-
rence of fractures in U.S. nursing homes: Resident and facility characteristics
and prescription medications. J Am Geriatr Soc 2007;55:327e333.
41. Scott-Cawiezell J, Pepper GA, Madsen RW, et al. Nursing home error and level
of staff credentials. Clin Nurs Res 2007;16:72e78.
42. Handler SM, Perera S, Olshansky EF, et al. Identifying modifiable barriers to
medication error reporting in the nursing home setting. J Am Med Dir Assoc
2007;8:567e574.
43. Pierson S, Hansen R, Greene S, et al. Preventing medication errors in long-term
care: Results and evaluation of a large scale web-based error reporting system.
Qual Saf Health Care 2007;16:296e302.
44. Gurwitz JH, Field TS, Radford MJ, et al. The safety of warfarin therapy in the
nursing home setting. Am J Med 2007;120:539e544.
45. Handler SM, Hanlon JT, Perera S, et al. Consensus list of signals to detect po-
tential adverse drug reactions in nursing homes. J Am Geriatr Soc 2008;56:
808e815.
46. Gurwitz JH, Field TS, Rochon P, et al. Effect of computerized provider order
entry with clinical decision support on adverse drug events in the long-term
care setting. J Am Geriatr Soc 2008;56:2225e2233.
47. Tjia JB, Briesacher A, McGee BA, et al. Medication discrepancies upon hospital
to skilled nursing facility transitions. J Gen Intern Med 2009;24:630e635.
48. Verrue CL, Mehuys E, Somers A, et al. Medication administration in nursing
homes: Pharmacists’ contribution to error prevention. J Am Med Dir Assoc
2010;11:275e283.
49. Greene SB, Williams CE, Pierson S, et al. Medication error reporting in nursing
homes: Identifying targets for patient safety improvement. Qual Saf Health
Care 2010;19:218e222.
50. Crespin DJ, Modi AV, Wei D, et al. Repeat medication errors in nursing homes:
Contributing factors and their association with patient harm. Am J Geriatr
Pharmacother 2010;8:258e270.
51. Milligan FJ, Krentz AJ, Sinclair AJ. Diabetes medication patient safety incident
reports to the National Reporting and Learning Service: The care home setting.
Diabet Med 2011;28:1537e1540.
52. Lapane KL, Hughes CM, Daiello LA, et al. Effect of a pharmacist-led multi-
component intervention focusing on the medication monitoring phase to
prevent potential adverse drug events in nursing homes. J Am Geriatr Soc
2011;59:1238e1245.
53. Wild D, Szczepura A, Nelson S. New barcode checks help reduce drug round
errors in care homes. Nursing Manage UK 2011;18:25e30.
54. Siddiqi N, Young J, House AO, et al. Stop delirium! A complex intervention to
prevent delirium in care homes: A mixed-methods feasibility study. Age
Ageing 2011;40:90e98.
55. Desai R, Williams CE, Greene SB, et al. Medication errors during patient tran-
sitions into nursing homes: Characteristics and association with patient harm.
Am J Geriatr Pharmacother 2011;9:413e422.
56. Field TS, Tjia J, Mazor KM, et al. Randomized trial of a warfarin communication
protocol for nursing homes: An SBAR-based approach. Am J Med 2011;124:
179.e1e179.e7.
57. Tjia J, Field TS, Mazor KM, et al. Dementia and risk of adverse warfarin-related
events in the nursing home setting. Am J Geriatr Pharmacother 2012;10:
323e330.
58. Huybrechts KF, Gerhard T, Crystal S, et al. Differential risk of death in older
residents in nursing homes prescribed specific antipsychotic drugs: Population
based cohort study. BMJ 2012;344:e977.
19
59. Mahmood A, Chaudhury H, Gaumont A, Rust T. Long-term care physical
environmentsdEffect on medication errors. Int J Health Care Qual Assur 2012;
25:431e441.
60. Desai RJ, Williams CE, Greene SB, et al. Analgesic medication errors in
North Carolina nursing homes. J Pain Palliat Care Pharmacother 2013;27:
125e131.
61. Lemay CA, Mazor KM, Field TS, et al. Knowledge of and perceived need for
evidence-based education about antipsychotic medications among nursing
home leadership and staff. J Am Med Dir Assoc 2013;14:895e900.
62. Handler SM, Boyce RD, Ligons FM, et al. Use and perceived benefits of mobile
devices by physicians in preventing adverse drug events in the nursing home.
J Am Med Dir Assoc 2013;14:906e910.
63. Desai RJ, Williams CE, Greene SB, et al. Anticoagulant medication errors in
nursing homes: Characteristics, causes, outcomes, and association with patient
harm. J Healthcare Risk Manag 2013;33:33e43.
64. Desai RJ, Williams CE, Greene SB, et al. Exploratory evaluation of medication
classes most commonly involved in nursing home errors. J Am Med Dir Assoc
2013;14:403e408.
65. Walsh JE, Lane SJ, Troyer JL. Impact of medication aide use on skilled nursing
facility quality. Gerontologist 2014;54:976e988.
66. Dilles T, Vander Stichele RH, Van Bortel LM, Elseviers MM. The development
and test of an intervention to improve ADR screening in nursing homes. J Am
Med Dir Assoc 2013;14:379.e1e379.e6.
67. Marcum ZA, Arbogast KL, Behrens MC, et al. Utility of an adverse drug event
trigger tool in Veterans Affairs nursing facilities. Consult Pharm 2013;28:99e109.
68. Baril C, Gascon V, Brouillette C. Impact of technological innovation on a nursing
home performance and on the medication-use process safety. J Med Syst 2014;
38:1e12, 22.
69. Jordan S, Gabe M, Newson L, et al. Medication monitoring for people with
dementia in care homes: The feasibility and clinical impact of nurse-led
monitoring. Sci World J 2014;2014:843621.
70. Handler SM, Cheung PW, Culley CM, et al. Determining the incidence of drug-
associated acute kidney injury in nursing home residents. J Am Med Dir Assoc
2014;15:719e724.
71. Lane SJ, Troyer JL, Dienemann JA, et al. Effects of skilled nursing facility
structure and process factors on medication errors during nursing home
admission. Health Care Manage Rev 2014;39:340e351.
72. Culley CM, Perera S, Marcum ZA, et al. Using a clinical surveillance system to
detect drug-associated hypoglycemia in nursing home residents. J Am Geriatr
Soc 2015;63:2125e2129.
73. Dilles T, Van Rompaey B, Van Bogaert P, Elseviers MM. Resident and nurse
reports of potential adverse drug reactions. Eur J Clin Pharmacol 2015;71:
741e749.
74. Daneman N, Bronskill SE, Gruneir A, et al. Variability in antibiotic use across
nursing homes and the risk of antibiotic-related adverse outcomes for indi-
vidual residents. JAMA Intern Med 2015;175:1331e1339.
75. Berry SD, Placide SG, Mostofsky E, et al. Antipsychotic and benzodiazepine drug
changes affect acute falls risk differently in the nursing home. J Gerontol Ser A
Biol Sci Med Sci 2016;71:273e278.
76. Alldred DP, Raynor DK, Hughes C, et al. Interventions to optimise prescribing
for older people in care homes. Cochrane Database Syst Rev 2013;2:
CD009095.
77. Mahmood A, Chaudhury H, Gaumont A. Environmental issues related to
medication errors in long-term care: Lessons from the literature. Herd 2009;2:
42e59.
78. Rozich JD, Haraden CR, Resar RK. Adverse drug event trigger tool: A practical
methodology for measuring medication related harm. Qual Saf Health Care
2003;12:194e200.
79. Boyce RD, Perera S, Nace DA, et al. A survey of nursing home physicians to
determine laboratory monitoring adverse drug event alert preferences. Appl
Clin Inform 2014;5:895e906.
80. Resar RK, Rozich JD, Classen D. Methodology and rationale for the mea-
surement of harm with trigger tools. Qual Saf Health Care 2003;12:
ii39eii45.
81. Bardet JD, Vo TH, Bedouch P, Allenet B. Physicians and community pharmacists
collaboration in primary care: A review of specific models. Res Social Adm
Pharm 2015;11:602e622.
82. Dilles T, Vander Stichele R, Van Rompaey B, et al. Nurses’ practices in phar-
macotherapy and their association with educational level. J Adv Nurs 2010;66:
1072e1079.
83. Handler SM, Hanlon JT. Detecting adverse drug events using a nursing home
specific trigger tool. Ann Longterm Care 2010;18:17e22.