Nurses’ Role in Identifying Mild

Cognitive Impairment in Older

Adults
Amanda Floetke Elliott, PhD, ARNP, Ann L. Horgas, PhD, RN,
and Michael Marsiske, PhD

Mild cognitive impairment (MCI), a relatively to recognize and provide care to patients with
new descriptive category, is believed to rep- dementia, it is also important to recognize pa-
resent a stage between normal aging and tients who are experiencing predementia cogni-
early dementia. Nurse practitioners, who tive changes. As with many other diseases, early
provide care for older adults across a variety detection is essential for implementing treat-
of settings, are in a key position to detect ment regimes in an effort to minimize the cata-
early cognitive changes. The purpose of this strophic effects of dementia. However, these
study is to describe an approach to identify- subtle changes are often not detected through
ing MCI using a variety of measures and a commonly used clinical screening tools for de-
consensus conference with neuropsycholo- mentia, such as the Mini-Mental State Exam
gists. The study was conducted in a sample of (MMSE).1 The term mild cognitive impairment
130 elderly participants (aged 82.5 years; (MCI) is most commonly used to describe this
81% female) residing in nursing homes, as- state between normal aging and early dementia,
sisted living facilities, and senior housing. although several alternative labels have been
A team of clinicians (neuropsychologists used in the literature (e.g., Age-associated Mem-
and nurses) reviewed cognitive, mental ory Impairment [AAMI], Cognitive Impairment
health, and demographic data in consensus Not Dementia [CIND], and Amnestic Mild Cog-
conference and classified study participants nitive Impairment [AMCI]). Research has shown
into 1 of 3 groups: cognitively intact (50.8%), that persons with MCI are at increased risk for
amnestic MCI (19.2%), or probable dementia developing dementia.2,3 Therefore, they may
(30%). Discriminant function analysis (DFA) benefit from early initiation of treatment strate-
was used to independently classify individu- gies to slow progression to dementia. This arti-
als into cognitive status groups based on test cle summarizes the MCI concept for practitio-
scores alone and to compare quantitatively ners who may encounter the syndrome in their
determined groups with consensus confer- practice but may be unfamiliar with its detec-
ence evaluations. The results indicate that tion, and demonstrates the procedures by which
the DFA correctly classified 95% of the par- MCI was identified in a study of residentially
ticipants. Further, results revealed a pattern heterogeneous older adults. The major focus of
in which persons with amnestic MCI have this study is to help make practitioners more
subtle memory impairments (similar to per- aware of the continuum of cognitive functioning
sons with dementia) but that more general they may encounter and to sensitize them to a
cognitive functioning remains high (similar broader set of tools (beyond simple mental sta-
to intact persons). Nurse practitioners’ height- tus inventories) they may wish to incorporate
ened awareness of subtle distinctions in the into their practice.
dimensions of cognitive status associated with Identifying MCI has not been an easy task
MCI can enhance their practice and assist because there has been little agreement in the
them in making more informed referrals for literature on definitions, classification, and mea-
dementia evaluations. (Geriatr Nurs 2008;29: surement tools to identify persons in this state.
38-47) As a result, most of the studies conducted to
date have used varying classification criteria
urse practitioners who work in a variety and measurements to assess MCI. This inconsis-

N of practice settings commonly provide

care to older adults with varying levels of
cognitive ability. Although nurses are educated
tency has made it difficult to directly compare
study results. For instance, across studies, the
rate of MCI ranges from 1% to 26% in elderly,

38 Geriatric Nursing, Volume 29, Number 1

 study, which was initially designed only to dis-
Table 1. tinguish demented individuals (with primary
Full Criteria for Identifying Mild memory impairment) from nondemented indi-
Cognitive Impairment viduals, we identified a broad “middle group” of
elders with focal memory impairment but who
● Cognitive complaints reported by the were otherwise quite cognitively normal. The
patient or their family nature of our sample, selected to study persis-
● Patient or informant report of a decline in tent pain, meant that almost 100% indicated
cognition and/or functional performances some form of functional impairment. Thus,
compared with previous abilities
functional impairment was not a useful discrim-
● Cognitive disorder evidenced by clinical
evaluation (impairment in memory or other inator of persons in our sample. Moreover, given
cognitive domain) ambiguity in the literature regarding what de-
● Absence of major repercussions on daily gree of functional impairment is a feature of
life (although may have difficulties with MCI, we chose to focus primarily on the cogni-
complex day-to-day activities) tive features of the syndrome.
● No evidence of dementia Additionally, we sought to identify differ-
Data from Portet, et al.10 ences between cognitive groups regarding de-
mographic and residential characteristics. In
previous research, the majority of studies have
population-based samples.4-7 Obtaining precise found individuals with MCI to be older and have
estimates of MCI is further complicated by the the same or less education than cognitively nor-
fact that many investigators use statistical crite- mal controls.4,6,13-16 Women are typically classi-
ria in identifying MCI cases (e.g., cognitive func- fied with MCI at a higher rate than men,5,13,14
tioning 1.5 SD below normative means), which and persons with MCI perform lower on the
means that in a normally distributed sample MMSE than normal control individuals, al-
about 7% of cases will be identified as having though these scores are usually above the cutoff
MCI. for dementia.13,14,16 Persons with MCI are also
Recently, researchers and clinicians have more likely to have greater functional deficits
converged on methods of detecting MCI.8,9 The than normal controls (i.e., difficulty with toilet-
MCI working group of the European Consor- ing, mobility, bathing, and use of the tele-
tium of AD10 has published updated criteria for phone).16-18
identifying MCI.9,10 This working group has sug- Early detection and increased accuracy in
gested that MCI should first be classified accord- identifying MCI is important because it enables
ing to general criteria (see Table 1), and then practitioners to 1) administer currently avail-
according to specific subtypes of MCI that may able treatments for cognitive impairment and to
be important for directing appropriate therapeu- initiate new treatment strategies as they are
tic strategies. For this purpose, 3 distinct sub- developed, 2) provide counseling and recom-
types of MCI have been identified: 1) amnestic mend support services for patients and families,
(focal memory impairment, the subtype most and 3) identify and treat reversible etiologies of
frequently recognized as a precursor to AD); 2) cognitive impairment. Currently, most research
multiple domain, slightly impaired (amnestic in this area has been conducted in specialists’
and nonamnestic, involving impairment in more settings and resultant diagnostic algorithms
than 1 cognitive domain); and 3) single, non- have not transferred well to the general practice
memory domain impaired (impairment in a sin- setting. The majority of persons with MCI, how-
gle cognitive domain without memory impair- ever, will present initially to general practice
ment).10-12 In this article, we focus on the clinicians (physicians and nurse practitioners),
detection of amnestic MCI. who will be primarily responsible for initiating
In primary care, where all formal criteria and pharmaceutical treatment.7,17 Evidence has sug-
measures are typically not available, it is still gested that it is feasible for general practitioners
possible to screen aging clients for probable to verify cognitive complaints and screen for
MCI to aid in making referral decisions to neu- MCI because this screening has resulted in a
rologists, psychiatrists, or neuropsychologists high degree of accuracy using 3 tests of cogni-
for further evaluation. Indeed, in the current tive functioning (i.e., verbal fluency, visuospatial

Geriatric Nursing, Volume 29, Number 1 39

copying task, and recall of a short story) com- participant, demographic information and cog-
pleted in 5-10 minutes.17 nitive, depressive, and functional status data
were compiled and presented at a consensus
Purpose and Research Questions conference. All data collection was completed
by either nurse practitioners or trained health
The goal of this study was to demonstrate
care professionals. A team of 3 PhD-prepared
how persons of varying cognitive impairment
neuropsychologists, 1 PhD-prepared nurse, 1
levels might be identified using relatively simple
neuropsychology doctoral student, and 3 nurse
screening tools available to nurses and other
practitioners (also nursing doctoral students)
clinicians. Described in this study are the pro-
reviewed each case and classified the individual
cess of case identification, the cognitive and
as cognitively intact, amnestic MCI, or probable
demographic characteristics that differentiate
dementia based on the data profile. The consen-
unimpaired from mildly impaired elders, mildly
sus conference resulted in the classification of
from fully impaired elders, and the relative im-
50.8% of participants as cognitively intact, 30%
portance of the tools used to identify individuals
of participants as cognitively impaired, and
in each cognitive status group. In this study, a
19.2% of participants as amnestic MCI. In an
more heterogeneous sample is used than in pre-
effort to represent the consensus conference
vious research, much of which has relied solely
experts’ decision making process to a nursing
on community-dwelling or clinic-based samples
practice audience, discriminant function analy-
of older adults. The present sample included
sis was performed to identify the relative impor-
residents in a mix of settings, including commu-
tance of study variables in classifying cognitive
nity-dwelling, assisted living, and long-term care
status. Further analyses were conducted to
facilities. The specific research aims were: 1) to
compare cognitive status groups for differences
identify the rate and characteristics of persons
in demographics and cognitive test performance
with amnestic MCI in participants across resi-
to obtain a representative description of per-
dential settings, 2) to identify specific cognitive
sons with MCI.
differences between persons with amnestic MCI
and those who were either cognitively intact or
Measures
who had dementia, and 3) to examine the rela-
tive importance of the tools used to identify Demographic Data and Depression. Demo-
categories of cognitive functioning. graphic data used for analyses included age,
education level (classified as less than high
Methods school, completed some or all high school, or
some college or higher), sex, residential status
Sample and Setting (assisted living facility, nursing home, or com-
munity dwelling), and marital status (coded as
This study involved a sample of elderly adults
currently married or unmarried). Depression
who lived in nursing homes, assisted living fa-
was measured with the Geriatric Depression
cilities, and retirement communities in North
Scale (GDS), a commonly used screening mea-
Central Florida. Participants were recruited as
sure for depression in older adults.19 A score of
part of a larger study of the effect of cognitive
5 on the GDS is the cutoff for depression (scale
status on pain. The sample for this study con-
range ⫽ 0-15).
sisted of 130 older adults (of the 158 enrolled
Cognition. Participants underwent 3 cognitive
participants) who had complete data on the full
tests: the MMSE, the Dementia Rating Scale-2
battery of cognitive measures. University insti-
(DRS-2),20 and the Hopkin’s Verbal Learning
tutional approval was obtained, and each partic-
Test (HVLT).21 The tests were administered by
ipant and/or his/her legally authorized represen-
trained research personnel. The MMSE was
tative provided informed consent. See Table 2
used solely as a cognitive screen for entry in the
for complete sample characteristics.
parent study and was not used as an MCI diag-
nostic test. The standard cutoff score of 24 was
Procedures
used to indicate that persons scoring 24 or
Participants underwent an initial interview, above on this measure did not have dementia.22
lasting 1-2 hours, in which demographic, cogni- The DRS-2 and HVLT were used as more sensi-
tive, and health data were collected. For each tive measures of cognitive functioning that as-

40 Geriatric Nursing, Volume 29, Number 1

 Table 2.
Description of Overall Sample and Cognitive Classification Groups (Intact,
MCI, and Impaired)
By Cognitive Status Group
n (%)
Full sample
(N ⴝ 130) Intact MCI Impaired
Variable n (%) (n ⴝ 66) (n ⴝ 25) (n ⴝ 39) Statistic p

Residential Status: 53 (41) 31 (47.0) 16 (64.0) 6 (15.4) ␹ (df ⫽ 4) ⫽ 40.8

2
.000
Assisted Living Facility
Nursing Home 47 (36) 12 (18.2) 6 (24.0) 29 (74.4)
Community Dwelling 30 (23) 23 (34.8) 3 (12.0) 4 (10.3)
Education
⬍8th Grade 12 (9) 4 (6.1) 1 (4.0) 7 (17.9) ␹2 (df ⫽ 4) ⫽ 15.2 .004
9–12th Grade 50 (39) 19 (28.8) 10 (40.0) 21 (53.8)
Some College or More 68 (52) 43 (65.2) 14 (56.0) 11 (28.2)
Sex
Male 24 (19) 11 (16.7) 8 (32.0) 5 (12.8) ␹2 (df ⫽ 2) ⫽ 4.0 .135
Female 106 (81) 55 (83.3) 17 (68.0) 34 (87.2)
Marital Status
Unmarried 93 (71) 43 (65.2) 24 (96.0) 21 (65.8) ␹2 (df ⫽ 2) ⫽ 9.2 .010
Married 37 (29%) 23 (34.8) 1 (4.00 13 (34.2)
Age (X ៮ ) (range ⫽ 65–97) 82.5 82.3 84.0 81.8 F (df ⫽ 2) ⫽ 0.7 .503
MMSE (X ៮) 24.7 28.0 26.2 18.0 F (df ⫽ 2) ⫽ 121.2 .000
(range ⫽ 9–30)*
Depression (X ៮) 3.9 3.5 4.0 4.5 F (df ⫽ 2) ⫽ 1.0 .362
(range ⫽ 0–15)
Hopkins Verbal Learning
Test (X ៮)
Total Learning† 17.6 23.3 16.1 8.9 F (df ⫽ 2) ⫽ 101.8 .000
Total Recall† 40.3 50.3 36.8 25.5 F (df ⫽ 2) ⫽ 104.5 .000
Delayed Recall‡ 38.3 50.2 27.2 25.5 F (df ⫽ 2) ⫽ 176.2 .000
Retention‡ 39.5 50.8 25.9 28.9 F (df ⫽ 2) ⫽ 79.1 .000
Recognition‡ 38.3 48.4 30.0 26.4 F (df ⫽ 2) ⫽ 54.4 .000
Dementia Rating Scale—
2 (X៮)
Initiation† 40.4 48.1 40.8 27.4 F (df ⫽ 2) ⫽ 65.7 .000
Conceptualization† 45.4 52.6 48.1 32.0 F (df ⫽ 2) ⫽ 108.3 .000
Memory† 42.1 53.0 38.0 26.6 F (df ⫽ 2) ⫽ 129.1 .000
Attention* 52.3 57.5 57.5 40.2 F (df ⫽ 2) ⫽ 67.8 .000
Construction* 45.4 49.5 49.3 36.3 F (df ⫽ 2) ⫽ 48.4 .000
MCI ⫽ mild cognitive impairment.
*Post hoc Bonferroni tests indicate impaired group significantly different from other 2 groups (Impaired ⬍ intact and
MCI).
†
Post hoc Bonferroni tests indicate monotonic pattern with each group significantly different from the others (Impaired ⬍
MCI ⬍ Intact).
‡
Post hoc Bonferroni tests indicate intact group was significantly different from other 2 groups (MCI and impaired ⬍
intact).

sessed specific aspects of cognitive perfor- these as well as a total score (total scores can
mance. The DRS-2 tests several components of range from 0 to 144). This measure has been
cognitive functioning (i.e., memory, construc- well validated, having a correlation of .86 in
tion, conceptualization, attention, and initiation/ persons with memory impairment on the Wech-
perseveration) and gives a score for each of sler Adult Intelligence Scale, and alpha coeffi-

Geriatric Nursing, Volume 29, Number 1 41

cients ranging from .75 to .95 for tool subscales The HVLT included several indices of memory
for controls, patients with mild dementia, and functioning. Congruent with common conceptu-
moderately severe dementia groups.23 The alizations of amnestic MCI (i.e., focal memory
HVLT tests memory and consists of a list of 12 loss but other cognitive functions remain in-
words read aloud to the individual, who is then tact), the total recall and total learning sub-
assessed for immediate recall (3 trials), delayed scales of the HVLT showed the expected pattern
recall (after 20 minutes), and word recognition. related to impairment; intact persons performed
This measure has been found to have high sen- significantly better than those with amnestic
sitivity (96%) and good specificity (80%) in de- MCI who, in turn, performed significantly better
tecting mild dementia.24 Normative age and ed- than cognitively impaired persons. The other
ucation adjusted t scores were used in analyses HVLT subtasks (delayed recall, retention per-
for each of the subcomponents of the HVLT and centage, and recognition discrimination) dif-
DRS-2.14 The total time to complete these mea- fered significantly between cognitively intact
sures was on average 40-45 minutes. participants and all others, suggesting that both
impaired groups had poorer memory ability, but
Results did not differ from one another.
The DRS-2 was used to characterize cognitive
Demographic Differences between losses more broadly in participants. Based on
Cognitive Status Groups clinical guidelines, it was expected that non-
memory measures would show impairment only
The 3 cognitive status groups (cognitively in-
in persons with probable dementia, but few dif-
tact, amnestic MCI, and impaired) demonstrated
ferences should emerge on these nonmemory
several statistically significant residential and
measures between cognitively intact persons
demographic differences. Details are shown in
and those with amnestic MCI. This expectation
Table 2. The majority (64%) of persons with
was partially supported. As with the MMSE, the
amnestic MCI resided in assisted living facili-
DRS-2 attention and construction subscales dis-
ties, whereas the majority (74%) of those with a
criminated those with probable dementia from
suspected dementia lived in a nursing home. In
all others, but it did not further distinguish the
this study, the cognitively intact elders were
amnestic MCI group from the cognitively intact
mostly split between community-dwelling (35%)
group. In contrast, the DRS-2 initiation, concep-
and assisted-living (47%) status. In terms of ed-
tualization, and memory subscales all showed a
ucation, the majority of cognitively intact and
more expected impairment pattern, such that
amnestic MCI participants had more than high
persons with amnestic MCI performed signifi-
school education (65% and 56%, respectively),
cantly worse than intact elders, and cognitively
whereas the majority of impaired persons had a
impaired persons performed significantly worse
high school education (54%). A significantly
than persons with amnestic MCI.
higher proportion of persons with amnestic MCI
were not married (96%) than in either the cog-
Factors Related to Participant
nitively intact or cognitively impaired groups Classification
(65% and 66%, respectively). The cognitive sta-
tus groups did not differ in age or sex. Consensus conferences rely on expert judg-
ment, implicit decision algorithms, and the
availability of teams of neuropsychologists. To
Cognitive Profile Differences between
Cognitive Status Groups evaluate the classification decisions and to see
whether the test scores alone could be used to
We also examined how the specific cognitive identify amnestic MCI, we conducted a discrimi-
profiles of the groups differed and whether nant function analysis (DFA). In DFA, a set of
these differences were statistically significant. classification instruments (here the cognitive
As Table 2 shows, impaired participants per- measures from the MMSE, DRS-2, and HVLT) is
formed significantly worse than the amnestic used to predict some previously defined classi-
MCI or intact participants on the MMSE, an fication (here, the 3-leveled cognitive status
index of global cognitive functioning. Note that variable from the consensus conference).
both amnestic MCI and intact groups had mean We used DFA to answer 2 questions. First,
MMSE scores above the cut-off for dementia. how well did a classification function replicate

42 Geriatric Nursing, Volume 29, Number 1

 Table 3.
Factors Related to Identifying Persons with Mild Cognitive Impairment
(Loading Matrix of Correlations Between Cognitive and Demographic
Variables and the Discriminant Function)
Characteristics Function 1* Pattern Coefficients Function 2† Pattern Coefficients

Cognitive Variables
MMSE .06 .72
HVLT Total Learning ⴚ.61 ⴚ.62
HVLT Total Recall .43 .66
HVLT Delayed Recall .68 ⴚ.81
HVLT Retention .01 ⫺.20
HVLT Recognition ⫺.21 .07
DRS Initiation .27 .16
DRS Conceptualization .36 .24
DRS Memory .49 ⫺.23
DRS Attention .14 .24
DRS Construction .22 .02
Age .05 .09
Education ⫺.20 .21
Depression ⫺.04 ⫺.05
Bold numbers indicate scores above 0.33, representing approximately 10% of the variance, are considered interpretable
predictors. DRS ⫽ Dementia Rating Scale; HVLT ⫽ Hopkins Verbal Learning Test; MMSE ⫽ Mini-Mental State Exam.
*Factor differentiating intact persons from all others (MCI and impaired).
†
Factor differentiating persons with MCI from all others (intact and impaired).

the classifications made by experts in a consen- The first function primarily distinguished the
sus conference? If there was high agreement cognitively intact from the cognitively impaired
between the expert classifications and DFA re- individuals (i.e., those with probable dementia
sults, this would suggest that the cognitive test who were more globally impaired). The second
scores alone could correctly classify individuals’ function distinguished those with amnestic MCI
cognitive status. Second, what variables differ- from all others. The results in Table 3 show that
entiate cognitive status groups? DFA provides the first function had positive coefficients above
pattern coefficients (similar to regression .33 for both the HVLT (primarily memory tests)
weights), which present an index of the relative and DRS-2 (tests of more generalized cogni-
importance of each instrument or tool for explain- tion). Positive coefficients indicate that individ-
ing the classification decisions. Table 3 presents uals with intact cognition scored better than
these pattern coefficients in the current sample. impaired participants on these variables. This
The coefficients show the relative importance of finding is consistent with the generalized nature
each cognitive and demographic variable for dis- of dementia such that impaired individuals dem-
tinguishing persons with amnestic MCI from those onstrate impairment across multiple domains of
who are cognitively intact and cognitively im- cognitive functioning. In contrast, function 2
paired. Scores above 0.33 (representing approxi- was defined specifically by memory measures,
mately 10% of the variance; shown in bold) are indicating the disproportionate memory loss
considered interpretable predictors.25 that characterizes persons with amnestic MCI.
Two functions, required to classify partici-
pants optimally, correctly classified 95.2% of the Discussion
sample (98.4% of cognitively intact participants,
95.8% of participants with amnestic MCI, and In this sample of older adults, 19.2% were
89.5% of participants with probable dementia), classified as having amnestic MCI when experts
indicating a high degree of agreement with the at a consensus conference applied the criteria
experts’ consensus conference judgments. described earlier. This rate is higher than those

Geriatric Nursing, Volume 29, Number 1 43

previously found for amnestic MCI, which range relatively minimal training and appear to pro-
from 2% to 6%.11,15 The rate of amnestic MCI in vide important information in detecting subtle
the present sample was more similar to the cognitive changes associated with amnestic
overall rate of all 3 subtypes of MCI, which MCI.
ranged from 17% to 34% in previous stud- As expected, older adults with intact cogni-
ies.11,26,27 tive functioning performed better on the mem-
ory testing HVLT and the memory and concen-
Demographic Differences between tration subscales of the DRS than those with
Cognitive Status Groups intact or mildly impaired cognition. In contrast,
individuals identified as having amnestic MCI
Several demographic differences were noted performed worse on memory tests (e.g., recall
between the 3 classifications of cognitive and retention subscales of the HVLT) than intact
groups (i.e., intact, amnestic MCI, and im- persons but had higher scores on the MMSE and
paired). Groups differed overall in terms of res- DRS-2 attention/construction subscales than
idential status, education level, and marital sta- persons with dementia. This finding is not sur-
tus but did not differ significantly in terms of prising given previous research findings that
age, sex, or rate of depression. Persons with persons with MCI had mean MMSE scores
cognitive impairment were most likely to reside above the cutoff for dementia.13,14,16 This find-
in nursing homes. Overall, 65% of cognitively ing highlights the limitations of the MMSE in
intact persons resided in a care facility (assisted detecting the subtle, memory-specific cognitive
living facility or nursing home) compared with losses associated with amnestic MCI.
88% of those with MCI and 90% of cognitively As could be expected based on typical MCI
impaired persons. However, this finding is not testing batteries, the cognitive test scores were
surprising given that the majority of participants the only significant variables that contributed to
in the parent pain assessment study were re- classification of cognitive status in the discrimi-
cruited from nursing homes or assisted living nant function analysis. Age, education, and de-
facilities. pression scores were not salient predictors.
Educational levels between persons with am- These findings provide support for the concept
nestic MCI and cognitively intact persons were of amnestic MCI as a unique diagnostic category
similar, and both of these groups had signifi- characterized by deficits in the recall and reten-
cantly higher educational levels than cognitively tion domains of memory. Although experts’
impaired persons. This finding suggests the pos- classifications of cognitive status remain the di-
sibility that persons with higher levels of educa- agnostic gold standard, the discriminant func-
tion may progress toward dementia more slowly tion algorithm in the present study provides
than those with lower levels of education and evidence for a set of tests than can be used to
are therefore more likely to have cognitive de- detect amnestic MCI in elderly adults.
clines identified in the MCI phase. This belief
reflects the greater cognitive reserve hypothesis
that supports the notion that persons with Limitations
higher levels of education and brain activity are
better able to maintain high levels of cognitive This study explored the rate of amnestic MCI
function. in a convenience sample from a parent study not
specifically designed for this purpose, and the
cognitive measures used were biased toward
Relative Importance of Measures that
Distinguished Cognitive Status Groups memory and simple screening and are not rep-
resentative of a comprehensive MCI test bat-
Discriminant function analysis correctly clas- tery. However, we were able to classify persons
sified 95.2% of the participants in agreement with amnestic MCI accurately using this much
with classifications of the consensus confer- shorter battery of tests than the traditional bat-
ence. This high agreement provides empirical teries. Therefore, it is encouraging that amnestic
support for the usefulness of the study mea- MCI could be identified using the screening
sures (e.g., DRS-2 and HVLT) in identifying am- measures in the present study. Future studies
nestic MCI. These 2 measures can be adminis- with the primary goal of distinguishing rates and
tered in a clinic or primary care setting with subtypes of MCI should include a broader cog-

44 Geriatric Nursing, Volume 29, Number 1

nitive battery and have memory and physical involved with developing quicker screening
function complaints corroborated by an infor- methods for MCI, similar to the MMSE, which
mant. has transformed screenings for dementia into
Finally, the label of amnestic MCI is used here easily administered tests during routine care. As
as a descriptive category not a medical diagno- sets of useful measures are discovered, they can
sis. Formal neurological follow-up is necessary be incorporated into nursing assessment classes
to establish a diagnosis of MCI. Again, our study for nurses in school or in-service training ses-
identified persons with amnestic MCI, 1 subtype sions for nurses in the workforce. Ideally,
of MCI. This finding can be attributed to several screenings for MCI should be completed with
causes: 1) we specifically recruited older adults greater frequency (i.e., every 3 months) in cli-
with pain and memory problems, 2) we have ents to monitor discrete changes in cognitive
more detail in our memory measures than in our functioning.
nonmemory measures, and 3) at the time the Researchers have begun to develop quicker
study and consensus conference were done, am- ways to detect MCI in older adults. Artero and
nestic MCI was the most commonly discussed Ritchie17 have had the best success (sensitivity
variant of MCI. Thus, sample selection factors 99%, specificity 73%) in identifying MCI in a
and measure selection factors may have re- general practice setting using computerized-
duced our sensitivity to nonamnestic variants of based tests of delayed auditory verbal recall,
MCI. verbal fluency, and visuospatial construction
that require 10-15 minutes to complete. Others
Implications have found preliminary success with the Tele-
phone Interview of Cognitive Status—Modified
The results of this study support the feasibil- (TICS-M), which also requires only 10 minutes
ity of primary care providers in detecting MCI in to complete and can be given over the tele-
their clinical population; regardless of setting. It phone.28,29 Graff-Radford and colleagues (2006)28
is our goal to aid clinicians in understanding the found a sensitivity of 86% and a specificity of
broader range of cognitive impairments that ex- 63% for the TICS-M when using a cutoff score
ist and are identifiable. In this study, amnestic of 29. Cook, Marsiske, and McCoy (under re-
MCI was easily detectable using simple tools view)29 found the TICS-M to have a sensitivity of
that were administered by advanced registered 82.4% and a specificity of 87.0% in detecting
nurse practitioners and trained health care pro- amnestic-MCI in their sample of older adults.
fessionals, which demonstrates that these meth- Although there is preliminary evidence for using
ods can be used and implemented into the stan- these quicker methods, it remains the role of
dard of care for older adult clients. The choice primary care providers to ask their clients about
of measurement tools to use is ultimately up to cognitive changes, to verify these complaints
each clinician’s own judgment and expertise. with the MCI screening method of their choice,
However, results from this study suggest that and to corroborate evidence of cognitive change
clinicians move beyond the traditional use of with a proxy informant. The clinician should
the MMSE because it was only able to discrim- explore cognitive and functional changes in the
inate persons with dementia from all others and last year and should repeat this assessment in 6
was not useful in discriminating persons with months and 1 year to minimize false-negative
MCI from cognitively intact persons. reports.17
Currently there are no “quick” testing batter- Although there are no treatments specifically
ies for identifying MCI, suggesting that this is an developed to treat MCI, there are important im-
area in which practitioners need to increase plications for early identification of cognitive
their focus and time with patients. The mea- changes. Pharmacological treatment options for
sures used in this study took on average 40-45 dementia may delay the progression of demen-
minutes to complete, a substantially shorter tia when initiated early in the disease process.30
time period than traditional MCI testing batter- Additionally, behavioral and cognitive training
ies, which can last up to 4 hours. However, a could be initiated as previous researchers have
40-minute screening inventory is not feasible for found these methods can improve cognitive
a typical health care appointment, and this high- functioning in persons with MCI.31,32 Delaying
lights the need for more practitioners to get the onset of dementia has large economic impli-

Geriatric Nursing, Volume 29, Number 1 45

cations; if its onset could be delayed 5 years 10. Portet F, Ousset PJ, Visser PJ, et al. Mild cognitive
through these treatment strategies, the preva- impairment (MCI) in medical practice: a critical
review of the concept and new diagnostic procedure.
lence of dementia would drop by 50%.30 Early Report of the MCI working group of the European
identification of cognitive changes can also ben- Consortium on Alzheimer’s disease. J Neurol Neurosur
efit individuals and their families by allowing Psychiatry 2006;77:714-18.
them additional time to engage in planning fi- 11. Petersen RC. Mild cognitive impairment as a
nancial matters, creating advanced directives, diagnostic entity. J Intern Med 2004;256:183-94.
12. Petersen RC, Doody R, Kurz A, et al. Current concepts in
and organizing support for future care needs.
mild cognitive impairment. Arch Neurol 2001;58:1985-92.
Finally, the high rate of amnestic MCI found 13. Bennett DA, Wilson RS, Schneider JA, et al. Natural
in this sample is important to consider clinically history of mild cognitive impairment in older persons.
because health care providers may encounter Neurology 2002;59:198-205.
increasing numbers of persons with MCI and 14. Busse A, Bischkopf J, Riedel-Heller SG, et al.
should thus be equipped with the tools to gather Subclassifications for mild cognitive impairment:
prevalence and predictive validity. Psychol Med 2003;
more information regarding the cognitive status 33:1029-38.
of their clients. It is important for health care 15. Norlund A, Rolstad S, Hellstrom P, et al. The Goteborg
professionals to become part of the team in MCI study: Mild cognitive impairment is a
identifying MCI as they hold specific knowledge heterogeneous condition. J Neurol Neurosur
of their clients’ cognitive abilities through inter- Psychiatry 2005;76:1485-90.
16. Tabert MH, Albert SM, Borukhova-Milov L, et al.
acting with and educating them. Nurse practitio-
Functional deficits in patients with mild cognitive
ners are in a position to administer tests to impairment. Neurology 2002;58:758-64.
evaluate the characteristics of MCI and refer 17. Artero S, Ritchie K. The detection of mild cognitive
clients demonstrating cognitive changes to re- impairment in the general practice setting. Aging Ment
ceive more thorough cognitive evaluations. Health 2003;7:251-58.
18. Graham JE, Rockwood K, Beattie BL, et al. Prevalence
and severity of cognitive impairment with and without
References dementia in an elderly population. Lancet 1997;349:
1793-6.
1. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental
19. Shiekh J, Yesavage J. Geriatric depression scale:
State”: A practical method for grading the cognitive
recent findings and development of a short version. In
state of patients for the physician. J Psychiatr Res
1975;12:189-98. Brink TL, editor. Clinical gerontology: a guide to
2. Aggarwal NT, Wilson RS, Beck TL, et al. Mild assessment and intervention. New York: Hawarth
cognitive impairment in different functional domains Press; 1986.
and incident Alzheimer’s disease. J Neurol Neurosur 20. Mattis S, Jurica PJ, Leitten CL. Dementia Rating
Psychiatry 2005;76:1479-84. Scale—2. Lutz, FL: Psychological Assessment
3. Morris JC, Storandt M, Miller JP, et al. Mild cognitive Resources; 2001.
impairment represents early-stage Alzheimer disease. 21. Brandt J. The Hopkins Verbal Learning Test:
Arch Neurol 2001;58:397-405. Development of a new memory test with six
4. Hänninen T, Hallikainen M, Tuomainen S, et al. equivalent forms. Int J Clin Neuropsyc 1991;5:125-42.
Prevalence of mild cognitive impairment: a population- 22. Gallo JJ, Wittink MN. Cognitive assessment. In Gallo
based study in elderly subjects. Acta Neurol Scand JJ, Bogner HR, Fulmer T, et al., editors. Handbook of
2002;106:148-54. geriatric assessment. 4th ed. Boston: Jones & Bartlett;
5. Larrieu S, Letenneur L, Orgogozo JM, et al. Incidence 2006.
and outcome of mild cognitive impairment in a 23. Mattis S. Dementia Rating Scale: Professional Manual.
population-based prospective cohort. Neurology 2002; Odessa, FL: Pyschological Assessment Resources; 1988.
59:1594-9. 24. Frank RM, Byrne GJ. The clinical utility of the
6. Manly JJ, Bell-McGuinty S, Tang M, et al. Hopkins Verbal Learning Test as a screening test for
Implementing diagnostic criteria and estimating mild dementia. Int J Geriatr Psychiatry 2000;15:317-24.
frequency of mild cognitive impairment in an urban 25. Tabachnick BG, Fidell LS. Using multivariate
community. Arch Neurol 2005;62:1739-46. statistics. New York: HarperCollins College; 1996.
7. Ritchie K, Artero S, Touchon J. Classification criteria 26. Burns A, Zaudig M. Mild cognitive impairment in older
for mild cognitive impairment: a population-based people. Lancet 2002;360:1963-5.
study. Neurology 2001;56:37-42. 27. Lopez OL, Jagust WJ, DeKosky ST, et al. Prevalence
8. Artero S, Petersen R, Touchon J, Ritchie K. Revised and classification of mild cognitive impairment in the
criteria for mild cognitive impairment: validation Cardiovascular Health Study Cognition Study: part 1.
within a longitudinal population study. Dement Geriatr Arch Neurol 2003;60:1385-9.
Cogn 2006;22:465-70. 28. Graff-Radford NR, Ferman TJ, Lucas JA, et al. A cost
9. Touchon J. Recent consensus efforts in the diagnosis effective method of identifying and recruiting persons
of mild cognitive impairment. Psychogeriatrics 2006;6: over 80 free from dementia or mild cognitive
S23-5. impairment. Alzheimer Dis Assoc Disord 2006;20:101-4.

46 Geriatric Nursing, Volume 29, Number 1

 29. Cook SE, Marsiske M, McCoy KJM. The use of the Mental Health and Aging, Tuscaloosa, AL. ANN L. HOR-
Modified Telephone Interview for Cognitive Status GAS, PhD, RN, is an Associate Professor and Associate
(TICS-M) in the detection of mild cognitive impairment. Dean for Research from the University of Florida College of
Alzheimer Dis Assoc Disord. Manuscript under review. Nursing, Gainesville, FL. MICHAEL MARSISKE, PhD, is
30. Burns A, O’Brien J. Clinical practice with anti- an Associate Professor and Associate Chair for Research
dementia drugs: a consensus statement from British from the University of Florida Department of Clinical and
Association for Psychopharmacology. Health Psychology, Gainesville, FL.
J Psychopharmacol 2006;20:732-55.
31. Belleville S, Gilbert B, Fontaine F, et al. Improvement ACKNOWLEDGMENT
of episodic memory in persons with mild cognitive This work was supported by grants from the John A. Hart-
impairment and healthy older adults: evidence from a
ford Foundation Building Academic Geriatric Nursing
cognitive intervention program. Dement Geriatr Cogn
Capacity Predoctoral Scholarship (to AFE) and Grant No.
2006;22:486-99.
NR05069-02 from the National Institute for Nursing Re-
32. Willis SL, Tennstedt SL, Marsiske M, et al. Long-term
search awarded (to ALH).
effects of cognitive training on everyday functional
outcomes in older adults. J Amer Med Assoc 2006;296:
2805-14. 0197-4572/08/$ - see front matter
AMANDA FLOETKE ELLIOTT, PhD, ARNP, is a Post Doc- © 2008 Mosby, Inc. All rights reserved.
toral Fellow from the University of Alabama Center for doi:10.1016/j.gerinurse.2007.04.015

7JTJU
XXXKPVSOBMTFMTFWJFSIFBMUIDPN

GPS
B
GVMM
TFMFDUJPO
PG

&MTFWJFS
QFSJPEJDBMT
4BVOEFST
.PTCZ
$IVSDIJMM
-JWJOHTUPOF
)BOMFZ

#FMGVT

ª #VJMEJOH
*OTJHIUT
#SFBLJOH
#PVOEBSJFT

Geriatric Nursing, Volume 29, Number 1 47