See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317285644

Psychometric evaluation of the Cohen-Mansfield Agitation Inventory in an

acute general hospital setting

Article in International Journal of Geriatric Psychiatry · May 2017

DOI: 10.1002/gps.4741

CITATIONS READS

17 822

7 authors, including:

Nuriye Kupeli Victoria Vickerstaff

University College London University College London
90 PUBLICATIONS 1,613 CITATIONS 155 PUBLICATIONS 3,249 CITATIONS

SEE PROFILE SEE PROFILE

Nicola White Kathryn Lord

University College London University of Bradford
73 PUBLICATIONS 1,298 CITATIONS 53 PUBLICATIONS 1,562 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Elizabeth L Sampson on 19 January 2018.

The user has requested enhancement of the downloaded file.

RESEARCH ARTICLE

Psychometric evaluation of the Cohen-Mansfield Agitation

Inventory in an acute general hospital setting
Nuriye Kupeli1 , Victoria Vickerstaff1, Nicola White1 , Kathryn Lord2, Sharon Scott3, Louise Jones1 and
Elizabeth L. Sampson1,4
1
Marie Curie Palliative Care Research Department, University College London, Division of Psychiatry, UK
2
School of Dementia Studies, Faculty of Health Sciences, University of Bradford, UK
3
St. Christopher’s Hospice, UK
4
Barnet Enfield and Haringey Mental Health Trust Liaison Psychiatry Team, North Middlesex University Hospital, UK
Correspondence to: Dr. N. Kupeli, E-mail: n.kupeli@ucl.ac.uk

Objectives: The Cohen-Mansfield Agitation Inventory (CMAI; (Cohen-Mansfield and Kerin, 1986)) is a
well-known tool for assessing agitated behaviours in people with dementia who reside in long-term
care. No studies have evaluated the psychometric qualities and factor structure of the CMAI in acute
general hospitals, a setting where people with demand may become agitated.
Method: Longitudinal study investigating pain, agitation and behavioural problems in 230 people with
dementia admitted to acute general hospitals in 2011–2012. Cohen-Mansfield Agitation Inventory
was completed as part of a battery of assessments including PAINAD to measure pain.
Results: A nine-item two-factor model of aggressive and nonaggressive behaviours proved to be the
best-fitting measurement model in this sample, (χ 2 = 96.3, df = 26, p<0.001; BIC [Bayesian
Information Criterion] = 4593.06, CFI [Comparative Fit Index] = 0.884, TLI [Tucker Lewis
Index] = 0.839, RMSEA [Root Mean Square Error of the Approximation] = 0.108). Although similar
to the original factor structure, the new model resulted in the elimination of item 13 (screaming).
Validity was confirmed with the shortened CMAI showing similar associations with pain as the original
version of the CMAI, in particular the link between aggressive behaviours and pain.
Conclusion: The factor structure of the CMAI was broadly consistent with the original solution although
a large number of items were removed. Scales reflecting physical and verbal aggression were combined
to form an Aggressive factor, and physical and verbal nonaggressive behaviours were combined to form
the Nonaggressive factor. A shorter, more concise version of the CMAI was developed for use in acute
general hospital settings. Copyright # 2017 John Wiley & Sons, Ltd.
Key words: dementia; acute general hospitals; agitation; BPSD; psychometric; CMAI
History: Received 11 January 2017; Accepted 19 April 2017; Published online in Wiley Online Library (wileyonlinelibrary.com)
DOI: 10.1002/gps.4741

Introduction disinhibition (Ware et al., 1990). Increasing numbers

of people admitted to acute general hospitals have
Dementia is a progressive, neurodegenerative disease clinically significant cognitive impairment or dementia
characterised by cognitive and physical decline. (Sampson et al., 2009), and BPSD, in particular
Challenging behavioural and psychological symptoms aggression, may predict hospital admission (Ware
(BPSD) are common throughout the moderate and et al., 1990). At any given time in the UK, 6% of
severe stages. Psychological symptoms of dementia people with dementia occupy inpatient beds at acute
include anxiety, depression, paranoia and hallucinations. general hospitals compared with 0.6% of older ageing
Behavioural symptoms include physical and verbal adults without dementia (Russ et al., 2012). A hospital
aggression, repetitive mannerisms, restlessness and admission can be a frightening experience for people

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
 N. Kupeli et al.

with dementia (Alzheimer’s Society, 2016) and Method

exposes people with dementia to an increased risk of
adverse events including an increased risk of mortality, Participants and procedure
functional decline and a longer length of stay
(Mukadam and Sampson, 2011; Watkin et al., 2012). Detailed methodology for the BePaid study is reported
Literature also suggests that the experience of pain elsewhere (Scott et al., 2011). In summary, patients
is related to behavioural problems, in particular were recruited as part of a longitudinal study
aggressive and agitated behaviours (Ahn and Horgas, investigating pain, agitation and behavioural problems
2013). in people with dementia who were admitted to two
The Cohen-Mansfield Agitation Inventory (CMAI; acute general hospitals in London, UK (N = 230).
(Cohen-Mansfield and Kerin, 1986)) is a 29-item tool Four trained researchers assessed all patients under
(Cohen-Mansfield et al., 1989) developed to assess the care of geriatricians within 72 h of admission to
agitated behaviours of people residing within nursing an acute general hospital ward from accident and
homes. The utility of the CMAI across different emergency services. All information on patients
cultures is demonstrated by the translations available including medical notes and discussions with family
(Cohen-Mansfield, 1991). Additionally, the Japanese, carers and health care professionals was used to assess
Dutch, Korean and Chinese versions of the CMAI for pain and BPSD. Patients who fulfilled the
have been validated (de Jonghe and Kat, 1996; following criteria were approached to participate in
Schreiner et al., 2000; Choy et al., 2001; Suh, 2004). the study:
The CMAI has been shown to have high internal • Aged 70 years or above with an unplanned acute
consistency (Finkel et al., 1992), and high test–retest medical admission
(Koss et al., 1997) and inter-rater reliability over time • Able to give written informed consent or agreement
(Chrisman et al., 1991; Whall et al., 1999). High to participate provided by available informal carer
correlations have been found between the CMAI and or “professional consultee”
other measures of BPSD (Finkel et al., 1992), but the • Abbreviated Mental Test Score (AMTS;
CMAI has been shown to be more sensitive to change (Hodkinson, 1972)) of ≤7/10 (routinely measured
when compared with other measures of agitation on admission)
(Zuidema et al., 2011).
Originally, the CMAI was found to contain three
factors: physical aggression, verbal aggression and Patients were excluded if they indicated verbally
nonaggressive behaviours (Cohen-Mansfield, 1986; or nonverbally that they did not wish to participate,
Cohen-Mansfield et al., 1989). Later studies were moribund, or non-English speaking. To ensure
produced a four-factor structure separating the that our study sample had a diagnosis of dementia,
nonaggressive agitated behaviours to reflect physical the Confusion Assessment Method (CAM; (Inouye
and verbal behaviour (Cohen-Mansfield, 1991; et al., 1990)) was used to screen potential
Rabinowitz et al., 2005). The CMAI has been participants for delirium. Those who were not
validated previously using data collected within delirious were consented, or assent was obtained
nursing homes (Cohen-Mansfield, 1986), outpatient from a carer or professional consultee, and were
clinics in general hospitals (Altunöz et al., 2015) assessed using the Mini-Mental State Examination
and day centres (Cohen-Mansfield et al., 1995). (MMSE; (Folstein et al., 1975)). Those who scored
However, as older ageing adults with cognitive ≤24 were entered into the study. Patients who
impairment are commonly admitted to acute initially screened positive for delirium were
general hospitals and interest in research in this reassessed 48 h later, and if this resolved, we then
area is increasing, it is important to determine if completed the MMSE (Folstein et al., 1975). If
these factor structures are applicable for people delirium was persistent, patients were ineligible as
assessed during inpatient acute hospital stays. To we could not establish a clear dementia diagnosis.
the best of our knowledge, the measurement However, those with delirium who already had a
properties of the CMAI have not been systematically documented specialist dementia diagnosis were
evaluated within acute general hospitals. Therefore eligible. Dementia diagnosis was confirmed using a
our aim was to explore the factor structure of the structured clinical assessment based on operationalised
CMAI to further validate this tool in a sample of DSM-IV (American Psychiatric Association, 1994)
people with dementia during their admission to criteria, comprising cognitive testing (MMSE;
acute hospital settings. (Folstein et al., 1975)), structured case notes review,

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
Psychometric evaluation of CMAI

discussion with family and the clinical team. Data from loadings were non-significant or if they loaded
a sample size of 230 participants were potentially significantly but weakly (i.e. <.40) onto more than
available for analysis. one factor (Cohen-Mansfield et al., 1989; Rabinowitz
et al., 2005). Fit indices were assessed to determine
how well the proposed model fitted the sample data.
Ethical approval The Chi2 statistic is used as a measure of fit between
the sample covariance and fitted covariance matrices
Ethical approval was obtained from the Central (Byrne, 1998). Although a non-significant Chi2 is
London Research Ethics Committee 3 (reference: desired, due to the sample size of the current study,
10/H0716/79). a significant Chi2 is expected based on standard
statistical theory of how sample size, power and
Measures significance are associated (Cohen, 1992). Therefore,
in addition to the Chi2 statistic, several fit indices were
Agitation. The CMAI is an observer-based measure. evaluated including the Bayesian Information
Each of the 29 items is rated on a 7-point Likert scale Criterion (BIC), Comparative Fit Index (CFI) and
(“Never” to “Several times per hour”) commonly over the Tucker Lewis Index (TLI). The model with the
the previous two weeks, but for the purpose of this lowest BIC is preferred (Raftery, 1995), and values
study, patients were assessed for the previous 3–5 days. >0.95 for the CFI and TLI indicate a reasonable fit
Behaviours include aimless wandering, physical (Hu and Bentler, 1999). The Root Mean Square Error
aggression such as hitting and kicking, verbal agitation of the Approximation (RMSEA) is another fit index
such as cursing or constant request for attention and which takes into account the error of approximation
hoarding. A total score ranging between 29 and 203 in the population (Byrne, 1998). Root Mean Square
can be computed, reflecting overall agitation; there Error of the Approximation values <0.06 indicate a
are four separate subscales indicating physically good model fit (Hu and Bentler, 1999). Based on
nonaggressive behaviours (PNAB; 9 items), physically previous literature (Ahn and Horgas, 2013) and to
aggressive behaviours (PAGB; 12 items), verbally validate the measure further, additional analyses were
nonaggressive behaviours (VNAB; 4 items) and conducted to examine associations between the
verbally aggressive behaviours (VAGB; 4 items) subscales of the CMAI and pain.
(Cohen-Mansfield, 1991).

Pain. Pain was measured objectively using the Pain Results

Assessment in Advanced Dementia (PAINAD) Scale
(Warden et al., 2003). This observational measure Sample characteristics
consists of five domains (breathing, negative
vocalisations, facial expression, body language and We used the first assessment from the longitudinal
consolability). Each domain is scored on a severity cohort study, recorded within 72 h of admission. A
scale of 0 to 2 points during movement and during total of 230 participants were recruited (117 from
rest (maximum score of 10). Scores ≥2 indicate the Hospital 1; 113 from Hospital 2). The mean and
presence of pain (Zwakhalen et al., 2012). median CMAI scores at baseline were 33 (SD = 5.5)
and 31 (29–35), respectively. The CMAI scores were
stable across the study visits. For cohort demographic
Statistical analysis and clinical information, see Table 1.

Items on the CMAI with a prevalence of <5% were

excluded. Exploratory factor analysis (EFA) was Reliability analysis
conducted using STATA (StataCorp, 2013). Oblique
Promax rotation was employed as extracted factors Cronbach’s alpha (α) for the overall CMAI and its
were expected to correlate. The number of factors subscales, PNAB, PAGB, VNAB and VAGB in the
extracted was based on the Kaiser criterion current study were 0.76, 0.42, 0.86, 0.56 and 0.57,
(eigenvalues >1) and the examination of a scree plot. respectively. In the current study, the PNAB did not
Exploratory factor analysis using STATA allows model have acceptable reliability with a score <0.5, and the
fit indices to be evaluated across several factor VNAB and VAGB were also below the ideal cutoff
solutions. Items were removed from the EFA if factor of 0.70.

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
 N. Kupeli et al.

Table 1 Patient demographics

Total cohort N (%)

DEMOGRAPHICS (N = 230)
Gender
Female 151 (65.7)
Male 79 (34.3)
Age, years
75–84 85 (36.9)
85–94 118 (51.3)
95+ 27 (11.7)
Ethnicity
White British 175 (76.1)
Black Caribbean 15 (06.5)
Other 40 (17.4)
Place of residence
Home 145 (66.2)
Residential home 26 (11.9)
Nursing home 39 (17.8)
Other 09 (04.1)
CLINICAL CHARACTERISTICS

FAST score, %
3–5 (objective functional deficit, difficulties with activities of daily living) 86 (37.4)
6a–c (help required putting on clothes, toileting or bathing) 39 (16.9)
6d–e (urinary and faecal incontinence) 74 (32.2)
7a–f (less than six words, can no longer walk, sit up, smile, hold up head) 31 (13.5)
CMAI, median (IQR) 31 (29–35)

Frequency of behaviours Exploratory Factor Analysis (EFA) of the CMAI

The most commonly reported behaviours from the Models extracting two factors were considered based
CMAI included general restlessness (n = 98; 43%), on the indication of the scree plot and using the Kaiser
cursing or verbal aggression (n = 46; 20%), pushing criterion (eigenvalues >1), two eigenvalues were
(n = 36; 16%), repetitive sentences or questions observed to exceed one. The EFA model revealed
(n = 32; 14%), trying to get to a different place (n = 32; two factors consisting of 5 items each (a total of 10
14%), screaming (n = 22; 10%), hitting (n = 23; 10%), items remaining for the scale); however, the fit of the
pacing and aimless wandering (n = 15; 7%), constant model was poor using standard SEM criteria
unwarranted request for attention (n = 17; 7%) and (χ 2 = 132.8, df = 34, p < .001; BIC = 5133.57,
grabbing onto people or things inappropriately (n = 15; CFI = .857, TLI = .810, RMSEA = .112). Using the
7%). From the 29 behaviours, 19 behaviours did not criteria outlined above, item 13 (screaming) was
occur very often (<5%) or did not occur at all in our eliminated from the model due to weak factor
sample of people in acute settings and thus were excluded loadings (<.40) and double loadings thus limiting
from further analyses; complaining and negativism, interpretation. Although item 1 (pacing and aimless
inappropriate dressing or disrobing, spitting, kicking, wandering) had weak factor loadings, this item was
scratching, repetitive mannerisms, making strange noises, retained as the loading on Factor 2 was only just below
handling things inappropriately, throwing things, biting, the cut-off criteria (<0.40) and its factor loading on
eating or drinking inappropriate substances, intentional Factor 2 made theoretical sense (the Promax rotated
falling, hurting self or others, hiding and hoarding factor solution presented in Table 2). The two-factor
things, tearing or destroying property and making solution was re-estimated after excluding item 13.
verbal or physical sexual advances. The most commonly The fit of the final two-factor model was slightly
reported behaviours selected for EFA included general outside of standard SEM limits (χ 2 = 96.3, df = 26,
restlessness, cursing or verbal aggression, pushing, p < .001; BIC = 4593.06, CFI = .884, TLI = .839,
repetitive sentences or questions, trying to get to a RMSEA = .108), but was an improvement from the
different place, screaming, hitting, pacing and aimless 10-item two-factor solution and thus was chosen for
wandering, constant unwarranted request for attention further analysis. As the scores for the original 29-item
and grabbing onto people or things inappropriately. four-factor solution can be combined to reflect overall

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
Psychometric evaluation of CMAI

Table 2 EFA solution of the CMAI

No. Item Factor 1 Factor 2 Communalities

4 Cursing or verbal aggression 0.48 0.29 0.42

7 Hitting (including self) 0.91 0.08 0.78
9 Grabbing onto people or things inappropriately 0.84 0.05 0.68
10 Pushing 0.73 0.00 0.53
13 Screaminga 0.39 0.33 0.37
1 Pacing and aimless wandering 0.06 0.32 0.09
5 Constant unwarranted request for attention or help 0.08 0.61 0.34
6 Repetitive sentences or questions 0.20 0.56 0.27
16 Trying to get to a different place 0.08 0.49 0.27
29 General restlessness 0.02 0.48 0.24
Eigenvalue 2.93 1.05

Note.
a
Item 13 was removed due to low and double factor loadings for the final model; Loadings >.40 are shown in bold; Factor labels are: 1 = Aggressive
behaviours; 2 = Nonaggressive behaviours.

agitation, the two factors were combined to test a one- the overall CMAI score (based on 29-item version)
factor EFA model reflecting all types of agitated and PAINAD at rest and during activity.
behaviours. However, the model fit indices for this Correlations between the subscales of the original
solution were very poor using standard SEM criteria scale CMAI and PAINAD demonstrated significant
(χ 2 = 211.6, df = 35, p < .001; BIC = 5206.93, associations between physically aggressive behaviours
CFI = 0.744, TLI = 0.670, RMSEA = 0.148), suggesting and pain during activity and verbally aggressive
that these factors are distinct forms of agitated behaviours correlated both with pain at rest and
behaviours. during activity. Similarly, significant correlations
The loading pattern of the final two-factor solution were present between the aggressive agitation
was similar to the original solution described by subscale of the shortened 9-item version of the
Cohen-Mansfield (Cohen-Mansfield, 1991), but our CMAI and pain both at rest and during activity.
model combined physically and verbally aggressive
behaviours together and nonaggressive physical and
verbal behaviours together. The final model consisted Discussion
of two factors; first factor was labelled “Aggressive
behaviours” as it included items such as cursing or Findings
verbal aggression, hitting (including self), grabbing
onto people or things inappropriately and pushing. We found that in a sample of people with dementia
The second factor was labelled “Nonaggressive admitted to the acute general hospital, the CMAI
behaviours” as the items referred to the physical and measures two types of agitated behaviours, aggressive
verbal non-aggressive agitated behaviours (pacing and nonaggressive. A large number of types of
and aimless wandering, constant unwarranted request agitation included in the 29-item CMAI were not
for attention or help, repetitive sentences or questions, observed in this sample. Through exploratory factor
trying to get to a different place and general analysis, we developed a shortened 9-item modified
restlessness). Internal reliabilities (α) for the 9-item, CMAI scale for further validation and use in acute
two-factor CMAI scales were 0.83 and 0.57. settings.†
Out of the 29 items, 19 items were excluded prior to
EFA as they were not present or occurred at a rate of
Inter-correlations with pain less than 5% in our acute setting sample. Several

Inter-correlations between the original 29-item †

Excluded items were: Inappropriate dress or disrobing; Spitting; Kicking;
CMAI and PAINAD and also the shortened 9-item Pushing; Throwing things; Strange noises; Screaming; Biting; Scratching;
CMAI and PAINAD are presented in Table 3. Intentional falling; Complaining; Negativism; Eating/drinking
Significant correlations are present between the inappropriate substances; Hurt self or other; Handling things
inappropriately; Hiding things; Hoarding things; Tearing things or
subscales of the original and new versions of the destroying property; Performing repetitious mannerisms; Making verbal
CMAI. There were significant correlations between sexual advances; Making physical sexual advances.

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
 N. Kupeli et al.

Table 3 Inter-correlations between the CMAI and pain (N = 230)

29-item CMAI CMAI total PNAB PAGB VNAB VAGB

PNAB .66**
PAGB .79** .21*
VNAB .55** .36** .13*
VAGB .84** .42** .67** .34**
Pain at rest .16* .09 .07 .13 .22**
Pain during activity .10* .01 .15* .09 .19**

9-item CMAI Aggressive agitation Nonaggressive agitation

Nonaggressive agitation .33**
Pain at rest .16* .09
Pain during activity .20* .05

Note.
**p < .001. *p < .05;
PNAB = physically nonaggressive behaviours; PAGB = physically aggressive behaviours; VNAB = verbally nonaggressive behaviours;
VAGB = verbally aggressive behaviours.

modifications were required to improve the factor to other studies of people with dementia admitted to
structure of the remaining 10 items including removal acute settings in the UK (Sampson et al., 2009;
of item 13, screaming, due to low and double factor Goldberg et al., 2012). Although the sample size was
loadings on both the aggressive and nonaggressive relatively large for a clinical study, it was not large
factors. This modification resulted in a shorter 9-item enough to conduct an independent EFA and
version of the CMAI which included only behaviours subsequent confirmatory factor analysis (CFA) to
seen in people with dementia admitted to an acute confirm the factor structure developed during our
general hospital. The new, shortened version of the EFA. Although ratings of agitated behaviours were
CMAI was found to be similar to the original factor partly based on observations of patients in the acute
structure, albeit reducing the model from four factors setting, researchers also completed their ratings based
to two factors and grouping verbally and physically on discussions with family carers and health care
aggressive behaviours and the nonaggressive behaviours professionals. Thus retrospective recall bias is another
together. However, the new shortened two-factor limitation of this study. Finally, due to the
model was the best-fitting measurement model in this communication problems presented by people with
sample, suggesting that the aggressive and nonaggressive dementia, the PAINAD (Warden et al., 2003) was used
behaviours are separate forms of agitation and thus as an observational tool of pain. Nevertheless, despite
should not be combined. its simplicity, this tool consists of clear and well-
Inter-correlations between the individual sub-scales defined criteria of behaviours commonly presented
of the CMAI and pain were very similar in the shorter by people with dementia during the experience of
9-item version and the original 29-item version. However, pain. A recent systematic review of tools assessing
this relationship appears to be more prominent between symptoms presented by people with dementia has also
aggressive behaviours and pain compared with that demonstrated that out of 11 tools assessing pain,
between nonaggressive behaviours and pain, highlighting PAINAD is one of two measures to have the strongest
the distinction between the two types of agitation psychometric properties (Ellis-Smith et al., 2016).
measured by the CMAI. These findings support previous
research on the link between pain and disruptive
behaviours, in particular the relationship between Implications
aggressive behaviours and pain (Ahn and Horgas, 2013).
We have developed the original version of the CMAI
by expanding its transferability from a nursing home
Strengths and limitations population to a sample admitted to acute general
hospitals. Our study builds on work describing the
Study limitations include generalisability with regards usefulness of the tool in those attending out-patient
to ethnicity: a high percentage of the sample were clinics (Altunöz et al., 2015). Those people with
classified as ‘white’ and female. However, dementia who present with acute illnesses, are sicker
demographics of the current sample are comparable and require admission may experience different levels

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
Psychometric evaluation of CMAI

of distress, pain and agitation. However, further Acknowledgement

psychometric testing is required in acute settings.
Redundant items such as eating or drinking This study is part of The Impact of acute
inappropriate substances and hiding and hoarding Hospitalisation on People with Dementia: The
things which are unlikely to occur in this environment Behaviour and Pain (BepAid) Study. This project is
were removed to produce a more clinically relevant funded jointly by the Alzheimer’s Society and the
tool. In addition, physically unwell patients with BUPA Foundation (Grant reference number: 131).
dementia may be more likely to be bed-bound making
the prevalence of other behaviours such as wandering
lower in this population. References
People with dementia may have communication
difficulties as the disease progresses and thus may Ahn H, Horgas A. 2013. The relationship between pain and disruptive behaviors in
exhibit agitated behaviours to express unmet needs, nursing home resident with dementia. BMC Geriatrics 13: 1–7.
Altunöz U, Kizil ETÖ, Kirici S, et al. 2015. Dimensions of agitation based on the
including pain (Bachino et al., 2001). The relationship Cohen-Mansfield Agitation Inventory in patients with dementia. Turkish Journal
between aggressive behaviours and pain emphasises of Psychiatry 26: 116–122.
Alzheimer’s Society. 2016. Fix Dementia Care—Hospitals. January. 2016th edn.
the importance of structured assessment of agitated Alzheimer’s Society: London.
behaviours using a scale specifically developed for American Psychiatric Association. 1994. Diagnostic and Statistical Manual of Mental
Disorders. American Psychiatric Association: Washington DC.
acute general hospital settings. This version of the Bachino C, Snow AL, Kunrk ME, Cody M, Wristers K. 2001. Principles of pain
CMAI is a short and concise measure allowing efficient assessment and treatment in non-communicative demented patients. Clinical
Gerontologist 23: 97–115.
assessment and monitoring of agitation in people with Byrne BM. 1998. Structural Equation Modelling with Lisrel, Prelis, and Simplis: Basic
dementia when admitted to a fast-paced environment Concepts, Applications, and Programming. Lawrence Erlbaum Associates, Inc.:
Mahwah, NJ.
like a hospital ward. This may be helpful in developing Choy CN, Lam LC, Chan WC, Li SW, Chiu HF. 2001. Agitation in Chinese elderly:
management strategies for agitation in dementia in validation of the Chinese version of the Cohen-Mansfield Agitation Inventory.
International Psychogeriatrics 13: 325–335.
acute general hospitals, which, currently, can be quite Chrisman M, Tabar D, Whall AL, Booth DE. 1991. Agitated behavior in the
limited (White et al., 2017) and monitoring response cognitively impaired elderly. Journal of Gerontological Nursing 17: 9–13.
Cohen J. 1992. A power primer. Psychological Bulletin 112: 155–159.
to interventions for the agitation supporting more Cohen-Mansfield J. 1986. Agitated behaviors in the elderly: II. Preliminary results in
methodologically robust work in this area. the cognitively deteriorated. Journal of the American Geriatrics Society 34: 722–727.
Cohen-Mansfield J. 1991. Instruction manual for the Cohen-Mansfield agitation
inventory (CMAI). In Research Institute of the Hebrew Home of Greater
Washington. Rockville, MD.
Cohen-Mansfield J, Kerin P. 1986. Agitation in nursing home elderly: a quantitative
development of the concept. Monograph.
Conflict of interest Cohen-Mansfield J, Marx MS, Rosenthal AS. 1989. A description of agitation in a
nursing home. Journal of Gerontology 44: M77–M84.
Cohen-Mansfield J, Werner P, Watson V, Pasis S. 1995. Agitation among elderly
None declared. persons at adult day-care centers: the experiences of relatives and staff members.
International Psychogeriatrics 7: 447–458.
de Jonghe JFM, Kat MG. 1996. Factor structure and validity of the Dutch version on
the Cohen-Mansfield Agitation Inventory (CMAI-D). Journal of the American
Key points Geriatrics Society 44: 888–889.

• People with dementia admitted to acute settings

Ellis-Smith C, Evans CJ, Bone AE, et al. 2016. Measures to assess commonly
experienced symptoms for people with dementia in long-term care settings: a
exhibit aggressive and nonaggressive agitated systematic review. BMC Medicine 14: 38.
Finkel SI, Lyons JS, Anderson RL. 1992. Reliability and validity of the Cohen-
behaviours. Mansfield Agitation Inventory in institutionalized elderly. International Journal of
• A short and concise tool of agitation will allow Geriatric Psychiatry 7: 487–490.
Folstein MF, Folstein SE, Mchugh PR. 1975. “Mini-mental state”: a practical method
efficient assessment and monitoring of people for grading the cognitive state of patients for the clinician. Journal of Psychiatric
with dementia admitted to acute settings. Research 12: 189–198.

• The newly developed and validated 9-item

Goldberg SE, Whittamore KH, Harwood RH, et al. 2012. The prevalence of mental
health problems among older adults admitted as an emergency to a general
version of the Cohen-Mansfield Agitation hospital. Age and Ageing 41: 80–86.
Hodkinson HM. 1972. Evaluation of a mental test score for assessment of mental
Inventory can be used to monitor response to impairment in the elderly. Age and Ageing 1: 233–238.
interventions designed to reduce agitated Hu LT, Bentler PM. 1999. Cutoff criteria for fit indexes in covariance structure
analysis: conventional criteria versus new alternatives. Structural Equation
behaviours in acute settings. Modeling: A Multidisciplinary Journal 6: 1–55.
Inouye SK, Van Dyck CH, Alessi CA, et al. 1990. Clarifying confusion: the confusion
assessment methoda new method for detection of delirium. Annals of Internal
Medicine 113: 941–948.
Koss E, Weiner M, Ernesto C, et al. 1997. Assessing patterns of agitation in
Alzheimer’s disease patients with the Cohen-Mansfield Agitation Inventory.
Disclosure Alzheimer Disease & Associated Disorders 11: 45–50.
Mukadam N, Sampson EL. 2011. A systematic review of the prevalence, associations
and outcomes of dementia in older general hospital inpatients. International
None declared. Psychogeriatrics 23: 344–355.

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017
 N. Kupeli et al.

Rabinowitz J, Davidson M, De Deyn PP, et al. 2005. Factor analysis of the Cohen- Ware CJ, Fairburn CG, Hope R. 1990. A community-based study of aggressive
Mansfield Agitation Inventory in three large samples of nursing home patients behaviour in dementia. International Journal of Geriatric Psychiatry 5: 337–342.
with dementia and behavioral disturbance. The American Journal of Geriatric Watkin L, Blanchard MR, Tookman A, Sampson EL. 2012. Prospective cohort
Psychiatry 13: 991–998. study of adverse events in older people admitted to the acute general hospital:
Raftery AE. 1995. Bayesian model selection in social research. Sociological risk factors and the impact of dementia. International Journal of Geriatric
Methodology 1: 111–163. Psychiatry 27: 76–82.
Russ TC, Shenkin SD, Reynish E, et al. 2012. Dementia in acute hospital inpatients: Whall AL, Black ME, Yankou DJ, et al. 1999. Nurse aides’ identification of onset and
the role of the geriatrician. Age and Ageing 41: 282–284. level of agitation in late stage dementia patients. American Journal of Alzheimer’s
Sampson EL, Blanchard MR, Jones L, Tookman A, King M. 2009. Dementia in the Disease and Other Dementias 14: 202–206.
acute hospital: prospective cohort study of prevalence and mortality. The British White N, Leurent B, Lord K, et al. 2017. The management of behavioural and
Journal of Psychiatry 195: 61–66. psychological symptoms of dementia in the acute general medical hospital: a
Schreiner AS, Yamamoto E, Shiotani H. 2000. Agitated behavior in elderly nursing longitudinal cohort study. International Journal of Geriatric Psychiatry 32:
home residents with dementia in Japan. The Journals of Gerontology: Series B: 297–230.
Psychological Sciences and Social Sciences 55B: P180–P186. Zuidema SU, Buursema AL, Gerritsen MG, et al. 2011. Assessing neuropsychiatric
Scott S, Jones L, Blanchard MR, Sampson EL. 2011. Study protocol: the behaviour symptoms in nursing home patients with dementia: Reliability and Reliable
and pain in dementia study (BePAID). BMC Geriatrics 11: 61. Change Index of the Neuropsychiatric Inventory and the Cohen-Mansfield
StataCorp. 2013. Stata Statistical Software: Release 13. StataCorp LP: Texas. Agitation Inventory. International Journal of Geriatric Psychiatry 26: 127–134.
Suh G-H. 2004. Agitated behaviours among the institutionalized elderly with Zwakhalen SM, Van Der Steen JT, Najim M. 2012. Which score most likely represents
dementia: validation of the Korean version of the Cohen-Mansfield Agitation pain on the observational PAINAD pain scale for patients with dementia? Journal
Inventory. International Journal of Geriatric Psychiatry 19: 378–385. of the American Medical Directors Association 13: 384–389.
Warden V, Hurley AC, Volicer L. 2003. Development and psychometric evaluation of
the Pain Assessment in Advanced Dementia (PAINAD) Scale. Journal of the
American Medical Directors Association 4: 9–15.

Copyright # 2017 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2017

View publication stats