119© NAPICU 2016
Journal of PsychiatricIntensive Care
Journal of Psychiatric Intensive Care, 12 (2): 119–127doi:10.20299/jpi.2016.009Received 15 July 2015 | Accepted 28 January 2016© NAPICU 2016
REVIEW ARTICLE
The use of a token economy for
behaviour and symptom management
in adult psychiatric inpatients: a critical
review of the literature
Krista Glowacki, Grace Warner, Cathy White
School of Occupational Therapy, Dalhousie University, CanadaCorrespondence to: Krista Glowacki, School of Occupational Therapy, ForrestBuilding, PO Box 15000 Halifax, Nova Scotia, B3H 4R2, Canada;krista.glowacki@dal.ca
Background: A token economy is a behavioural modification and rewardbased intervention in which tokens are given for predefined terms. Thisreview aims to answer the question: What is the effectiveness of the use ofa token economy for the reduction of negative behaviours and symptomsin adult psychiatric inpatients?Method: A systematic review of studies using a token economy for adultswith mental illness, within an inpatient setting was undertaken for theperiod 1999–2013. References cited in relevant literature were alsoexamined.Results: The Oxford CEBM Levels of evidence was used to determinequality. Grade A and B recommended studies were included in the review.A total of seven studies were included in the analysis. All of the studiesshowed the effectiveness of a token economy for reducing negativebehaviours and symptoms in the short-term.Conclusions: The use of a token economy, on the basis of reward andencouragement, should be considered within inpatient psychiatric settings.The literature shows the effectiveness on behavioural changes in reductionof violence and aggression. The literature on negative symptom reductionis scarce and cannot be generalised. There is no evidence to support thetransfer outside of an inpatient/secure setting.
Key words: token economy; psychiatric inpatient; symptom manage-ment; behaviour management
Financial support: This research received no specific grant from any funding agency,commercial or not-for-profit sectors.Declaration of interest: None.
120 © NAPICU 2016
GLOWACKI ET AL.
Introduction
With the shift toward community-based mental healthcare, inpatient psychiatry units are seeing an increase inacuity of the patients who come through their doors (Bow-ers, 2005). Common reasons for admission include dangerto self or others, severe mental disorder such as psychosis,and extreme behaviours such as agitation, mania,unpredictability, confusion, disorientation, emotionallability, distress/tears, acting out and delusions (Bowers,2005). Patients may exhibit negative symptoms such asslow and superficial responses, social withdrawal, andlack of energy (Hopko et al. 2003; Gholipour et al. 2012),or negative behaviours, including agitation and aggres-sion particularly toward staff members (Lepage et al.2003; Park & Lee, 2012). Thus, the creation of a safe andsecure environment becomes paramount.
As Bowers (2005) discussed, ongoing risk assessmentand monitoring and observation of the patients are routineaspects of the care, which may lead to the need to employskills in negotiation, persuasion, coaxing, distraction andde-escalation. When patients do escalate, disrupting themilieu and placing themselves and/or others at risk, be-haviour management strategies such as exerting physicalcontrol, restraints and coercive use of medications may beemployed to mediate the behaviour. One approach tobehaviour modification that has received limited recentattention in the literature is the use of a token economy.
Background
A token economy, developed for use within inpatientpsychiatry settings, is a behaviour modification interven-tion that can be used to shape behaviours including acquir-ing new skills, reducing undesired behaviours, increasingtreatment compliance, and improving overall manage-ment of patients on psychiatry units (LePage et al. 2003;McMonagle & Sultana, 2000). This intervention is basedon operant conditioning. Patients can earn ‘tokens’ whichhave no innate value, and can exchange them for some-thing that does have value to them, such as goods, servicesor privileges in the facility when they exhibit a desiredbehaviour (Seegert, 2003; McMonagle & Sultana, 2000).The first principle of the token economy is the law of causeand effect based on the idea that reinforcement is the mosteffective means in changing behaviour. The second princi-ple is the law of contiguity association, in that two eventswill be associated with one another if they happen together(Dickerson et al. 2005; McMonagle & Sultana, 2000). Inthe original economy, both reward and punishment tech-niques could be implemented (Kreyenbuhl et al. 2010).Punishment is now viewed as inappropriate within ahealthcare setting, causing the decline of this intervention.Punishment is considered a negative consequence, includ-
ing the removal of tokens. There are common mis-conceptions about all token economies, including thebelief that the intervention is abusive, it does not fosterindividual treatment, and does not generalise. These mis-conceptions prevail among health care practitioners andfurther contribute to its lack of use (LePage et al. 2003).
A token economy can facilitate improvement in behav-iour and function. It is an economically friendlyintervention, and can be beneficial in facilities with lim-ited resources (LePage, 1999; Seegert, 2003; Coelho et al.2008; Comaty et al. 2001; McMonagle & Sultana, 2000;Kreyenbuhl et al. 2010). It is relatively simple in its overallconceptualisation for those involved, and is beneficial forreducing challenging or disruptive behaviours (LePage,1999; Coelho et al. 2008). Token economies can be used toincrease functioning and to foster recovery, a key focus oftoday’s mental health care (Hassell, 2009).
A systematic review of the use of token economies waspublished in 2000, analysing literature up to 1999(McMonagle & Sultana, 2000). McMonagle & Sultana(2000) concluded by recommending the token economyas a cost-effective alternative to psychosocial interven-tions in institutions with financial struggles. The articlealso recommends further in-depth research in a controlledsetting using randomised trials to further explore effec-tiveness. This systematic review of the literature examinescurrent research (1999–2013) on the use of a current tokeneconomy in adult inpatient psychiatric settings. The ques-tion guiding the review is: What is the effectiveness of theuse of a token economy for the reduction of negativebehaviours and symptoms in adult psychiatric inpatients?
Method
Inclusion criteria
Types of studies. Peer reviewed articles including:randomised controlled trials, prospective cohort studies,retrospective cohort studies and pre–post design.
Types of participants. Adults ages 18 and older admittedto a psychiatric facility as an inpatient in a forensic, acute,or rehabilitation unit, with a mental health disorder asidentified in the Diagnostic Statistical Manual of MentalDisorder, 5th Edition (DSMV).
Types of interventions. Intervention included a tokeneconomy in which tokens or vouchers are given as rewardsfor behaviour specified prior to entering the programme/economy. Rewards may be given for positive behaviour orabstinence of negative behaviour. The goal is to achievebehavioural change by means of use of non-monetary andnon-consumable tokens, which can be exchanged for avariety of goods, privileges or services in the facility.
121© NAPICU 2016
A TOKEN ECONOMY
Types of outcome measures. To determine if the therapy iseffective, there must be a reduction in one of the twoidentified outcomes after the implementation of the inter-vention. The identified outcomes are negative behavioursor negative symptoms. Negative behaviours include: vio-lence, aggression, and drug abuse. Negative symptomsinclude: flat affect, lack of pleasure in life, lack of partici-pation, lack of ability to begin and sustain activities, andlack of socialisation and interaction with others. Out-comes can be measured by observation data, frequencydata, incident reports, patient charts, group participationnumbers/percentages and number of positive urine sam-ples. Statistical information was extracted from each studyinclusive of average test scores and standard deviation,statistical significance and effect size in changes or differ-ences.
Search strategy
Electronic searches were undertaken, limiting results tothe English language and publication in the period 1999–2013 (due to the McMonagle & Sultana (2000) reviewincluding research prior to 1999). The databases CINAHL,EMBASE, OTseeker, PubMed, PsycInfo and GoogleScholar were used. The search terms used in CINAHL(EBSCOhost) were: (1) “token economy” OR (tokens ORvouchers) and psychiatric OR (mental* N2 (health OR ill*
OR disorder*)) and inpatient* OR hospital* OR ward*OR unit OR patient* OR forensic*; (2) “token economy”OR tokens OR vouchers and psychiatric OR mental*NEAR/2 (health OR ill* OR disorder*) and inpatient* ORhospital* OR ward* OR unit OR patient* OR forensic* andbehavi* OR violen* OR aggressi* OR negative; (3) “To-ken economy” and adult; (4) “Token economy” andpsychiatric OR (mental* N2 (health OR ill* OR disor-der*)); (5) Voucher-based and mental health. Other similarsearch terms were used in the other databases. An exami-nation of references cited in relevant literature was alsoundertaken.
Exclusion criteria
Research done before 1999, participants under the age of18, outpatient settings, and diagnoses not in the DSMVwere excluded. Specific study types not included were:systematic reviews, open forum blogs, hospital unit re-views and descriptive articles of intervention without ameasureable outcome (see Fig. 1).
Data extraction & quality review
Articles were identified through electronic searches andabstracts were reviewed. Those that did not meet theinclusion criteria were then excluded. Of the abstractreviews, 20 articles were identified and the full manuscripts
Identified through
searching database:
n = 342
Identified through
examination of
references:
n = 3
Excluded after
abstract review:
n = 325
Full-text articles assessed
for eligibility:
n = 20
Included after
manuscript review:
n = 7
Excluded after manuscript review: n = 13
Reasons: Population not inpatients: n = 4
Study design: n = 7
Outcome measured not negative behaviour change: n = 1
Full text not accessible: n = 1
Fig. 1. Articles included and excluded.
122 © NAPICU 2016
GLOWACKI ET AL.
of the papers were read and assessed for quality andeligibility. The Oxford Centre for Evidence-Based Medi-cine Levels of Evidence was used to determine quality,and only grade A and B studies were included in the review(OCEBM, 2009). Grade A studies are considered thehighest quality and grade B studies are the second highestquality. After the full manuscript reviews, 13 were ex-cluded, leaving 7 studies to be included in the systematicreview. Figure 1 indicates reasons why studies were ex-cluded.
Results
Data were extracted from seven studies and compared todetermine the effectiveness of a token economy (Table 1).Each study included a rewards and incentive based tokeneconomy for adults within an inpatient psychiatry setting.Of the seven studies, three were randomised controlledtrials, one was a prospective cohort, two were pre–postdesigns and one was a retrospective cohort. All of thesestudies were categorised using the OCEBM (2009) todetermine study quality.
The studies classified as grade A of the OCEBM (2009)were the three randomised controlled trials (Hopko et al.2003; Gholipour et al. 2012; Park & Lee, 2012).Randomisation methods were difficult to assess as entireinpatient units were used. Two of the studies looked only atmale units and were done outside of North America(Gholipour et al. 2012; Park & Lee, 2012), and in one, allmales on the unit were diagnosed with schizophrenia(Gholipour et al. 2012). Thus, cultural and gender differ-ences should be considered and the results be used withcaution to generalise to North American culture and prac-tice, and to mixed units. Two of the studies had smallsample sizes (Hopko et al. 2003; Gholipour et al. 2012).The next grade of studies, level B of the OCEBM (2009),were pre–post designs and prospective cohorts (LePage,1999; Comaty et al. 2001; LePage et al. 2003). Two of thestudies only analysed one unit of a hospital, limiting thegeneralisation, and there was no control group as all patientsin the unit participated in the token economy (LePage,1999; LePage et al. 2003). Lastly, a retrospective study,also grade B of the OCEBM (2009) was used (Hassell,2009). This included an analysis of medical records and nopower analysis was used to determine if sample size wasadequate. In the same study various healthcare profession-als implemented the token economy and no informationwas given on inter-relater reliability or training. It shouldalso be noted that in all of the studies the intervention wascombined with individualised pharmacological treatment.
The effect of interventions on outcomes
The effectiveness of the intervention being analysed willbe reported as a narrative synthesis. This part will be
separated into two sections. The first will describe theeffectiveness of the intervention on negative behaviours,and the second will describe the effectiveness on negativesymptoms.
Behaviour. Behavioural change was an outcome meas-ured in five of the studies (LePage, 1999; Comaty et al.2001; LePage et al. 2003; Hassell, 2009; Park & Lee,2012). These studies included violent and aggressive be-haviour in a physical or non-physical manner that couldharm or threaten other individuals or themselves, mainlyreported as ‘incidents’. All of the studies’ findings supportthe efficacy of the token economy for reducing negativebehaviours and unit incidents related to these behaviourson inpatient psychiatric units. It should be noted that onestudy did not have a control, so improvements in behav-iour and function cannot be attributed to the token economyalone (Hassell, 2009). Further, one study and its resultsprovides data to support the positive long-term impact atoken economy can have on the safety and function of anacute care unit (LePage et al. 2003).
Negative symptoms. The effect of the intervention onnegative symptoms was examined in two studies (Hopkoet al. 2003; Gholipour et al. 2012). The findings from bothstudies support the efficacy of the token economy fornegative symptom reduction in an inpatient setting. How-ever, the study by Hopko et al. (2003) only includedinpatients diagnosed with depression, and the study byGholipour et al. (2012) only included males with schizo-phrenia. See Table 1 for a summary of the outcomes,results, and the statistical data from each study.
Discussion
This systematic review supports the efficacy of a tokeneconomy for reducing negative behaviours in adults withmental health illness in an inpatient psychiatry setting. Allfive of the studies analysing behavioural change showedstatistical significance in the reduction of negative behav-iours. While the literature reviewed on the efficacy of theeconomy for reducing negative symptoms supported theuse of the intervention, the number of studies was limited,providing insufficient evidence to support its use. Further,the symptom reduction studies only targeted one diagno-sis, not representative of most psychiatric inpatient units.The research on symptom reduction alone is scarce, andthis outcome measure needs to be explored further. All ofthe studies looked at the effects within an inpatient setting,and the regime of hospital units is important to con-sider. This includes structure, schedules, expectations andstaff. None of the studies were able to look at the directeffects on behaviour or symptoms outside of an inpatientsetting, so this is an important caution if considering use of
123© NAPICU 2016
A TOKEN ECONOMYTa
ble
1.
Data
ext
ract
ed fro
m s
eve
n s
tudie
s on the u
se o
f a toke
n e
conom
y
Co
maty
et
al. 2
001
Gh
olip
ou
r et
al. 2
012
Hassell, 2
009
Ho
pko
et
al.
2003
LeP
ag
e, 1
999
LeP
ag
e e
t al.
2003
Park
& L
ee
, 2012
Ob
jecti
ve(s
)1. T
o c
om
pare
and
To c
om
pare
the e
ffect
-1. T
o in
vest
igate
diff
er-
To e
xam
ine e
ffect
ive-
To e
valu
ate
the im
pact
1.
To e
xpand o
n the
To e
xam
ine th
e e
ffect
-analy
se d
ata
of th
ose
iveness
of exe
rcis
eence
s betw
een
ness
of bri
ef B
AT
Dof a toke
n e
conom
y in
pre
vious
findin
gs
of
iveness
of th
e S
TT
Ew
ho c
om
ple
ted the
and toke
n-b
ehavi
our
inpatie
nts
with
seri
ous
com
pare
d to S
P.an a
cute
psy
chia
tric
LeP
age (
1999)
of
on v
iole
nt behavi
ou
rspro
gra
mm
e a
nd those
thera
py
on n
egativ
em
enta
l illn
ess
and
unit
on n
egativ
ere
duce
d le
vels
of
am
ong p
sych
iatr
icw
ho d
id n
ot.
sym
pto
ms
in a
dults
dually
dia
gnose
deve
nts
.aggre
ssio
n fro
m a
inpatie
nts
.2. T
o d
ete
rmin
e the
with
sch
izophre
nia
.in
patie
nts
.to
ken e
conom
y.effect
iveness
of th
e2. T
o b
etter
unders
tand
2.
To c
om
pare
the
pro
gra
mm
e for
those
the in
fluence
of a toke
nnum
ber
of ass
aults
who c
om
ple
ted the
eco
nom
y pro
gra
mm
eon a
unit
12 m
onth
spro
gra
mm
e p
ost
-on funct
ionin
g a
fter
befo
re the im
ple
ment-
dis
charg
e.
hosp
ital d
isch
arg
e.
atio
n o
f a toke
neco
nom
y to
the 2
4m
onth
s after
imple
m-
enta
tion o
f a toke
neco
nom
y.
Desig
nP
rosp
ecti
ve c
oh
ort
Ran
do
mis
ed
Retr
osp
ecti
ve
Ran
do
mis
ed
Pre
–p
ost
desig
n. T
wo
Pre
–p
os
t de
sig
n.
Ran
do
mis
ed
de
sig
n a
naly
sing d
ata
co
ntr
olled
tri
al w
ithco
ho
rt s
tud
y a
naly
sing
co
ntr
olled
tri
al co
m-
four-
month
peri
ods
Data
was
analy
sed for
co
ntr
olled
tria
l.co
mpari
ng a
PD
gro
up,
two in
terv
entio
n g
roups
data
fro
m m
edic
al
pari
ng tw
o in
terv
entio
nw
ere
ass
ess
ed. T
he
12 m
onth
s pre
-im
ple
-P
art
icip
ants
were
a B
D g
roup a
nd a
(toke
n thera
py
or
reco
rds
of in
patie
nts
gro
ups:
BAT
D, and S
P.fir
st w
as
with
out th
em
enta
tion o
f th
e toke
neith
er
in a
n e
xperi
-gro
up w
ho d
id n
ot co
m-
exe
rcis
e)
and o
ne
underg
oin
g the P
IP.
toke
n e
conom
y and
eco
nom
y and a
gain
inm
enta
l gro
up in
volv
ed
ple
te th
e in
terv
entio
n.
contr
ol g
roup r
ece
ivin
gT
he d
ata
com
pare
dth
e s
eco
nd w
as
with
a fo
llow
-up p
eri
od o
fin
the toke
n e
conom
yP
atie
nts
who d
id n
ot
regula
r in
patie
nt
two p
opula
tions,
in-
imple
menta
tion o
f th
e24 m
onth
s post
-or
a c
ontr
ol g
roup
com
ple
te the
treatm
ent.
patie
nts
with
dia
gnose
sto
ken e
conom
y.im
ple
menta
tion.
with
regula
r in
patie
nt
pro
gra
mm
e w
ere
of one s
eri
ous
menta
ltr
eatm
ent.
dis
charg
ed fro
m the
illness
and in
patie
nts
pro
gra
mm
e if
they
with
a d
ual d
iagnose
s.co
uld
not m
eet cr
iteri
ae.g
. dem
onst
ratin
g a
speci
fic r
educt
ion in
ma
lad
ap
tive
be
hav-
iours
in the first
1–
3 w
ee
ks.
Sam
ple
Part
icip
ants
were
both
Part
icip
ants
were
male
Part
icip
ants
were
male
Part
icip
ants
were
Part
icip
ants
in the
Part
icip
ants
in the
Part
icip
ants
in the
sexe
s on a
rest
rict
ive
inpatie
nts
, 20–50
inpatie
nts
. They
eith
er
inpatie
nts
with
ast
udy
were
adm
itted to
study
were
adm
itted
study
were
male
inpatie
nt unit,
aged
years
of age, m
inim
um
had a
dia
gnosi
s of a
pri
nci
pal d
iagnosi
s of
a n
eo-a
dult
inpatie
nt
to a
genera
l adm
issi
on
aged 3
0–60 w
ith a
18 y
ears
or
old
er
and
3 y
ears
of si
ngle
ment-
single
seri
ous
menta
lm
ajo
r depre
ssio
n w
ithpsy
chia
tric
unit,
and
adult
inpatie
nt
his
tory
of vi
ole
nce
,a
du
al-d
iag
no
sis
of
al i
llne
ss d
iag
no
ses
illn
ess
or
a d
ua
lno h
isto
ry o
f psy
chosi
s.in
volu
nta
rily
com
mitt
ed
psy
chia
tric
unit
with
adm
itted to a
menta
l reta
rdatio
n a
nd
of sc
hiz
ophre
nia
.dia
gnosi
s, it
was
their
BAT
D n
= 1
0w
ith v
ari
ous
menta
lva
rious
menta
l illn
ess
psy
chia
tric
ho
spita
l.a D
SM
IV
Dia
gnose
sE
xerc
ise a
s in
ter-
first
hosp
italis
atio
n o
nS
P n
= 1
5.
illness
dia
gnose
s.dia
gnose
s.E
xperi
menta
l gro
up
of a s
eve
re P
D o
r B
D.
ventio
n n
= 1
5th
e u
nit;
they
staye
d in
Pre
-im
ple
menta
tion
Pre
-im
ple
menta
tion
Contr
ol g
roup
Com
ple
tion o
fTo
ken-b
ehavi
our
hosp
ital a
t le
ast
30
n =
593
n =
31
6n
= 2
2.
pro
gra
mm
e &
PD
thera
py a
s in
terv
entio
ndays
and p
art
icip
ate
dP
ost
-im
ple
menta
tion
Post
-inte
rventio
nn =
17
n =
15
in the P
IP. S
MI n =
101
n =
59
6.
n =
55
3.
Co
mp
letio
n o
fC
on
tro
l n =
15
.D
ua
l dia
gn
osi
s D
Dpro
gra
mm
e &
BD
n =
46.
n =
1N
on-c
om
ple
tion
of th
e p
rogra
mm
en =
17.
Contin
ued
124 © NAPICU 2016
GLOWACKI ET AL.
Tab
le 1
.C
on
tinu
ed Co
maty
et
al. 2
001
Gh
olip
ou
r et
al. 2
012
Hassell, 2
009
Ho
pko
et
al.
2003
LeP
ag
e, 1
999
LeP
ag
e e
t al.
2003
Park
& L
ee
, 2012
Inte
rven
tio
nR
ew
ard
s in
the form
of
The thera
peutic
inte
r-Im
media
te p
osi
tive
Act
ivity
and g
oal
Toke
n e
conom
yTo
ken e
conom
yIn
the S
TT
E g
roup,
pla
stic
toke
ns
were
ventio
n o
f to
ken-
rein
forc
em
ent in
the
settin
g a
ctiv
ities
with
ain
cluded a
cla
ssro
om
invo
lved r
ece
ivin
gpart
icip
ants
were
giv
en to p
atie
nts
for
behavi
our
thera
py
form
of poin
ts for
the
train
ed thera
pis
t. A
train
ing s
ess
ion a
nd a
stam
ps
on a
daily
tim
eg
iven p
osi
tive
re
in-
posi
tive b
ehavi
our
was
giv
en to the r
es-
perf
orm
ance
of
pro
gre
ss a
nd m
ast
er
tria
l of th
e toke
n-
sheet co
ntr
olle
d b
yfo
rcem
ents
thro
ug
he.g
. appro
pri
ate
pect
ive in
terv
entio
nta
rgete
d b
ehavi
our
inlo
g w
as
kept and
base
d e
conom
y.st
aff.
Patie
nts
earn
ed
stic
kers
as
toke
ns.
pers
onal h
ygie
ne,
gro
up for
2-h
our
AD
Ls,
medic
atio
npatie
nts
would
meet
Toke
ns
were
pri
mari
lyst
am
ps
for
behavi
-T
hey
were
giv
en fo
rcl
eanin
g a
bedro
om
,se
ssio
ns,
3 d
ays
aco
mplia
nce
and g
roup
with
a c
linic
ian 3
tim
es
giv
en a
s re
info
rcem
ent
ours
that fa
cilit
ate
dth
e a
bst
inence
of
bein
g o
n tim
e e
tc.
week
for
3 m
onth
s.attendance
. Po
ints
per
week
for
20
for
posi
tive b
ehavi
ours
treatm
ent or
were
aggre
ssiv
e b
ehavi
our
Toke
ns
were
sto
red in
were
rem
ove
d fo
rm
inute
s to
meet and
e.g
. gro
up p
art
icip
atio
nuse
ful f
or
rem
ain
ing
for
a 2
4-h
our
day.
a ‘b
ank’
for
each
patie
nt
unacc
epta
ble
or
in-
adju
st g
oals
. To
ken
sand k
eepin
g a
room
in the c
om
munity
Toke
ns
could
be r
e-
at th
e n
urs
ing s
tatio
n.
appro
pri
ate
behavi
our.
were
pro
vided fo
rcl
ean. T
oke
ns
could
e.g
. part
icip
atin
g in
deem
ed fo
r re
ward
sF
ines
or
penalti
es
were
Poin
ts c
ould
be r
e-
ach
ievi
ng
BAT
D g
oa
ls,
be lo
st fo
r m
ajo
r sa
fety
gro
up thera
py,
medic
-e.g
. coffee, fr
ied
giv
en a
nd toke
ns
deem
ed fo
r ite
ms
in the
wh
ich
co
uld
be
ex-
viola
tions
e.g
. vio
lence
,atio
n a
dhere
nce
and
chic
ken, outd
oor
rem
ove
d for
mal-
PIP
sto
re e
.g. c
loth
ing.
changed fo
r th
ings
thre
ats
or
dest
royi
ng
show
eri
ng. M
ajo
ract
iviti
es
or
sleepin
gadaptiv
e b
ehavi
ours
e.g
. off-g
rounds
pro
pert
y. T
oke
ns
could
viola
tion to s
afe
tyin
.div
ided in
to 3
cate
-pass
es,
phone c
ard
sbe e
xchanged for
rule
s re
sulte
d in
gori
es
of se
veri
ty. T
he
and s
nack
s.th
ings
e.g
. off-g
round
toke
ns
bein
g take
nhig
hest
fin
es
were
for
pass
es,
movi
es,
aw
ay.
To
kens
could
Majo
r M
ala
daptiv
esn
ack
s etc
.be u
sed in
a toke
nV
iola
tion
s (i
ncl
ud
ing
store
for
snack
s or
phys
ical a
ssault,
movi
es
or
exc
hanged
pro
pert
y dam
age,
for
a d
esi
rable
eve
nt
subst
ance
abuse
, etc
.).
e.g
. an o
ff-u
nit
pass
.
Ou
tco
mes
Ass
ess
ment w
as
done
Ass
ess
ment w
as
done
Data
were
analy
sed in
Ass
ess
ment w
as
done
Data
were
analy
sed in
Data
were
analy
sed
Data
were
an
aly
sed
at 3 w
eeks
and 3
2pre
– and 3
month
sth
e p
eriod o
f fo
ur
years
.at base
line p
rior
to the
a 1
23 d
ay
peri
od
12 m
onth
s pre
-and the e
xperi
menta
lw
eeks
into
the p
ro-
post
-inte
rventio
n.
Change in
psy
cho-
first
tre
atm
ent se
ssio
nbefo
re the toke
nin
terv
entio
n a
nd 2
4and c
ontr
ol g
roups
gra
mm
e. D
ata
at 32
Negativ
e s
ympto
ms
soci
al f
unct
ionin
gand a
t post
-inte
r-eco
nom
y and a
122
month
s post
-co
mpare
d b
efo
re a
nd
weeks
post
-inte
r-w
ere
measu
red u
sing
(few
er
mala
daptiv
eve
ntio
n o
n D
ay
14 o
rday
peri
od a
fter
the
inte
rventio
n. T
he
after
the 1
2 w
ee
kve
ntio
n is
rele
vant.
the s
cale
for
the
behavi
ours
and m
ore
at dis
charg
e if
earl
ier.
toke
n e
conom
yperc
enta
ge d
iffere
nce
toke
n e
conom
yN
um
ber
of to
tal f
ines
ass
ess
ment of
posi
tive b
ehavi
ours
)D
epre
ssiv
e s
ympto
ms
imple
menta
tion.
of th
e n
um
ber
of
peri
od. N
um
ber
of
giv
en fo
r m
ala
daptiv
enegativ
e s
ympto
ms
from
tim
e o
f adm
issi
on
ass
ess
ed u
sing the
1.
The p
erc
enta
ge
ass
aults
, defin
ed a
saggre
ssiv
e b
ehav-
behavi
ours
, num
ber
(SA
NS
).to
dis
charg
e m
easu
red
Be
ck D
epre
ssio
ndiff
ere
nce
of th
e tota
lst
aff in
juri
es
from
iours
incl
udin
g v
erb
al
of m
inor
fines,
and
by
the G
lobal A
ssess
-In
vento
ry (
BD
I).
num
ber
of negativ
epatie
nts
, patie
nt-
attack
, pro
pert
ynum
ber
of m
ajo
r fin
es
ment of F
unct
ionin
gin
cidents
on u
nit.
patie
nt in
juri
es
and
dam
age, phys
ical
com
pare
d b
etw
een
(GA
F)
score
.2.
The n
um
ber
of
self-
inju
ries
that
attack
again
st s
elf
and
the 3
gro
ups.
patie
nt and s
taff
required m
edic
al
phys
ical a
ttack
inju
ries.
attentio
n.
again
st o
ther.
125© NAPICU 2016
A TOKEN ECONOMY
Resu
lts
The n
on-c
om
ple
ters
The toke
n-b
ehavi
our
Both
the S
MI gro
up
The B
AT
D g
roup
Tota
l num
ber
of
Tota
l num
ber
of
Tota
l num
ber
of
show
ed
sig
nifi
can
tlyth
era
py g
roup s
how
ed
and the D
D g
roup
show
ed s
ignifi
cantly
inci
dents
reduce
din
cidents
reduce
daggr
ess
ive b
ehav
iours
more
fin
es
than b
oth
few
er
negativ
esh
ow
ed s
ignifi
cant
better
score
s on the
signifi
cantly
. This
wa
ssi
gnifi
cantly
after
use
for
the e
xperi
menta
lof th
e g
roups
that
sym
pto
ms
after
inte
r-im
pro
vem
ents
on the
BD
I in
dex
than the S
Pfr
om
129 tota
l inci
dents
of th
e toke
n e
conom
y.gro
up d
ecr
ease
d fr
om
finis
hed the toke
nve
ntio
n. T
he r
esu
ltsG
AF
after
hosp
italis
-gro
up (
usu
al i
npatie
nt
to 7
3 in
cidents
. The
The n
um
ber
of
178 to 1
41aft
er
the
eco
nom
y in
terv
entio
n,
were
sta
tistic
ally
atio
n a
nd the toke
ntr
eatm
ent)
, sh
ow
ing
num
ber
of st
aff a
nd
patie
nt–
patie
nt in
juries
inte
rventio
n. T
his
was
thus
show
ing h
igher
signifi
cant.
eco
nom
y pro
gra
mm
e.
negativ
e s
ympto
mpatie
nt in
juri
es
als
odecr
ease
d s
ignifi
c-a s
ignifi
cant d
ecr
ease
rate
s of m
ala
daptiv
eT
he im
pro
vem
ent is
an
reduct
ion. T
here
wa
ssi
gnifi
cantly
reduce
d.
antly
. The n
um
ber
of
as
com
pare
d to the
behavi
ours
while
inin
dic
atio
n o
f psy
cho-
als
o a
larg
e e
ffect
patie
nt–
staff a
nd s
elf-
contr
ol g
roup.
hosp
ital.
soci
al i
mpro
vem
ent,
size
.in
juri
es
reduce
d, but
meanin
g few
er
mal-
the r
esu
lts w
ere
not
ad
ap
tive
be
havi
ou
rs.
sta
tistic
ally
sig
nifi
can
t.
Sig
nif
ican
ce
Tota
l fin
es:
p <
0.0
01
p <
0.0
01.
Post
-inte
rventio
n a
nd
BAT
D g
roup:
After
toke
n e
conom
y:A
fter
toke
n e
conom
y:A
fter
toke
n e
conom
y:S
tatis
tical
Min
or
fines:
p <
0.0
06
Mean a
nd s
tandard
at dis
charg
e for
both
Mean a
nd s
tandard
Tota
l inci
dents
😛
atie
nt–
patie
nt
Ave
rage a
nd s
tandard
signifi
cance
Majo
r fin
es:
p <
0.0
02
dev
iatio
n:
gro
ups
p <
0.0
1. A
ndevi
atio
n:
43%
decr
ease
inju
ries:
48%
change
dev
iatio
n:
(p <
0.0
5)
Mean a
nd s
tandard
pre
– 77 ±
14
incr
ease
in the G
AF
pre
– 35.1
± 7
.4 a
nd
(p <
0.0
1)
(p <
0.0
5)
Exp
eri
menta
l gro
up:
devi
atio
n w
as
only
post
– 41 ±
11.
indic
ate
s im
pro
vem
ent.
post
– 19.1
± 1
3.1
.S
taff in
jury
: 54%
Self-
inju
ry: 1
7%
Pre
– 53.9
5 ±
3.1
2giv
en fo
r le
ngth
of
Mean a
nd s
tandard
This
was
signifi
cantly
decr
ease
(p
< 0
.01)
change (
not
Post
– 41.9
0 ±
2.3
4hosp
ital s
tay
whic
h is
devi
atio
n:
gre
ate
r th
an the
Patie
nt In
jury
: 45%
signifi
cant)
Contr
ol g
roup:
not th
e o
utc
om
eS
MI gro
up a
tch
ange o
f th
e S
Pdecr
ease
(p
< 0
.05).
Sta
ff in
juri
es:
21%
Pre
– 54.0
4 ±
1.5
5lo
oke
d a
t fo
r th
e p
ur-
adm
issi
on:
gro
up (
p <
0.0
5).
Mean a
nd s
tandard
change (
not
Post
– 56.2
2 ±
0.9
2pose
s of th
is p
aper.
35.6
8 ±
8.6
5 a
nd
Cohen E
ffect
siz
edevi
atio
n n
ot giv
en.
signifi
cant)
Gro
up d
iffere
nce
s:dis
charg
e:
larg
e (
d =
0.7
3).
Tota
l inju
ries:
33%
p <
0.0
01.
51.7
0 ±
8.5
4.
change (
p <
0.0
5).
DD
gro
up a
t adm
issi
on:
Mean a
nd s
tandard
35.3
5 ±
8.6
5 a
nd a
tdevi
atio
n n
ot g
iven.
dis
charg
e:
54.2
2 ±
7.0
8.
Co
nclu
sio
ns
The fin
din
gs
support
The fin
din
gs
support
The fin
din
gs
show
The fin
din
gs
support
The fin
din
gs
support
The fin
din
gs
support
The fin
din
gs
support
& Im
plicati
on
sth
e e
ffic
acy
of th
eth
e e
ffic
acy
of th
est
atis
tical s
ignifi
cance
the e
ffic
acy
of a toke
nth
e e
ffic
acy
of th
eth
e e
ffic
acy
of th
eth
e e
ffic
acy
of to
ken
toke
n e
conom
y fo
rto
ken e
conom
y fo
rfo
r im
pro
ving p
sych
o-
eco
nom
y fo
r sy
mpto
mto
ken e
conom
y fo
rto
ken e
conom
y fo
reco
nom
y use
for
red
uci
ng
ma
lad
ap
tive
red
uci
ng
ne
ga
tive
soci
al b
eh
avi
ou
r, b
ut
red
uct
ion
in in
pa
tien
tsre
duci
ng tota
l negativ
ere
duci
ng in
juri
es
and
reduci
ng a
ggre
ssiv
ebehavi
ours
short
-term
.sy
mpto
ms
short
-term
.th
is c
annot be
dia
gnose
d w
ithin
cidents
on a
n a
cute
negativ
e b
ehavi
ours
behavi
ours
. The
It a
lso s
how
s th
eT
he p
opula
tion w
as
attri
bute
d to the toke
ndepre
ssio
n, but it
unit
and for
reduci
ng
rela
ted to v
iole
nce
on
popula
tion w
as
limite
dbenefit
s fo
r 2 d
iffere
nt
limite
d to m
ale
s w
itheco
nom
y alo
ne s
ince
cannot be g
enera
lised
patie
nt/st
aff in
jury
.an in
patie
nt acu
teto
male
s in
an in
-cl
inic
al p
opula
tions.
schiz
ophre
nia
.th
ere
was
no c
ontr
ol
due to s
mall
sam
ple
psy
chia
tric
unit.
patie
nt psy
chia
try
gro
up to c
om
pare
.si
ze. T
he p
opula
tion
settin
g w
ith a
his
tory
was
limite
d to
of aggre
ssiv
ein
div
idu
als
with
MD
D.
be
havi
ou
r.
AD
L: a
ctiv
ities
of d
aily
livi
ng; B
AT
D: b
ehavi
oura
l act
ivatio
n tr
eatm
ent f
or d
epre
ssio
n; B
D: b
ehavi
oura
l dis
ord
er;
DD
: dual d
iagnosi
s M
DD
: majo
r depre
ssiv
e d
isord
er;
PD
: psy
chia
tric
dis
ord
er;
PIP
: poin
ts in
centiv
e p
rogra
mm
e; S
MI:
seri
ous
menta
l illn
ess
; SP
: support
ive p
sych
oth
era
py;
ST
TE
: short
-term
toke
n e
conom
y
126 © NAPICU 2016
GLOWACKI ET AL.
this intervention in a community setting. Comaty et al.(2001) looked at a three-year follow-up after dischargeand re-hospitalisation rates, but the data was not conclu-sive; thus the authors were unable to say whether a longerstay in the community was a result of the token economyitself.
Using a token economy may be a means of reducingviolence and aggression on inpatient units. Improved safetyin inpatient units for staff and patients may allow healthcarepractitioners to focus their attention on treatment andrehabilitation, and less on control of the unit. Behaviouralfocus could be less on the reduction of negative behav-iours, and more on the promotion of positive behaviour forrehabilitation and recovery. Further, reduction in negativesymptoms may also increase participation and collabora-tion of patients toward rehabilitation goals.
Healthcare practitioners should be cautious in using atoken economy. The token economies used in the studiesof this paper varied in rewards, desired behaviour andstructure. In saying this, the general concept of eacheconomy was the same, earning tokens for positive behav-iour towards rehabilitation that can be accumulated forpre-determined rewards. The economy has potential forabuse if implemented improperly, so programmes shouldlimit punishment and response costs for behaviour. This isinclusive of the removal of tokens for negative behaviour.Based on a recovery-oriented, patient-centred approach,the following recommendations derive from the literature:
1. Participants in the programme should be given theoption of enrolling at admission, and participationshould stay voluntary throughout (LePage et al. 2003;Park & Lee, 2012).
2. Staff and patients involved in the token economyshould collaborate to pre-determine rewards (Chiouet al. 2006; Dunn et al. 2008; Park & Lee, 2012).Rewards should also be individualised (Park & Lee,2012).
3. Thorough staff training should be done to ensureconsistency in programme implementation (LePage1999; LePage et al. 2003).
4. The token/voucher should be given immediately,thereby verifying positive behaviour (Chiou, et al.2006; LePage et al. 2003).
5. The economy should be used in congruence withindividualised treatment programmes (LePage, 1999;Gholipour et al. 2012; Park & Lee, 2012).
It is important to identify the limitations to this literaturereview. A small number of studies were analysed, asthere are limited recent research studies on this interven-tion. The quality is also limited by the fact that onlythree studies were randomised controlled trials. Further,this was a brief review of the literature available. The
research was only analysed by one person, and onlyEnglish studies were used.
Areas for future research have been identified from thisreview. The first is using a date after hospital discharge tofocus on psychosocial outcomes of participants and theirfunctioning in society outside of an institutional environ-ment. Usually hospitalisation is temporary, and it isimportant to find a transferable intervention for commu-nity care and living. Further, multiple units in one hospital,or comparison of multiple hospital studies could be donewith comparison of control and experimental groups.Research should also be done on symptom reduction for awider range of diagnoses within a hospital or inpatientunit. Research could also be done on the comparison ofdifferent types of rewards within the economies.
Conclusions
The findings of this literature review support the efficacyof a token economy for reducing negative behaviours in aninpatient psychiatric setting. This intervention should beconsidered by healthcare professionals working in aninpatient setting for reduction of negative behaviours, orincreased safety of the units. When implementing a tokeneconomy, the recommendations made on adapting theeconomy based on a recovery-oriented patient-centredpractice should also be considered. Few studies were doneon the efficacy of the economy for reducing negativesymptoms, but all that were done determined that the tokeneconomy was effective in symptom reduction. The use of atoken economy for symptom reduction should be usedwith caution until further research is done on the topic.
Since the overall number of research studies was lim-ited, the findings identify the need for further research onthis intervention and its effects on reducing negative be-haviours and symptoms. While more research is needed,the current findings should encourage healthcare practi-tioners to consider the use of a rewards-based tokeneconomy as a treatment intervention in an inpatient psy-chiatry setting.
Acknowledgements
School of Occupational Therapy at Dalhousie University.The Waterford Hospital in St. John’s Newfoundland, wherethe research first started based on a clinical need.
ReferencesBowers, L. (2005) Reasons for admission and their implications for
the nature of acute inpatient psychiatric nursing. Journal ofPsychiatric & Mental Health Nursing, 12(2): 231–236. doi:10.1111/j.1365-2850.2004.00825.x
Chiou, J., Chou, M., Hsu, M. and Lin., M. (2006) Significantexperiences of token therapy from the perspective of psychoticpatients. Journal of Nursing Research, 14(4): 315–322.
127© NAPICU 2016
A TOKEN ECONOMY
Coelho, C.M., Palha, A.P., Gonçalves, D.C. and Pachana, N. (2008)Rehabilitation programs for elderly women inpatients with schizo-phrenia. Journal of Women & Aging, 20(3/4): 283–295.
Comaty, J.E., Stasio, M. and Advokat, C. (2001) Analysis of out-come variables of a token economy system in a state psychiatrichospital: a program evaluation. Research in DevelopmentalDisabilities, 22(3): 233–253.
Dickerson, F.B., Tenhula, W.N., and Green-Paden, L.D. (2005) Thetoken economy for schizophrenia: review of the literature andrecommendations for future research. Schizophrenia Research,75(2–3): 405–416.
Dunn, K., Sigmon, S., Thomas, C., Heil, S. and Higgins, S. (2008)Voucher-based contingent reinforcement of smoking abstinenceamong methadone-maintained patients: a pilot study. Journal ofApplied Behavior Analysis, 41: 527–538.
Gholipour, A.A., Abolghasemi, S.H., Gholinia, K.K and Taheri, S.(2012) Token reinforcement therapeutic approach is more effec-tive than exercise for controlling negative symptoms ofschizophrenic patients: a randomized controlled trial. Interna-tional Journal of Preventive Medicine, 3(7): 466–470.
Hassell, J.T. (2009) An evolution of the points incentive program: atoken economy program for veterans hospitalized on an inpa-tient treatment unit. PhD Thesis, Alliant International University,Los Angeles. Dissertation Abstracts International, 71.
Hopko, D.R., Lejuez, C.W., LePage, J.P., Hopko, S.D. and McNeil,D.W. (2003) A brief behavioral activation treatment for depres-
sion: a randomized pilot trial within an inpatient psychiatrichospital. Behavior Modification, 27(4): 458–469.
Kreyenbuhl, J., Buchanan, R.W., Dickerson, F.B. and Dixon, L.B.(2010) The schizophrenia Patient Outcomes Research Team(PORT): updated treatment recommendations 2009. Schizo-phrenia Bulletin, 36(1): 94–103.
LePage, J. (1999) The impact of a token economy on injuries andnegative events on an acute psychiatric unit. Psychiatric Serv-ices, 50(7): 941–944.
Lepage, J., Delben, K., Pollard, S., McGhee, M., VanHorn, L.,Murphy, J., Lewis, P., Aboraya, A. and Mogge, N. (2003) Reduc-ing assaults on an acute psychiatric unit using a token economy:a 2 year follow-up. Behavioral Interventions, 18: 179–190.
McMonagle, T. and Sultana, A. (2000) Token economy for schizo-phrenia. Cochrane Database of Systematic Reviews, 2000(3):CD001473.
OCEBM Levels of Evidence Working Group (2009) The OxfordLevels of Evidence. Oxford Centre for Evidence-Based Medi-cine, Levels of Evidence Working Group.
Park, J.S. and Lee, K. (2012) Modification of severe violent andaggressive behavior among psychiatric inpatients through theuse of a short-term token economy. Journal of Korean Academyof Nursing, 42(7): 1062–1069.
Seegert, C.R. (2003) Token economies and incentive programs:behavioral improvement in mental health inmates housed instate prisons. The Behavior Therapist, 26(1): 208, 210–211.
Reproduced with permission of copyright owner. Further reproductionprohibited without permission.
