10.1192/bjp.186.6.500Access the most recent version at DOI: 2005, 186:500-506.BJP 

DOMINIC H. LAM, PAUL McCRONE, KIM WRIGHT and NATALIE KERRbipolar disorder: 30-month studyCost-effectiveness of relapse-prevention cognitive therapy for

Referenceshttp://bjp.rcpsych.org/content/186/6/500#BIBLThis article cites 0 articles, 0 of which you can access for free at:

permissionsReprints/

permissions@rcpsych.ac.ukwrite to To obtain reprints or permission to reproduce material from this paper, please

to this article atYou can respond http://bjp.rcpsych.org/cgi/eletter-submit/186/6/500

from Downloaded

The Royal College of PsychiatristsPublished by on January 16, 2012http://bjp.rcpsych.org/

http://bjp.rcpsych.org/site/subscriptions/ go to: The British Journal of PsychiatryTo subscribe to

BackgroundBackground Wehave reported theWehavereported the

advantageous clinical outcome of addingadvantageous clinical outcome of adding

cognitive therapy tomedication inthecognitive therapy tomedication in the

prevention of relapse of bipolardisorder.prevention of relapse of bipolardisorder.

AimsAims This 30-month studycomparesThis 30-month studycompares

the cost-effectiveness of cognitive therapythe cost-effectiveness of cognitive therapy

with standard care.with standard care.

MethodMethod Werandomly allocated103Werandomly allocated103

individualswith bipolar1disorder toindividualswith bipolar1disorder to

standard treatment and cognitive therapystandard treatment and cognitive therapy

plus standard treatment.Service use andplus standard treatment.Service use and

costsweremeasured at 3-monthintervalscostsweremeasured at 3-monthintervals

and cost-effectivenesswas assessedusingand cost-effectivenesswas assessedusing

thenet-benefit approach.thenet-benefit approach.

ResultsResults The group receivingcognitiveThe group receivingcognitive

therapyhad significantly better clinicaltherapyhad significantly betterclinical

outcomes.The extra costswere offset byoutcomes.The extra costswere offset by

reduced service use elsewhere.Thereduced service use elsewhere.The

probabilityofcognitive therapybeingcost-probabilityofcognitive therapybeingcost-

effectivewashigh androbustto differenteffectivewashigh androbustto different

therapyprices.therapyprices.

ConclusionsConclusions Combination of cognitiveCombination of cognitive

therapyandmoodstabiliserswas superiortherapyandmoodstabiliserswas superior

tomood stabilisers alone interms oftomood stabilisers alone interms of

clinical outcome and cost-effectiveness forclinical outcome and cost-effectiveness for

thosewith frequent relapses of bipolarthosewith frequent relapses of bipolar

disorder.disorder.

Declaration of interestDeclaration of interest None.None.

Bipolar disorder, characterised by frequentBipolar disorder, characterised by frequent

relapses, imposes a high economic burdenrelapses, imposes a high economic burden

on society (Rice & Miller, 1995; Guptaon society (Rice & Miller, 1995; Gupta

& Guest, 2002; Patel, 2003). In our study& Guest, 2002; Patel, 2003). In our study

of relapse prevention (Lamof relapse prevention (Lam et alet al, 2003,, 2003,

2005), the group receiving cognitive2005), the group receiving cognitive

therapy had significantly better clinicaltherapy had significantly better clinical

outcomes than those receiving standardoutcomes than those receiving standard

National Health Service (NHS) care. How-National Health Service (NHS) care. How-

ever, the cost of such intervention should beever, the cost of such intervention should be

evaluated in the context of other NHSevaluated in the context of other NHS

services used. An economic evaluation willservices used. An economic evaluation will

enable healthcare providers to make pro-enable healthcare providers to make pro-

curement and policy decisions. This papercurement and policy decisions. This paper

reports health economic data for the firstreports health economic data for the first

12 months (when patients were receiving12 months (when patients were receiving

cognitive therapy and up to two boostercognitive therapy and up to two booster

sessions), as well as for the whole 30-monthsessions), as well as for the whole 30-month

study period. It is hypothesised that cogni-study period. It is hypothesised that cogni-

tive therapy may be cost-effective becausetive therapy may be cost-effective because

the costs may be offset by the less-frequentthe costs may be offset by the less-frequent

use of other NHS services.use of other NHS services.

METHODMETHOD

ProcedureProcedure

The details of the study have been reportedThe details of the study have been reported

elsewhere (Lamelsewhere (Lam et alet al, 2003, 2005). Only a, 2003, 2005). Only a

summary of the procedure will be givensummary of the procedure will be given

here. After the study had been fully ex-here. After the study had been fully ex-

plained to participants, written informedplained to participants, written informed

consent was obtained. Those who fulfilledconsent was obtained. Those who fulfilled

the study criteria were randomly allocatedthe study criteria were randomly allocated

to cognitive therapy (to cognitive therapy (nn¼51) or the compar-51) or the compar-

ison group (treatment as usual;ison group (treatment as usual; nn¼52) using52) using

sequentially numbered and sealed opaquesequentially numbered and sealed opaque

envelopes. The allocation sequence wasenvelopes. The allocation sequence was

generated by a computer program prior togenerated by a computer program prior to

recruitment. The cognitive therapists wererecruitment. The cognitive therapists were

all clinical psychologists. Independentall clinical psychologists. Independent

assessors, who were masked to the groupassessors, who were masked to the group

status, performed assessments every 6status, performed assessments every 6

months to collect information on primarymonths to collect information on primary

clinical outcomes and resource usage.clinical outcomes and resource usage.

ParticipantsParticipants

Participants, aged 18–70 years, were out-Participants, aged 18–70 years, were out-

patients of the Maudsley and Bethlempatients of the Maudsley and Bethlem

NHS Trust with DSM–IV bipolar I disorderNHS Trust with DSM–IV bipolar I disorder

(American Psychiatric Association, 1994).(American Psychiatric Association, 1994).

They experienced frequent relapses despiteThey experienced frequent relapses despite

the prescription of mood stabilisers.the prescription of mood stabilisers.

Prophylactic medication was prescribed atProphylactic medication was prescribed at

adequate doses according to theadequate doses according to the BritishBritish

National FormularyNational Formulary (British Medical Asso-(British Medical Asso-

ciation & Royal Pharmaceutical Society ofciation & Royal Pharmaceutical Society of

Great Britain, 2001). Because of relapseGreat Britain, 2001). Because of relapse

prevention, individuals were currently notprevention, individuals were currently not

fulfilling criteria for a bipolar episode: Beckfulfilling criteria for a bipolar episode: Beck

Depression Inventory (BDI; BeckDepression Inventory (BDI; Beck et alet al,,

1961) score1961) score 5530, Mania Rating Scale30, Mania Rating Scale

(Bauer(Bauer et alet al, 1991) score, 1991) score 557. This avoided7. This avoided

therapists having to use the majority oftherapists having to use the majority of

therapy sessions to treat an acute episode.therapy sessions to treat an acute episode.

In order to identify a sub-group vulnerableIn order to identify a sub-group vulnerable

to relapses, participants had to have had atto relapses, participants had to have had at

least two episodes in the previous 2 years orleast two episodes in the previous 2 years or

three episodes in the previous 5 years.three episodes in the previous 5 years.

Exclusion criteria were: actively suicidalExclusion criteria were: actively suicidal

(BDI suicide item scored 3) and currently(BDI suicide item scored 3) and currently

fulfilling criteria for substance use dis-fulfilling criteria for substance use dis-

orders. There were no significant differ-orders. There were no significant differ-

ences between the two groups in any ofences between the two groups in any of

the initial demographic characteristics orthe initial demographic characteristics or

clinical features (Table 1).clinical features (Table 1).

Assessment and primary clinicalAssessment and primary clinicaloutcomeoutcome

The Structured Clinical Instrument forThe Structured Clinical Instrument for

DSM–IV (SCID; FirstDSM–IV (SCID; First et alet al, 1996) was used, 1996) was used

to determine any episode that fulfilledto determine any episode that fulfilled

DSM–IV criteria for major depression,DSM–IV criteria for major depression,

mania or hypomania. The number of daysmania or hypomania. The number of days

with bipolar episodes was defined as dayswith bipolar episodes was defined as days

during which individuals fulfilled DSM–IVduring which individuals fulfilled DSM–IV

criteria for bipolar episodes from the SCIDcriteria for bipolar episodes from the SCID

interview. Hospital computerised recordsinterview. Hospital computerised records

were used to confirm the exact length ofwere used to confirm the exact length of

hospital stays.hospital stays.

Service utilisationService utilisation

Participants were interviewed using theParticipants were interviewed using the

Client Service Receipt Inventory (CSRI;Client Service Receipt Inventory (CSRI;

Beecham & Knapp, 2001) at baseline andBeecham & Knapp, 2001) at baseline and

at 3-monthly follow-up visits. The CSRIat 3-monthly follow-up visits. The CSRI

asks about specific health and social careasks about specific health and social care

service use. Services measured includedservice use. Services measured included

contacts with mental healthcare servicescontacts with mental healthcare services

(psychiatrists, psychologists, community(psychiatrists, psychologists, community

mental health nurses, day centres, counsel-mental health nurses, day centres, counsel-

lors and other therapists), general practi-lors and other therapists), general practi-

tioners, social workers, hospital servicestioners, social workers, hospital services

(out-patient care, day hospital contacts(out-patient care, day hospital contacts

5 0 05 0 0

BR IT I SH JOURNAL OF P SYCHIATRYBR IT I SH JOURNAL OF P SYCHIATRY ( 2 0 0 5 ) , 1 8 6 , 5 0 0 ^ 5 0 6( 2 0 0 5 ) , 1 8 6 , 5 0 0 ^ 5 0 6

Cost-effectiveness of relapse-prevention cognitiveCost-effectiveness of relapse-prevention cognitive

therapy for bipolar disorder: 30-month studytherapy for bipolar disorder: 30-month study

DOMINIC H. LAM, PAUL McCRONE, KIM WRIGHT and NATALIE KERRDOMINIC H. LAM, PAUL McCRONE, KIM WRIGHT and NATALIE KERR

COST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR BIPOLAR DISORDERCOST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR B IPOLAR DISORDER

and accident and emergency attendances),and accident and emergency attendances),

support groups and residential care.support groups and residential care.

Details of in-patient stays for mentalDetails of in-patient stays for mental

health and physical health reasons andhealth and physical health reasons and

medication were checked from case notes.medication were checked from case notes.

Medication use was only recorded every 6Medication use was only recorded every 6

months, and therefore it was assumed thatmonths, and therefore it was assumed that

the same level of medication was used inthe same level of medication was used in

the 3 months prior to this.the 3 months prior to this.

Service costsService costs

Unit and hospital costs (which aim toUnit and hospital costs (which aim to

reflect the long-term marginal costs) forreflect the long-term marginal costs) for

most services were obtained from a recog-most services were obtained from a recog-

nised national source (Netten & Curtis,nised national source (Netten & Curtis,

2000), where staff costs are calculated by2000), where staff costs are calculated by

dividing the total cost (salary, oncosts,dividing the total cost (salary, oncosts,

overheads, capital, land and training)overheads, capital, land and training)

of the service over 1 year by anof the service over 1 year by an

appropriate unit of activity. Medicationappropriate unit of activity. Medication

costs were taken from thecosts were taken from the British NationalBritish National

FormularyFormulary (British Medical Association &(British Medical Association &

Royal Pharmaceutical Society of Great Brit-Royal Pharmaceutical Society of Great Brit-

ain, 2001). The cost of a cognitive therapyain, 2001). The cost of a cognitive therapy

session was assumed to be equal to 1 h ofsession was assumed to be equal to 1 h of

a psychologist’s time (£61). Unit costs werea psychologist’s time (£61). Unit costs were

multiplied by the service utilisation data tomultiplied by the service utilisation data to

generate service costs per patient.generate service costs per patient.

Statistical analysesStatistical analyses

Clinical outcomeClinical outcome

The main clinical outcome (days with bi-The main clinical outcome (days with bi-

polar episodes), which was a continuouspolar episodes), which was a continuous

scale, was assessed for group differencesscale, was assessed for group differences

using a multivariate analysis of covariance,using a multivariate analysis of covariance,

covarying the number of previous bipolarcovarying the number of previous bipolar

episodes and medication compliance. Allepisodes and medication compliance. All

analyses were on an intention-to-treatanalyses were on an intention-to-treat

basis. The primary measure of cost-basis. The primary measure of cost-

effectiveness was the number of bipolar-effectiveness was the number of bipolar-

free days (days without a bipolar episode)free days (days without a bipolar episode)

in the period following randomisation toin the period following randomisation to

12-month and 30-month follow-up. Bi-12-month and 30-month follow-up. Bi-

polar episodes are not a sensitive measurepolar episodes are not a sensitive measure

of relapse prevention, as they can varyof relapse prevention, as they can vary

tremendously in length.tremendously in length.

Resource use dataResource use data

Comparisons were made between theComparisons were made between the

cognitive therapy and comparison groupscognitive therapy and comparison groups

forfor use of community services (i.e.use of community services (i.e.

non-in-non-in-patient services combined), psychi-patient services combined), psychi-

atric in-patient care, general in-patient careatric in-patient care, general in-patient care

and medication. Tests of significance wereand medication. Tests of significance were

only conducted when comparing total costsonly conducted when comparing total costs

for each of the 3-month time periods.for each of the 3-month time periods.

Hospital use and medication data wereHospital use and medication data were

available for most participants. Infor-available for most participants. Infor-

mation on the use of community servicesmation on the use of community services

was less complete. Where missing, the costwas less complete. Where missing, the cost

of community services was imputed byof community services was imputed by

taking the mean of the costs for the othertaking the mean of the costs for the other

time periods.time periods.

Total cost differences between theTotal cost differences between the

groups were tested for statistical signifi-groups were tested for statistical signifi-

cance using a regression model with costcance using a regression model with cost

as the dependent variable and the group in-as the dependent variable and the group in-

dicator as the independent variable. Regres-dicator as the independent variable. Regres-

sion analysis allowed us to deal with thesion analysis allowed us to deal with the

expected non-normality of the costs dis-expected non-normality of the costs dis-

tribution. Non-parametric bootstrappingtribution. Non-parametric bootstrapping

with 1000 resamples was used to addresswith 1000 resamples was used to address

the skewness in the cost data (Mooney &the skewness in the cost data (Mooney &

Duval, 1993). Confidence intervals wereDuval, 1993). Confidence intervals were

constructed at the 90% level because weconstructed at the 90% level because we

assumed that it is more acceptable to makeassumed that it is more acceptable to make

a type II error with economic data thana type II error with economic data than

with clinical data.with clinical data.

Cost-effectiveness analysisCost-effectiveness analysis

The incremental cost-effectiveness of cogni-The incremental cost-effectiveness of cogni-

tive therapy compared with standard caretive therapy compared with standard care

was determined using the net-benefitwas determined using the net-benefit

approach (Briggs, 2001). There is theoreti-approach (Briggs, 2001). There is theoreti-

cal, but unknown, value (represented bycal, but unknown, value (represented by

the termthe term ll below) that society would placebelow) that society would place

on a 1-unit improvement in outcome, ason a 1-unit improvement in outcome, as

measured by the number of bipolar-freemeasured by the number of bipolar-free

days. The net benefit to society of cognitivedays. The net benefit to society of cognitive

therapy can be defined as:therapy can be defined as:

NBNB¼((ll66E)E)77SC,SC,

where NBwhere NB¼net benefit, Enet benefit, E¼effectiveness (i.e.effectiveness (i.e.

days free of bipolar episodes over 12 and 30days free of bipolar episodes over 12 and 30

months) and SCmonths) and SC¼service costs. For exam-service costs. For exam-

ple, if a bipolar-free day is assumed to haveple, if a bipolar-free day is assumed to have

a value of £10 and if a particular individuala value of £10 and if a particular individual

has 50 of these, then theirhas 50 of these, then their grossgross benefit isbenefit is

£500. However, it is assumed that is£500. However, it is assumed that is

achieved at a cost, and if that is, say,achieved at a cost, and if that is, say,

£300 then a net benefit of £200 is achieved.£300 then a net benefit of £200 is achieved.

The trial provided us with data on effective-The trial provided us with data on effective-

ness and service costs. Therefore to esti-ness and service costs. Therefore to esti-

mate a net benefit for each individual wemate a net benefit for each individual we

had to make an assumption regarding thehad to make an assumption regarding the

level oflevel of ll..

We estimated net benefits for all parti-We estimated net benefits for all parti-

cipants by assuming different values forcipants by assuming different values for llranging between £0 and £50 in £10 incre-ranging between £0 and £50 in £10 incre-

ments. A regression model was then usedments. A regression model was then used

to determine the mean difference in netto determine the mean difference in net

benefit between the cognitive therapy andbenefit between the cognitive therapy and

standard care (treatment as usual) groupsstandard care (treatment as usual) groups

for every value offor every value of ll, controlling for baseline, controlling for baseline

costs. For each model, 1000 regressioncosts. For each model, 1000 regression

coefficients for the cognitive therapy/stand-coefficients for the cognitive therapy/stand-

ard care variable were generated usingard care variable were generated using

5 015 01

Table1Table1 Initial characteristics of thosereceiving cognitive therapy and standard treatment (comparisongroup)Initial characteristics of thosereceiving cognitive therapy and standard treatment (comparison group)

Cognitive therapy groupCognitive therapy group Comparison groupComparison group

Age, years: mean (s.d.)Age, years: mean (s.d.) 46.4 (12.1)46.4 (12.1) 41.5 (10.8)41.5 (10.8)

Female participants, %Female participants, % 5454 5757

Age at onset, years: mean (s.d.)Age at onset, years: mean (s.d.) 28.2 (11.4)28.2 (11.4) 26.2 (9.5)26.2 (9.5)

BDI score: mean (s.d.)BDI score: mean (s.d.) 12.8 (9.4)12.8 (9.4) 14.3 (10.7)14.3 (10.7)

HRSD score: mean (s.d.)HRSD score: mean (s.d.) 5.7 (5.4)5.7 (5.4) 6.5 (6.0)6.5 (6.0)

MRS score: mean (s.d.)MRS score: mean (s.d.) 2.0 (3.2)2.0 (3.2) 1.8 (2.1)1.8 (2.1)

Previous depression episodes: mean (s.d.)Previous depression episodes: mean (s.d.) 5.8 (8.0)5.8 (8.0) 5.1 (4.2)5.1 (4.2)

Previousmanic episodes: mean (s.d.)Previous manic episodes: mean (s.d.) 5.5 (6.1)5.5 (6.1) 3.9 (2.8)3.9 (2.8)

Previous hypomanic episodes: mean (s.d.)Previous hypomanic episodes: mean (s.d.) 1.3 (2.7)1.3 (2.7) 0.2 (0.5)0.2 (0.5)

Previous hospitalisations: mean (s.d.)Previous hospitalisations: mean (s.d.) 6.3 (5.9)6.3 (5.9) 5.1 (6.3)5.1 (6.3)

Proportion (Proportion (nn//NN) of patients on:) of patients on:

Onemood stabiliserOnemood stabiliser 80% (41/51)80% (41/51) 90% (47/52)90% (47/52)

Twomood stabilisersTwomood stabilisers 20% (10/51)20% (10/51) 10% (5/52)10% (5/52)

AntidepressantsAntidepressants 26% (13/51)26% (13/51) 35% (13/52)35% (13/52)

Major tranquillisersMajor tranquillisers 51% (26/51)51% (26/51) 40% (21/52)40% (21/52)

Social classSocial class11: % (: % (nn//NN))

11 12% (6/51)12% (6/51) 12% (6/52)12% (6/52)

22 35% (18/51)35% (18/51) 37% (19/52)37% (19/52)

33 39% (20/51)39% (20/51) 38% (20/52)38% (20/52)

44 10% (5/51)10% (5/51) 6% (3/52)6% (3/52)

55 4% (2/51)4% (2/51) 8% (4/52)8% (4/52)

BDI, Beck Depression Inventory; HRSD,Hamilton Rating Scale for Depression; MRS,Mania Rating Scale.BDI, Beck Depression Inventory; HRSD,Hamilton Rating Scale for Depression; MRS,Mania Rating Scale.1. Office of Population Censuses and Surveys, 1991.1. Office of Population Censuses and Surveys, 1991.

LAM ET ALLAM ET AL

bootstrapping, and the proportion of thesebootstrapping, and the proportion of these

greater than zero indicated the probabilitygreater than zero indicated the probability

that cognitive therapy was cost-effectivethat cognitive therapy was cost-effective

(i.e. resulted in a mean incremental net(i.e. resulted in a mean incremental net

benefit greater than zero) for that value ofbenefit greater than zero) for that value of

ll. These probabilities were subsequently. These probabilities were subsequently

used to generate a cost-effectivenessused to generate a cost-effectiveness

acceptability curve.acceptability curve.

Sensitivity analysisSensitivity analysis

The only addition to the standard packageThe only addition to the standard package

of care was sessions of cognitive therapy.of care was sessions of cognitive therapy.

The unit cost of this was based on thatThe unit cost of this was based on that

for a clinical psychologist from a nationalfor a clinical psychologist from a national

source. However, it may be that the actualsource. However, it may be that the actual

unit cost could be different if other pro-unit cost could be different if other pro-

fessionals (such as mental health nurses)fessionals (such as mental health nurses)

provide the service, if more or less experi-provide the service, if more or less experi-

enced psychologists deliver the service orenced psychologists deliver the service or

if differences in supervision and generalif differences in supervision and general

infrastructure affect the costs. To take intoinfrastructure affect the costs. To take into

account such possibilities we recalculatedaccount such possibilities we recalculated

the total costs by assuming that (i) the unitthe total costs by assuming that (i) the unit

cost of cognitive therapy was 50% lowercost of cognitive therapy was 50% lower

(£30.50 per hour) and (ii) 50% higher(£30.50 per hour) and (ii) 50% higher

(£91.50 per hour).(£91.50 per hour).

RESULTSRESULTS

Summary of primary clinicalSummary of primary clinicaloutcomeoutcome

Those receiving cognitive therapy spentThose receiving cognitive therapy spent

62.3 fewer days with bipolar episodes than62.3 fewer days with bipolar episodes than

the comparison group (26.6 days,the comparison group (26.6 days,

s.d.s.d.¼46.046.0 v.v. 88.4 days, s.d.88.4 days, s.d.¼108.9; 95%108.9; 95%

CI of difference 27.31–96.99) in the firstCI of difference 27.31–96.99) in the first

365 days. Over the whole 30 months, they365 days. Over the whole 30 months, they

also spent 110 fewer days (95.3 days,also spent 110 fewer days (95.3 days,

s.d.s.d.¼152.1152.1 v.v. 201.0 days, s.d.201.0 days, s.d.¼95.3; 95%95.3; 95%

CI of difference 32–189 days) with bipolarCI of difference 32–189 days) with bipolar

episodes out of about 900 days in total. Theepisodes out of about 900 days in total. The

differences were significant after control-differences were significant after control-

ling for the number of previous bipolarling for the number of previous bipolar

episodes and medication compliance. Theepisodes and medication compliance. The

actuarial cumulative relapse rates foractuarial cumulative relapse rates for

bipolar episodes in the cognitive therapybipolar episodes in the cognitive therapy

and comparison groups were 64% (30/47)and comparison groups were 64% (30/47)

and 84% (43/51), respectively. Afterand 84% (43/51), respectively. After

controlling for the previous number ofcontrolling for the previous number of

episodes and medication compliance duringepisodes and medication compliance during

the whole 30 months, the differences werethe whole 30 months, the differences were

significant for bipolar episodes (hazardsignificant for bipolar episodes (hazard

ratioratio¼0.50, 95% CI 0.29–0.85,0.50, 95% CI 0.29–0.85, PP¼0.012).0.012).

A total of 47% (21/45) in the compari-A total of 47% (21/45) in the compari-

son groupson group v.v. 38% (18/47) of those receiving38% (18/47) of those receiving

cognitive therapy were admitted to hospitalcognitive therapy were admitted to hospital

for bipolar episodes. The difference was notfor bipolar episodes. The difference was not

significant. There was also a non-significantsignificant. There was also a non-significant

trend for patients receiving cognitivetrend for patients receiving cognitive

therapy to have fewer days in hospitaltherapy to have fewer days in hospital

over the whole 30-month period (meanover the whole 30-month period (mean

31.3 days, s.d.31.3 days, s.d.¼84.684.6 v.v. mean 35.7 days,mean 35.7 days,

s.d.s.d.¼69.8).69.8).

Service utilisationService utilisation

Table 2 summarises service use for the cog-Table 2 summarises service use for the cog-

nitive therapy and comparison groups innitive therapy and comparison groups in

the 3 months prior to baseline and eachthe 3 months prior to baseline and each

follow-up assessment. During the 3 monthsfollow-up assessment. During the 3 months

prior to randomisation (baseline), 14%prior to randomisation (baseline), 14%

(7/51) of the cognitive therapy group(7/51) of the cognitive therapy group

received psychiatric in-patient treatmentreceived psychiatric in-patient treatment

and 16% (8/52) of the comparison group;and 16% (8/52) of the comparison group;

92% of individuals also received some92% of individuals also received some

community services. Virtually all partici-community services. Virtually all partici-

pants were prescribed medication.pants were prescribed medication.

There were few differences in the pro-There were few differences in the pro-

portions of those using these groups ofportions of those using these groups of

services in the follow-up periods. However,services in the follow-up periods. However,

in the 3 months preceding the 6- and 9-in the 3 months preceding the 6- and 9-

month follow-up assessments, around twicemonth follow-up assessments, around twice

as many from the comparison group wereas many from the comparison group were

admitted compared with those receivingadmitted compared with those receiving

cognitive therapy. However, this was re-cognitive therapy. However, this was re-

versed in the period before the 18-monthversed in the period before the 18-month

follow-up. Community services continuedfollow-up. Community services continued

to be used by the majority of participantsto be used by the majority of participants

throughout the study. Medication also con-throughout the study. Medication also con-

tinued to be used by many, but the numberstinued to be used by many, but the numbers

did decline slightly.did decline slightly.

Service costService cost

Table 3 shows the average costs at baselineTable 3 shows the average costs at baseline

and for each follow-up period. There wasand for each follow-up period. There was

much variation in resource use throughoutmuch variation in resource use throughout

the study period. Significance tests werethe study period. Significance tests were

only carried out on the difference betweenonly carried out on the difference between

the groups in total costs. For most periods,the groups in total costs. For most periods,

there were no statistically significant differ-there were no statistically significant differ-

ences; the exceptions were for the period upences; the exceptions were for the period up

to the 9-month assessment, when the cog-to the 9-month assessment, when the cog-

nitive therapy group was significantly lessnitive therapy group was significantly less

costly, and at the 18-month assessment,costly, and at the 18-month assessment,

when the cognitive therapy group used sig-when the cognitive therapy group used sig-

nificantly more resources. These findingsnificantly more resources. These findings

were consistent with the clinical outcomeswere consistent with the clinical outcomes

(Lam(Lam et alet al, 2003, 2005)., 2003, 2005).

The costs for the 12- and 30-monthThe costs for the 12- and 30-month

periods following randomisation areperiods following randomisation are

shown in Table 4. For the first 12 monthsshown in Table 4. For the first 12 months

of the study and the whole of the 30of the study and the whole of the 30

5 0 25 0 2

Table 2Table 2 Use of services during the 3 months prior to baseline and follow-up assessmentsUse of services during the 3 months prior to baseline and follow-up assessments

Month 0Month 0 Month 3Month 3 Month 6Month 6 Month 9Month 9 Month12Month 12 Month 15Month 15 Month 18Month 18 Month 21Month 21 Month 24Month 24 Month 27Month 27 Month 30Month 30

CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU CTCT TAUTAU

PsychiatricPsychiatric

in-patient,in-patient,

nn (%)(%)

77

(14)(14)

88

(16)(16)

33

(6)(6)

55

(10)(10)

55

(10)(10)

1010

(20)(20)

33

(7)(7)

77

(15)(15)

77

(15)(15)

66

(13)(13)

66

(13)(13)

44

(9)(9)

66

(13)(13)

22

(4)(4)

55

(11)(11)

44

(9)(9)

44

(9)(9)

33

(7)(7)

11

(2)(2)

44

(9)(9)

66

(13)(13)

55

(11)(11)

GeneralGeneral

in-patient,in-patient,

nn (%)(%)

00

(0)(0)

11

(2)(2)

00

(0)(0)

22

(4)(4)

11

(2)(2)

00

(0)(0)

00

(0)(0)

00

(0)(0)

11

(2)(2)

11

(2)(2)

11

(2)(2)

11

(2)(2)

11

(2)(2)

11

(2)(2)

22

(4)(4)

00

(0)(0)

22

(4)(4)

00

(0)(0)

22

(4)(4)

00

(0)(0)

11

(2)(2)

22

(4)(4)

CommunityCommunity

services,services,

nn (%)(%)

4747

(92)(92)

4646

(92)(92)

3838

(84)(84)

3434

(85)(85)

3737

(84)(84)

3535

(88)(88)

3131

(89)(89)

2121

(88)(88)

3737

(90)(90)

3131

(89)(89)

3030

(91)(91)

2323

(89)(89)

3030

(83)(83)

2727

(84)(84)

2525

(89)(89)

1616

(76)(76)

2828

(82)(82)

1818

(78)(78)

1818

(75)(75)

1010

(77)(77)

2121

(72)(72)

1919

(79)(79)

MedicationMedication11,,

nn (%)(%)

5050

(98)(98)

5252

(100)(100)

^̂ ^̂ 4444

(86)(86)

3838

(73)(73)

^̂ ^̂ 4141

(80)(80)

3131

(60)(60)

^̂ ^̂ 3838

(75)(75)

3030

(58)(58)

^̂ ^̂ 3232

(63)(63)

2222

(42)(42)

^̂ ^̂ 2828

(55)(55)

2424

(46)(46)

CT, cognitive therapy;TAU, treatment as usual.CT, cognitive therapy;TAU, treatment as usual.1. The numbers vary because of missing data.1. The numbers vary because of missing data.

COST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR BIPOLAR DISORDERCOST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR B IPOLAR DISORDER

5 0 35 0 3

Table3

Table3

Servicecosts(»)for

cognitivetherapyandstandard

treatm

entg

roup

sduring

the3mon

thspriorto

baselineandfollow-upassessments

Servicecosts(»)for

cognitivetherapyandstandard

treatm

entg

roup

sduring

the3mon

thspriorto

baselineandfollow-upassessments

Mon

th0

Mon

th0

Mon

th3

Mon

th3

Mon

th6

Mon

th6

Mon

th9

Mon

th9

Mon

th12

Mon

th12

Mon

th15

Mon

th15

Mon

th18

Mon

th18

Mon

th21

Mon

th21

Mon

th24

Mon

th24

Mon

th27

Mon

th27

Mon

th30

Mon

th30

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

CT

CT

TAU

TAU

Psychiatricin-patient,

Psychiatricin-patient,

mean(s.d.)

mean(s.d.)

824

824

(226

0)(226

0)

754

754

(2276)

(2276)

528

528

(188

0)(188

0)

262

262

(1235)

(1235)

968

968

(2621)

(2621)

660

660

(209

9)(209

9)

801

801

(2430)

(2430)

323

323

(1531)

(1531)

582

582

(1922)

(1922)

617

617

(1754)

(1754)

438

438

(158

6)(158

6)

548

548

(1744)

(1744)

404

404

(1941)

(1941)

682

682

(1958)

(1958)

498

498

(1783)

(1783)

542

542

(1785)

(1785)

438

438

(2072)

(2072)

754

754

(2746)

(2746)

623

623

(2403)

(240

3)

9090 (601

)(601

)

729

729

(2745)

(2745)

742

742

(2579)

(2579)

((nn))

(50)

(50)

(51)

(51)

(51)

(51)

(49)

(49)

(51)

(51)

(48)

(48)

(46)

(46)

(45)

(45)

(48)

(48)

(46)

(46)

(46)

(46)

(47)

(47)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(45)

(45)

(46)

(46)

(45)

(45)

Generalin-patient,

Generalin-patient,

mean(s.d.)

mean(s.d.)

3131 (221)

(221)

00 (0)

(0)

5151 (255)

(255)

00 (0)

(0)

00 (0)

(0)

2020 (137)

(137)

00 (0)

(0)

00 (0)

(0)

2020 (137)

(137)

3131 (210)

(210)

1515 (105)

(105)

2525 (173)

(173)

2121 (140

)(140

)

3131 (210)

(210)

00 (0)

(0)

3636 (173)

(173)

00 (0)

(0)

3131 (155)

(155)

00 (0)

(0)

3737 (175)

(175)

4646 (232)

(232)

2121 (141)

(141)

((nn))

(50)

(50)

(51)

(51)

(51)

(51)

(49)

(49)

(51)

(51)

(48)

(48)

(46)

(46)

(45)

(45)

(48)

(48)

(46)

(46)

(46)

(46)

(47)

(47)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(45)

(45)

(46)

(46)

(45)

(45)

Com

mun

ityservices,

Com

mun

ityservices,

mean(s.d.)

mean(s.d.)

575

575

(863)

(863)

360

360

(407)

(407)

491

491

(868

)(868

)

355

355

(418)

(418)

348

348

(379)

(379)

296

296

(451)

(451)

393

393

(475)

(475)

229

229

(280

)(280

)

465

465

(716)

(716)

326

326

(456)

(456)

293

293

(347)

(347)

333

333

(296)

(296)

312

312

(424)

(424)

340

340

(339)

(339)

385

385

(487

)(487

)

349

349

(365)

(365)

357

357

(462)

(462)

191

191

(227)

(227)

430

430

(564

)(564

)

229

229

(240

)(240

)

343

343

(371)

(371)

241

241

(317)

(317)

((nn))

(50)

(50)

(51)

(51)

(40)

(40)

(45)

(45)

(40)

(40)

(44)

(44)

(24)

(24)

(35)

(35)

(35)

(35)

(41)

(41)

(26)

(26)

(33)

(33)

(32)

(32)

(36)

(36)

(21)

(21)

(28)

(28)

(23)

(23)

(34)

(34)

(13)

(13)

(24)

(24)

(24)

(24)

(29)

(29)

Medication,

Medication,

mean(s.d.)

mean(s.d.)

106

106

(175)

(175)

104

104

(139)

(139)

^̂^̂

8383 (139)

(139)

102

102

(179)

(179)

^̂^̂

7575 (139)

(139)

8383 (128

)(128

)

^̂^̂

6262 (110)

(110)

7676 (140

)(140

)

^̂^̂

5656 (162)

(162)

5858 (95)

(95)

^̂^̂

6060 (148

)(148

)

7070 (146

)(146

)

((nn))

(52)

(52)

(51)

(51)

(52)

(52)

(51)

(51)

(52)

(52)

(51)

(51)

(52)

(52)

(51)

(51)

(52)

(52)

(51)

(51)

(52)

(52)

(51)

(51)

Total,

Total,

mean(s.d.)

mean(s.d.)

1539

1539

(2610)

(2610)

1218

1218

(2302)

(2302)

978

978

(1723)

(1723)

676

676

(1267)

(1267)

1637

1637

(2961)

(2961)

914

914

(1574)

(1574)

2015

2015

(348

5)(348

5)

324

324

(340

)(340

)

1216

1216

(2454)

(2454)

1135

1135

(206

8)(206

8)

705

705

(124

0)(1240)

890

890

(199

0)(199

0)

442

442

(495)

(495)

888

888

(1613)

(1613)

1250

1250

(2545)

(2545)

956

956

(1894)

(1894)

791

791

(1224)

(1224)

816

816

(2251)

(2251)

519

519

(608

)(608

)

359

359

(403)

(403)

561

561

(654)

(654)

987

987

(258

9)(258

9)

((nn))

(50)

(50)

(51)

(51)

(51)

(51)

(49)

(49)

(51)

(51)

(48)

(48)

(46)

(46)

(45)

(45)

(48)

(48)

(46)

(46)

(46)

(46)

(47)

(47)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(46)

(45)

(45)

(46)

(46)

(45)

(45)

90%CIo

fdifference

90%CIo

fdifference

intotalcost

intotalcost

771124

to48

91124

to48

977892to

209

892to

209

771677

to43

1677

to43

773062

to3062

to77687

687

77979to

721

979to

721

7740

8to

945

408to

945

56to

975

56to

975

771524

to669

1524

to669

77619to

898

619to

898

7748

2to

120

482to

120

77187to

1441

187to

1441

CT,cognitivetherapy;TA

U,treatmentas

usual.

CT,cognitivetherapy;TA

U,treatmentas

usual.

LAM ET ALLAM ET AL

months, the group receiving cognitivemonths, the group receiving cognitive

therapy had lower service costs. However,therapy had lower service costs. However,

the differences were not statisticallythe differences were not statistically

significant.significant.

Cost-effectiveness analysisCost-effectiveness analysis

Figure 1 shows cost-effectiveness accept-Figure 1 shows cost-effectiveness accept-

ability curves, which show the probabilityability curves, which show the probability

that cognitive therapy is more cost-effectivethat cognitive therapy is more cost-effective

than standard care for a range of differentthan standard care for a range of different

values placed on a day free of bipolar symp-values placed on a day free of bipolar symp-

toms. Even with a zero value, the probabil-toms. Even with a zero value, the probabil-

ity of cognitive therapy being cost-effectiveity of cognitive therapy being cost-effective

is in excess of 0.85 for the first 12 monthsis in excess of 0.85 for the first 12 months

and 0.80 for the whole study period of 30and 0.80 for the whole study period of 30

months. However, at a value of £10 andmonths. However, at a value of £10 and

above per day free from bipolar episode,above per day free from bipolar episode,

the probability of cognitive therapy beingthe probability of cognitive therapy being

cost effective is in excess of 0.90 for the firstcost effective is in excess of 0.90 for the first

12 months and 0.85 for the whole study12 months and 0.85 for the whole study

period of 30 months.period of 30 months.

Sensitivity analysisSensitivity analysis

If the cost of the intervention is reduced byIf the cost of the intervention is reduced by

50%, the total mean cost for the cognitive50%, the total mean cost for the cognitive

therapy group falls to £3952 over the 12-therapy group falls to £3952 over the 12-

month period following randomisationmonth period following randomisation

but if the cost is increased by 50% the meanbut if the cost is increased by 50% the mean

rises to £4815. These changes represent arises to £4815. These changes represent a

10% shift in the average cost. Over the10% shift in the average cost. Over the

30-month follow-up period the total cost30-month follow-up period the total cost

for the cognitive therapy group falls tofor the cognitive therapy group falls to

£9925 with the lower unit cost and in-£9925 with the lower unit cost and in-

creases to £10 729 with the upper bound.creases to £10 729 with the upper bound.

This represents a smaller proportional shiftThis represents a smaller proportional shift

(4%).(4%).

Figure 2 shows the impact of these newFigure 2 shows the impact of these new

costs on the cost-effectiveness acceptabilitycosts on the cost-effectiveness acceptability

curves. For clarity only the most extremecurves. For clarity only the most extreme

results are shown. The best result for cogni-results are shown. The best result for cogni-

tive therapy is where the therapy cost istive therapy is where the therapy cost is

lower by 50% and the outcomes and costslower by 50% and the outcomes and costs

5 0 45 0 4

Table 4Table 4 Mean (s.d.) service costs (») for12 months and 30 months following randomisation (1999/2000)Mean (s.d.) service costs (») for12 months and 30 months following randomisation (1999/2000)

TimeTime 0^12 months0^12 months 0^30months0^30months

Cognitive therapyCognitive therapy

((nn¼45)45)

Standard treatmentStandard treatment

((nn¼46)46)

Cognitive therapyCognitive therapy

((nn¼43)43)

Standard treatmentStandard treatment

((nn¼40)40)

InterventionIntervention 863 (310)863 (310) NANA 854 (314)854 (314) NANA

Psychiatric in-patientPsychiatric in-patient 1786 (4918)1786 (4918) 2968 (5937)2968 (5937) 5151 (12476)5151 (12 476) 5940 (10739)5940 (10 739)

General in-patientGeneral in-patient 53 (353)53 (353) 77 (298)77 (298) 248 (1273)248 (1273) 183 (602)183 (602)

Community servicesCommunity services 1263 (1487)1263 (1487) 1974 (2218)1974 (2218) 3178 (3142)3178 (3142) 4921 (6169)4921 (6169)

MedicationMedication 419 (600)419 (600) 338 (496)338 (496) 921 (1017)921 (1017) 680 (1015)680 (1015)

TotalTotal 4383 (5264)4383 (5264) 5356 (6599)5356 (6599) 10352 (13 464)10 352 (13464) 11724 (12 061)11724 (12 061)

Fig. 1Fig. 1 Cost-effectiveness acceptability curves; CBT, cognitive^behavioural therapy;TAU, treatment as usual.Cost-effectiveness acceptability curves; CBT, cognitive^behavioural therapy;TAU, treatment as usual.

Fig. 2Fig. 2 Sensitivity analysis; CBT, cognitive^behavioural therapy;TAU, treatment as usual.Sensitivity analysis; CBT, cognitive^behavioural therapy;TAU, treatment as usual.

COST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR BIPOLAR DISORDERCOST- EFFECTIVENES S OF COGNIT IVE THERAPY FOR B IPOLAR DISORDER

are measured only for the first 12 months.are measured only for the first 12 months.

In this situation there is a 93% chance thatIn this situation there is a 93% chance that

cognitive therapy is more cost-effectivecognitive therapy is more cost-effective

than standard NHS care even if a zero valuethan standard NHS care even if a zero value

is placed on a bipolar-free day. The worstis placed on a bipolar-free day. The worst

case for cognitive therapy is with therapycase for cognitive therapy is with therapy

costs raised by 50% and outcomes andcosts raised by 50% and outcomes and

costs measured over the longer period ofcosts measured over the longer period of

30 months. Even here there is a 75.2%30 months. Even here there is a 75.2%

chance of cognitive therapy being the morechance of cognitive therapy being the more

cost-effective option with a zero valuecost-effective option with a zero value

placed on a bipolar-free day, and the figureplaced on a bipolar-free day, and the figure

rapidly rises to more than 90% as the valuerapidly rises to more than 90% as the value

of a bipolar-free day increases.of a bipolar-free day increases.

DISCUSSIONDISCUSSION

Clinical outcomeClinical outcome

In this study, the comparison group re-In this study, the comparison group re-

ceived standard treatment, which consistedceived standard treatment, which consisted

of mood stabilisers and psychiatric follow-of mood stabilisers and psychiatric follow-

up, and the other group received cognitiveup, and the other group received cognitive

therapy in addition to standard treatment.therapy in addition to standard treatment.

Combination of cognitive therapy andCombination of cognitive therapy and

mood stabilisers produced better clinicalmood stabilisers produced better clinical

outcomes, particularly in the first 12outcomes, particularly in the first 12

months. However, as therapy became moremonths. However, as therapy became more

distant, the effect of therapy was less robustdistant, the effect of therapy was less robust

(Lam(Lam et alet al, 2005)., 2005).

Cost effectivenessCost effectiveness

The group receiving cognitive therapy in-The group receiving cognitive therapy in-

curred £1000 less service costs for the firstcurred £1000 less service costs for the first

12 months and £1300 less over the whole12 months and £1300 less over the whole

30 months. However, the difference in total30 months. However, the difference in total

service cost between the cognitive therapyservice cost between the cognitive therapy

and comparison group was not significant.and comparison group was not significant.

As expected, the bulk of the service costAs expected, the bulk of the service cost

was for psychiatric in-patient care. Psychi-was for psychiatric in-patient care. Psychi-

atric in-patient care was very expensive,atric in-patient care was very expensive,

leading to highly skewed data. The analysisleading to highly skewed data. The analysis

showed that cognitive therapy was highlyshowed that cognitive therapy was highly

cost-effective compared with standard carecost-effective compared with standard care

alone. Even if no value is placed on aalone. Even if no value is placed on a

bipolar-free day, the probability of cogni-bipolar-free day, the probability of cogni-

tive therapy being more cost-effective thantive therapy being more cost-effective than

standard treatment was more than 80%standard treatment was more than 80%

during the first year and the whole studyduring the first year and the whole study

period of 30 months. The probability ofperiod of 30 months. The probability of

cognitive therapy being cost-effective wascognitive therapy being cost-effective was

slightly lower if the whole 30-month periodslightly lower if the whole 30-month period

was considered. However, if society iswas considered. However, if society is

willing to attribute a value of even £5 towilling to attribute a value of even £5 to

one bipolar-free day, the probability ofone bipolar-free day, the probability of

cognitive therapy being cost-effective in-cognitive therapy being cost-effective in-

creases to beyond 85% for both timecreases to beyond 85% for both time

periods. There are very few cost-periods. There are very few cost-

effectiveness analyses of interventions foreffectiveness analyses of interventions for

bipolar disorder; the few there are usebipolar disorder; the few there are use

outcome measures and methods that differoutcome measures and methods that differ

from the analyses presented here.from the analyses presented here.

ComparisonsComparisons are not therefore possible.are not therefore possible.

Finally, the probability of cognitive therapyFinally, the probability of cognitive therapy

being more cost effective than standardbeing more cost effective than standard

psychiatric care is robust in the sensitivitypsychiatric care is robust in the sensitivity

analysis. Even if the cost of cognitive ther-analysis. Even if the cost of cognitive ther-

apy is increased by 50% (from £863 toapy is increased by 50% (from £863 to

£1295), the probability of cognitive therapy£1295), the probability of cognitive therapy

being cost-effective is still high.being cost-effective is still high.

Clinical implicationsClinical implications

Our results support the addition of cogni-Our results support the addition of cogni-

tive therapy for relapse prevention in bi-tive therapy for relapse prevention in bi-

polar disorder, particularly for those whopolar disorder, particularly for those who

are vulnerable to relapses despite the pre-are vulnerable to relapses despite the pre-

scription of mood stabilisers. Clinically,scription of mood stabilisers. Clinically,

the combined treatment was significantlythe combined treatment was significantly

more effective. The cost of adding cognitivemore effective. The cost of adding cognitive

therapy to the routine treatment with moodtherapy to the routine treatment with mood

stabilisers and psychiatric follow-up wasstabilisers and psychiatric follow-up was

offset by fewer costs for other services.offset by fewer costs for other services.

LimitationsLimitations

There were a number of limitations in thisThere were a number of limitations in this

study. First, data on service use werestudy. First, data on service use were

collected using a self-report questionnaire.collected using a self-report questionnaire.

Although this allows a far greater breadthAlthough this allows a far greater breadth

of service use to be measured, it may notof service use to be measured, it may not

be as accurate. However, the recall periodbe as accurate. However, the recall period

was relatively short (3 months) and datawas relatively short (3 months) and data

on therapy and in-patient episodes wereon therapy and in-patient episodes were

collected from other sources. We did notcollected from other sources. We did not

address the reliability of the service useaddress the reliability of the service use

measures. However, other studies havemeasures. However, other studies have

found self-report to be an appropriatefound self-report to be an appropriate

way of measuring resource use (Calsynway of measuring resource use (Calsyn etet

alal, 1993; Goldberg, 1993; Goldberg et alet al, 2002). Second,, 2002). Second,

although a broad range of services was in-although a broad range of services was in-

cluded, there were others that were notcluded, there were others that were not

measured, such as informal care frommeasured, such as informal care from

family and friends; also participant timefamily and friends; also participant time

was not costed. Third, we did not havewas not costed. Third, we did not have

the health economy data prior to randomis-the health economy data prior to randomis-

ation. However, there was no significantation. However, there was no significant

difference between the two groups in termsdifference between the two groups in terms

of previous hospitalisation, which incurredof previous hospitalisation, which incurred

most of the health costs. Finally, the studymost of the health costs. Finally, the study

showed that cognitive therapy had a highshowed that cognitive therapy had a high

probability of being cost-effective but ofprobability of being cost-effective but of

course it is unknown what value societycourse it is unknown what value society

places on this. However, the cost-effective-places on this. However, the cost-effective-

ness acceptability curves do show the rangeness acceptability curves do show the range

5 0 55 0 5

CLINICAL IMPLICATIONSCLINICAL IMPLICATIONS

&& Cognitive therapy is a useful addition to standard National Health Service (NHS)Cognitive therapy is a useful addition to standard National Health Service (NHS)treatment of patients with bipolar disorder, at no extra overall cost.treatment of patients with bipolar disorder, at no extra overall cost.

&& The costof a course of cognitive therapy is offsetby lower costs for other services.The costof a course of cognitive therapy is offsetby lowercosts for other services.

&& Even if the cost of cognitive therapy is increased by 50% (from »863 to »1295), theEven if the cost of cognitive therapy is increased by 50% (from »863 to »1295), theprobability of cognitive therapy being cost-effective is still high.probability of cognitive therapy being cost-effective is still high.

LIMITATIONSLIMITATIONS

&& Therewas no protocol for standard NHS treatment.Decisions on drugs andTherewas no protocol for standard NHS treatment.Decisions on drugs andfrequency of psychiatric follow-upwere left to the clinicians responsible for day-to-frequency of psychiatric follow-upwere left to the clinicians responsible for day-to-day care.day care.

&& Service usewas based on self-report and hospital records, both of which haveService usewas based on self-report and hospital records, both of which haveproblems.problems.

&& There is no general agreement on the value of a bipolar-free day.There is no general agreement on the value of a bipolar-free day.

DOMINICH.LAM, PhD, Psychology Department, PAULMcCRONE, PhD,Centre for the Economics of MentalDOMINIC H.LAM, PhD, Psychology Department, PAULMcCRONE, PhD,Centre for the Economics of MentalHealth,Health Services Research Department,KIMWRIGHT,BA, Psychology Department,NATALIEKERR,Health,Health Services Research Department,KIMWRIGHT,BA, Psychology Department,NATALIEKERR,MSc, Psychological Medicine, Institute of Psychiatry, LondonMSc, Psychological Medicine, Institute of Psychiatry, London

Correspondence:Dr Dominic H.Lam,Psychology Department, Institute of Psychiatry,DeCrespignyCorrespondence:Dr Dominic H.Lam,Psychology Department, Institute of Psychiatry,DeCrespignyPark,London SE5 8AF.E-mail:D.LamPark,London SE5 8AF.E-mail:D.Lam@@iop.kcl.ac.ukiop.kcl.ac.uk

(First received 18 May 2004, final revision 9 November 2004, accepted 17 November 2004)(First received 18 May 2004, final revision 9 November 2004, accepted 17 November 2004)

LAM ET ALLAM ET AL

of values beyond which further increasesof values beyond which further increases

have a negligible impact on the probability.have a negligible impact on the probability.

Although valuing a bipolar-free day mayAlthough valuing a bipolar-free day may

have more practical meaning than valuinghave more practical meaning than valuing

a point change on a particular outcomea point change on a particular outcome

scale, it is still a rather nebulous concept.scale, it is still a rather nebulous concept.

Further research should be conducted to de-Further research should be conducted to de-

termine the views of users, family memberstermine the views of users, family members

and clinicians as to whether bipolar-freeand clinicians as to whether bipolar-free

days are meaningful as an outcome measuredays are meaningful as an outcome measure

and, if so, exactly how they might beand, if so, exactly how they might be

valued.valued.

ACKNOWLEDGEMENTACKNOWLEDGEMENT

This study was funded from the South London andThis study was funded from the South London andMaudsley Hospital’s research and development fund.Maudsley Hospital’s research and development fund.

REFERENCESREFERENCES

American Psychiatric Association (1994)American Psychiatric Association (1994) DiagnosticDiagnosticand Statistical Manual of Mental Disordersand Statistical Manual of Mental Disorders (4th edn)(4th edn)(DSM^IV).Washington,DC: APA.(DSM^IV).Washington,DC: APA.

Bauer, M. S., Crits-Christoph, P., Ball,W. A.,Bauer, M. S., Crits-Christoph, P., Ball,W. A., et alet al(1991)(1991) Independent assessment of manic and depressiveIndependent assessment of manic and depressive

symptoms by self-rating.symptoms by self-rating. Archives of General PsychiatryArchives of General Psychiatry,,4848, 807^812., 807^812.

Beck, A.T.,Ward,C.H., Mendelson, M.,Beck, A.T.,Ward,C.H., Mendelson, M., et alet al (1961)(1961)An inventory for measuring depression.An inventory for measuring depression. Archives ofArchives ofGeneral PsychiatryGeneral Psychiatry,, 44, 53^63., 53^63.

Beecham, J. & Knapp, M.Beecham, J. & Knapp, M. (2001)(2001) Costing psychiatricCosting psychiatricinterventions. Ininterventions. In Measuring Mental Health Needs,Measuring Mental Health Needs, 2nd2ndednedn (ed.G.Thornicroft), pp. 200^224. London: Gaskell.(ed.G.Thornicroft), pp. 200^224. London: Gaskell.

Briggs, A.H.Briggs, A.H. (2001)(2001) ABayesian approach to stochasticABayesian approach to stochasticcost-effectiveness analysis.cost-effectiveness analysis. International Journal ofInternational Journal ofTechnology Assessment in Health CareTechnology Assessment in Health Care,, 1717, 69^82., 69^82.

British Medical Association & RoyalBritish Medical Association & RoyalPharmaceutical Society of Great BritainPharmaceutical Society of Great Britain (2001)(2001)British National FormularyBritish National Formulary (September). London &(September). London &Wallingford: BMJ Books & Pharmaceutical Press.Wallingford:BMJ Books & Pharmaceutical Press.

Calsyn, R. J., Allen,G., Morse,G. A.,Calsyn, R. J., Allen,G., Morse,G. A., et alet al (1993)(1993) CanCanyou trust self-report data provided by homelessyou trust self-report data provided by homelessmentally ill individuals?mentally ill individuals? Evaluation ReviewEvaluation Review,, 1717, 353^366., 353^366.

First, M. B., Gibbon, M., Spitzer, R. L.,First, M. B., Gibbon, M., Spitzer, R. L., et alet al (1996)(1996)User Guide for the Structured Clinical Interview forUser Guide for the Structured Clinical Interview forDSM^IV Axis I Disorders.DSM^IV Axis I Disorders.Washington,DC: AmericanWashington,DC: AmericanPsychiatric Association.Psychiatric Association.

Goldberg, R.W., Seybolt, D. C. & Lehman, A.Goldberg, R.W., Seybolt, D.C. & Lehman, A. (2002)(2002)Reliable self-report of health service use by individualsReliable self-report of health service use by individualswith serious mental illness.with serious mental illness. Psychiatric ServicesPsychiatric Services,, 5353,,879^881.879^881.

Gupta, R. D. & Guest, J. F.Gupta, R. D. & Guest, J. F. (2002)(2002) Annual cost ofAnnual cost ofbipolar disorder to UK society.bipolar disorder to UK society. British Journal ofBritish Journal ofPsychiatryPsychiatry,, 180180, 227^233., 227^233.

Lam, D.H.,Watkins, E.,Hayward, P.,Lam, D.H.,Watkins, E.,Hayward, P., et alet al (2003)(2003) AArandomized controlled study of cognitive therapy ofrandomized controlled study of cognitive therapy ofrelapse prevention for bipolar affective disorder ^relapse prevention for bipolar affective disorder ^outcome of the first year.outcome of the first year. Archives of General PsychiatryArchives of General Psychiatry,,6060, 145^152., 145^152.

Lam, D.H.,Hayward, P.,Watkins, E.,Lam, D.H.,Hayward, P.,Watkins, E., et alet al (2005)(2005)Outcome of a two-year follow-up of a cognitive therapyOutcome of a two-year follow-up of a cognitive therapyof relapse prevention in bipolar disorder.of relapse prevention in bipolar disorder. AmericanAmericanJournal of PsychiatryJournal of Psychiatry,, 162162, 324^329., 324^329.

Mooney,C. & Duvall, R.Mooney,C. & Duvall, R. (1993)(1993) Bookstrapping: A Non-Bookstrapping: A Non-Parametric Approach to Statistical InferenceParametric Approach to Statistical Inference. London: Sage.. London: Sage.

Netten, A. & Curtis, L.Netten, A. & Curtis, L. (2000)(2000) Unit Costs of HealthUnit Costs of Healthand Social Careand Social Care.Canterbury: Personal Social Services &.Canterbury: Personal Social Services &Research Unit.Research Unit.

Office of Population Censuses and SurveysOffice of Population Censuses and Surveys (1991)(1991)Standard Occupational Classification, vol. 3.Standard Occupational Classification, vol. 3. London:London:Stationery Office.Stationery Office.

Patel, A.Patel, A. (2003)(2003) The cost of mood disorders.The cost of mood disorders. PsychiatryPsychiatry,,55, 28^31., 28^31.

Rice, D. P. & Miller, L. S.Rice, D. P. & Miller, L. S. (1995)(1995) The economic burdenThe economic burdenof affective disorders.of affective disorders. British Journal of PsychiatryBritish Journal of Psychiatry,, 166166(suppl. 27), 34^42.(suppl. 27), 34^42.

5 0 65 0 6