psychological interventions for adults with asthma: a systematic review
TRANSCRIPT
ARTICLE IN PRESS
Respiratory Medicine (2007) 101, 1–14
0954-6111/$ - sdoi:10.1016/j.r
�CorrespondiE-mail addr
REVIEW
Psychological interventions for adults with asthma:A systematic review
Janelle Yorke�, Sharon L. Fleming, Caroline Shuldham
Royal Brompton and Harefield NHS Trust, Sydney Street, London SW3 6NP, UK
Received 20 January 2006; accepted 4 April 2006
KEYWORDSAsthma;Psychological;Systematic review
ee front matter & 2006med.2006.04.003
ng author. Tel.: +44 207ess: [email protected]
SummaryAim: The purpose of this study was to conduct a systematic review of randomizedcontrolled trials where the efficacy of psychological interventions in modifyinghealth and behavioural outcomes for adults with asthma was investigated.Method: A review of randomized controlled trials was designed. The literaturesearch was conducted until May 2005.Results: Fourteen studies, involving 617 adults, were included in the review. Theuse of ‘as needed’ medications was reduced by relaxation therapy (OR 4.47, CI1.22–16.44), quality of life, measured using the Asthma Quality of Life Ques-tionnaire, showed a positive effect following cognitive behavioural therapy (WMD0.71, CI 0.23–1.19), and peak expiratory flow outcome data indicated a significantdifference in favour of bio-feedback therapy (SMD 0.66, CI 0.09–1.23).Conclusions: Some promising results did emerge from meta-analyses performed.However, due to heterogeneity and the low quality of included studies, this reviewwas unable to draw firm conclusions for the role of psychological interventions inasthma. We recommend that larger and well-conducted randomized trials use validoutcome measures to evaluate the effectiveness of psychological interventions foradults with asthma.& 2006 Elsevier Ltd. All rights reserved.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Elsevier Ltd. All rights reserved.
3 518 846; fax: +44 2073 518 853.k (J. Yorke).
ARTICLE IN PRESS
J. Yorke et al.2
Types of studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Types of participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Types of interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Outcome measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Methodological quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Methodological quality of included studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Study participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Method of randomisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Interventions used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Primary outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Health service utilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Secondary outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Asthma symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Lung function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Medication use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Absenteeism from school/work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Psychological health status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Patient satisfaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Withdrawals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Reviewers conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Introduction
Asthma has a psychological component, includingemotion,1 so the treatment of asthma in adultsincreasingly needs to focus on the whole person,taking account of psychological as well as physiolo-gical elements. It is suggested that psychologicalinterventions may be appropriate for adults withasthma. Several psychological interventions may beemployed to ameliorate health problems associatedwith asthma. These include behavioural therapies,cognitive therapies, cognitive behavioural therapy,relaxation techniques, psychodynamic psychothera-pies, and counselling, in both individual and groupformats. These have been identified as the proce-dure by which a therapist systematically attempts toinfluence a patient by psychological means so thatthe patient’s symptoms decrease or there is apositive change in behaviour.2
This means that evidence to support decisionsabout the type, format and frequency of psycholo-gical techniques is needed. The aim of thesestrategies is to help reduce panic or fear, improvebreathing and respiratory function and impactpositively on general health and quality of life.Literature is growing on the relationship betweenpsychosocial factors and asthma,3,4 and review
methodologies are being used to assess the impactof a range of psychosocial interventions in asthma.For example, reviews have been undertaken onself-management education for asthmatic adults,5
psycho-educational interventions for adults andchildren,6 family therapy for asthma in children7
and psychological interventions for children withasthma.8
These reviews do not answer questions specifi-cally about psychological interventions for adultswith asthma and therefore a systematic review ofthe effectiveness of such interventions is alsorequired. When managing patients, clinical staffneed to have reliable information on whetherpsychological techniques work, and if so whichare the most effective, for which patients. Clini-cally, psychological interventions are required inmedical conditions for general adjustment to thedemands of symptom management and whenpsychological factors impede conventional coursesof medical treatment, such as in those patientswith genuine psychopathology. In particular, psy-chological co-morbidity has been demonstrated tohave high prevalence (49%) in patients referredwith difficult-to-control asthma.9 However, thesesame authors also report that psychological co-morbidity is often not diagnosed.
ARTICLE IN PRESS
Psychological interventions for adults with asthma 3
It is also important to know whether interven-tions work best alone or in combination with eachother, and whether it is better that patients aretaught individually or in a group. If possible it wouldalso be useful for staff and patients to know whatbenefits might be expected, and whether they areshort-lived or last in the longer term. Of note,psychological interventions receive little or nomention in international guidelines, such as theBritish Thoracic Society (BTS),10 Global Initiativefor asthma (GINA)11 for the management of asthma.Therefore, we conducted a systematic review toprovide health care professionals with an evidencebase in relation to psychological intervention foradults with asthma.
Methods
Objective
The purpose of this study was to conduct asystematic review and meta-analysis of randomisedcontrolled trials where the efficacy of psychologicalinterventions in modifying health and behaviouraloutcomes for adults with asthma was investigated.
Types of studies
Randomized controlled trials (RCT) comparing theeffects of psychological interventions for adultswith asthma were considered for inclusion. Cross-over trials were considered inappropriate forstudies using psychological interventions as theinfluence of a treatment might continue after theintervention has been stopped and were thereforeexcluded.
Individual and group formats were included butpatient education programmes were only acceptedwhere psychotherapy formed the major part of theintervention. Breathing retraining, yoga and mas-sage therapies were not incorporated, as thesetherapies were not considered to be primarilypsychological in nature.
Types of participants
Adults, both male and female, over the age of 16years of age, with asthma that has been diagnosedby a physician, although the accuracy can bevariable, or diagnosed using internationally estab-lished criteria.10 Treatments in both in and out-patient settings were included.
Types of interventions
Psychotherapy models were grouped according totheir theoretical frameworks or methods of opera-tion.
Approaches included:
1.
Behavioural therapies. 2. Cognitive therapies. 3. Cognitive behaviour therapy. 4. Relaxation techniques (including progressiverelaxation, autogenic training, hypnosis, andbiofeedback).
5.
Psychodynamic psychotherapies (including psy-choanalysis, psychosomatic therapy).6.
Counselling.Approaches excluded:
1.
Family therapy: this is the subject of anotherreview recently updated by our group.72.
Educational approaches: such approaches arealready the subject of several reviews.5 For thisreason, patient education programmes wereonly included where they comprised only partof a more complex psychological intervention.3.
Breathing re-training exercises: a review hasalready been completed on this topic.12Outcome measures
Primary:
�
Health service utilisation (e.g. hospitalisation,emergency room visits and, GP visits)Secondary:
�
Asthma symptoms � Lung function measures � Medication use � Absenteeism from school/work � Psychological health status (e.g. coping skills,anxiety, depression, asthma related behaviour,locus of control, self-esteem, self efficacy,quality of life and psychological status)
Search strategy
The primary source of studies was from the CochraneAirways Review Group trials register. In addition, thepsychological database PSYCHINFO was searched. Adetailed description of the search terms applied tothese databases is available in the Cochrane Databaseof Systematic Reviews.10 Bibliographies of each trialidentified were searched for additional relevant
ARTICLE IN PRESS
J. Yorke et al.4
trials. Authors of all identified RCTs were contactedand asked for information on further published orunpublished work. All searches were restricted tocases aged 16 years of age or more. Studies found upto the end of May 2005 were included.
Study selection
Two reviewers (SF and CS or JY and CS) establishedindependently whether each study met the inclu-sion criteria. Disagreements were resolved bydiscussion and 14 RCTs were included.
Methodological quality
The methodological quality of the studies (alloca-tion concealment) was assessed independently bytwo reviewers (SF and CS or JY and CS) usingthe Cochrane criteria for allocation concealment(Table 1) and a modified 0–5 scale developedby Jadad et al.14 (Table 2). Modification of thisscale was essential as, due to the nature of thepsychological interventions, it would be difficult toconduct double-blinded trials. Therefore in step 2and 5 ‘double-blind’ has been changed to ‘blind’.
Data analysis
Data were extracted and entered into RevMan 4.2.For continuous outcomes, we pooled data with afixed effect mean difference and 95% confidenceintervals (CIs). Where heterogeneity was present(40%) we performed a random effects analysis toincorporate statistical heterogeneity into thepooled estimate. Where this altered the signifi-cance of the effect we have reported both sets ofresults. Where data were not available as Ns,
Table 2 Modified Jadad scale.3
� Was the study described as randomised (1 ¼ yes;0 ¼ no� Was the outcome assessment blinded (1 ¼ yes;0 ¼ no)?� Was there a description of withdrawals and dropouts (1� Was the method of randomisation well described and a� Was the method of blinding well described and approp� Deduct one point if methods for randomisation or blind
Table 1 Allocation concealment.
Grade A: adequate concealmentGrade B: uncertainGrade C: clearly inadequate concealmentGrade D: not used
means and SDs or SEMs, we have attempted to deriveeffect estimates based on the mean difference andan estimate for the variance based upon thepublished P value. This was subsequently enteredas generic inverse variance data (GIV).
Dichotomous outcomes were entered as simpleevent rates for treatment and control groups. Wepooled data with a fixed effect odds ratio (OR).Where heterogeneity was present (40%) we per-formed a random effects analysis to incorporatestatistical heterogeneity in to the pooled estimate.Where this altered the significance of the effect wehave reported both sets of results. For significantoutcomes, we calculated a number needed to treat(benefit) (NNT(b)) or number needed to treat(harm) (NNT(h)), based upon the OR with Visual Rx.
Results
The literature search identified 85 papers, of which14 (Table 3) met our inclusion criteria15–29 with twopublications from one study reporting differentoutcomes.16,17 All were randomised and conductedover a variety of durations (3 days–12 months). Afull reference list for excluded papers can be foundin the Cochrane Database of Systematic Reviews.13
Methodological quality of included studies
The methodological quality of the studies was poorwith only two recent studies allocated a Jadadscore of four (Table 3).24,26 Allocation concealmentwas often not described (Table 3).
Study participants
A total of 617 participants were included in thisreview (Table 3). The studies were generally smallwith only one having more than a 100 peopleincluded23 and the smallest had 12.17 In somestudies a description of withdrawals was notgiven.15,19,25 Others gave the numbers who with-drew but no details of their characteristics. The
)?
¼ yes;0 ¼ no)?ppropriate (1 ¼ yes;0 ¼ no)?riate (1 ¼ yes;0 ¼ no)?ing were inappropriate.
ARTICLE IN PRESS
Table
3Cha
racteristics
ofinclud
edstud
ies.
Stud
yQua
lity
assessmen
tMetho
ds
Sample
size
sPa
rticipan
tch
arac
teristics
Deter
15
Jadad
:1
RCT,
3grou
ps,
metho
dof
rand
omisationno
tdescribed
.Eligible:90
Age
mea
n43
.5yrs.
AC:B
Ran
dom
ised
:34
Sex:
M14
,F17
.E1
:9,E2
:10
,C:12
Diagn
osis:64
%allergy,
87%infection,
93%
‘psych
olog
ical
trou
bles’.
DO:no
tstated
Epstein1
6Jadad
:2
RCT,
2grou
ps,
block
design.
Eligible:10
7Age
:25
-55yrs.
AC:B
Ran
dom
ised
:68
Sex:
E:M
27%,E:
F73
%,C:M
43%,C:F
57%.
E:34
,C:34
Diagn
osis:na
tion
alasthmaex
pertpan
elCRITER
IADO:E:
17,C:18
Erskine1
7Jadad
:2
RCT,
2grou
ps,
match
edin
pairs
acco
rdingto
age,
sex,
FEV1,an
dseve
rity
ofasthma
Eligible:no
tstated
Age
:16
–46
,mea
n30
yrs
AC:B
Ran
dom
ised
:12
Sex:
notdescribed
E:6,
C:6
Diagn
osis:AT
Scriteria
DO:E:
1,C:1
Seve
rity:mod
erateor
seve
re.
Ewer
18
Jadad
:3
RCT,
5grou
ps,
computerized
(autho
rco
rrespon
den
ce)
Eligible:no
tstated
Age
:18
–44
yrs
AC:A
Ran
dom
ised
:44
Sex:
M15
,F24
E1:10,
E2:12
,C:7
Diagn
osis:Ph
ysician
DO:E:
0,C:5
Seve
rity:mildto
mod
erate.
Hen
ry19
Jadad
:1
RCT,
2grou
ps,
metho
dof
rand
omizationno
tdescribed
Eligible:no
tstated
Age
:18
–58
,mea
n:39
.66
AC:B
Ran
dom
ised
:24
Sex:
M3,
F21
E:12
,C:12
Diagn
osis:Ph
ysician
DO:no
tstated
Seve
rity:mod
erateto
seve
re
Hoc
kemey
er21
Jadad
:3
RCT,
3grou
ps,
sealed
enve
lopes
Eligible:60
Age
:18
-51yrs,mea
n:20
.7AC:B
Ran
dom
ised
:60
Sex:
M25
,F29
E:30
,E:
30Diagn
osis:AT
Sgu
idelines
DO:E:
3,C:3
Seve
rity:no
tdescribed
Lehrer
22
Jadad
:2
RCT,
3grou
ps,
metho
dno
tstated
Eligible:10
6Age
:18
–65
yrs
AC:B
Ran
dom
ised
:10
6Se
x:M
51,F55
E1:38
,E2
:38
,C:30
Diagn
osis:Ph
ysician
DO:E1
:12
,E2
:13
,C:9
Seve
rity:mildto
mod
erate
Lehrer
23
Jadad
:1
RCT,
3grou
ps,
metho
dof
rand
omizationno
tdescribed
Eligible:no
tstated
Age
18–65
yrs,
mea
n37
.8AC:D
Ran
dom
ised
:17
Sex:
M5,
F12
E1:6,E2
:5,C:6
Diagn
osis:Ph
ysician
Psychological interventions for adults with asthma 5
ARTICLE IN PRESS
Table
3(con
tinu
ed)
Stud
yQua
lity
assessmen
tMetho
ds
Sample
size
sPa
rticipan
tch
arac
teristics
DO:2
Seve
rity:no
tdescribed
Lehrer
24
Jadad
:4
RCT,
4grou
ps,
restricted
rand
omization
Eligible:no
tstated
Age
:mea
n37
.3yrs(10.2),
AC:B
Ran
dom
ised
:94
Sex:
M30
,F64
E1:23:
E2:22
C1:
24,C2:
25Diagn
osis:Ph
ysician
DO:E1
:6,
E2:5,
C1:
5,C2:2
Seve
rity:mild
Paye
tte2
5Jadad
:1
RCT,
2grou
ps,
metho
dof
rand
omizationno
tdescribed
Eligible:no
tstated
Age
:22
–67
yrs,
mea
nno
tprovided
AC:D
Ran
dom
ised
:18
Sex:
M5,
F13
E:11
,C:7
Diagn
osis:Ph
ysician
DO:no
tstated
Seve
rity:no
tstated
Put
26
Jadad
:4
RCT,
2grou
ps,
enve
lopemetho
dEligible:10
1Age
:Emea
n43
yrs,
Cmea
n48
yrs
AC:B
Ran
dom
ised
:25
Sex:
M1,
F12
E:12
,C:11
Diagn
osis:AT
Scriteria
DO:2
Seve
rity:mild7,
mod
erate15
,seve
re1
Ross
27
Jadad
:3
RCT,
2grou
ps,
metho
dof
rand
omisationno
tdescribed
Eligible:86
Age
:Emea
n37
.87yrs(10.49
),Cmea
n40
.7yrs(12.57
)AC:B
Ran
dom
ised
:48
Sex:
M0,
F48
E:24
,C:24
Diagn
osis:Ph
ysician
DO:23
,E:
9,C:14
Seve
rity:on
a10
point
like
rtscale:
Emea
n6.80
(1.97),Cmea
n3.89
(1.90)
Sommarug
a28
Jadad
:2
RCT,
2grou
ps,
computerised
metho
d(autho
rco
rrespon
den
ce)
Eligible:no
tstated
Age
:mea
n48
yrs
AC:C
Ran
dom
ised
:40
Sex:
M21
,F19
E:20
,C:16
Diagn
osis:AT
Scriteria
DO:4
Seve
rity:no
tstated
Wag
aman
29
Jadad
:2
RCT,
4grou
ps,
stratified
acco
rding
tohy
pno
ticsuscep
tibility
Eligible:45
Age
:19–
65yrs,
mea
n41
.0AC:C
Ran
dom
ised
:30
Sex:
M3,
F18
E1:7,
E2:7,
C1:
7,C2:
7Diagn
osis:Ph
ysician
DO:9
Seve
rity:mod
erateto
mod
eratelyseve
re(Nationa
lInstituteof
Hea
lthGuidelines)
AC:alloca
tion
conc
ealm
ent
RCT:
rand
omised
controlled
trial
E:ex
perimen
talgrou
pC:co
ntrolgrou
pM:male
F:female
DO:dropou
tsSD:stan
dard
dev
iation
.
J. Yorke et al.6
ARTICLE IN PRESS
Psychological interventions for adults with asthma 7
severity of asthma varied from mild to severe,however not all studies reported this.
Method of randomisation
Studies were randomised with patients allocated tocontrol and experimental groups; however themethod of randomisation was not always men-tioned.15,19,22,23,25,27 Therefore it is difficult in allthe studies to gauge whether the method ofrandomisation was appropriate. This is reflectedin the Jadad scores (Table 3).
Interventions used
Diverse interventions were used (Table 4). Rarelywas the theoretical underpinning of the therapyprovided. Eight studies used some form of relaxa-tion technique as their intervention15–19,21,22,29
however techniques ranged from autogenic therapyto hypnosis and progressive muscle relaxation.Other psychological interventions included bio-feedback22,24,25 and CBT.26–28 In asthma, all ofthese techniques are used in combination with drugtherapies.
Table 4 Effect of psychological interventions on outcom
Study Outcome and effect
Healthcare utilisationHospital admission rates
Deter15 �
Sommaruga28 +Emergency room visits
Sommaruga28 +Asthma symptomsAsthma Quality of Life QuestionnairTotal
Put26 +Ross27 �
Pooled effect +Epstein16 �
Asthma symptom checklistAll variables
Lehrer22 +Lehrer23 �
HyperventilationPut26 �
ObstructionPut26 +
FatiguePut26
Irritation
Primary outcome
Health service utilisationThree studies15,28,29 examined this outcome. Deterand Allert15 found no significant decrease in health-care utilisation in the group receiving relaxationcompared to the control group. Sommaruga et al.28
found numbers of hospitalisation days and numberof emergency visits were decreased for both theintervention group (an asthma rehabilitation pro-gramme) and the control group (Po0:05). Numberswere too small in a study on hypnosis29 to test forstatistical differences for this outcome. Due tovarying interventions and insufficient reporting ofdata a pooled effect on the primary outcome couldnot be performed.
Secondary outcomes
Asthma symptomsThese were measured in a variety of ways in anumber of studies of which five used a form ofrelaxation therapy,16–18,22,29 two used bio-feedbacktechniques20,21 and three employed CBT.26–28
The Asthma Quality of Life Questionnaire (AQLQ),30
which includes asthma symptom categories, wasadapted as an outcome measure in three studies. Ameta-analysis could be performed with the two CBT
e measures.
Analysis Intervention
Between groups Relaxation therapyWithin group CBT
Within group CBT
e
Between groups CBTBetween groups CBT
Between groups Relaxation therapy
Across all groups Relaxation therapyWithin group Bio-feedback
Between groups CBT
Between groups CBT
Between groups CBT
ARTICLE IN PRESS
Table 4 (continued )
Study Outcome and effect Analysis Intervention
Put26 � Between groups CBTDyspnoea
Put26 � Between groups CBTHyperventilation
Put26 � Between groups CBTAnxiety
Put26 � Between groups CBTHome diaries/self reportSleep
Ewer18 + Within sub-group Relaxation therapyWheeze
Ewer18 + Within sub-group Relaxation therapyEpstein16 � Between groups Relaxation therapy
ActivityEwer18 + Within sub-group Relaxation therapy
CoughEwer18 � Within sub-group Relaxation therapy
PhlegmEwer18 � Within sub-group Within sub-group
Other symptom measuresLehrer24 + Within group Bio-feedback
Lung FunctionPEF
Lehrer23 � Between groups Bio-feedbackLehrer22 + Between groups Bio-feedbackPooled effect +
FEV1Ewer18 � Between groups Relaxation therapyHenry19 � Between groups Relaxation therapyEpstein16 � Between groups Relaxation therapyPooled effect �
Payette25 � Between groups Bio-feedbackHockemeyer21 � Between groups Bio-feedback
FEV1% predictedHenry19 + Within group Relaxation therapyHockemeyer21 � Between groups Bio-feedback
FEV1/FVCHockemeyer21 + Between groups Relaxation therapyPayette25 � Between groups Bio-feedback
Medication useEpstein16 � Between groups Relaxation therapyWagaman29 � Between groups Relaxation therapyPooled effect + Within sub-group Relaxation therapyEwer18 +Deter15 + Between groups Relaxation therapyLehrer24 + Within group Bio-feedbackLehrer23 � Between groups Bio-feedback
School/work absenteeismSommaruga28 + Within group CBT
Psychological Health StatusAnxiety Between groups CBT
Put26 �(ASC) Between groups CBTRoss27 + (SPRAS) Between groups CBTRoss27 + (number of panic attacks) Between groups CBTRoss27 + (ASI) Between groups Relaxation therapy
Trait anxietyEpstein16 —(Spielberger) Between groups Relaxation therapy
J. Yorke et al.8
ARTICLE IN PRESS
Table 4 (continued )
Study Outcome and effect Analysis Intervention
Sommaruga28 + (CBA) Within group CBTState anxiety
Epstein16 —(Spielberger) Between groups Relaxation therapyDepression
Epstein16 —(BDI)Ross27 + (BDI) Between groups CBTSommaruga28 + (CBA) Within group CBT
Health locus of controlSommaruga28 � Within group CBT
Respiratory Illness QuestionnaireExternal control
Sommaruga28 + Within group CBTPsychological stigma
Sommaruga28 + Within group CBTVegetative state
Henry19 + Within group Relaxation therapyEmotional state
Henry19 � Within group Relaxation therapyBehavioural state
Henry19 � Within group Relaxation therapyReactivity to stress
Henry19 � Within group Relaxation therapyNegative emotional state
Put26 + Between groups CBTPerceived Stress Scale
Hockemeyer21 � Between groups CBTParticipant satisfaction
Hockemeyer21 + Between groups CBTWithdrawals
Sommaruga28 � Between groups CBTPut26 �
Pooled effect �
Erskine17 � Between groups Relaxation therapyEwer18 �
Hockemeyer21 �
Epstein16 �
Pooled effect �
+: statistically significant effect; �: no statistically significant effect; CBT: cognitive behavioral therapy; PEF: peak expiratoryflow; FEV1: forced expiratory volume in 1 s; FEV1%: forced expiratory volume in 1 s percent predicted; FEV1/FVC: ratio of forcedexpired volume in one second to forced vital capacity; ASC: Asthma Symptom Checklist; SPRAS: Sheehan Patient-Rated AnxietyScale; BDI: Beck Depression Index.
Figure 1 Quality of life scores (AQLQ).
Psychological interventions for adults with asthma 9
ARTICLE IN PRESS
J. Yorke et al.10
studies.26,27 This analysis indicated a significantdifference in favour of CBT for the Total AQLQ(WMD 0.71, CI 0.23–1.19) (Fig. 1). One study16 foundno significant improvement in the relaxation groupwith a Total AQLQ mean 5.25 (SD 1.39) compared tothe control group mean 4.89 (SD 1.45) (data availablethrough author correspondence).
The Asthma Symptom Checklist (ASC)31 was usedas an outcome measure in a number of studies.Cognitive behavioural therapy26 resulted in asignificant improvement in the intervention groupcompared to the control group for the sub-categories of obstruction (P ¼ 0:04), fatigue(Po0:001) and irritation (P ¼ 0:03) but not fordyspnoea, hyperventilation and anxiety.
Lung functionA number of studies included lung function as anoutcome to measure the effectiveness of theirintervention. Seven of these16–19,21,22,29 used aform of relaxation as the intervention. A meta-analysis including three of these studies16,18,20
could be performed indicating no significant differ-ence in favour of relaxation therapy for forcedexpired volume in 1 s (FEV1) (SMD �0.01, CI �0.41to 0.40) (Fig. 2).
Lehrer23 and Lehrer24 used bio-feedback therapyand presented peak expiratory flow (PEF) outcomedata that could be pooled for a meta-analysis. Thisindicated a significant difference in favour of bio-feedback therapy (SMD 0.66, CI 0.09–1.23) (Fig. 3).
Figure 2
Figure 3
Hockemeyer and Smyth21 measured FEV1/FVC(forced vital capacity ratio) following relaxationand found a statistically significant improvement inthe intervention group mean 102.0 compared tothe control mean 93.7 (Fð1; 54Þ ¼ 4:57, P ¼ 0:038).The predicted FEV1% was also measured by Hock-emeyer and Smyth,21 no statistically significantimprovement was found in the intervention groupmean 111.5 compared to the control mean 99.8(Fð1; 54Þ ¼ 3:41, P ¼ 0:71). Payette25 also found nosignificant differences between groups for FVC,FEV1, FEV1/FVC after biofeedback training.
Medication useSix studies examined intervention effects onmedication use.15,16,18,23,24,29 A pooled effect fromtwo studies16,29 demonstrated a positive responseto relaxation treatment by decreasing use ofmedication (OR 4.47, CI 1.22–16.44) (Fig. 4). Thistranslates to a number needed to treat of 3 (95% CI2–28). Deter and Allert,15 using relaxation, found asignificant difference in the numbers of people inthe experimental group (n ¼ 4) who required lessbronchodilators when compared to the control(n ¼ 0; Po0:05).
Absenteeism from school/workSchool/work absences were significantly decreasedpost-intervention for both the asthma rehabilita-tion group and the control group (data to support
FEV1.
PEF.
ARTICLE IN PRESS
Figure 4 Medication decrease or discontinuation.
Psychological interventions for adults with asthma 11
this are not provided, apart from Po0:05) in onestudy.28
Psychological health statusAnxiety. Anxiety was used as an outcome measurein a number of studies using a variety of interven-tions and measurement techniques.
Anxiety was measured using the State-traitAnxiety Inventory32 by Epstein et al.16 A narrativereport of no significant difference between thegroups was provided. Following author correspon-dence, data was provided for Trait Anxiety (inter-vention mean 42.0, SD 13.11 and control mean38.63, SD 11.90) and State Anxiety (interventionmean 39.09, SD 13.0 and control mean 42.85, SD
15.24).Anxiety is a sub-scale on the ASC31 used by Put26
following CBT. No significant differences betweengroups were found. Ross27 measured anxiety using apanic attack diary, the Sheehan Patient-RatedAnxiety Scale (SPRAS)33 and the Anxiety SensitivityIndex (ASI).34
Depression. Two studies used the Beck DepressionInventory (BDI)35 as an outcome.16,27 Epsteinet al.16 provided a narrative description of nostatistical difference between groups (meansand SDs were provided following author correspon-dence for the BDI (intervention group: mean7.84, SD 6.98 and control group: mean 7.0, SD
6.74)). Ross et al.27 examined the effects of CBTand found that the results of 2� 2 ANOVA analysisdemonstrated no statistical improvement in BDIlevels (Fð1:22Þ ¼ 2:94, Po0:10) between the twogroups (except for external chance (Po0:03) in thecontrol group).Health locus of control. Health locus of control,including internal beliefs, and external controlthrough powerful others and chance was measuredusing the Health Locus of Control Scale36 in a studyinvestigating the effects of CBT.28 There were nosignificant differences between baseline and 1 yearfollow-up in the intervention group or controlgroup.
A negative emotionality scale, which incorpo-rates measures of negative affectivity as a person-ality trait, including irritability, nervousness, andemotional instability was used by Put et al.26
People in the experimental group had a significantdecrease in scores compared to the control group(Fð2; 42Þ ¼ 10:8, P ¼ 0:0002).
Hockemeyer and Smyth21 measured participantstress levels using the Perceived Stress Scale37
ANCOVA (controlling for age, age at diagnosis, andperceived stress levels at baseline) did not revealany significant differences between groups inperceived stress levels at the end of a 4 weekCBT programme (Fð1; 54Þ ¼ 1:48, P ¼ 0:23).
Patient satisfactionRoss et al.27 assessed participant satisfaction withtheir CBT programme using a scale of 1–4.Participants’ mean (SD) ratings of their satisfactionare recorded in Fig. 5. Hockemeyer and Smyth21
assessed participant satisfaction with a self-deliv-ered workbook on relaxation, CBT exercises, andwriting therapy. This was assessed using a pre-developed tool.38 A greater satisfaction withtherapy was found in the intervention group mean33.73 (SD 6.91) compared to the placebo groupmean 26.67 (SD 9.03) (Po0:01).
WithdrawalsData relating to drop-outs was provided by sevenstudies.16–18,21,24,26,28 A pooled effect for partici-pant withdrawals was performed for relaxa-tion16–18,21 which demonstrated no significantdifference between the intervention group andcontrol group (OR 0.67, CI 0.32–1.38) (Fig. 5).Other withdrawal rates are provided in Fig. 5.
Discussion
This systematic review evaluated 14 trials of variedpsychological interventions for adults with asthma.
ARTICLE IN PRESS
Figure 5 Withdrawals.
J. Yorke et al.12
The ability to make firm conclusions as to theeffectiveness of psychological interventions waslimited by poor study quality, insufficient reportingof data and varied outcome measures. In addition,the psychological interventions themselves werevaried, did not necessarily have a clear theoreticalunderpinning, and were not always well described.Additionally, the origin of the need for psychologi-cal intervention was rarely described making itdifficult to discern whether the aim of treatmentwas for general adjustment to asthma or forpsychological co-morbidity. Whilst psychologicalco-morbidity is recognised as being difficult tocharacterise and often not diagnosed,9 it isimperative that studies evaluating the effects ofpsychological interventions define this character-istic as the aims and objectives will be differentfor each patient group. As such, any results andconclusions must be viewed with caution. Theseissues were also apparent in a recent reviewof psychological interventions for children withasthma.8
Some meta-analyses were performed which gen-erally showed positive results. For instance, qualityof life, as measured by the AQLA,30 was improvedfollowing CBT. This is an important finding thatrequires further research in this patient population.Quality of life is an important outcome measure inpatients with respiratory disease as whilst manyinterventions may not significantly improve physio-logical parameters, they have been found to causea meaningful improvement in the patient, such asquality of life.40 Further use of the AQLQ would bebeneficial in assessing the effect of CBT and otherpsychological interventions on this outcome. Arecently published RCT6 examining the effects ofa nurse-led psycho-educational intervention alsomeasured quality of life. This paper concluded thattreatment had a positive effect on quality of lifebut not physical functioning, symptom control, andother variables. However, the main component ofthe intervention in this RCT was educational andtherefore not included in our systematic review.
The ASC31 was also adopted more than once as anoutcome measure.18,21,29 Whilst these studies usedrelaxation therapy no meta-analysis could beperformed due to small sample sizes29 and thepresentation of within group analysis only.18 Assuch, no conclusion as to the benefits of relaxationtherapy on the ASC outcomes can be provided.Future use of the ASC is recommended to enablethe effectiveness of varied psychological interven-tions to be assessed using meta-analyses.
Health care utilisation is increasingly being usedas a primary outcome in drug trials and otherstudies on patients with asthma. This being thecase, the primary outcome of this review reflectsthis. However, few trials included in this reviewmeasured health care use. In addition, it is assumedthat self-report measures were used, and these maynot give accurate data. As stated in a similar reviewinvolving children,8 health care utilisation is animportant indicator of the effect of many interven-tions as utilisation may be expected to decrease ifthere is an improvement in other variables. Thisreview is unable to make any conclusions as to thebenefit of psychological interventions in reducinghealth care utilisation. This outcome needs to beincluded in future trials of psychological interven-tion for adults with asthma.
Lung function was measured as an outcome in anumber of studies and two separate meta-analyseswere performed. Although bio-feedback was foundto improve PEF significantly, relaxation therapy didnot have such an impact on FEV1. Whilst lungfunction results are valuable in the assessment ofclinical variables, the relationship between theseand psycho-social variables remains questionable.39
The psychological outcomes examined werenumerous and diverse and there seems to be noconsensus as to which psychological outcomes areconceptually linked to asthma or to the psycholo-gical interventions being studied. The interventionsused varied as did the anxiety measurement tools.This prevented any pooled result being analysed. Inaddition, depression was measured using the Beck
ARTICLE IN PRESS
Psychological interventions for adults with asthma 13
Depression Inventory35 by two studies however;one used relaxation as the intervention13 and theother CBT.27
As highlighted in the similar review for children8
the aim of holistically orientated asthma manage-ment, incorporating psychological interventions, isnot solely to affect health in itself, but rather tofacilitate the patient’s adjustment to the illness37
which should include coping. The coping style ofpatients is an important predictor of asthmamorbidity.40 With increasing emphasis on patientself-management of asthma10 coping should beconsidered as an outcome measure for trials ofpsychological interventions. We, therefore recom-mend that valid outcome measures for evaluatingthe effectiveness of psychological interventions foradults with asthma need to address adjustment toand coping with asthma, as well as other psycho-logical indicators such as anxiety, depression,behaviour change, and quality of life.
Reviewers conclusions
This review highlights that the effects of psycholo-gical interventions are difficult to investigate andpresent challenges for the design of good RCTs.Researchers have to recruit sufficient numbers ofsubjects to show an effect if there is one, ensureappropriate randomisation and blinding techniques,and follow up subjects for a reasonable period.
RCTs evaluating this area are diverse. They studya mixed group of psychological techniques, whichare difficult to classify due to the different methodsused to deliver the intervention. This resulted inheterogeneous interventions even when the tech-nique was given the same classification by studyauthors. The diversity of the interventions was alsocomplicated by a multiplicity of outcomes and thetools used to measure these. In addition, thisreview highlights the need to classify patientsaccording to the presence or absence of psycholo-gical co-morbidity. We recommend that studiesevaluating the effects of psychological interven-tions state the origin of the need for treatment asthese two different approaches have differentstudy aims and objectives.
In addition, this body of work does not seem tohave a clear direction where current work isinfluenced by previous studies. Most of thesestudies were done by trialists who, with theexception of Lehrer’s team, did only one study.Research funding should target a range of goodquality research, including well-designed RCTs, todetermine the effectiveness and cost effectiveness
of psychological techniques that have a soundtheoretical base, with common taxonomy andoutcome indicators. As evidenced by this review,no recommendations for clinical practise as to theefficacy of psychological interventions for asthmacan be made. However, the mention of these non-pharmacological options in international guide-lines for the management of asthma may act as astimulus to research in this area.
Acknowledgements
The authors thank members of the Cochraneairways group; Toby Lasserson, Steve Milan, KarenBlackhall, Elizabeth Arnold, Phillipa Mills andBettina Reuben for their support and encourage-ment. Thanks to the following for translatingstudies: Gianni Ferrara (Italian), Keiji Hayashi(Japanese), Toby Lasserson (French and German),Sam Lasserson (Spanish), Makiko Meguro (Spanishand Japanese). Claudia Pagliari advised on thepsychological intervention classification and inclu-sion criteria. Rachel Churchill advised on theprotocol and the inclusion criteria. Two anonymouspeer reviewers gave valuable input for which weare grateful.
References
1. Lehrer PM, Isenberg S, Hochron SM. Asthma and emotion: areview. J Asthma 1993;30:5–21.
2. Barendregt JT. A psychological investigation of the effect ofgroup psychotherapy in patients with bronchial asthma.J Psychosom Res 1957;2:115–9.
3. Hyland ME. The mood-peak flow relationship in adultasthmatics: a pilot study of individual differences anddirection of causality. B.J. Med. Psychol 1990;63:379–84.
4. Affleck G, Apter A, Tennen H, Reisine S, Barrows E, WillardA, et al. Mood states associated with transitory changes inasthma symptoms and peak expiratory flow. Psychosom. Med2000;62:61–8.
5. Gibson PG, Powell H, Coughlan J, Wilson AJ, Abramson M,Haywood P, et al. Self-management education and regularpractitioner review for adults with asthma. In: The Cochranelibrary. Chichester, UK: Wiley; 2005.
6. Smith JR, Mildenhall S, Noble MJ, Shepstone L, Koutantji M,Mugford M, et al. The coping with asthma study: arandomized controlled trial of a home based, nurse ledpsychoeducational intervention for adults at risk of adverseasthma outcomes. Thorax 2005;60:1003–11.
7. Yorke J, Shuldham C. Family therapy for asthma in children(Cochrane review). In: The Cochrane library. Chichester, UK:Wiley; 2005 Issue 3.
8. Yorke J, Fleming S, Shuldham C. Psychological Interventionsfor children with asthma (Cochrane review). In: TheCochrane library. Chichester, UK: Wiley; 2005 Issue 4.
9. Heaney LG, Conway E, Kelly C, Gamble J. Prevalence ofpsychiatric morbidity in a difficult asthma population:
ARTICLE IN PRESS
J. Yorke et al.14
relationship to asthma outcome. Respir Med 2005;99:1152S–9S.
10. BTS/SIGN British Guideline on the management of asthma: anational clinical guideline. http://libraries.nelh.nhs.uk/guidelinesFinder/ accessed 23rd August 2005. An updatefrom 2003 guidelines published in Thorax 2004; 58(Suppl 1):i1–i94.
11. GINA Global Initiative for Asthma Management and Preven-tion. NIH publications 02-3659. 2003 (available at: http://www.ginasthma.com).
12. Holloway E, Ram FSF. Breathing exercises for asthma(Cochrane review). In: The Cochrane library. Chichester,UK: Wiley; 2004 Issue 4.
13. Yorke J, Fleming S, Shuldham C. Psychological Interventionsfor adults with asthma (Cochrane review). In: The Cochranelibrary. Chichester, UK: Wiley; 2005 Issue 5.
14. Jadad AR, Moore A, Carroll D, et al. Assessing the quality ofreports of randomised controlled trials: is blinding neces-sary? Control Clin Trials 1996;17:1–12.
15. Deter H-C, Allert G. Group therapy for asthma patients: aconcept for the psychosomatic treatment of patients in amedical clinic—a controlled study. Psychother Psychosom1983;40:95–105.
16. Epstein GN, Halper JP, Barret EAM, Birdsall C, McGee M,Baron KP, et al. A pilot study of mind-body changes in adultswith asthma who practice mental imagery. Alternat TherapHealth Med 2004;14(4):66–71.
17. Erskine J, Schonell M. Relaxation therapy in bronchialasthma. J Psychosom Res 1979;23:131–9.
18. Ewer TC, Stewart DE. Improvement in bronchial hyper-responsiveness in patients with moderate asthma aftertreatment with a hypnotic technique: a randomised con-trolled trial. BMJ 1986;293:1129–32.
19. Henry M, Gonzales de Rivera JL, Gonzales-Martin IJ, Abreu J.Improvement of respiratory function in chronic asthmaticpatients with autogenic therapy. J Psychosom Res 1993;37:265–70.
20. Henry M, Gonzalez de Rivera JL, de las Cuevas C, et al.Reduced neurovegetative reactivity patterns to stress inchronic asthmatics with autogenic therapy. Reduccion de lareactividad neurovegetativa ante el estres en asmaticoscronicos con terapia autogena. Psiquis 1991;12:26–32.
21. Hockemeyer J, Smyth J. Evaluating the feasibility andefficacy of a self administered manual-based stress manage-ment intervention for individuals with asthma: results froma randomized controlled study. Behav Med 2002;27:161–72.
22. Lehrer PM, Hochron SM, Mayne T, et al. Relaxation and musictherapies for asthma among patients prestablized on asthmamedication. J Behav Med 1994;17:1–24.
23. Lehrer P, Carr RE, Smetankine A, et al. Respiratory sinusarrhythmia versus neck/trapezius EMG and incentive in-spirometry biofeedback for asthma: a pilot study. ApplPsychophysiol Biofeedback 1997;22:95–109.
24. Lehrer PM, Vaschillo E, Vaschillo B, Lu S, Scardella A,Siddique M, et al. Biofeedback for asthma. Chest 2004;126(2):352–61.
25. Payette BA. The effect of EMG biofeedback trainingon selected physiological and personality variables inthe adult asthma patient. Diss Abstr Int 1977;38(5-B):2419.
26. Put C, van den Bergh O, Lemaigre V, Demedts M, Verleden G.Evaluation of an individualised asthma programme directedat behavioural change. Eur Respir J 2003;21:109–15.
27. Ross CJM, Davis TMA, MacDonald GF. Cognitive behavioraltreatment combined with asthma education for adults withasthma and coexisting panic disorder. Clin Nurs Res2005;14(2):131–57.
28. Sommaruga M, Spanevello A, Migliori GB, Neri M, Callegari S,Majani G. The effects of a cognitive behavioural interven-tion in asthmatic patients. Mondali Arch Chest Dis 1995;50:398–402.
29. Wagaman MJ. Physiological and psychological effects ofvarious hypnotic suggestions with asthma patients. DissAbstr Int 2000;61(1-B):185.
30. Juniper EF, Guyatt GH, Epstein RS, Ferrie PJ, Jaeschke R,Hiller TK. Evaluation of impairment of health related qualityof life in asthma: development of a questionnaire for use inclinical trials. Thorax 1992;47:76–83.
31. Kinsman RA, Luparello T, O’Banion K, Spector S. Multi-dimensional analysis of the subjective symptomatology ofasthma. Psychosom Med 1973;35:250–67.
32. Spielberger CD, Edwards CD, Lushene RE. Manual for thestate-trait inventory (self evaluation questionnaire). PaloAlto, CA: Consulting Psychologists Press; 1970.
33. Sheehan D. The sheehan patient rated anxiety scale. Tampa,FL: University of South Florida Psychiatry Centre, ResearchDivision; 1986.
34. Peterson R, Reiss S. Anxiety sensitivity index manual, 2nded. Worthington, OH: International Diagnostic Systems;1992.
35. Beck AT, Steer RA, Garbin MG. Psychometric properties ofthe beck depression inventory: twenty-five years of evalua-tion. Clin Psych Rev 1988;8(1):77–100.
36. Wallstone BS, Wallstone KA, Kaplan GD. Development andvalidation of the Health Locus Control scales. J Consult ClinPsychol 1976;44:580–5.
37. Cohen S, Kamarck T, Mermelstein R. A global measureof perceived stress. J Health Social Behav 1983;24:385–96.
38. Borkovec TJ, Nau SD. Credibility of analogue therapyrationales. J Behav Ther Exp Psychiatry 1972;3:257–60.
39. ZuWallack RL, Haggerty MC, Jones P. Clinically meaningfuloutcomes in patients with chronic obstructive pulmonarydisease. Am J Med 2004;117(Suppl 12A):49S–59S.
40. Aalto AM, Harkapaa K, Aro AJ, Rissanen P. Ways ofcoping with asthma ineveryday life: validation of theasthma specific coping scale. J Psychosom Res 2002;53:1061–9.