systematic meta-review and health economic meta-analysis ... · settings. core components are...
TRANSCRIPT
Systematic meta-review and health economic meta-analysis of supported self-management for asthma: a healthcare perspective Pinnock, H. , Parke, H.L. , Panagioti, M. , Daines, L. , Pearce, G. , Epiphaniou, E. , Bower, P. , Sheikh, A. , Griffiths, C. and Taylor, S.J.C. Published PDF deposited in Coventry University Repository Citation: Pinnock, H. , Parke, H.L. , Panagioti, M. , Daines, L. , Pearce, G. , Epiphaniou, E. , Bower, P. , Sheikh, A. , Griffiths, C. and Taylor, S.J.C. (2017) Systematic meta-review and health economic meta-analysis of supported self-management for asthma: a healthcare perspective. BMC Medicine, 15:64. DOI: 10.1186/s12916-017-0823-7 http://dx.doi.org/10.1186/s12916-017-0823-7 BioMed Central This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders.
RESEARCH ARTICLE Open Access
Systematic meta-review of supported self-management for asthma: a healthcareperspectiveHilary Pinnock1*, Hannah L. Parke2, Maria Panagioti3, Luke Daines1, Gemma Pearce4, Eleni Epiphaniou2,Peter Bower3, Aziz Sheikh1, Chris J. Griffiths2, Stephanie J. C. Taylor2 and for the PRISMS and RECURSIVE groups
Abstract
Background: Supported self-management has been recommended by asthma guidelines for three decades;improving current suboptimal implementation will require commitment from professionals, patients and healthcareorganisations. The Practical Systematic Review of Self-Management Support (PRISMS) meta-review and ReducingCare Utilisation through Self-management Interventions (RECURSIVE) health economic review were commissionedto provide a systematic overview of supported self-management to inform implementation. We sought toinvestigate if supported asthma self-management reduces use of healthcare resources and improves asthmacontrol; for which target groups it works; and which components and contextual factors contribute to effectiveness.Finally, we investigated the costs to healthcare services of providing supported self-management.
Methods: We undertook a meta-review (systematic overview) of systematic reviews updated with randomisedcontrolled trials (RCTs) published since the review search dates, and health economic meta-analysis of RCTs. Twelveelectronic databases were searched in 2012 (updated in 2015; pre-publication update January 2017) for systematicreviews reporting RCTs (and update RCTs) evaluating supported asthma self-management. We assessed the qualityof included studies and undertook a meta-analysis and narrative synthesis.
Results: A total of 27 systematic reviews (n = 244 RCTs) and 13 update RCTs revealed that supported self-managementcan reduce hospitalisations, accident and emergency attendances and unscheduled consultations, and improvemarkers of control and quality of life for people with asthma across a range of cultural, demographic and healthcaresettings. Core components are patient education, provision of an action plan and regular professional review. Self-management is most effective when delivered in the context of proactive long-term condition management. The totalcost (n = 24 RCTs) of providing self-management support is offset by a reduction in hospitalisations and accident andemergency visits (standard mean difference 0.13, 95% confidence interval −0.09 to 0.34).
Conclusions: Evidence from a total of 270 RCTs confirms that supported self-management for asthma can reduceunscheduled care and improve asthma control, can be delivered effectively for diverse demographic and culturalgroups, is applicable in a broad range of clinical settings, and does not significantly increase total healthcare costs.Informed by this comprehensive synthesis of the literature, clinicians, patient-interest groups, policy-makers andproviders of healthcare services should prioritise provision of supported self-management for people with asthma as acore component of routine care.(Continued on next page)
* Correspondence: [email protected] UK Centre for Applied Research, Allergy and Respiratory ResearchGroup, Usher Institute of Population Health Sciences and Informatics,University of Edinburgh, Doorway 3, Medical School, Teviot Place, EdinburghEH8 9AG, UKFull list of author information is available at the end of the article
© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Pinnock et al. BMC Medicine (2017) 15:64 DOI 10.1186/s12916-017-0823-7
(Continued from previous page)
Systematic review registration: RECURSIVE: PROSPERO CRD42012002694; PRISMS: PROSPERO does not registermeta-reviews
Keywords: Supported self-management, Asthma, Systematic meta-review, Health economic analysis, Meta-analysis
BackgroundAsthma is common, affecting 334 million people world-wide, and is responsible for substantial morbidity and anincreasing burden on healthcare services globally [1]. Inthe UK, there are over 6 million primary care consulta-tions, and 100,000 hospital admissions each year, at anestimated cost of £1 billion per year [2].For a quarter of a century [3], national and international
guidelines have recommended – unequivocally – thatpeople with asthma should be provided with self-management education reinforced by a personalisedasthma action plan and supported by regular review [4, 5],though mode of delivery, personnel delivering the support,the targeted group and the intensity of the interventionvary [6]. The 2014 UK National Review of Asthma Deathsprovided a stark reminder of the importance of ensuringthat people with asthma respond in a timely and appropri-ate manner to deteriorating symptoms: only 23% haddocumented evidence of having been provided with self-management education and 45% of people who died hadnot sought or received medical attention in their finalattack [7].However, despite self-management being highlighted as
a core component of all models of care for people withlong-term conditions (LTCs) [8–10] and the conceptbeing well established in the context of asthma [4, 5], inpractice only a minority of people with asthma have anaction plan [11]. Effective implementation requires awhole systems approach, combining active engage-ment of patients with the training and motivation ofprofessionals embedded within an organisation inwhich self-management is valued [12]. Patient organi-sations, healthcare professionals, policy-makers, com-missioners and providers of healthcare services thus needan up-to-date systematic overview of the evidence toinform decisions about prioritisation of supported self-management and to underpin implementation strategieswithin diverse healthcare systems.The data presented in this paper are derived from
two parallel programmes of work on supported self-management in LTCs commissioned by the NationalInstitute of Health Research: Practical SystematicReview of Self-Management Support (PRISMS) [13]and Reducing Care Utilisation through Self-managementInterventions (RECURSIVE) [14]. In the context ofasthma, we aimed to answer questions of importance toclinicians, patient-interest groups, managers responsible
for developing healthcare services and policy-makers: cansupported self-management reduce the use of healthcareresources and improve asthma control? More specifically,in which target groups has it been shown to work, whichcomponents are important, in what healthcare contexts,and at what cost?
MethodsWe used established methodology for undertaking ameta-review of systematic reviews (PRISMS) and a sys-tematic review of randomised controlled trials (RCTs)(RECURSIVE) [15]. The PRISMS and RECURSIVE re-views were undertaken during 2012–2013 with initialsearches completed in November 2012 and May 2012,respectively. We updated the PRISMS searches in March2015 with a pre-publication update in January 2017, andthe RECURSIVE searches in September 2015. RECUR-SIVE is registered on PROSPERO: CRD42012002694.(PRISMS could not be registered because PROSPEROdoes not register meta-reviews.)
Search strategyTable 1 summarises the PICOS criteria, search strat-egies, sources and search dates; further details are inAdditional file 1. The PRISMS search strategy in-volved searching nine electronic databases using theterms: ‘self-management support’ AND ‘asthma’ AND‘systematic review’. We defined self-management as‘the tasks that individuals must undertake to live withone or more chronic conditions. These tasks includehaving the confidence to deal with medical manage-ment, role management and emotional managementof their conditions’ [16]. For the update, we searchednot only for systematic reviews published after ourinitial search date but also for RCTs published afterthe search dates used by the included systematicreviews (see Additional file 2 for the details of thesedates). Included systematic reviews were groupedaccording to the populations studied (children, adultsor ethnic minority groups) and the search dates ofthe reviews extracted. Dates for the update RCTsearch were set from the date of the latest reviewsearch within each population group.The RECURSIVE search strategy in nine databases
comprised the terms: ‘self-management support’ AND‘long-term condition’ AND ‘healthcare use’ AND ‘rando-mised controlled trial’. (RECURSIVE included asthma
Pinnock et al. BMC Medicine (2017) 15:64 Page 2 of 32
and other LTCs in a single search.) We also specificallysought health economic publications linked to includedRCTs.
Identification of relevant papersTable 2 summarises the PRISMS and RECURSIVE pro-cesses. Following training (repeated cycles of duplicatescreening of 100 titles, team discussion and clarification ofexclusion rules), one reviewer (HLP or GP for PRISMS;LD for the update; MP for RECURSIVE) reviewed titlesand abstracts and selected possibly relevant studies. A ran-dom sample of titles and abstracts (10% in PRISMS; 40%in RECURSIVE) was examined by a second reviewer (HPfor PRISMS; PB or NS for RECURSIVE) workingindependently as a quality check. The agreement was 97%for the initial search and 99% for the update in PRISMSand 87% for the initial search and 88% for the update inRECURSIVE.After a similar training process, the full texts of all
potentially eligible studies were assessed against the
eligibility criteria (see Additional file 3) by one reviewer(HLP for PRISMS; LD for update; MP for RECURSIVE).Second reviewers undertook a 10% check for PRISMS(HP) and a 30% check for RECURSIVE (PB or NS),achieving 83% and 85% agreement, respectively. Disagree-ments were because unclear papers were included by thereviewer pending discussion with a lead investigator.Uncertainties and disagreements were resolved by fullteam discussion.
Assessment of methodological qualityWe used the R-AMSTAR (Revised Assessment of MultipleSystematic Reviews [17]) quality appraisal tool to assess themethodological quality of the systematic reviews includedin the PRISMS study. This reflects both the quality of thereview process and the rigour with which the reviewassessed the quality of the studies it included. We used theCochrane Risk of Bias tool to assess the quality of RCTs in-cluded in the updated search [15]. Quality assessment wasundertaken by HLP or LD and independently by a second
Table 1 PICOS search strategy and sources for the reviews
PRISMS systematic meta-review RECURSIVE systematic review
Population Adults/children with asthma, from all social anddemographic settings. Multi-condition studies ifasthma data reported.
Adults (≥18 years) with asthma (within a wider searchof long-term conditions), excluding studies in thedeveloping world.
Intervention Self-management support interventions. Self-management support interventions.
Comparator Typically ‘usual care’ or less intense self-managementinterventions.
Typically ‘usual care’ or less intense self-managementinterventions.
Outcomes Unscheduled use of healthcare services (admissions,A&E attendances, unscheduled consultations), healthoutcomes (asthma control), quality of life, processoutcomes (ownership of action plans, self-efficacy).
Healthcare utilisation with comprehensive measures ofcosts or major cost drivers (i.e. hospitalisation, A&Eattendances), quality of life.
Settings Any healthcare setting. Any healthcare setting.
Study design Systematic reviews of RCTs.RCTs published after the date of the last search in theincluded systematic reviews (see Additional file 2).
RCTs
Dates Initial database search: January 1993 (3 years before thepublication of the earliest systematic review identifiedin scoping work) to July 2012. Manual and forwardcitations were completed in November 2012.Update search: March 2015. Pre-publication updateJanuary 2017.
Initial database search: inception to May 2012.Update search: September 2015.
Databases MEDLINE, EMBASE, CINAHL, PsycINFO, AMED, BNI,Cochrane Database of Systematic Reviews, Database ofAbstracts of Reviews of Effects, and ISI Proceedings(Web of Science).
CENTRAL, CINAHL, EconLit, EMBASE, Health EconomicsEvaluations Database, MEDLINE, MEDLINE In-Process &Other Non-Indexed Citations, NHS Economic EvaluationDatabase, and the PsycINFO.
Manual searching Systematic Reviews, Health Education and Behaviour,Health Education Research, Journal of BehaviouralMedicine, and Patient Education and Counseling.
Systematic Reviews.
Forward citations On all included systematic reviews. Bibliographies ofeligible reviews.
None.
In progress studies Abstracts were used to identify recently published trials. Abstracts were used to identify recently published trials.
Other exclusions Previous versions of updated reviews.Papers not published in English.
Not applicable.
A&E accident and emergency, RCT randomised controlled trial
Pinnock et al. BMC Medicine (2017) 15:64 Page 3 of 32
reviewer (HP) with disagreements resolved by discussionwithin the team (EE, GP, HLP, ST and HP).To reflect both quality and size of the review, we
developed a star weighting system based on (a) the R-AMSTAR score (≥31 was defined as ‘high-quality’) and(b) the number of participants (≥1000 participants wasdefined as ‘large’):
*** Large high-quality review** Either small high-quality review or large
low-quality review* Small low-quality review
In the RECURSIVE study, quality assessment offormal economic evaluations was undertaken usingthe Drummond checklist [18, 19]; RCTs reportinghealthcare utilisation were assessed by judging allo-cation concealment (the quality component most
associated with treatment effect [20]) as adequate orinadequate according to the Cochrane Risk of Biastool [15].
OutcomesThe primary outcome in the PRISMS meta-review wasunscheduled use of healthcare resources (specifically un-scheduled consultations, accident and emergency (A&E)department attendances and hospital admissions). Otheroutcomes of interest were asthma control, asthma-related quality of life and process outcomes (specifically,ownership of action plans). Healthcare utilisation ratesand costs were the primary focus of the RECURSIVE re-view, especially major cost drivers (i.e. hospitalisationrates and costs) and comprehensive summaries includingmultiple sources of cost. The results of formal cost-effectiveness, cost-utility and cost–benefit analyses werealso of interest.
Table 2 PRISMS and RECURSIVE processes for selection of studies, quality assessment, data extraction, analysis and interpretation
PRISMS systematic meta-review RECURSIVE systematic review
Title and abstract screening Initial training.One reviewer selected studies for full-text screening.Quality check: Random sample of 10% checkedindependently by second reviewer.Agreement: 97% for the initial search and 99%for the update.Uncertainties resolved by discussion.
Initial training.One reviewer selected studies for full-text screening.Quality check: Random sample of 40% checkedindependently by second reviewer.Agreement: 87% for the initial search and 88%for the update.Uncertainties resolved by discussion.
Full-text screening Following training, one reviewer selected possiblyrelevant studies for inclusion.Quality check: Random sample of 10% checkedindependently by second reviewer.Agreement: 83%.Uncertainties resolved by discussion.
Following training, one reviewer selected possiblyrelevant studies for inclusion.Quality check: Random sample of 30% checkedindependently by second reviewer.Agreement: 85%.Uncertainties resolved by discussion.
Quality assessment Duplicate quality assessment using:R-AMSTAR [17] for systematic reviews (‘high-quality’defined as ≥31), combined with size of the review(‘large’ defined as ≥1000 participants) to give star rating(1* to 3*).Cochrane Risk of Bias tool for RCTs [15].Disagreements resolved by discussion.
Duplicate quality assessment using:Drummond for economic evaluations [18, 19].Allocation concealment for RCTs.Disagreements resolved by discussion.
Data extraction Data extraction by one reviewer.Quality check: 100% checked for accuracy by a secondreviewer.Disagreements resolved by discussion.
Data extraction by one reviewer.Quality check: Random sample of 40% extractedindependently by second reviewer.Disagreements resolved by discussion.
Analysis Reviews/RCTs categorised according to the question(s)that they answered:• Does supported self-management reduce healthcareutilisation and improve control?
• For which target groups does it work?• Which components contribute to effectiveness?• In what healthcare contexts does supported self-management work?Meta-Forest plots for pooled statistics of the primaryoutcome (healthcare utilisation).Narrative synthesis within categories.
Meta-analysis: Standardised mean differences (randomeffects model) to examine the effects of self-management support interventions on hospitalisationrates, A&E attendances, quality of life and total costs.Permutation plots of the data from trials reporting bothutilisation (hospitalisation rates, A&E attendances ortotal costs) and health outcomes (quality of life).
Interpretation Monthly teleconferences to enable synergies between PRISMS and RECURSIVE.End-of-project stakeholder conference to discuss findings and implications for commissioning and providing servicesfor people with LTCs.
A&E accident and emergency, LTC long-term condition, R-AMSTAR Revised Assessment of Multiple Systematic Reviews, RCT randomised controlled trial
Pinnock et al. BMC Medicine (2017) 15:64 Page 4 of 32
Extraction of dataData for the PRISMS review were extracted by HLP andLD (update) using a piloted data extraction form, andchecked independently by HP for integrity and accuracy.Disagreements were resolved by team discussion. Weextracted data on review rationale, the self-managementintervention under review, review methodology, summarydetails of included RCTs (participant demographics, com-parison groups, settings, service arrangements, compo-nents, duration/intensity of the intervention, follow-uparrangements) and the results of meta-analyses and narra-tive syntheses. We extracted the findings and conclusionsas synthesised by the authors of the systematic reviews,specifically avoiding going back to the individual primarystudies. The RCTs in the update review were extractedusing similar headings.A piloted data extraction sheet was devised for RE-
CURSIVE that included descriptive data (characteristicsof studies, populations and interventions) and quantita-tive data (for use in meta-analyses). All the descriptivedata and approximately 40% of the quantitative datawere double-extracted by two members of the researchteam working independently.
Data analysisMeta-analysis is inappropriate at the meta-review levelowing to the overlap of included RCTs between reviews.However, for the primary outcome, where two or moresystematic reviews (including the RECURSIVE meta-analyses) present pooled statistics, we displayed theresults graphically by creating ‘meta-Forest plots’. Weundertook narrative syntheses to answer our key ques-tions: Does supported self-management reduce use ofhealthcare resources and improve asthma control? Forwhich target groups does it work? Which componentscontribute to effectiveness? and In what contexts doessupported self-management work? We categorised thereviews and RCTs included in the PRISMS meta-reviewaccording to the question(s) that they answered (seeTables 3 and 4: column 3) and synthesised the findingswithin these categories.th=tlb=The final question (What is the effect of self-
management on healthcare utilisation and costs?) wasanswered by a meta-analysis of the RECURSIVE RCTdata. The primary analysis explored whether self-management support could reduce utilisation withoutcompromising outcomes. Standardised mean differences(SMD) were computed using a random effects modelmeta-analysis due to anticipated heterogeneity. Fourmeta-analyses examined the effects of self-managementsupport interventions on hospitalisation rates, A&Eattendances, quality of life and total costs, respectively.We then constructed permutation plots of the data fromthe subset of trials reporting both utilisation (hospitalisation
rates, A&E attendances or total costs) and health outcomes(quality of life). Further details about the analytic approachare described in the RECURSIVE report [14]. Forest plotsand permutation plots [21] for the subset of studies report-ing both health outcomes and utilisation outcomes wereconstructed in STATA version 14.
Interpretation and end-of-project workshopThe PRISMS and RECURSIVE teams worked independ-ently, but held regular teleconferences to enable synergiesbetween the findings of the parallel reviews to be devel-oped. Frequent meetings of the multidisciplinary teamsaided interpretation of the emerging findings. Finally, weheld an end-of-project stakeholder conference at whichthe findings and over-arching conclusions from PRISMSand RECURSIVE were presented to 34 multidisciplinarystakeholders, including people with LTCs, clinicians, com-missioners, providers of healthcare services and policy-makers. Small discussion groups discussed and advised onpractical implications for commissioning and providingservices for people with LTCs.
Lay involvementThe PRISMS project (which reviewed evidence from 14LTCs) benefited from a lay collaborator who was involvedfrom the inception of the project. She and other lay repre-sentatives from a range of LTC interest groups (includingAsthma UK) contributed to an initial stakeholder work-shop at which the choice of LTCs studied in the projectand self-management interventions of interest werediscussed. Lay members also participated in the end-of-project workshop (described above), which aidedinterpretation and guided dissemination. The PRIMERpatient and public involvement group at the University ofManchester, UK, collaborated with the RECURSIVEproject.
Updating of searches prior to publicationWe updated our PRISMS searches in January 2017 byundertaking forward citation of the original included re-views using Web of Science. Forward citation has beenshown to be an efficient and effective method of identi-fying relevant papers in systematic reviews of complexand heterogeneous evidence [22]. We considered it wasvery unlikely that a subsequent systematic review orRCT would be published without citing at least one ofthe previously published reviews. One reviewer (HP)undertook focused data extraction of key findings, whichwere checked by MP. The additional data were addedinto the syntheses as appropriate. Had we identifiedstudies that substantially changed our conclusions weplanned to undertake full duplicate data extraction,quality assessment and revise our synthesis.
Pinnock et al. BMC Medicine (2017) 15:64 Page 5 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns
Referenceand
weigh
ting*;RCTs,n
;Participants,n;R-
AMSTAR;Daterang
eof
includ
edRC
Ts
Com
parison
Relevanceto
meta-
review
questio
ns:
Interven
tions
includ
edTarget
grou
p(s)
Synthe
sis
Mainresults
Whatistheim
pact?
Target
grou
ps?
Which
compo
nents?
Con
text?
Bailey2009
[25]**
4RC
Ts617participants
R-AMSTAR36
RCTs
2000–2008
Culturally
orientated
prog
rammes
vs.usualcare
orlim
ited/
gene
riced
ucation.
FU(m
ode):12mo,rang
e4–12
mo
Impact
Target:Ethnicgrou
psEducation,actio
nplans,trigge
rsand
avoidance,
collabo
ratio
nwith
healthcare
services.
Lang
uage
-app
ropriate
asthmaed
ucators.
Minority
grou
ps:
Puerto
Rican,African-
American,H
ispanic,
Indian
sub-continen
t.Adu
ltsandchildren.
Meta-analysis
Narrativeanalysis
Redu
cedho
spitalisationin
children
(RR0.32,95%
CI0.15–0.70;1
RCT)
butno
trepo
rted
inadults.
Improved
QoL
inadults(W
MD0.25,
95%
CI0.09–0.41;2
RCTs).
2of
2RC
Tsrepo
rted
aredu
ctionin
A&E
visitsandho
spitalisations:one
repo
rted
nodifferencein
‘use
ofhe
althcare
resources’;2of
3repo
rted
improved
QoL
(adu
lts).
Bernard-Bo
nnin
1995
[26]**
11RC
Ts1290
participants
R-AMSTAR27
RCTs
1981–1991
Interactiveteaching
onself-managem
ent
vs.stand
ardcare.
Impact
Target:C
hildren
Interactiveteaching
(one
-to-on
eor
grou
p)to
supp
ortasthma
self-managem
ent.
Children1–18
y.Overallseverity
classifiedas
‘mild
tomod
erate’.
Meta-analysis
Narrativeanalysis
Redu
cedho
spitalisation(ES0.06
±−0.08)andem
erge
ncyvisits
(ES0.14
±0.09);5RC
Ts.
Childrenwith
high
baselinenu
mbe
rsof
hospitalisations
andem
erge
ncy
visitshadgreatestsubseq
uent
redu
ctionin
morbidity.
Bhog
al2006
[23]**
4RC
Ts355participants
R-AMSTAR41
RCTs
1990–2004
Symptom
-based
written
PAAPs
vs.p
eakflow-based
PAAP.
FU(m
ode):3
mo,rang
e3–24
mo
Target:C
hildren
Com
pone
nts:PEFvs.
symptom
mon
itorin
g
Asthm
aed
ucation
plus
PAAPs
forbo
thparentsandchildren.
Gen
erallycontaine
d3
step
s:often
employing‘traffic
lights’.
Mon
itorin
gvaried:
either
daily
orwhe
nsymptom
atic.
Children6–19
ywith
mild
tosevere
asthma.
Meta-analysis
Symptom
-based
PAAPs
redu
ced
unsche
duledcare
comparedto
peak
flow-based
PAAPs
(RR0.73,95%
CI
0.55–0.99;4RC
Ts).
Nodifferencein
hospitaladm
ission
s(RR1.51,95%
CI0.35–6.65.
Peak
flow-based
PAAPs
redu
cedthe
numbe
rof
symptom
aticdays/w
eek
(MD0.45
days/w
eek,95%
CI0.04–
0.26;2
RCTs).Nosign
ificant
differ-
ence
foradultor
child
QoL.
Zemek
2008
[24]**
5RC
Ts423participants
R-AMSTAR41
RCTs
1990–2005
WrittenPA
APs
vs.noPA
AP.
Symptom
-based
vs.
PEF-basedPA
AP.
FU(m
ode):3
mo,rang
e0.5–24
mo
Impact:
Target:C
hildren
Com
pone
nts:PA
AP
Educationforparents
andchildren,plus
PAAPs,w
ith3step
s:oftenem
ploying
‘traffic
lights’.
Mon
itorin
gvaried:
either
daily
orwhe
nsymptom
atic.
Scho
ol-age
dchildren
with
mild
tosevere
asthma.
Meta-analysis
APEF-basedPA
APredu
ced
unsche
duledcare
comparedto
noplan
(WMD−0.50,95%
CI−
0.83
to−0.17;1
RCT).
APEF-basedPA
APcomparedto
noplan
redu
cedsymptom
scores
(WMD
−11.80,95%
CI−
18.22to
−5.38)and
numbe
rof
scho
oldays
missed
(WMD−1.03,95%
CI−
1.85
to−0.21;
1RC
T).
Boyd
2009
[27]***
38RC
Ts7843
participants
R-AMSTAR39
RCTs
1985–2007
Educationtargeting
children/parents
vs.low
intensity
education.
Impact:
Target:C
hildren,A&E
attend
ees
Educationplus
therapyreview
,self-
mon
itorin
g,PA
APs,
andtrigge
ravoidance.
Children0–18
ywho
hadattend
edA&E
for
asthmawith
inthe
previous
12mo.
Meta-analysis
Subg
roup
analyses
Educationredu
cedA&E
attend
ances
(RR0.73,95%
CI0.65–0.81;17RC
Ts),
admission
s(RR0.79,95%
CI0.69–
0.92;18RC
Ts)andun
sche
duled
Pinnock et al. BMC Medicine (2017) 15:64 Page 6 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
FU(m
ode):12morang
e4–12
mo
Rang
eof
settings
and
profession
alsand
mod
eof
delivery.
consultatio
ns(RR0.68,95%
CI0.57–
0.81;7
RCTs).
Noeffect
onQoL
(WMD0.13,95%
CI
0.73–0.99;2RC
Ts).
Subg
roup
analyses
(typeandtim
ing
ofinterven
tion,tim
ingof
outcom
eassessmen
tor
ageof
participants)
didno
tchange
finding
s.
Bussey
Smith
2009
[28]*
9RC
Ts957participants
R-AMSTAR26
RCTs
1986
-2005
Com
puterised
education
vs.traditio
nalself-
managem
ent
FU(m
ode):12mo,rang
e3–12
mo
Impact:
Com
pone
nts:
Techno
logy-based
interven
tions
Interactive
compu
terised
educationalasthm
aprog
rammes
(gam
estailoredto
the
individu
al,w
eb-based
education,interactive
commun
ication
devices).
Patients3–75
y.7RC
Tsin
children,
2in
adults;4
RCTs
inurbanor
inne
r-city
popu
latio
ns.
Narrativeanalysis
1of
4im
proved
hospitalisation,and
1of
5redu
cedun
sche
duledcare.
5of
9stud
iesfoun
dstatistical
improvem
entsin
asthmasymptom
scomparedto
control.
Chang
2010
[29]**
1RC
T113participants
R-AMSTAR40
RCT2010
Educationby
IHWs
vs.edu
catio
nno
IHW.
FU:12mo
Impact:
Target:Ethnicgrou
psInitialclinical
consultatio
n,reinforced
byho
me
visitsfro
matraine
dIHW.Personalised,
child-friend
ly,culturally
approp
riate
education
materials.
African-American
and
Hispanic
commun
ities.
Children1–17
y;mean~7y.
Narrativeanalysis
Therewas
noeffect
onho
spitalisations
(OR1.58,95%
CI
0.37–6.79)
orA&E
attend
ances(OR
0.30,95%
CI−
0.17
to0.77;1
RCT).
Daysabsent
from
scho
olwere
redu
cedby
21%
intheinterven
tion
grou
p(95%
CI5–36%
;1RC
T).
Carer
asthmaQoL
was
not
sign
ificantlydifferent
(MD0.25,95%
CI−
0.39
to0.89).
Coffm
an2009
[30]**
18asthmaRC
Ts8077
participants
R-AMSTAR29
RCTs
1987-2007
Scho
ol-based
asthma
educationvs.usualcare.
Impact:
Target:Schoo
lchildren
Scho
ol-based
educationon
asthma,
med
ication,
mon
itorin
g,avoiding
trigge
rs.D
elivered
bynu
rses,health
educators,pe
ercoun
sellors,teachers,
±compu
ter
prog
rammes.
Children4–17
y.Severity:mild
tosevere,m
ajority
were
Blackor
Latin
o.
Narrativeanalysis
Unsched
uled
healthcare
was
not
repo
rted
.Scho
olabsences
sign
ificantly
redu
cedin
5of
13RC
Ts.D
ayswith
symptom
swereredu
cedin
3of
8RC
Ts.N
ightswith
symptom
sim
proved
in1of
4RC
Ts:1
foun
dim
provem
entin
thecontrolg
roup
.QoL
improved
in4of
6RC
Ts.
Gibson2002
[31]***
36RC
Ts6090
participants
R-AMSTAR39
RCTs
1986
–2001
Self-managem
ent
prog
rammes
vs.usualcare.
Impact:
Com
pone
nts:Regu
lar
review
Con
text:LTC
care
Education(100%);
self-mon
itorin
gof
symptom
sor
PEF
(92%
);regu
larreview
byamed
icalpracti-
tione
r(67%
);PA
AP
(50%
).Subg
roup
analyses
basedon
theseservicemod
els.
Adu
ltsandchildren.
Rang
eof
settings,
includ
ingho
spital,
emerge
ncyroom
,ou
tpatients,
commun
itysetting,
gene
ralp
ractice.
Meta-analysis
Subg
roup
analysis
Self-managem
entredu
ced
hospitalisations
(RR0.64,95%
CI
0.50–0.82;12
RCTs),A&E
visits(RR
0.82,95%
CI0.73–0.94;13RC
Ts]and
unsche
duledconsultatio
ns(RR0.68,
95%
CI0.56–0.81;7
RCTs).
Self-managem
entredu
ceddays
off
work/scho
ol(RR0.79,95%
CI0.67–
0.93;7
RCTs)andim
proved
QoL
(SMD0.29,95%
CI0.11–0.47;6
RCTs).
Optim
alself-managem
ent(sup
-po
rted
byaPA
APandregu
lar
Pinnock et al. BMC Medicine (2017) 15:64 Page 7 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
review
)redu
cedho
spitalisations
(RR
0.58,95%
CI0.43–0.77;9
RCTs),and
A&E
visits(RR0.78,95%
CI0.67–0.91;
9RC
Ts).
Gibson2004
[32]***
26RC
Ts6090
participants
R-AMSTAR39
RCTs
1987–2002
Differen
tcompo
nents
ofwrittenPA
APs
vs.usualcare.
Com
pone
nts:PA
APs
Com
pletePA
APs
specified
whe
n/ho
wto
increase
treatm
ent
(n=17);incomplete
omitted
advice
onincreasing
ICS(n=4);
non-specific(n=5)
onlyhadge
neral
instructions.
Adu
ltsandchildren.
Variety
ofsettings,
includ
ingho
spital,
emerge
ncyroom
,ou
tpatients,
commun
itysetting,
gene
ralp
ractice.
Actionpo
ints
%pred
ictedvs.%
best
Treatm
entadvice
Non
-spe
cific
plans
Bene
fitswerefoun
dforanynu
mbe
rof
actio
npo
ints(2
to4).
Both
%pred
ictedand%
best
redu
cedho
spitalisations,b
uton
ly%
person
albe
stredu
cedA&E
visits.
PAAPs
which
includ
edadvice
onincreasing
ICSandstartin
goral
steroids
redu
cedho
spitalisations
and
A&E
visits.
Efficacyof
incompleteandno
n-specificPA
APs
was
inconclusive.
Mou
llec2012
[33]**
18RC
Ts3006
participants
R-AMSTAR27
RCTs
1990–2010
Interven
tions
toim
prove
inhaledsteroidadhe
rence
vs.usualcare.
FU(m
ode):12mo,rang
e0.25–24mo
Con
text:LTC
care
Allstud
iesinclud
edself-managem
ent;
someinclud
edcom-
pone
ntsof
CCM:deci-
sion
supp
ort,de
livery
system
design
,clinical
inform
ationsystem
s.
Mod
erateto
severe
asthma(one
RCT
includ
edCOPD
).Age
d35–50y.Wom
enover-rep
resented
.
Meta-analysis
Effect
size
foradhe
renceto
ICS
comparedby
numbe
rof
compo
nentsof
theCCM
inthe
stud
y:1CCM
compo
nent
(n=13):sm
allES
0.29
(95%
CI0.16–0.42)
2CCM
compo
nents(n=5):large
ES0.53
(95%
CI0.40–0.66)
3CCM
compo
nents(nostud
ies)
4CCM
compo
nents(n=4)
very
largeES
0.83
(95%
CI0.69–0.98).
New
man
2004
[34]**
18asthmaRC
Ts(of6
3RC
Ts)
2004
participants
R-AMSTAR23
RCTs
1997
–2002
Self-managem
ent
interven
tions
vs.stand
ardcare/basic
inform
ation.
Impact:
Individu
al/group
interven
tions,focused
onsymptom
mon
itorin
g,trigge
ravoidanceand
adhe
renceto
med
ication.Afew
used
techniqu
esto
addressbarriersto
effectiveself-
managem
ent.
Adu
ltswith
3LTCs
(includ
ingasthma).
Narrativeanalysis
andcomparison
betw
een
interven
tions
7of
11stud
iesrepo
rted
aredu
ction
inun
sche
duledhe
althcare.
6of
12stud
iesrepo
rted
improved
QoL.
3of
8stud
iesrepo
rted
redu
ctions
inseverityof
symptom
s,allu
sed
educationandactio
nplans.
8of
14repo
rted
improved
adhe
rence.
Postma2009
[35]**
7RC
Ts2316
participants
R-AMSTAR23
RCTs
2004–2008
CHWs
vs.usualcare.
FU(m
ode):12mo,
rang
e4–24
mo
Impact:
Target:Ethnicgrou
ps,
children
CHWsfro
mthesame
commun
ityas
participants.
Educationon
asthma,
lifestyleandtrigge
ravoidance,with
resourcesto
redu
ceallergen
expo
sure.
Children5–9ywith
allergiesandlow-
income.Mainly
African-American
and
Hispanic.
Narrativereview
3of
6stud
iesrepo
rted
redu
ced
hospitalisationandredu
ced
unsche
duledconsultatio
ns.
4of
6repo
rted
redu
cedA&E
attend
ances
‘Con
sisten
tandsign
ificant
decrease
incaregiver-repo
rted
asthmasymp-
tomsam
onginterven
tionsubjects
comparedwith
controlsub
jectsin
6stud
ies.’
Powell2009[36]***
15RC
TsCom
pone
nts:PA
AP,
regu
larreview
Self-
vs.p
hysician
adjustmen
tof
Adu
ltswith
asthma
recruitedfro
ma
Self-
vs.p
hysician
managem
ent
Of6stud
ies:4repo
rted
nodifferencein
hospitalisation,1
Pinnock et al. BMC Medicine (2017) 15:64 Page 8 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
2460
participants
R-AMSTAR34
RCTs
1990–2001
Self-managem
entvs.
physician-review
edmanagem
ent.
Com
parison
ofmod
ified
PAAPs.
Con
text:LTC
care
med
ication
(n=6stud
ies).
PEFvs.sym
ptom
sPA
APs
(n=6).
Other
variatio
ns(n=3).
rang
eof
prim
ary,
commun
ity,A
&Eand
second
arycare.
Symptom
svs.PEF-
mod
ified
PAAPs
repo
rted
nodifferencein
A&E
visits,
3repo
rted
inconsistent
effectson
unsche
duledconsultatio
ns.
Of6stud
ies,6repo
rted
nodifferencein
hospitalisation,5
repo
rted
inconsistent
effectson
A&E
visits.
Omittingregu
larreview
(1RC
T)or
redu
cing
intensity
ofed
ucation
(1RC
T)increasedun
sche
duled
consultatio
ns.Verbal(vs.w
ritten)
PAAPs
hadno
effect
onho
spitalisations
orA&E
visits(1
RCT).
Ring
2007
[37]***
14RC
Ts4588
participants
R-AMSTAR35
RCTs
1993–2005
Interven
tions
encouraginguseof
PAAPs
vs.usualcare.
Con
text:O
rganisation
ofcare
Interven
tions
prom
otingPA
AP
owne
rshipor
use.
Diverse
interven
tions
(edu
catio
nal,
prom
pting,
asthma
clinics,asthma
managem
entsystem
s,qu
ality
improvem
ent).
Adu
ltsor
children
with
mod
erateto
severe
asthma;some
post-exacerbation.
Narrativeanalysis
4of
5stud
iesof
education,1of
2stud
iesof
teleph
oneconsultatio
ns,1
of2stud
iesof
asthmaclinicsand1
of2stud
iesof
asthmamanagem
ent
system
srepo
rted
increasedPA
AP
owne
rship.
1stud
yof
self-managem
ente
ducation,
1of
2stud
iesof
teleph
oneconsulta-
tions
and1of
2stud
iesof
asthma
managem
entsystemsincreased
understand
ing/useof
PAAPs.
Tapp
2007
[38]***
13RC
Ts2157
participants
R-AMSTAR39
RCTs
1979–2009
Asthm
aed
ucation
atA&E
visit
vs.usualcare.
FU(m
ode):6
mo,
rang
e2–18
mo
Impact:
Target:PostA&E
attend
ance
Asthm
aed
ucation
provided
byasthma
orA&E
nurses
with
inaweekof
A&E
visit
includ
edPA
APs,
trigge
rs,m
onito
ring,
inhalersand
med
ication.
Adu
ltsrecruited
durin
gA&E
attend
ance.
Meta-analysis
Narrativeanalysis
Theinterven
tionredu
cedho
spital
admission
s(RR0.50,95%
CI0.27–
0.91;5
RCTs),A&E
visits(RR0.66,95%
CI0.41–1.07;8
RCTs).
Effect
onQoL
(2RC
Ts)w
asinconsistent.The
rewas
noeffect
ondays
offwork/scho
ol.
Toelle2004
[39]**
7RC
Ts967participants
R-AMSTAR38
RCTs
1990–2001
WrittenPA
AP
vs.noplan.
Symptom
vs.
PEF-basedPA
AP.
FU(m
ode):12mo,
rang
e6–12
mo
Com
pone
nts:PA
AP
Peak
flow-based
writtenPA
APor
symptom
-based
writtenPA
APde
liv-
ered
inprim
aryor
tertiary
care.
Adu
lts28–45yand
childrenin
1RC
T.Meta-analysis
Subg
roup
analysis
Unsched
uled
healthcare:assessedin
1RC
T,no
trepo
rted
bysystem
atic
review
.Nodifferencebe
tweensymptom
andpe
akflow-based
PAAPs
inho
spitalisations
(RR1.17,95%
CI
0.31–4.43;3RC
Ts)or
A&E
atten-
dances
(RR1.17,95%
CI0.31–4.43;3
RCTs).
Symptom
-based
PAAPs
weremore
effectiveat
redu
cing
unsche
duled
consultatio
ns(RR1.34,95%
CI
1.01–1.77;2RC
Ts).
Welsh
2011
[40]***
12RC
Ts2342
participants
R-AMSTAR41
Hom
e-based
self-managem
ent
vs.rou
tinecare
orge
neral
education.
Impact:
Target:C
hildren
Lang
uage
-app
ropriate
education(asthm
a,trigge
rs,m
edication,
inhalers,self-
Children(m
ostly
<12
y)recruitedfro
mrecent
healthcare
visit.
Mainlyethn
icand/or
Meta-analysis
Narrativeanalysis
Nodifferencebe
tweengrou
psin
meannu
mbe
rof
A&E
visits(M
D0.04,
95%
CI−
0.20
to0.27;2
RCTs).
Pinnock et al. BMC Medicine (2017) 15:64 Page 9 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
RCTs
1986–2010
FU(m
ode):12mo,rang
e6–24
mo
managem
entwith
PAAPs).Alsoho
me-
work,techno
logy
de-
vices,24-hou
rho
tline
.
deprived
commun
ities
inUSA
.2of
5stud
iesrepo
rted
hospitalisation:on
efoun
da
redu
ctionandon
ean
increase
inthe
interven
tiongrou
p.Effect
onA&E
visits(6
RCTs)was
inconsistent.
Overallno
effect
onQoL
was
foun
din
5stud
ies.
Bravata2009
[41]***
63RC
Ts13,476
participants
R-AMSTAR40
RCTs
1966–2006
Self-managem
entQIvs.
othe
rQIstrateg
ies.
Impact:
Target:C
hildren
Self-mon
itorin
gor
self-managem
ent.Pa-
tient/careg
iver
educa-
tion.Provider
education.Organisa-
tionalchang
eandin-
terven
tions
with
multip
leQIstrateg
ies.
Children<18
y.Meta-analysis
Interven
tions
targetingparents/
caregiversredu
cedho
spitalisation
ratesby
1.2%
peryear
(95%
CI0.1–
2.4;n=5).
Self-managem
entinterven
tion
stud
iesim
proved
symptom
-free
days
by2.8%
(95%
CI0.6–5.0),which
equalled0.8days
permon
th(n=7);
andredu
cedmon
thlyscho
olabsen-
teeism
by0.4%
(95%
CI0–0.7),which
equalled0.1daype
rmon
th(n=16).
Long
erdu
ratio
nof
interven
tionin-
creasedtheeffect
onscho
olabsences.
Den
ford
2014
[43]***
38RC
Ts7883
participants
R-AMSTAR36
RCTs
1993–2000
Asthm
aself-care
vs.usual/le
ssintensive
interven
tion.
FU(m
ode):12mo,rang
e3–18
mo
Impact:
Com
pone
nts:
Behaviou
rchange
Com
mon
est
behaviou
ralchang
etechniqu
esinclud
ing:
self-mon
itorin
g(n=
30),instruction(n=
27),go
al-settin
g(n=
26)andinhalertech-
niqu
e(n=24).
Adu
lts≥18
ywith
adiagno
sisof
asthma.
Meta-analysis
Interven
tiongrou
pparticipantshad
redu
cedasthmasymptom
s(SMD
−0.38,95%
CI−
0.52
to0.24;27RC
Ts)
andun
sche
duledhe
althcare
use(OR
0.71,95%
CI0.56–0.9;23
RCTs).
Increasedadhe
renceto
preven
tative
med
icationcomparedto
control(OR
2.55,95%
CI2.11–3.10;16RC
Ts).
deJong
h2012
[42]**
1asthmaRC
T(of4)
16participants
R-AMSTAR35
RCTs
1993–2009
Mob
ileph
onemessaging
forself-
managem
entvs.
usualcare.
FU:range
4–12
mo
Com
pone
nts:Mob
ileph
onemessaging
Self-managem
ent
interven
tions
delivered
bymob
ileph
onemessaging
.
Participantsof
all
ages,g
ende
ror
ethn
icity.
Includ
edanyLTC
(one
asthmastud
y).
Narrativesynthe
sis
Inthesing
leasthmastud
y,there
werefewer
admission
s(2
vs.7)bu
tmoreun
sche
duledconsultatio
ns(21
vs.15)
intheinterven
tiongrou
pcomparedto
theusualcaregrou
p.Thepo
oled
asthmasymptom
score
show
edasign
ificant
difference
betw
eengrou
ps,favou
ringthe
interven
tiongrou
p(M
D−0.36,95%
CI−
0.56
to−0.17).
Kirk
2012
[44]**
10asthmaRC
Ts2195
participants
R-AMSTAR23
RCTs
1995–2010
Self-care
supp
ort
vs.usualcare.
FU(m
ode):12mo,rang
e3–24
mo
Impact:
Target:C
hildren
Interven
tions
aiming
tohe
lpchildrentake
controlo
fand
managetheir
cond
ition
,promote
theircapacity
forself-
care
and/or
improve
theirhe
alth.
Children≤18
ywith
aLTC:
asthma(10RC
Ts),
cysticfib
rosis(2)or
diabetes
(1).
Narrativesynthe
sis
Of8RC
Ts,2
repo
rted
fewer
asthma
admission
s,5repo
rted
fewer
A&E
attend
ancesand2of
3repo
rted
fewer
unsche
duledconsultatio
ns.
Con
trol
improved
in5of
8RC
Ts.
Qol
improved
in2of
5RC
Ts.
Marcano
Belisario
2013
[45]**
Self-managem
entapps
Com
pone
nts:
Smartpho
neApp
sSelf-managem
ent
supp
ortinterven
tions
Adu
ltswith
clinician-
diagno
sedasthma.
Narrativesynthe
sis
Of2RC
Ts,2
repo
rted
nodifference
inho
spitaladm
ission
s;1repo
rted
Pinnock et al. BMC Medicine (2017) 15:64 Page 10 of 32
Table
3Summarytableof
finding
sof
PRISMSsystem
aticreview
sandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
2RC
Ts408participants
R-AMSTAR39
RCTs
2000–2013
vs.traditio
nalself-
managem
ent.
FU:6
mo
provided
bysm
artpho
neapp.
fewer
A&E
attend
ancescomparedto
control;1foun
dno
differencein
unsche
duledGPconsultatio
nsor
out
ofho
ursconsultatio
ns,b
utredu
ced
prim
arycare
nurseconsultatio
ns;1
repo
rted
nodifferencein
MDin
Asthm
aCon
trol
Questionn
aire
scores
betw
eentheinterven
tionand
controlg
roup
at6mon
ths;1foun
dim
proved
QoL
intheinterven
tion
grou
p.
Press2012
[46]***
5RC
Ts(of15
stud
ies)
1459
participants
R-AMSTAR34
RCTs
1950–2010
Interven
tions
targeted
atethn
icminority
grou
psvs.usualcare.
FU(m
ode):6
mo,rang
e0.25–32mo
Impact:
Target:Ethnic
grou
ps
Interven
tions
targetingethn
icpo
pulatio
nsin
US.15
wereed
ucation-
based,
9weresystem
-levelinterventions,5
werecultu
rally
tai-
loredand
commun
ity-based
,10
wereho
spital-b
ased
.
Adu
lts≥18
y.Ethn
icminority
grou
ps:
African-Americans
(10stud
ies,Latin
os(4
stud
ies).
Narrativesynthe
sis
Aned
ucationinterven
tionredu
ced
A&E
attend
ance
in2of
4RC
Tsand
hospitaladm
ission
sin
2of
3RC
Ts.
Symptom
swereno
tredu
cedin
any
ofthe3RC
Tsthat
measuredcontrol.
QoL
was
improved
in3of
4RC
Tsthat
used
anasthma-relatedQoL
outcom
e.
Stinson2009
[47]*
4asthmaRC
Ts(of9
stud
ies)
826asthma
participants
R-AMSTAR28
RCTs
1993–2008
Internet-based
self-
managem
entvs.usual
care.
FU(m
ode):12mo,rang
e3–12
mo
Target:C
hildren
Com
pone
nts:
Internet-based
Any
Internet-based
oren
abledself-
managem
ent
interven
tion.
Children6–12
yor
adolescents13–18y
with
LTCs:asthm
a(4
RCTs),pain
(1),
encopresis(1),brain
injury
(1)or
obesity
(1).
Narrativesynthe
sis
1RC
Trepo
rted
nodifferencein
hospitalisations
comparedto
control,
1RC
Trepo
rted
sign
ificant
redu
ctions
inA&E
visitsand1of
2RC
Tsshow
edfewer
unsche
duledconsultatio
ns.
4ou
tof
4repo
rted
sign
ificant
improvem
entin
ameasure
ofcontrol.
1of
4asthmaRC
Tsrepo
rted
asign
ificant
bene
fiton
QoL.
Abb
reviations:A
&Eaccide
ntan
dem
erge
ncy,CC
Mchroniccare
mod
el,C
HW
commun
ityhe
alth
workers,C
Icon
fiden
ceinterval,C
OPD
chronicob
structivepu
lmon
arydisease,
ESeffect
size,FUfollow-up,
ICSinha
led
corticosteroid,IHW
indige
nous
healthcare
workers,LTC
long
-term
cond
ition
,MDmeandifferen
ce,m
omon
ths,ORod
dsratio
,PAAPpe
rson
alised
asthmaactio
nplan
,PEF
peak
expiratory
flow,Q
Iqua
lityim
prov
emen
t,QoL
quality
oflife,
RRriskratio
,SMDstan
dardised
meandifferen
ce,W
MDweigh
tedmeandifferen
ce,y
years
Pinnock et al. BMC Medicine (2017) 15:64 Page 11 of 32
Table
4Summarytableof
finding
sof
update
rand
omised
controlledtrialsandtheirrelevanceto
themeta-review
questio
nsReferenceandweigh
ting;
Participants,n
;Riskof
bias
Com
parison
Relevanceto
meta-review
questio
ns:
Stud
ytype
and
interven
tions
includ
edTarget
grou
p(s)
Mainresults
[1o]isthede
fined
prim
aryou
tcom
e
Whatistheim
pact?
Target
grou
ps?
Which
compo
nents?
Con
text?
Al-She
yab2012
[48]
n=261
HIGHriskof
bias
Ado
lescen
tAsthm
aActionprog
rammevs.
standard
care.
FU:3
mo
Target:A
dolescen
tsCom
pone
nts:Peer
education
Cluster
RCT.
TripleA.Peerleadersfro
myear
11weretraine
dto
deliver
prog
rammeto
years8,9and10.
Ado
lescen
tsin
Jordanian
high
scho
ol.Igrou
phad
fewer
females,few
ersymptom
sandhigh
erEnglishproficiency.
Com
paredto
controlimprovem
ents
QoL
scoreim
proved
[I:5.42
(SD0.14)
vsC:4.07(SD0.14)MD1.35
(95%
CI
1.04–1.76)].
Baptist2013
[49]
n=70
HIGHriskof
bias
Person
alised
asthma
self-regu
latio
ninter-
ventionvs.edu
catio
nsession.
FU12
mo
Target:O
lder
adults
Com
pone
nts:Health
educator
RCT.
6-sessionprog
ramme
(group
teleph
one).
Patientsselected
anasthma-specificgo
al,and
addressedpo
tential
barriers.
Con
trol
issing
lesession
basiced
ucation+2
teleph
onecalls.
Age
d≥65
y.Ph
ysician
diagno
sisof
asthma,no
restrictio
nin
severity.
Majority
Caucasian.
Nobe
tween-grou
pdifferences
inA&E
visitsor
hospitalisations.H
ealth
care
utilisatio
nwas
lower
at6mobu
tno
t12
mo.ACQwas
similarat
1moand
6mo.At12
mo,Ip
articipantswere
4.2tim
esmorelikelyto
have
anACQ
score<0.75.
[1o]QoL
(mAQLQ
)was
sign
ificantly
high
erin
theIthanin
Cat
alltim
epo
ints(1,6
and12
mo).
Ducharm
e2011
[50]
n=219
LOW
riskof
bias
‘Take-ho
meplan’p
ost
A&E
visitwith
PAAP+
prescriptio
ninform
ationvs.
prescriptio
nbu
tno
PAAP/inform
ation.
FU:28days
Target:C
hildren,
A&E
attend
ees
Com
pone
nts:PA
APwith
prescriptio
n
RCT.
Interven
tioniswritten
PAAPwith
a‘form
atted’
prescriptio
nforICS(i.e.
includ
inginform
ation
abou
tuse)
issued
byA&E
doctor
ondischarge
followingasthma
exacerbatio
n.
Canadianchildren1–17
yrecruiteddu
ringA&E
attend
ance
foracute
asthma(78%
wereun
der
theageof
6y).
Nobe
tween-grou
pdifferences
inun
sche
duledcare
at28
days.C
om-
paredto
control,at
28days
children
giventhePA
APhadbe
tter
asthma
control(prop
ortio
nwith
Asthm
aQuiz
Score<2I:58%
vs.C
:41%
;RR1.36,
95%
CI1.04–1.86).
Nobe
tween-grou
pdifferences
inchild/careg
iver
QoL
at28
days.
[1o]Adh
eren
ceto
ICSde
clined
from
90%
(day
1)to
50%
atday14,w
ithno
sign
ificant
grou
pdifference.
Goe
man
2013
[51]
n=114
Low
riskof
bias
Person
-cen
tred
educationvs.w
ritten
inform
ation.
FU:12mo
Target:O
lder
adults
Com
pone
nts:Person
alised
education
RCT.
Person
allytailored
educationsessionwith
asthmaed
ucator
basedon
respon
sesto
aqu
estio
nnaire;inh
aler
techniqu
e.
≥55
y,commun
ity-based
asthmaticswith
norestric-
tionin
asthmaseverity.
[1o]At12
moIp
articipantshadbe
tter
asthmacontrolthanC(ACQMD0.3,
95%
CI0.06–0.5,p=0.01)andbe
tter
asthma-relatedQoL
(p=0.01).
Nosign
ificant
differencein
numbe
rof
steroidcourses(p=0.17).
At12
mo,moreIp
articipants(n=36,
61%)ow
nedaPA
APcomparedto
C(n=21,38%
;p=0.015).
[1o]Similaradhe
renceto
ICSat
12mo
(p=0.015).
Halterm
an2014
[52]
n=638
LOW
riskof
bias
Person
alised
prom
pts
forclinicians
and
parents,practice
training
andfeed
back
vs.w
rittengu
idelines.
FU:6
mo
Target:C
hildren,
deprived
commun
ities
Com
pone
nts:Feed
back
Con
text:C
ommun
ity-
based,
clinicaltraining
Cluster
RCT.
Interven
tionpractices
received
person
alised
clinicianandparent
prom
pts+blankPA
AP;
practicetraining
;feedb
ack.
Con
trol
practices
sent
guidelines.
Urban,p
rimarycare
practices
inde
prived
commun
ities.
Parents/children2–12
ywith
persistent,p
oorly
controlledasthma.
Recruitedfro
mwaitin
groom
over
4ystudy.
11%
inbo
thgrou
pshadan
A&E
visit
orho
spitalisation.
[1o]Com
paredto
controlp
ractices,at
2mochildrenin
thePA
IR-UPpractices
hadmoresymptom
-free
days
[I:10.2days/2
weeks
(SD4.8)
vs.
C:9.5days/2
weeks
(SD5.1);M
D0.78,
95%
CI0.29–1.27]bu
tthedifference
was
notsign
ificant
at6mo.
Nightswith
symptom
sremaine
dsign
ificant
at6mo[I:1.4(SD3.0)
vs.
Pinnock et al. BMC Medicine (2017) 15:64 Page 12 of 32
Table
4Summarytableof
finding
sof
update
rand
omised
controlledtrialsandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
C:1.8(SD3.2);M
D−0.43;95%
CI
−0.77
to−0.09].
Horne
r2014
[53]
n=183
UNCLEARriskof
bias
Asthm
aplan
forkids
vs.teachingon
gene
ralh
ealth
and
well-b
eing
.FU
:12mo
Target:C
hildren,
rural
commun
ities
Cluster
RCT.
Prog
rammede
livered
in16
×15
min
sessions,
3days/w
eekfor5.5weeks,
byscho
olnu
rses
durin
glunchbreak+ho
mevisit.
Grade
s2–5(age
s7–11
y)with
physiciandiagno
sisof
asthma.
Nobe
tween-grou
pdifferencefor
admission
sor
A&E
visits.
Nobe
tween-grou
pdifferencein
QoL
scores.
Inhalerskillim
proved
inthe
interven
tiongrou
pcomparedto
controlafter
4mo,with
repo
rted
high
erself-efficacy.
Joseph
2013
[54]
n=422
UNCLEARriskof
bias
Web
-based
asthma
managem
ent
interven
tionvs.
control.
FU:12mo
Target:A
dolescen
ts,urban
deprived
,ethnicgrou
psCom
pone
nts:Web
-based
,be
haviou
ralchang
e
RCT.
Internet-based
prog
ramme
targeted
atAfrican-
Americans/urban
adolescentswith
traits
(low
motivation;
low
perceivedem
otional
supp
ort;resistance
tochange
;reb
elliousne
ss).
Grade
s9–12
(age
s14–18
y)with
physiciandiagno
sis
ofasthmaandrepo
rt>4days
ofrestricted
activity
inthepast30
days
atbaseline.
Nodifferencein
repo
rted
A&E
visits/
hospitalisations
at12
mo.
[1o]Com
paredto
C,at12
motheI
participantshadfewer
symptom
-days
(RR0.8,95%
CI0.6–1.0).
Nodifferencein
nigh
tswith
symptom
s,scho
oldays
missed,
days
ofrestrictedactivity
ordays
hadto
change
plans.
Stud
entscharacterised
with
rebe
lliou
snessor
low
perceived
emotionalsup
portrepo
rted
fewer
symptom
-days.
Khan
2014
[55]
n=91
HIGHriskof
bias
Asthm
aed
ucation+
individu
alised
written
PAAPvs.asthm
aed
ucation(excluding
PAAP).
Target:Ethnicgrou
psCom
pone
nts:Written
PAAP
RCT.
Both
grou
psreceived
individu
alasthma
educationdu
ringan
OPD
visitfro
mapaed
iatrician+
mon
thlyFU
.Intervention
grou
ptraine
din
usinga
PAAP.
1–14
y.RecruitedviaA&E
OPD
with
partlycontrolled
asthma(daytim
eor
nocturnalsym
ptom
s,activity
limitatio
n,lung
functio
n<0%
bestor
exacerbatio
nin
previous
year).
[1o]Tren
dforim
proved
outcom
esat
6mobu
tno
sign
ificant
betw
een-
grou
pdifferencein
prop
ortio
nof
childrenattend
ingA&E
(I:36%
vs.
C:52%
;p=0.141).
Therewas
nobe
tween-grou
pdifferencein
unsche
duleddo
ctor
visits,asthm
aattacks,missedscho
oldays
ornigh
t-tim
eaw
aken
ings.
Rhee
2011
[56]
n=112
UNCLEARriskof
bias
Peer-ledasthma
educationprovided
bype
ersat
adaycamp
vs.adu
lt-ledcamp.
Target:A
dolescen
ts.
Com
pone
nts:Peer
leaders
RCT.
Asthm
aself-managem
ent
skills+psycho
socialskills
taug
htat
adaycampby
peer
leaders+mon
thly
peer
teleph
onecontact.
Con
trol:Sim
ilared
ucation
delivered
byadults.N
oteleph
one.
13–17y(includ
inglow-
incomefamilies).Mild/
mod
erate/severe
asthma.
Asthm
adiagno
sisfor1y.
Ableto
unde
rstand
spoken
andwritten
English.
[1o]Bo
thgrou
psrepo
rted
sign
ificantly
increasedQoL
over
time(F=4.31,
p=0.002),w
ithIg
roup
having
sign
ificantlyhigh
erQoL
at6mo(M
D11.38,95%
CI0.96–21.79,p=0.03)and
9mo(M
D12.97,95%
CI3.46–22.48,p
=0.008).
Both
grou
psrepo
rted
improved
attitud
eto
asthma(F=11.94,p=
0.001),w
ithgreaterim
provem
entin
Iat
6mo(M
D4.11,95%
CI0.65–7.56,
p=0.02).
Rikkers-Mutsaerts2012
[57]
n=90
UNCLEARriskof
bias
Internet-based
self–
managem
entvs.usual
care.
FU:12mo
Target:A
dolescen
ts.
Com
pone
nts:Internet-
based
RCT.
Internet-based
self-
mon
itorin
gwith
algo
rithm
-based
advice.
Prog
rammeinclud
eded
ucation(web
-based
+grou
p),self-m
onito
ring
(FEV
1+ACQ),PA
APand
3–6moreview
.
12–18ywith
mild
tosevere
persistent
asthma
onregu
larICSmed
ication
andpo
orlycontrolledat
recruitm
ent.
Nobe
tween-grou
pdifferences
inexacerbatio
ns,p
hysicians’visitsor
teleph
onecontacts.
[1o]QoL
was
better
inIg
roup
at3
mo(PAQLQ
I:6.00
vs.C
:5.68;MD0.40,
95%
CI0.17–0.62)bu
tno
tat
12mo(I:
5.93
vs.C
:6.05;MD0.05,95%
CI0.50–
0.41).
Asthm
acontrolw
asim
proved
inI
grou
pat
3mo(ACQI:0.96
vs.C
:1.19;
MD−0.32,95%
CI−
0.56
to−0.08)bu
t
Pinnock et al. BMC Medicine (2017) 15:64 Page 13 of 32
Table
4Summarytableof
finding
sof
update
rand
omised
controlledtrialsandtheirrelevanceto
themeta-review
questio
ns(Con
tinued)
notat
12mo(I:0.83
vs.C
:0.79;MD
−0.05,95%
CI−
0.35
to0.25).
Shah
2011
[58]
150GPs
and201children
LOW
riskof
bias
GPtraining
(PACE
stud
y)vs.n
otraining
.FU
:12mo
Targets:Children
Com
pone
nts:GPtraining
Cluster
RCT.
GPs
participated
in2×3-h
worksho
pson
commun
icationand
educationstrategies
tofacilitatequ
ality
asthma
care.
150GPs
and221children
with
asthmain
theircare.
Nobe
tween-grou
pdifferencein
hospitalisation/A&E
visits(I:18%
vs.
C:12%
;differen
ce6%
,95%
CI−
4to
15).
Nobe
tween-grou
pdifferences
inscho
olabsenceor
parent
absenteeism
forchild’sasthma.
[1o]Morepatientsin
Igroup
GPs
had
aPA
AP(I:61%
vs.C
:46%
;differen
ce15%,95%
CI2–28).
vanGaalen2013
[59]
n=107
HIGHriskof
bias
Internet-based
self–
managem
entvs.
control(FU
ofSM
ASH
INGtrial).
FU:30mo
Target:A
dults
Com
pone
nts:Internet-
based
RCT(FUstud
y).
Education+PA
AP,
self-mon
itorin
gand
regu
larreview
.The200patientsin
original12-m
otrialw
ere
invitedforFU
after18
mo.
Adu
ltswith
asthmaaged
18–50y,usingICS.
107/200(54%
)participated
:Igrou
p:47/
101(47%
);Cgrou
p:60/99
(61%
).ParticipantsACQwas
similar,bu
tAQLQ
was
greaterthan
inno
n-participants.
At30
moafterbaseline,therewas
aslightlyattenu
ated
improvem
entfor
both
QoL
(AQLQ
adjusted
betw
een-
grou
pMD0.29,95%
CI0.01–0.57)and
ACQ(adjustedMDof
−0.33,95%
CI
−0.61
to−0.05)scores
infavour
ofthe
interven
tion.
Nobe
tween-grou
pdifferences
inFEV 1.
Won
g2012
[60]
n=80
HIGHriskof
bias
Symptom
-based
writtenPA
APvs.verbal
coun
selling
.FU
:6mo
Target:C
hildren,
ethn
icgrou
psCom
pone
nts:Written
PAAP
Sing
leblinde
dRC
T.Interven
tionwas
symptom
-based
PAAP
givenou
tat
initialcontact.
Outcomes
measuredat
baseline,3,6and9mo.
Malaysian
children(m
ixof
Malay,C
hine
seandIndian)
with
allseverities
ofasthma.Age
d6–17
y.Recruitm
entprocessno
tde
scrib
ed.
At6motherewas
nodifferencein
A&E
visits/unsched
uled
care
[interven
tion4(SD10.8)vs.con
trol
6(SD21.1);p=0.35].
At6motherewas
nodifferencein
prop
ortio
ncontrolled(ACT≥20
I:81%
vs.C
:87%
;p=0.50),with
noexacerbatio
ns(ACT≥20
I:89%
vs.C
:82%;p
=0.62)or
inQoL
[mean
PAQLQ
I:6.11
(SD0.88)vs.6.11
(SD1.09);
p=0.99].
Abb
reviations:A
&Eaccide
ntan
dem
erge
ncy,ACQ
Asthm
aCon
trol
Que
stionn
aire,A
CTAsthm
aCon
trol
Test,A
QLQ
Asthm
a-relatedQua
lityof
Life
Que
stionn
aire,C
control,CI
confiden
ceinterval,FEV
1forced
expiratory
volumein
onesecond
,FUfollow-up,
GPge
neralp
ractition
er,I
interven
tion,
ICSinha
ledcorticosteroid,m
AQLQ
miniA
sthm
a-relatedQua
lityof
Life
Que
stionn
aire,M
Dmeandiffer-
ence,m
omon
ths,PA
APpe
rson
alised
asthmaactio
nplan
,PAQLQ
paed
iatricasthma-relatedqu
ality
oflife,QoL
quality
oflife,
RCTrand
omised
controlledtrial,RR
riskratio
,SDstan
dard
deviation,
yyears
Pinnock et al. BMC Medicine (2017) 15:64 Page 14 of 32
ResultsDescription of the studies in the meta-reviewFigure 1 illustrates the PRISMA flow chart for bothreviews. After removal of duplicates, 9633 referenceswere identified from the initial PRISMS search and anadditional 6321 from the update search. From these, 25systematic reviews [23–47] were included in the PRISMSmeta-review, representing data from 240 unique RCTs.The year of review publication ranged from 1995 to2013, and included RCTs dated from 1979 to 2013.In addition we included 13 RCTs published since thelast search dates of the included reviews (2010 forchildren, 2012 for adults and 2011 for ethnic groups;see Additional file 2 for details) [48–60]. (For claritywe refer to these as “update RCTs”.) A further twosystematic reviews (which included a further fourRCTs) [61, 62] and six RCTs [63–68] were added afterthe pre-publication update. The RECURSIVE study in-cluded 24 RCTs with publication dates from 1993 to2015 [49, 69–91].
After excluding overlap, this represents 270 unique tri-als undertaken in at least 29 high- or middle-incomecountries: Argentina, Australia, Belgium, Brazil, Canada,Chile, Denmark, Finland, France, Germany, India,Israel, Italy, Jordan, Malaysia, Malta, Netherlands, NewZealand, Norway, Russia, Spain, Sweden, Switzerland,Taiwan, Trinidad, Turkey, UK, USA and Venezuela.In the 18 systematic reviews that reported the duration
of follow-up in their included RCTs [23–25, 27–29, 33,35, 38–40, 42–47, 61], the modal duration (in 10 of thereviews) was 12 months, with only 3% of reported RCTsfalling outside the range of 3–24 months. The updateRCTs had a similar profile, with 6 of 13 update RCTshaving a duration of 12 months (range 3–30 months).
Study quality and weight of evidenceTaking into consideration both study quality and totalpopulation size, 10 PRISMS reviews received anevidence weighting of three stars [27, 31, 32, 36–38, 40,41, 43, 46], 13 were weighted two star [23–26, 29, 30,
Fig. 1 PRISMA flowchart. Note: The initial RECURSIVE search included all long-term conditions: papers reporting asthma randomised controlledtrials (RCTs) were identified from 184 studies included in the full RECURSIVE report [14]
Pinnock et al. BMC Medicine (2017) 15:64 Page 15 of 32
33–35, 39, 42, 44, 45] and two were weighted one star[28, 47]. Of the PRISMS update RCTs, four were judgedto be at low risk of bias [50–52, 58], five at high risk ofbias [48, 49, 55, 59, 60] and in four the risk of bias wasunclear [53, 54, 56, 57]. Allocation concealment wasjudged as adequate in six of the 24 asthma studiesincluded in the RECURSIVE review [74, 76, 80, 83–85].Study quality is indicated in the first columns of Tables 3,4 and 5, with details of the quality assessments inAdditional file 4.
Overview of presentation of resultsTables 3, 4, 5 and 6 provide summaries of the studiesincluded in the PRISMS meta-review, update RCTs, theRECURSIVE review and pre-publication update withmore detailed tables in Additional file 5.
Can supported self-management reduce the use ofhealthcare resources and improve asthma control?Use of healthcare resourcesFigure 2 is a meta-Forest plot illustrating the meta-analyses (including three PRISMS 3* reviews and RECUR-SIVE) that report relative risks of admissions, A&E atten-dances and/or unscheduled consultations [27, 31, 38].Treatment event rates from the meta-analyses are inTable 7. These results suggest similar effects in adults[38], children [27] and mixed populations [31].Hospitalisations were reported in 12 reviews [25–29,
31, 35, 38, 40, 41, 44, 46]. Six meta-analyses (four 3*,two 2*) showed that self-management support interven-tions led to fewer hospital admissions [25–27, 31, 38, 41].Six narrative reviews of variable quality, reporting hetero-geneous interventions, showed inconsistent effects onhospitalisations [28, 29, 35, 40, 44, 46].Ten reviews reported A&E attendances [25–27, 29, 31,
35, 38, 40, 44, 46]. Four meta-analyses (three 3* [27, 31,38], one 2* [26]) reported a reduction in A&E atten-dances in the self-management intervention comparedto control groups. Four narrative reviews (one 3* [46],three 2* [25, 35, 44]) showed a reduction in A&E atten-dances in at least half of their included RCTs; one 3*review showed inconsistent results [40], and one 2*review showed no benefit on A&E attendances [29].Of the eight reviews that reported unscheduled care
[24, 27, 28, 31, 34, 35, 43, 44], three 3* meta-analysesreported fewer unscheduled consultations in participantswho received a self-management intervention whencompared to control [27, 31, 43]. Furthermore, three 2*narrative reviews reported that self-managementreduced unscheduled care in at least half their includedtrials [34, 35, 44]. The remaining two small or poorquality reviews had inconsistent results [24, 28].
Asthma controlOf the 10 reviews that reported measures of control[24, 28, 30, 31, 34, 35, 38, 41, 44, 46], three meta-analyses (two 3* [31, 41], one 2* [24]) and three narrativereviews [28, 35, 44] reported a reduction in symptoms inparticipants who received self-management interventionscompared to control groups. The other four narrativereviews (two 3* [30, 34], two 2* [38, 46]) had inconsistentresults [30, 34, 38] or showed no benefit on symptomcontrol [46]. The broader concept of quality of life wasreported as improved in some reviews [25, 30, 34, 46], butnot others [27, 29, 40, 44].Six reviews reported a reduction in days missed
from school or work [24, 29–31, 38, 41]. Two 3*meta-analyses [31, 41], two small reviews each withonly one RCT [24, 29] and five of the 13 RCTs in a2* narrative synthesis of school-based interventions[30] concluded that self-management interventionsreduced absenteeism. A single RCT reported in a 3*narrative review in adults concluded that asthmaeducation following A&E attendance had no effect onabsenteeism [38].
In which target groups has supported self-managementbeen shown to work?The systematic reviews encompassed a broad range ofpopulations in diverse healthcare and demographic set-tings with consistently positive findings. For example,the reviews included all ages [28, 31] or only children[24, 26, 27, 29, 30, 35, 40, 41] or adults [34, 38, 43, 46].Some focused on lower socioeconomic groups [35, 40]or ethnic minority communities [25, 29, 35]. The reviewsand RCTs identified in the PRISMS update typically builton this extensive generic evidence base and investigatedinterventions targeting specific groups such as urban[52, 54], rural [53], deprived communities [46, 52, 54],cultural groups [46, 54, 55, 60], adolescents [48, 54, 56,57] or older adults [49, 51]. Table 8 summarises the keystrategies used in trials to tailor interventions, or theirmode of delivery, to different groups.
Cultural groupsFour reviews explored the impact of self-management incultural groups [25, 29, 35, 46]. A 2* meta-analysisreported that culture-specific programmes reducedhospitalisations in children and improved quality of life inadults compared to generic interventions [25]. A 3* narra-tive synthesis found only two RCTs testing culturally tai-lored interventions, one of which improved quality of life[46]. The involvement of community health workers re-duced use of healthcare resources in two thirds, and im-proved symptoms in all seven RCTs included in a 2*narrative review [35]. An inpatient visit from a lay educa-tor to Black or Latino children improved self-efficacy and
Pinnock et al. BMC Medicine (2017) 15:64 Page 16 of 32
Table
5Summarytableof
stud
iesinclud
edin
theRECURSIVEhe
alth
econ
omicanalysis
Reference;Cou
ntry;
Allocatio
nconcealm
ent
Stud
ytype
;Participants,n;
Interven
tion(s)
Com
parison
Target
grou
p(s)
Health
econ
omicresults
Form
alhe
alth
econ
omic
evaluatio
n,cost-effectiven
ess
(societaland
health
service
perspe
ctive)
Qualityof
life/
asthmacontrol
Health
care
utilisatio
n(hospitalisation)
Totalh
ealth
care
costs
Unsched
uled
care
Baptist2013
[49]
US
Con
cealmen
tno
tadeq
uate
RCT
n=70
Person
alised
6-session
self-regu
latio
ned
ucation.
Usualcare.
FU:12mo
Older
adultswith
asthma(>65
y).
Meanage74
y.14%
male.
Prop
ortio
nwith
ACQ<0.75
was
greaterin
Igroup
than
Cgrou
p[I:
13(41.9%
)vs.C
:5(15.6%
)].
Igroup
hadfewer
hospitalisations
(I:0
vs.C
:4;p
=0.04).
n/a
Nodifferencein
A&E
visits(I:1vs.
C:2;p
=0.58).I
grou
phadfewer
unsche
duledvisits
(I:6vs.C
:14;
p=0.048).
n/a
Castro2003
[69]
US
Con
cealmen
tno
tadeq
uate
RCT
n=96
Education,
psycho
socialsupp
ort,
PAAPandco-
ordinatio
nof
care.
Usual(private)
prim
arycare.
FU:12mo
Inpatients,adults
with
asthma.
Meanage38
y.15%
male.
Nobe
tween-
grou
pdifference
inmeanAQLQ
[I:4.0(SD1.3)
vs.C
:3.9(SD1.5);
p=0.55].
Igroup
hadfewer
re-adm
ission
s/patient
[I:0.4(SD0.9)
vs.C
:0.9(SD1.5);p
=0.04].
Igroup
hadlower
costs/patient
[I:$5726(SD5679)
vs.C
:$12,188
(SD19,352);MD
$6,462;p
=0.03].
Nobe
tween-
grou
pdifferences
innu
mbe
rA&E
visits/patient
[I:1.9
(SD4.3)
vs.C
:1.4
(SD=1.5);
p=0.52].
n/a
Clark
2007
[70]
US
Con
cealmen
tno
tadeq
uate
RCT
n=808
Self-regu
latio
ninterven
tion;nu
rse
teleph
one-de
livered
.
Usualcare.
FU:12mo
Adu
ltwom
enwith
asthma.
Meanage49
y.100%
female
Nobe
tween-
grou
pdifference
inmeanAQLQ
[I:2.1(SD0.9)
vs.C
:2.1(SD0.9].
Nobe
tween-grou
pdifferencein
admission
s/patient
[I:0.2(SD0.7)
vs.C
:0.1
(SD0.5)]
n/a
Igroup
had
greaterredu
ction
inun
sche
duled
visits[m
ean
change
:I:−
0.63
(SD2.4)
vs.C
:−0.24
(SD1.5)].
n/a
deOliveira
1999
[71]
Brazil
Con
cealmen
tno
tadeq
uate
RCT
n=52
Outpatient
education
prog
ramme,includ
ing
awrittenPA
AP.
Usualcare.
FU:6
mo
Adu
lts;m
oderateto
severe
asthma.
Meanage38
y.15%
male.
Nobe
tween-
grou
pdifferences
inQoL
score[I:28
(SD17)
vs.C
:50(SD15);
p=0.0005].
Nobe
tween-grou
pdifferences
inadmission
s/patient
[I:0vs.C
:0.5(SD0.8);
p=0.08].
n/a
Igroup
hadfewer
A&E
visits/patient
[I:0.7(SD1.0)
vs.
C:2
(SD2)].
n/a
Gallefoss
2001
[72]
Norway
Con
cealmen
tno
tadeq
uate
RCT
n=78
Group
-plusindividu
alself-managem
ent
educationwith
awrit-
tenPA
AP.
Usualprim
ary
care.
FU:12mo
Adu
ltswith
asthma.
Meanage44
y.21%
male.
Better
QoL
(SGRQ
)in
Igroup
at12
mo[I:20
(SD15)
vs.C
:36.5(SD18);
MD16.3,95%
CI
16.3–24.4]
n/a
Nobe
tween-
grou
pdifferences
intotalcosts(in
NOK)
[I:10,500
(SD20,500)v
s.C:
16,000
(SD35,400);
p=0.510].
n/a
Increm
entalSGRQ
gain
16.3;health
costsdifference
NOK1900;allcost
diffNOK−5500.
Gruffydd
-Jon
es2005
[73]
UK
Con
cealmen
tno
tadeq
uate
RCT
n=174
Targeted
nurse-led
teleph
onereview
s,in-
clud
ingPA
APs.
Usualprim
ary
care.
FU:12mo
Adu
ltswith
asthma.
Meanage50
y.40%
male.
Nobe
tween-
grou
pdifference
inmeanchange
inACQ[I:−0.11
(95%
CI−
032to
0.11)vs.C
:−0.18
(95%
CI−
0.38
to0.02);p=0.349].
n/a
Nobe
tween-
grou
pdifference
intotalcosts[I:
£209.85(SD
220.94)vs.C:
£333.85(SD
410.64);MD
£122.35;p=0.071].
n/a
n/a
Pinnock et al. BMC Medicine (2017) 15:64 Page 17 of 32
Table
5Summarytableof
stud
iesinclud
edin
theRECURSIVEhe
alth
econ
omicanalysis(Con
tinued)
Hon
koop
2015
[74]
Nethe
rland
sAde
quate
concealm
ent
RCT
n=611
Nurse-ledcare
tosymptom
control(I)
(orFeNOcontrolled).
Usualcare
(partially
controlled).
FU:12mo
Adu
ltswith
asthma.
Meanage40
y.28%
male.
Nobe
tween-
grou
pdifference
inEQ
5D(QALYs)
(I:0.91
vs.C
:0.89;
MD0.01,95%
CI
−0.02
to0.04).
n/a
Nobe
tween-
grou
pdifference
intotalcosts[I:
$4591vs.C
:$4180;MD$411,
95%
CI−
904to
1797;p
>0.05].
n/a
n/a
Kaup
pine
n1998
[75]
Finland
Con
cealmen
tno
tadeq
uate
RCT
n=162
Intensiveed
ucation
(use
ofinhaleddrug
s,PEF,mon
itorin
gand
PAAP).
Con
ventional
education.
FU:12mo
Adu
lts,new
lydiagno
sedasthma.
Meanage43
y.44%
male.
Nobe
tween-
grou
pdifference
in15D[I:0.93
(95%
CI0.90–0.94)
vs.C
:0.91(95%
CI
0.89
to0.94);p=
0.47].
n/a
Igroup
had
greatertotalcosts
than
control[I:
£345
(95%
CI247–
1758)v
s.C:£294
(95%
CI0–8078);
p<0.001].
n/a
Intensive
education:
increm
entalg
ain
of0.02
15D.
Increm
ental
differencein
health
costsof
£51.
Kriege
r2015
[76]
US
Ade
quate
concealm
ent
RCT
n=366
Com
mun
ityhe
alth
worker-supp
orted
self-managem
ent.
Usualcare.
FU:12mo
Adu
ltswith
asthma.
Meanage41
y.27%
male.
Interven
tion
improved
QoL.
Meanchange
inmAQLQ
(I:0.95
vs.
C:0.36;MD0.50,
95%
CI0.28–0.71;
p<0.001).
Nodifferencein
meanchange
innu
mbe
rof
urge
ntcare
episod
es.(I:
−1.50
vs.C
:−1.60;
difference0.09,95%
CI−
0.59
to0.73;
p=0.78.)
n/a
n/a
n/a
Lahd
ensuo1996
[77]
Finland
Con
cealmen
tno
tadeq
uate
RCT
n=122
Self-managem
ent,
includ
ingbreathing
exercises,ed
ucation
andPEFmon
itorin
g.
Tradition
altreatm
ent.
FU:12m
Adu
ltswith
asthma.
Meanage43
y.48%
male.
Interven
tion
improved
QoL
SGRQ
(sym
ptom
domain)
[I:16.6
(SD15.9)vs.C
:8.4
(SD18.4);p=
0.009].
n/a
n/a
Igroup
hadfewer
unsche
duledcare
visits/patient/year
(I:0.5vs.C
:1;
p=0.04).
n/a
Levy
2000
[78]
UK
Con
cealmen
tno
tadeq
uate
RCT
n=211
Structured
education
with
PAAPby
A&E
specialistnu
rses.
Usualprim
ary
care.
FU:6
mo
Adu
ltswith
asthma.
Meanage40
yrs.
43%
male.
Nobe
tween-
grou
pdifference
inSG
RQ(I:30.25
vs.C
:28.73;M
D1.52,95%
CI−
4.05
to7.09).
Nobe
tween-grou
pdifferencein
hospital
consultatio
ns[m
edian
(IQR)
I:0(1–3)v
s.C:0
(1–6);p=0.17].
n/a
Nobe
tween-
grou
pdifference
inGPconsulta-
tions
[med
ian
(IQR)
I:0(1–7)vs.
C:0
(1–7);
p=0.14].
n/a
Mancuso
2011
[79]
US
Con
cealmen
tno
tadeq
uate
RCT
n=296
Self-managem
ent
workboo
k,be
haviou
ralcon
tract,
teleph
onecalls.
Inform
ation/PEF
training
.FU
:12mo
Adu
ltsattend
ingA&E
with
asthma.
Meanage43
y.23%
male.
Nobe
tween-
grou
pdifference
inchange
inAQLQ
at1y(I:
0.04
vs.C
:0.18;
MD0.22,95%
CI
−0.15
to0.60).
n/a
n/a
Nobe
tween-
grou
pdifference
inprop
ortio
nwith
A&E
visits(I:13%
vs.C
:11%
).
n/a
McLean2003
[80]
Canada
Ade
quate
concealm
ent
RCT
n=225
Usualph
armacist
care.
FU:7
mo
Adu
ltswith
asthma.
Meanage38
y.47%
male.
Interven
tion
improved
QoL
asmeanAQLQ
(I:
Nobe
tween-grou
pdifferencein
hospitali-
satio
ns(I:0.078vs.C
:0.16;p
=0.94).
Interven
tion
redu
cedtotal
costs(costspe
r
Nobe
tween-
grou
pdifference
inA&E
visits
n/a
Pinnock et al. BMC Medicine (2017) 15:64 Page 18 of 32
Table
5Summarytableof
stud
iesinclud
edin
theRECURSIVEhe
alth
econ
omicanalysis(Con
tinued)
Pharmacist-ledself-
managem
ent,with
PAAP.
5.13
vs.C
:4.40;p
=0.0001).
patient
I:$150
vs.
C:$351).
(I:0.04
vs.C
:0.21;
p=0.48).
Mou
dgil2000
[81]
UK
Con
cealmen
tno
tadeq
uate
RCT
n=689
Individu
aled
ucation
andop
timisationof
drug
therapy.
Usualprim
ary
care.
FU:12mo
Adu
ltswith
asthma.
Meanage35
y.47%
male.
Greater
improvem
entin
QoL
inIg
roup
(MDin
change
inAQLQ
0.22
,95%
CI0.15–0.29).
Nobe
tween-grou
pdifferencein
hospitali-
satio
ns(OR0.51,95%
CI0.22–1.14).
n/a
Nobe
tween-
grou
pdifference
inA&E
visits(OR
0.63,95%
CI0.23–
1.68).
n/a
Pilotto2004
[82]
Australia
Con
cealmen
tno
tadeq
uate
Cluster
RCT
n=170
Nurse-run
asthma
clinicsinclud
ing
provisionof
PAAPs.
Usualprim
ary
care.
FU:9
mo
Adu
ltswith
asthma.
Meanage50
y.48%
male.
Nobe
tween-
grou
pdifference
inSG
RQ(I:27.3vs.
C:27.0;MD−0.5
(−4.0to
2.9).
Nobe
tween-grou
pdifferencein
numbe
radmitted
(I:2vs.C
:0;
p=0.499).
n/a
Nobe
tween-
grou
pdifference
innu
mbe
rattend
-ingA&E
(I:2vs.C
:0;p=0.499).
n/a
Pinn
ock2003
[83]
UK
Ade
quate
concealm
ent
RCT
n=278
Nurse-delivered
,routineteleph
one
review
.
Usualprim
ary
care.
FU:3
mo
Adu
ltswith
asthma.
Meanage57
y.41%
male.
Nobe
tween-
grou
pdifference
inmAQLQ
(I:5.17
vs.C
:5.17;MD
0.22,95%
CI−
0.15
to0.60).
Nopatientsin
either
grou
phadaho
spital
admission
forasthma.
n/a
Nopatientsin
either
grou
phad
anA&E
attend
ance
for
asthma
n/a
Price2004
[84]
UK
Ade
quate
concealm
ent
Cluster
RCT
n=1553
Use
ofPA
APs
with
adjustable
mainten
ance
dosing
.
Usualcare.
FU:3
mo
Adu
ltswith
asthma.
Meanage48
y.41%
male.
Nobe
tween-
grou
pdifference
inprop
ortio
nwith
improved
QoL
(I:22.5%
vs.C
:23.6%).
Nobe
tween-grou
pdifferencein
hospital
admission
s(I:2vs.C
:2).
Interven
tion
redu
cedtotal
costs(cost/day/
patient
I:£1.13vs.
C:£1.31;M
D−
£0.17,95%
CI
-£0.11
to-£0.23).
Nobe
tween-
grou
pdifference
inA&E
visits(I:5
vs.C
:11).
n/a
Ryan
2012
[85]
UK
Ade
quate
concealm
ent
RCT
n=288
Mob
ileph
one
supp
ortedself-
managem
ent.
Pape
r-based
PAAPs.
FU:6
mo
Adu
ltswith
asthma.
Meanage52
y.41%
male.
Nobe
tween-
grou
pdifference
inmeanchange
inmAQLQ
(differen
ce−0.10,
95%
CI−
0.16
to0.34).
Nobe
tween-grou
pdifferencein
hospital
admission
sforasthma
(I:3vs.C
:1).
n/a
Nobe
tween-
grou
pdifference
inA&E
atten-
dances
forasthma
(I:3vs.C
:0).
n/a
Sche
rmer
2002
[86]
Nethe
rland
sCon
cealmen
tno
tadeq
uate
RCT
n=193
Guide
dself-
managem
entwith
educationandPEF
mon
itorin
g.
Usualprim
ary
care.
FU:24mo
Adu
ltswith
asthma.
Meanage39
y.42%
male.
Nobe
tween-
grou
pdifference
intotalA
QLQ
(I:39
vs.C
:29;MD
10,95%
CI−
3to
23).
Noho
spital
admission
sin
either
treatm
entgrou
p.
Nobe
tween-
grou
pdifference
intotalcosts(I:
€1084vs.C
:€1097;MD−€13).
NoA&E
visitsin
either
treatm
ent
grou
p.
Increm
entalQ
ALY
gain
0.015.
Increm
entaltotal
cost−€13.
Increm
ental
health
cost€11.
Increm
ental
health
ICER
€33/
QALY.
Shelledy
2009
[87]
US
Con
cealmen
tno
tadeq
uate
RCT
n=166
Nurse-(N)vs.
respiratory
therapist-
Usualprim
ary
care.
FU:6
mo
Adu
lts:A
&Eor
admitted
with
asthma.
Meanage44
y.22%
male.
RTIg
roup
had
greaterchange
inSG
RQ[I(RT)−11.0
vs.I(N)−6.0vs.C
:−2.5,p<0.05).
Igroup
hadfewer
hospitalisations
[I(RT):
0.04
vs.I(N):0vs.C
:0.20;p
<0.05).
Igroup
hadlower
hospitalisation
costs[I(RT):$202
vs.I(N):$0
vs.C
:$1065;p<0.05].
Nobe
tween-
grou
pdifference
inA&E
visits[I(RT):
0.09
vs.I(N):0.26
vs.C
:0.37)].
n/a
Pinnock et al. BMC Medicine (2017) 15:64 Page 19 of 32
Table
5Summarytableof
stud
iesinclud
edin
theRECURSIVEhe
alth
econ
omicanalysis(Con
tinued)
(RT)
leded
ucation
andmanagem
ent.
Sund
berg
2005
[88]
Swed
enCon
cealmen
tno
tadeq
uate
RCT
n=97
Interactivecompu
ter-
baseded
ucationplus
nursesupp
ort.
Usualcare.
FU:12mo
Youn
gadultswith
asthma.Meanage19
y. 55%
male.
Nobe
tween-
grou
pdifference
inLiving
with
Asthm
aQuestion-
naire
(I:163.6vs.
C:166.2,p
>0.05).
Nobe
tween-grou
pdifferencein
hospital
admission
s(1
admis-
sion
ineach
grou
p).
n/a
Nobe
tween-
grou
pdifference
inA&E
visits(I:17
vs.C
:16).
n/a
vande
rMeer2011
[89]
Nethe
rland
sCon
cealmen
tno
tadeq
uate
RCT
n=200
Internet-based
self-
managem
ent
prog
ramme,includ
ing
electron
icPA
AP.
Usualou
tpatient
care.
FU:12mo
Adu
ltswith
asthma.
Meanage37
y.55%
male.
Nobe
tween-
grou
pdifference
inEQ
5D(I:0.93
vs.
C:0.89;difference
0.006,95%
CI
−0.042to
0.054).
Nobe
tween-grou
pdifferencein
hospital
admission
s(m
ean
cost:I:$571vs.C
:$589;M
D−17;
p=0.95).
Nobe
tween-
grou
pdifference
intotalh
ealth
care
costs(I:$2555vs.
C:$2518;M
D−$37;p=0.94).
n/a
Increm
entalQ
ALY
gain
0.024.
Increm
entaltotal
cost$641.
Increm
ental
health
cost$37.
Increm
ental
health
ICER
$1541/QALY.
Yilm
az2002
[90]
Turkey
Con
cealmen
tno
tadeq
uate
RCT
n=80
Outpatient
clinic,
specialedu
catio
nprog
ramme.
Usualprim
ary
care.
FU:36mo
Adu
ltswith
asthma.
Meanage29
y.37%
male.
Igroup
had
greater
improvem
entsin
AQLQ
(I:197.1vs.
C:176.7;
p=0.009).
Nobe
tween-grou
pdifferencein
hospitali-
satio
ns(I:0vs.C
:4);p
>0.05.
n/a
Igroup
hadfewer
A&E
visits(I:0vs.
C:7;p
=0.01).
n/a
Yoon
1993
[91]
Australia
Con
cealmen
tno
tadeq
uate
RCT
n=76
Brief,grou
p-based,
educationwith
aPA
AP.
Usualou
tpatient
care.
FU:10mo
Inpatient
adults.
Meanageno
trepo
rted
.28%
male.
Nobe
tween-
grou
pdifference
inQoL
[I:4.0(SD
4.38)vs.C
:3.96
(SD=3.34);
p>0.05).
Igroup
hadfewer
participantswith
hospitaladm
ission
s(I:1vs.C
:7;p
<0.001).
n/a
Nobe
tween-
grou
pdifference
inA&E
visits(I:3
vs.C
:7).
n/a
Abb
reviations:A
&Eaccide
ntan
dem
erge
ncy,ACQ
Asthm
aCon
trol
Que
stionn
aire,A
QLQ
Asthm
aQua
lityOfLife
Que
stionn
aire,C
control,CI
confiden
ceinterval,EQ5D
EuroQol
Five
Dim
ension
sQue
stionn
aire,FeN
Ofractio
nale
xhaled
nitricoxide,
FUfollow-up,
GPge
neralp
ractition
er,I
interven
tion,
ICER
increm
entalcost-effectiven
essratio
,IQRinterqua
rtile
rang
e,mAQALminiA
sthm
aQua
lityOfLife
Que
stionn
aire,M
Dmean
differen
ce,m
omon
th,N
nurse,
n/ano
tavailable,
PAAPpe
rson
alised
asthmaactio
nplan
,PEF
peak
expiratory
flow,Q
ALY
quality
-adjustedlifeyears,QoL
quality
oflife,
RCTrand
omised
controlledtrial,RT
respira
tory
therap
ist,SD
stan
dard
deviation,
SGRQ
StGeo
rge’sRe
spira
tory
Que
stionn
aire,y
year
Pinnock et al. BMC Medicine (2017) 15:64 Page 20 of 32
Table
6Focuseddata
extractio
nfro
madditio
nalstudies
iden
tifiedby
forw
ardcitatio
npriorto
publication
Reference;RC
Ts,n;
Participants,n;D
ate
rang
eRC
Ts
Com
parison
Relevanceto
meta-review
questio
ns:
Interven
tions
includ
edTarget
grou
p(s)
Synthe
sis
Mainresults
Whatistheim
pact?
Target
grou
ps?
Which
compo
nents?
Con
text?
System
aticreview
s
Coe
lho2016
[61]
17RC
Ts;5879
participants
RCTs
2005–2013
Scho
ol-based
asthma
educationvs.usual
care.
FU:m
inim
um1mo
Target:Schoo
lchildren
Educationalinterventions
toindividu
als,grou
psor
classesby
healthcare
profession
als,teache
rs,
educatorsand/or
IT.
Scho
olchildrenwith
asthmaand/or
who
lescho
ol.
Narrativeanalysis
6/17
show
edaredu
ctionin
unsche
duledcare;5/17show
eda
redu
ctionof
theasthmasymptom
s;5/17
redu
cedscho
olabsenteeism;
7/17
improved
QoL
ofthe
individu
als;8/17
show
edthat
asthma
educationim
proved
know
ledg
e.
McLean2016
[62]
5RC
Ts595participants
RCTs
2011–2013
Interactivedigital
interven
tions
vs.usual
care.
FU:10weeks
to12
mo
Impact
Com
pone
nts:
Techno
logy-based
interven
tions
Interactiveinterven
tion
(i.e.en
terin
gdata,
receivingtailored
feed
back,m
aking
choices)accessed
throug
han
appthat
provides
self-
managem
entinform
ation.
Adu
lts(≥16
y)with
asthma.
Meta-analysis
Meta-analyses
(3stud
ies)show
edno
sign
ificant
differencein
asthma
control(SM
D0.21,95%
CI−
0.05
to0.42)or
asthmaQoL
(SMD0.05,95%
CI−
0.22
to0.32)bu
the
teroge
neity
was
very
high
.Removalof
theou
tlier
stud
yredu
ced
heteroge
neity
andindicated
sign
ificant
improvem
entforbo
thasthmacontrol(SM
D0.54,95%
CI
0.22–0.86)
andasthmaQoL
(SMD
0.45,95%
CI0.13–0.77).
Rand
omised
trials
Hoskins
2016
[63]
48participants
Goal-settin
g+SM
/PA
APs
vs.usualcare.
Com
pone
nts:Goal-settin
gPracticeasthmanu
rses
traine
din
goal-settin
gapproach.
Prim
arycare
patients
dueareview
.Cluster
feasibility
RCT.
FU:6
mo
Difficulty
recruitin
g:10/124
practices
participated
and48
patients.No
betw
een-grou
pdifferencein
QoL
[mAQLQ
I:6.20
(SD0.76,95%
CI
5.76–6.65)
vs.C
:6.1(SD0.81,95%
CI
5.63–6.57),M
D0.1].
Moraw
ska2016
[64]
107participants
Gen
ericparenting
skillsvs
usualcare.
Com
pone
nts:Parenting
skills
Parentingskillsfor
managingLTCs+asthma
‘take-hom
etip
ssheets’.
Parentsof
children
2–10
ywith
asthma
and/or
eczema.
RCT.FU
:6mo
Betw
een-grou
pim
provem
entin
parents’self-efficacyandchilds’
‘eczem
abe
haviou
r’,bu
tno
teq
uivalent
asthmaou
tcom
es.
Parent
andfamily
gene
ricQoL
improved
(p=0.01).
Plaza2015
[65]
230participants
Traine
dpractices
(I)vs.spe
cialistun
it(Is)
vs.usualcare
(C).
Impact:
Com
pone
nts:Education
prog
ramme
Basicinform
ationon
asthma,inhaler
techniqu
e;provisionof
aPA
AP.
Adu
ltswith
persistent
asthma.
Cluster
RCT.FU
:12
mo
Igroup
shadfewer
unsche
duled
visits[I:0.8(SD1.4)
andIs :0.3(SD
0.7)
vs.C
:1.3(SD1.7);p
=0.001],and
greaterim
provem
entsin
asthma
control(p=0.042)
andQoL
(p=0.019).
Pinnock et al. BMC Medicine (2017) 15:64 Page 21 of 32
Table
6Focuseddata
extractio
nfro
madditio
nalstudies
iden
tifiedby
forw
ardcitatio
npriorto
publication(Con
tinued)
Rice
2015
[66]
711participants
PAAP+inpatient
lay
educator
vs.PAAP.
Com
pone
nts:Inpatient
layed
ucator
EncourageFU
attend
ance,b
uild
self-
efficacy,setgo
als,
overcomebarriers.
Children2–17
yadmitted
with
asthma.
RCT.
FU:1
mo
Nodifferencein
attend
ance
atFU
appo
intm
ent.Ig
roup
hadgreater
preven
teruse(OR2.4,95%
CI
1.3–4.2),PAAPow
nership(OR2.0,
95%
CI1.3–3.0)a
ndim
proved
self-efficacy(p=0.04).
Yeh2016
[67]
76participants
Family
prog
ramme
(+PA
AP)
vs.usualcare
(+PA
AP).
Com
pone
nts:Family
empo
wermen
tFamily
empo
wermen
tto
redu
ceparentalstress,
increase
family
functio
ning
.
Children6–12
ywith
asthma.
RCT.
FU:3
mo
Ifam
ilies
hadredu
cedparentalstress
inde
x(p=0.026)
andim
proved
family
environm
entscores
(p<
0.0001),im
proved
lung
functio
n,less
disturbe
dsleep,
less
coug
hbu
tno
differencein
whe
eze.
Zairina
2016
[68]
72participants
Telehe
alth
supp
orted
PAAPvs.usualcare.
Com
pone
nts:Telehe
alth
Telehe
alth
(FEV
1,symptom
s)mon
itored
weekly.
Preg
nant
wom
enwith
mod
erate/severe
asthma
RCT.
FU:6
mo
Telehe
alth
improved
ACQ[M
D0.36
(SD0.15,95%
CI−
0.66
to−0.07)]and
mAQLQ
[MD0.72
(SD0.22;95%
CI
0.29–1.16)].
Nodifferencein
perin
atalou
tcom
es.
Abb
reviations:A
CQAsthm
aCon
trol
Que
stionn
aire,A
QLQ
Asthm
aQua
lityOfLife
Que
stionn
aire,C
control,CI
confiden
ceinterval,FEV
1forced
expiratory
volumein
onesecond
,FUfollow-up,
Iintervention,
LTClong
-term
cond
ition
,mAQALminiA
sthm
aQua
lityOfLife
Que
stionn
aire,M
Dmeandifferen
ce,m
omon
th,O
Rod
dsratio
,PAAPpe
rson
alised
asthmaactio
nplan
,QoL
quality
oflife,
RCTrand
omised
controlledtrial,SD
stan
dard
de-
viation,
SMDstan
dardised
meandifferen
ce,y
year
Pinnock et al. BMC Medicine (2017) 15:64 Page 22 of 32
action plan ownership 1 month post-discharge [66]. Incontrast, three generic interventions in US minority popu-lations showed no improvement [46]. Update RCTs, someunderpowered, in indigenous populations had inconsistentoutcomes [29, 48, 55, 60].
A&E attendeesTwo 3* meta-analyses demonstrated reduced use ofhealthcare resources (admissions, A&E attendancesand unscheduled consultations) in adults recruitedduring A&E attendance (13 RCTs) [38] and in
children with a history of A&E attendance in theprevious 12 months (38 RCTs) [27]. Neither reviewfound improved markers of asthma control [27, 38],though an update RCT in paediatric A&E attendees(low risk of bias) found that children dischargedwith an action plan had fewer symptoms at 28 dayscompared with usual care [50].
Specific age groupsSchool-based interventions [30], often using informa-tion technology-based programmes [30] or delivered
Table 7 Treatment event rates from the meta-analyses
Events/total participants Percentage of participants with the event
Intervention Control Intervention Control
Proportion hospitalised
Boyd 2009 [27]*** 276/2009 351/2010 13.7 17.4
Gibson 2002 [31]*** 85/1200 139/1218 7.1 11.4
Tapp 2007 [38]*** 40/286 74/286 14.0 25.9
RECURSIVE 80/1727 124/1734 4.6 7.2
Proportion with A&E attendances
Boyd 2009 [27]*** 337/1505 462/1503 22.4 30.7
Gibson 2002 [31]*** 291/1457 354/1445 20.0 24.5
Tapp 2007 [38]*** 74/472 104/474 15.7 22.0
RECURSIVE 153/1171 227/1170 13.1 19.4
Proportion with unscheduled visits
Boyd 2009 [27]*** 128/515 181/494 24.9 36.6
Gibson 2002 [31]*** 112/784 170/772 14.3 22.0
Abbreviations: A&E accident and emergency
Fig. 2 Meta-Forest plot of healthcare resource use from meta-analyses. This meta-Forest plot displays the summary data from the PRISMS systematicreviews that reported relative risk (RR). Note that meta-analysis is inappropriate at meta-review level owing to the overlap of included randomisedcontrolled trials between reviews
Pinnock et al. BMC Medicine (2017) 15:64 Page 23 of 32
Table 8 Tailoring of self-management support for targeted populations
Group Key strategies Description of tailoring ofself-management intervention
Relevant systematic reviews/update RCTs
Evidence
Cultural groups Cultural tailoring Culturally orientated self-management programmes includingindividual sessions with language-appropriate asthma educators,videos/workbooks featuring culturallyappropriate role models, educationappropriate to socioeconomiccontext, strategies for use of localhealthcare services, asthma actionplans.
**Bailey 2009 [25]Adults and children fromminority groups
Culture-specific programmesare more effective thangeneric programmes inimproving QoL, knowledgeand asthma control but notall asthma outcomes.
Culturally tailored, community-basedinterventions in which healthcareproviders (pharmacists, asthmaeducator, social workers, respiratorynurses) provided language-appropriate education programmesincluding health literacy-focusedteaching, use of videos, asthmaphysiology and management, inhalertechnique, PAAP.
***Press 2012 [46]Adults from minority groupsin the USA
The 5 (of 15) educationstudies that were culturallytailored showed reduced useof unscheduled care andimproved QoL, but this is notcompared to non-tailoredinterventions.
Internet-based programmedeveloped to deliver education anda behaviour change intervention toAfrican-Americans adolescents.Strategies include voice-overs toaccommodate literacy limitationsand advice delivered by a ‘discjockey’.
(RCT) Joseph 2013 [54]Young teens
The intervention reducedsymptom-free days but hadno effect on A&E visits/hospitalisations.
Communityworkers
Community health worker from thesame/very similar community asparticipating families providedindividually tailored education athome visits. Topics included asthma,lifestyle and trigger avoidance, withresources to reduce allergenexposure and smoking cessationsupport.
**Postma 2009 [35]Ethnic minority children withasthma
Interventions involvingcommunity health workersreduced emergency andurgent care use in somebut not all studies.
Indigenous healthcare workersprovided personalised, child-friendly,culturally appropriate education ma-terials at home visits to reinforceclinical consultations.
**Chang 2010 [29]Ethnic minority children withasthma
The involvement ofindigenous healthcareworkers in asthmaprogrammes (1 RCT)improved control and QoLbut not unscheduled care.
A&E attendees Education duringthe A&Eattendance
Education sessions conducted byasthma or A&E nurses, or, less often,respiratory specialists or aphysiotherapist. Content varied,usually including triggers, PAAPsand/or inhaler technique.
***Tapp 2007 [38]Adult A&E attendees
Education delivered in A&Ereduced subsequent hospitaladmissions but not A&Eattendances. Effect on QoLwas inconsistent.
PAAP, completed by the A&Ephysician, coupled with theprescription provided on dischargefrom A&E.
(RCT) Ducharme 2011 [50]Children 1–17 y, A&Eattendees
Provision of a PAAP increasedpatient adherence to steroids(oral/inhaled), and improvedasthma control.
Education afterA&E
Education delivered by a healthcareprofessional or asthma educatorshortly after an A&E attendance,including triggers and PAAPs, to thechild and their carers.
***Boyd 2009 [27]Children, A&E attendees
Asthma education reducedA&E attendances andadmissions, but had no effecton QoL.
Schoolchildren School-basedprogrammes
School-based group education, themajority including education forclassmates without asthma.
**Coffman 2009 [30]Children
The intervention improvesknowledge, self-efficacy andself-management behaviours,but inconsistent effect onasthma control.
Pinnock et al. BMC Medicine (2017) 15:64 Page 24 of 32
by peers [48, 56], improved quality of life and, insome cases, reduced absenteeism [30, 48, 56, 61].Generic parenting skills initiatives improved self-efficacy in families struggling to manage young chil-dren with asthma, with inconsistent effect on asthmaoutcomes [64, 67].Two update RCTs reported interventions in older
people that improved control and quality of life [49,51], and one reduced use of unscheduled care [49]. A
key feature of both complex interventions was astructured approach to tailoring in order to meetpersonal goals or address individual problems.
Which components of supported self-management areimportant?A 3* meta-analysis (36 RCTs; 6090 participants of all agesrecruited from primary and secondary care settings)defined optimal self-management as education including
Table 8 Tailoring of self-management support for targeted populations (Continued)
16 short group educational sessions,including strategies for problemsolving, delivered in the schoollunch break.
Horner 2014 [53]Grades 2–5 (7–11 y)
Compared to generic healtheducation, the interventionimproved self-efficacy buthad no effect on admissions,A&E visits or QoL.
Peer-ledprogrammes
Year 11 pupils were trained todeliver the school-based asthmaeducational lessons to youngerpupils.
Al-Sheyab 2012 [48]Adolescents
Compared to children incontrol schools, knowledgeand QoL improved. Alsoincreased self-efficacy toresist smoking.
Asthma self-management skills andpsychosocial skills taught at a daycamp by peer leaders followed bymonthly peer telephone contact.
Rhee 2011 [56]Adolescents 13–17 y
The intervention group hadimproved QoL and positive‘attitude to illness’ comparedto those attending adult-ledcamps.
Technology-based Internet-based interventions,delivered at home, clinicor school, which delivereda psycho-educational programmeinvolving information andskills training modulestargeting improved healthoutcomes.
**Stinson 2009 [47]Children 4–17 y
The majority of studiesreported improvement insymptoms, but impact onother outcomes wasinconsistent.
Theoretically based asthmacomputer programmewith core modules (adherence,inhaler use, smoking reduction),with tailored sub-modulesto address specificbehavioural traits.
Joseph 2013 [54]9–12 grade (14–18 y)
The intervention improvedsymptom control, but had noeffect on A&E visits/hospitalisations.
Internet-based self-managementprogramme covering education,self-monitoring and an electronicaction plan, and encouraging regularmedical review. Supported by 2face-to-face groups.
Rikkers-Mutsaerts 2012 [57]Adolescents 12–18 y
QoL and asthma controlimproved compared to usualcare, but no difference in useof healthcare resources.
Elderly Goal-setting Six-session programme, conductedby a health educator in groups(n = 3) and telephone calls (n = 3).Participants selected anasthma-specific goal, identifiedproblems and addressedpotential barriers.
(RCT) Baptist 2013 [49]≥65 y
Compared to educationalone, the interventionimproved asthma control andQoL, but not unscheduledcare.
Addressingindividualconcerns
Specific concerns, identified with thePatient Assessment and ConcernsTool (PACT), were addressed in anhour-long session. Both groups hadstandard education (inhalertechnique, PAAP).
(RCT) Goeman 2013 [51]≥55 y
Compared to usual care,asthma control and QoL wasimproved by educationtailored to individual patientconcerns and unmet needs.
Abbreviations: A&E accident and emergency, PAAP personalised asthma action plan, QoL quality of life, RCT randomised controlled trial
Pinnock et al. BMC Medicine (2017) 15:64 Page 25 of 32
advice on self-monitoring and a written action planthat was supported by regular professional review[31]. There is evidence that reducing the intensity ofself-management education or level of clinical reviewmay reduce its effectiveness [36].
Components of an action planThe components of an action plan were further de-fined in two 3* and three 2* reviews [23, 24, 32, 36,39]. In adults, self-monitoring based on peak flow orsymptoms is equally effective [32, 36, 39]. In a com-parison in children, symptom-based plans were moreeffective at reducing unscheduled healthcare [23], andequally effective at improving most measures ofasthma control; the exception was days with symp-toms, which were reduced more by peak-flow-basedthan symptom-based plans [23]. A 3* review con-cluded that action plans with between two and fouraction points, including recommendations on increas-ing inhaled corticosteroids and initiating oral cortico-steroids, were consistently effective in reducingadmissions and A&E attendances [32].
Behavioural change techniquesOne 3* meta-analysis demonstrated that self-managementinterventions that incorporated specific behaviour changetechniques reduced unscheduled care and improved con-trol [43]. Meta-regression of the data from the 38 RCTs(7883 participants) concluded that active involvement ofparticipants in the intervention was a key factor in redu-cing unscheduled healthcare [43]. More specifically, iden-tifying individual behavioural traits (e.g. rebelliousness,low perceived emotional support) in adolescents en-abled targeted use of behavioural change techniques[54]. A goal-setting approach proved challenging toimplement in primary care settings [63].
TechnologyTwo 1* narrative reviews investigated computer- orinternet-based interactive self-management programmes[28, 47]. The effect on healthcare utilisation was incon-sistent, confirmed by a recent review identified in thepre-publication update [62], though both showedimprovement in symptoms [28] and/or quality of life[28, 47]. Two update RCTs of web-based self-management programmes for adolescents also showedimproved asthma control [54, 57], and an extendedfollow-up of RCT participants concluded that theseeffects could be sustained 18 months after conclusion ofthe trial [59]. Several school-based programmes usedtechnology-based interventions to improve control andreduce absenteeism [30]. Supported self-managementusing mobile phone technology currently has a limitedand inconclusive evidence base [42, 45], though a recent
RCT in pregnancy demonstrated improved asthma con-trol and quality of life [68].
Which contextual factors influence effectiveness?Resonating with the concept of ‘optimal’ self-management(education, an action plan and regular review) [31], a 3*meta-analysis identified that omitting regular review(1 RCT) or reducing intensity of education (1 RCT)was associated with a smaller reduction in unsched-uled consultations [36]. A 2* meta-analysis analysedthe findings of 18 RCTs (3006 participants) accordingto the components of the Chronic Care Model [92].Interventions that included all four components had agreater effect on adherence to inhaled corticosteroidscompared to trials including self-management unsup-ported by the organisational components [33].
Organisational role in promoting supported self-managementA 3* narrative review of 14 RCTs (4588 participants) con-cluded that proactive organisational systems can increaseaction plan ownership by promoting uptake of asthmareviews and implementing (and monitoring) structuredmanagement systems for asthma care [37]. A recent RCTof a structured approach to self-management education inboth primary care and specialist units improved asthmacontrol and reduced unscheduled care [65], and a largecluster RCT at low risk of bias showed an increased adher-ence to guidelines and reduced asthma symptoms bysystematically providing individualised prompts to generalpractitioners and parents of children with asthma [52].Automatically linking an action plan to prescriptions givento patients being discharged from A&E improved clinicianmanagement and patient uptake of steroid courses [50].
What is the effect of supported self-management onhealthcare utilisation and costs?The RECURSIVE meta-analysis confirmed that self-management support interventions for people withasthma are associated with significant improvements inquality-of-life outcomes (SMD 0.26, 95% CI 0.12–0.39),significant small decreases in hospitalisation rates andcosts (SMD −0.21, 95% CI −0.40 to −0.01), significantsmall decreases in A&E visits (SMD −0.25, 95% CI −0.49to −0.01), and non-significant small increases in totalhealthcare costs (SMD 0.13, 95% CI −0.09 to 0.34).Figure 3 shows a Forest plot of the total costs.
What is the evidence that supported self-management forasthma can reduce costs without compromising outcomes?Figure 4 shows the overall permutation plot of thestudies (n = 21) reporting data on both quality of life andhealthcare utilisation. The majority of the studies onquality of life versus costs related to hospitalisations andA&E attendances were in the right-down quadrant,
Pinnock et al. BMC Medicine (2017) 15:64 Page 26 of 32
indicating cost-effectiveness (reduced healthcare utilisa-tion and improved quality of life). However, in terms oftotal costs (n = 7), the picture was mixed with morestudies around zero and the right-up quadrant, indicat-ing that similar costs or small cost increases are neces-sary to achieve better quality of life.
What is the evidence that supported self-management forasthma is cost-effective?Four studies applied formal economic analyses; twoshowed that self-management support interventions weredominant (i.e. significantly better health outcomes withsignificantly lower costs) [72, 86], and two produced non-
Fig. 3 Meta-analysis of total costs. CI confidence interval, ES effect size
Fig. 4 Permutation plot. Quality of life (x-axis), hospitalisations (y-axis blue) and total costs (y-axis red). In this permutation plot, the effects ofself-management interventions on outcomes (quality of life) and utilisation (hospitalisations and total costs) can be visualised simultaneously byplacing them in quadrants of the cost-effectiveness plane depending on the pattern of outcomes. Such plots identify studies in the appropriatequadrant (i.e. those that reduce costs without compromising outcomes) and those in problematic quadrants (i.e. those that reduce costs but alsocompromise outcomes, or those that compromise both outcomes and costs).
Pinnock et al. BMC Medicine (2017) 15:64 Page 27 of 32
significant ratios between costs and benefits at levels likelyto appeal to decision-makers (better outcomes with non-significant increases in costs) [75, 89] (see Additional file 5for more details).Thus, the benefits derived by supported self-management
interventions are associated with reductions in key areas ofhealthcare utilisation such as hospitalisations and A&Eattendances and can be delivered at similar levels of totalcosts to usual care.
DiscussionSummary of findingsExtensive evidence (n = 270 RCTs) derived from a broadrange of demographic and healthcare settings revealsthat supported self-management can reduce hospitalisa-tions, A&E attendances and unscheduled consultations,and improve markers of control and quality of life forpeople with asthma. Core components of effective self-management are education, provision of an action planand the support of regular professional review. Effective-ness has been demonstrated in diverse cultural, clinicaland demographic groups, with evidence that tailoredprogrammes have greater impact than generic interven-tions. A range of modes of delivery (including telehealth-care) may be employed to suit preferences and context.The cost of providing self-management support is offsetby a reduction in hospitalisations and unscheduledhealthcare.
Interpretation of findingsThe literature on asthma self-management is particularlywell developed and may thus be an exemplar for otherLTCs [13, 14]. The 16 systematic reviews reporting ef-fectiveness were typically large (five included data from>5000 participants [27, 30, 31, 41, 43]) and had consist-ently positive results, suggesting a mature evidence base,unlikely to be influenced by further trials. Outcomes insubgroups were more often the subject of the updateRCTs as the field moves on from demonstrating overalleffectiveness to investigating the impact in specifictarget groups [48–58, 60, 61, 72], demographic contexts[52–54, 66], or mode of delivery [54, 59, 62, 72].Self-management support for asthma is a complex
intervention and successful interventions were multi-component, including education, trigger avoidance,teaching self-monitoring, optimal treatment strategies,promotion of adherence and behaviour change tech-niques, many of which are common to self-managementin other LTCs [6]. Appropriately in a variable condition[4], the hallmark of asthma self-management is theprovision of an action plan with advice on recognisingand responding to deterioration in control [4, 32].People with asthma, however, have broader concerns asthey accommodate the condition within their lives and
the action plan needs to be embedded in support for‘living with asthma’ [93].Individuals with LTCs adjust medical regimes and self-
management strategies to fit into their own lives andhealth beliefs [13]. Meta-reviews, for example in type 2diabetes [94, 95], hypertension [96] and asthma [25],have emphasised the importance of culturally tailoredinterventions. Self-management support can be providedby many different professionals, often specialist nurses[38, 63] or LTC educators [25, 27, 95], but in some con-texts the key personnel were community health workers[35, 97] or peer counsellors [30, 56, 66]. Traditionallyeducation is delivered face-to-face, but increasinglytechnology-based interventions are being developed asalternatives [27, 28, 30, 42, 45, 47, 54, 57, 59, 62, 68].Self-management support interventions are an integral
component of high-quality care for people with LTCs[8–10]. Several of the systematic reviews demonstratedthe synergy between self-management education andregular clinical review [31, 33, 36], and supported self-management is most effective when delivered within aproactive asthma management programme [33, 37, 65],or integrated within organisational routines [50, 52].Only a minority of trials had follow-up periods over12 months, and studies are needed to confirm long-termsustainability. Costs associated with self-managementinterventions are similar to usual care.
Strengths and limitationsMeta-reviews have some intrinsic strengths and limita-tions. The methodology enables the efficient review of alarge body of evidence and thus provision of a compre-hensive overview to inform policy and practice. However,it relies on the quality of the included systematic reviews(e.g. comprehensive search strategies, accurate data ex-traction and synthesis). We used the validated R-AMSTAR instrument to assess the quality of included sys-tematic reviews [17]. In contrast to GRADE [98] (now rec-ommended by the Cochrane Handbook [15]), R-AMSTAR assesses the overall quality of the review, ratherthan assessing the quality of evidence individually for eachoutcome.Re-synthesising materials that have already been syn-
thesised risks further loss of detail and has the potentialfor erroneous assumptions, especially if the primaryfocus of the review did not directly align with the ques-tions of the meta-review. Overlap between the RCTsincluded in the systematic reviews may result in undueemphasis on commonly cited papers.Whilst some reviews and update RCTs directly com-
pared interventions with or without specific components[23–25, 32, 36, 39, 43], or a specific mode of delivery[28, 29, 41, 45], often the different interventions werecompared to usual care, allowing only indirect
Pinnock et al. BMC Medicine (2017) 15:64 Page 28 of 32
comparison [31, 33, 35, 37, 42, 46, 47]. A further limita-tion is that ‘usual care’ is rarely defined in RCTs [99],and the definition is even more unclear at meta-review level. Typically usual care is enhanced in thecontext of a trial, reducing the apparent impact of anintervention [100].Systematic reviews are only as current as their
most-recent search, and meta-reviews add anadditional time delay. In the PRISMS meta-review wetherefore not only updated our search for systematicreviews, but also searched for RCTs published afterthe date of the last search used by the includedsystematic reviews. In addition, prior to publicationwe undertook forward citation on all the includedsystematic reviews, which identified two recentsystematic reviews and six RCTs [61–68]. None ofthese changed our conclusions, confirming the matur-ity of the evidence base.The two reviewers who undertook the screening
and data extraction were not working independently;however, both projects ensured all the reviewers werefully trained and instituted random checks at everystage. Restricting inclusion to reviews with extractableRCT data maintained the quality of evidence, butmay have resulted in some lower-grade but useful evi-dence being rejected.RECURSIVE was not restricted to formal cost-
effectiveness studies – it had a broader focus and in-cluded studies reporting data on healthcare utilisationonly, without a full effectiveness analysis includingcosts and quality of life. Some of the RCTs in theRECURSIVE meta-analysis used a more comprehen-sive definition of ‘total costs’ (e.g. based on societalperspective) compared to others; to account for thisinconsistency, we also present the results on keysources of costs such as hospitalisation and A&Eattendance rates.The PRISMS and RECURSIVE teams worked inde-
pendently, but met regularly throughout the studiesto optimise synergies. A further strength was themultidisciplinary team, including backgrounds in pub-lic health, general practice, epidemiology and healthpsychology, enabling a balanced interpretation.
ConclusionsSupported self-management for asthma can reduceunscheduled care, improve asthma control and qualityof life, and does not lead to significant increases intotal healthcare costs. Effective self-managementshould be tailored to cultural, clinical and demo-graphic characteristics and is most effective whendelivered in the context of proactive LTC manage-ment. Healthcare organisations should prioritise and
promote the provision of supported self-managementfor people with asthma.
Additional files
Additional file 1: Detailed search terms: PRISMS and RECURSIVE(all databases). (DOCX 88 kb)
Additional file 2: Dates of initial and update searches. (DOCX 21 kb)
Additional file 3: Detailed PICOS table and inclusion/exclusion criteria.(DOCX 22 kb)
Additional file 4: Quality assessment and weighting. (DOCX 43 kb)
Additional file 5: Characteristics of included studies and key outcomes.(DOCX 169 kb)
AbbreviationsA&E: Accident and emergency; LTC: Long-term condition; RCTs: Randomisedcontrolled trials; SMD: Standardised mean difference
AcknowledgementsWe thank Ms Christine Hunter, lay collaborator to the PRISMS project; thePRIMER patient and public involvement group at the University of Manchester;representatives from Asthma UK; and other stakeholder groups whocontributed to the development of the project and the project workshops.
The following are members of the PRISMS group:Stephanie JC Taylor, Hilary Pinnock, Chris J Griffiths, Trisha Greenhalgh, AzizSheikh, Eleni Epiphaniou, Gemma Pearce, Hannah L Parke, AnnaSchwappach, Neetha Purushotham, Sadhana Jacob.
The following are members of the RECURSIVE group:Peter Bower, Maria Panagioti, Gerry Richardson, Elizabeth Murray, AnneRogers, Anne Kennedy, Stanton Newman, Nicola Small.
FundingPRISMS and RECURSIVE were funded by the National Institute for HealthResearch Health Services and Delivery Research Programme (projectnumbers 11/1014/04 and 11/1014/06. The funding body had no role in thedesign of the study, collection, analysis, nor interpretation of data, nor inwriting the manuscript. HP was supported by a Primary Care Research CareerAward from the Chief Scientist’s Office of the Scottish Government at thetime of the study. LD is supported by an Academic Fellowship in GeneralPractice from the Scottish School of Primary Care.
Availability of data and materialsNot applicable: all data used in this meta-review are derived from publishedstudies and thus already available.
Authors’ contributionsST and HP initiated the idea for the PRISMS study, led the development of theprotocol, securing of funding, study administration, data analysis, interpretationof results and writing of the paper. CG and AS were grant holders on thePRISMS review who contributed to the development of the protocol, thesecuring of funding, the interpretation of results and the writing of the paper.EE, HLP and GP were systematic reviewers who undertook searching, selectionof papers and data extraction with ST and HP. LD updated the PRISMS review.PB developed the idea for the RECURISVE study, secured funding and hadprimary responsibility for the interpretation of the results and writing the paper.MP and PB reviewed articles, extracted the data, undertook the data analysisand wrote the RECURSIVE paper. MP performed the RECURSIVE update for thismeta-review. All authors had full access to all the data, and were involved ininterpretation of the data. HP wrote the initial draft of the paper with HLP, LD,MP and ST to which all the authors contributed. ST and HP are study guarantorsfor PRISMS; PB and MP are study guarantors for RECURSIVE. All authors read andapproved the final manuscript.
Competing interestsThe submitted work was funded by a grant from the National Institute forHealth Research Health Services and Delivery Research Programme. None ofthe authors have financial relationships with any organisations that might
Pinnock et al. BMC Medicine (2017) 15:64 Page 29 of 32
have an interest in the submitted work. HP chairs the self-managementevidence review group for the British Thoracic Society/Scottish IntercollegiateGuideline Network Asthma guideline; the authors declare that they have noother relationships or activities that could appear to have influenced thesubmitted work.
Consent for publicationsNot applicable: no individual person’s data.
Ethics approval and consent to participateNot applicable: meta-review of published data.
Department of Health disclaimerThe views and opinions expressed therein are those of the authors and donot necessarily reflect those of the HS&DR programme, NIHR, NHS or theDepartment of Health.
Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.
Author details1Asthma UK Centre for Applied Research, Allergy and Respiratory ResearchGroup, Usher Institute of Population Health Sciences and Informatics,University of Edinburgh, Doorway 3, Medical School, Teviot Place, EdinburghEH8 9AG, UK. 2Centre for Primary Care and Public Health, Barts and TheLondon School of Medicine and Dentistry, Queen Mary University of London,London, UK. 3NIHR School for Primary Care Research, Centre for Primary Care,Manchester Academic Health Science Centre, University of Manchester,Manchester, UK. 4Centre for Technology Enabled Health Research (CTEHR),Coventry University, Coventry, UK.
Received: 28 September 2016 Accepted: 20 February 2017
References1. Global Asthma Network. The global asthma report 2014. Global Asthma
Network. 2014. http://www.globalasthmareport.org/resources/Global_Asthma_Report_2014.pdf. Accessed 22 Dec 2016
2. Mukherjee M, Stoddart A, Gupta RP, Nwaru BI, Farr A, Heaven M, et al. Theepidemiology, healthcare and societal burden and costs of asthma in theUK and its member nations: analyses of standalone and linked nationaldatabases. BMC Med. 2016;14:113.
3. British Thoracic Society, Research Unit of the Royal College of Physicians ofLondon, King’s Fund Centre. National Asthma Campaign. Guidelines formanagement of asthma in adults: I-chronic persistent asthma. BMJ.1990;301:651–3.
4. British Thoracic Society/Scottish Intercollegiate Guideline Network. Britishguideline on the management of asthma. 2016 update. http://sign.ac.uk/guidelines/fulltext/153/index.html. Accessed 22 Dec 2016.
5. Global Initiative for Asthma. Global strategy for asthma management andprevention 2015 update. http://www.ginasthma.org. Accessed 22 Dec 2016.
6. Pearce G, Parke H, Pinnock H, et al. The PRISMS taxonomy of self-management support: derivation of a novel taxonomy and initial testing ofutility. J Health Serv Res Policy. 2016;21:73–82.
7. Royal College of Physicians. Why asthma still kills: the National Review ofAsthma Deaths (NRAD) Confidential Enquiry report. London: Royal Collegeof Physicians; 2014.
8. Singh D, Ham C. Improving care for people with long-term conditions: areview of UK and international frameworks. NHS Institute for Innovation andImprovement. Department of Health: Birmingham; 2006.
9. World Health Organization. Innovative care for chronic conditions. Buildingblocks for action. Global report. Geneva: WHO; 2002.
10. Department of Health. The Mandate. A mandate from the Government tothe NHS Commissioning Board: April 2013 to March 2015. London:Department of Health; 2012.
11. Asthma UK. Time to take action on asthma. London: Asthma UK; 2014.12. Pinnock H, Epiphaniou E, Pearce G, et al. Implementing supported self-
management for asthma: a systematic review of implementation studies.BMC Med. 2015;13:127.
13. Taylor SJC, Pinnock H, Epiphaniou E, Pearce G, Parke H. A rapid synthesis of theevidence on interventions supporting self-management for people with long-term conditions. (PRISMS Practical Systematic Review of Self-ManagementSupport for long-term conditions). Health Serv Deliv Res. 2014;2:54.
14. Panagioti M, Richardson G, Murray E, Rogers A, Kennedy A, Newman S, et al.Reducing care utilisation through self-management interventions(RECURSIVE): a systematic review and meta-analysis. Health Serv Deliv Res.2014;2:54.
15. Higgins JPT, Green S. Cochrane handbook for systematic reviews ofinterventions version 5.1.0 http://handbook.cochrane.org. Accessed 22 Dec 2016.
16. Adams K, Greiner AC, Corrigan JM, editors. The 1st annual crossing thequality chasm summit – a focus on communities. Washington, D.C: TheNational Academic Press; 2004. p. 57.
17. Kung J, Chiappelli F, Cajulis OO, Avezova R, Kossan G, Chew L, et al. Fromsystematic reviews to clinical recommendations for evidence-based healthcare: validation of Revised Assessment of Multiple Systematic Reviews(R-AMSTAR) for grading of clinical relevance. Open Dent J. 2010;4:84–91.
18. Drummond MF, O’Brien, Stoddart LG, Torrance GW. Methods for theeconomic evaluation of health care programmes. 2nd ed. Oxford: OxfordUniversity Press; 1997.
19. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL.Methods for the economic evaluation of health care programmes.3rd ed. Oxford: Oxford University Press; 2005.
20. Pildal J, Hrobjartsson A, Jorgensen KJ, Hilden J, Altman DG, Gotzsche PC.Impact of allocation concealment on conclusions drawn from meta-analyses of randomized trials. Int J Epidemiol. 2007;36:847–57.
21. Nixon J, Khan K, Kleijnen J. Summarising economic evaluations in systematicreviews: a new approach. BMJ. 2001;322:1596–8.
22. Greenhalgh T, Peacock R. Effectiveness and efficiency of search methods insystematic reviews of complex evidence: audit of primary sources. BMJ.2005;331:1064–5.
23. Bhogal SK, Zemek RL, Ducharme F. Written action plans for asthma inchildren. Cochrane Database Syst Rev. 2006;3, CD005306.doi:10.1002/14651858.CD005306.pub2.
24. Zemek RL, Bhogal SK, Ducharme FM. Systematic review of randomizedcontrolled trials examining written action plans in children. What is theplan? Arch Pediatr Adolesc Med. 2008;162:157–63.
25. Bailey EJ, Cates CJ, Kruske SG, Morris PS, Brown N, Chang AB. Culture-specific programs for children and adults from minority groups who haveasthma. Cochrane Database Syst Rev. 2009;2, CD006580.doi:10.1002/14651858.CD006580.pub4.
26. Bernard-Bonnin AC, Stachenko S, Bonin D, Charette C, Rousseau E. Self-management teaching programs and morbidity of pediatric asthma: ameta-analysis. J Allergy Clin Immunol. 1995;95:34–41.
27. Boyd M, Lasserson TJ, McKean MC, Gibson PG, Ducharme FM, Haby M.Interventions for educating children who are at risk of asthma-relatedemergency department attendance. Cochrane Database Syst Rev.2009;2, CD001290. doi:10.1002/14651858.CD001290.pub2.
28. Bussey-Smith KL, Rossen RD. A systematic review of randomized controltrials evaluating the effectiveness of interactive computerized asthmapatient education programs. Ann Allergy Asthma Immunol. 2007;98:507–16.
29. Chang AB, Taylor B, Masters IB, Laifoo Y, Brown ADH. Indigenous healthcareworker involvement for Indigenous adults and children with asthma.Cochrane Database Syst Rev. 2010;5, CD006344. doi:10.1002/14651858.CD006344.pub3.
30. Coffman JM, Cabana MD, Yelin EH. Do school-based asthma educationprograms improve self-management and health outcomes? Pediatrics.2009;124:729–42.
31. Gibson PG, Powell H, Wilson A, Abramson MJ, Haywood P, Bauman A, et al.Self management education and regular practitioner review for adults withasthma. Cochrane Database Syst Rev. 2002;3, CD001117.doi:10.1002/14651858.CD001117.
32. Gibson PG, Powell H. Written action plans for asthma: an evidence-basedreview of the key components. Thorax. 2004;59:94–9.doi:10.1136/thorax.2003.011858.
33. Moullec G, Gour-Provencal G, Bacon SL, Campbell TS, Lavoie KL. Efficacy ofinterventions to improve adherence to inhaled corticosteroids in adultasthmatics: impact of using components of the chronic care model. RespirMed. 2012;106:1211–25.
34. Newman S, Steed L, Mulligan K. Self-management interventions for chronicillness. Lancet. 2004;364:1523–37.
Pinnock et al. BMC Medicine (2017) 15:64 Page 30 of 32
35. Postma J, Karr C, Kieckhefer G. Community health workers andenvironmental interventions for children with asthma: a systematic review.J Asthma. 2009;46:564–76.
36. Powell H, Gibson PG. Options for self-management education for adultswith asthma. Cochrane Database Syst Rev. 2002;3, CD004107.doi:10.1002/14651858.CD004107.
37. Ring N, Malcolm C, Wyke S, et al. Promoting the use of personal asthmaaction plans: a systematic review. Prim Care Respir J. 2007;16:271–83.
38. Tapp S, Lasserson TJ, Rowe BH. Education interventions for adults whoattend the emergency room for acute asthma. Cochrane Database Syst Rev.2007;3:CD003000. doi:10.1002/14651858.CD003000.pub2.
39. Toelle B, Ram FSF. Written individualised management plans for asthma inchildren and adults. Cochrane Database Syst Rev. 2004;1, CD002171.doi:10.1002/14651858.CD002171.pub2.
40. Welsh EJ, Hasan M, Li P. Home-based educational interventions for childrenwith asthma. Cochrane Database Syst Rev. 2011;10, CD008469.doi:10.1002/14651858.CD008469.pub2.
41. Bravata DM, Gienger AL, Holty JE, Sundaram V, Khazeni N, Wise PH, et al.Quality improvement strategies for children with asthma: a systematicreview. Arch Pediatr Adolesc Med. 2009;163:572–81.
42. de Jongh T, Gurol-Urganci I, Vodopivec-Jamsek V, Car J, Atun R. Mobilephone messaging for facilitating self-management of long-term illnesses.Cochrane Database Syst Rev. 2012;12, CD007459. doi:10.1002/14651858.CD007459.pub2.
43. Denford S, Taylor RS, Campbell JL, Greaves CJ. Effective behavior changetechniques in asthma self-care interventions: systematic review and meta-regression. Health Psychol. 2014;33:577–87.
44. Kirk S, Beatty S, Callery P, Gellatly J, Milnes L, Pryjmachuk S. Theeffectiveness of self-care support interventions for children and youngpeople with long-term conditions: a systematic review. Child CareHealth Dev. 2012;39:305–24.
45. Marcano Belisario JS, Huckvale K, Greenfield G, Car J, Gunn LH. Smartphoneand tablet self management apps for asthma. Cochrane Database Syst Rev.2013;11, CD010013. doi:10.1002/14651858.CD010013.pub2.
46. Press VG, Pappalardo AA, Conwell WD, Pincavage AT, Prochaska MH, AroraVM. Interventions to improve outcomes for minority adults with asthma: asystematic review. J Gen Int Med. 2012;27:1001–15.
47. Stinson J, Wilson R, Gill N, Yamada J, Holt J. A systematic review of internet-based self-management interventions for youth with health conditions.J Pediatr Psychol. 2009;34:495–510.
48. Al-sheyab N, Gallagher R, Crisp J, Shah S. Peer-led education for adolescentswith asthma in Jordan: a cluster-randomized controlled trial. Pediatrics.2012;129:e106–12.
49. Baptist AP, Ross JA, Yang Y, Song PXK, Clark NM. A randomized controlledtrial of a self-regulation intervention for older adults with asthma. J AmGeriatrics Soc. 2013;61:747–53.
50. Ducharme FM, Zemek RL, Chalut D, McGillivray D, Noya FJ, Resendes S, etal. Written action plan in pediatric emergency room improves asthmaprescribing, adherence, and control. Am J Respir Crit Care Med.2011;183:195–203.
51. Goeman D, Jenkins C, Crane M, Paul E, Douglass J. Educational interventionfor older people with asthma: a randomised controlled trial. Pat Ed Counsel.2013;93:586–95.
52. Halterman JS, Szilagyi PG, Fisher SG, Fagnano M, Tremblay P, Conn KM, etal. Randomized controlled trial to improve care for urban children withasthma: results of the school-based asthma therapy trial. Arch PediatricAdolesc Med. 2011;165:262–8.
53. Horner SD, Brown A. Evaluating the effect of an asthma self-managementintervention for rural families. J Asthma. 2013;51:168–77.
54. Joseph CL, Ownby DR, Havstad SL, Saltzgaber J, Considine S, Johnson D,Research team members. Evaluation of a web-based asthma managementintervention program for urban teenagers: reaching the hard to reach.J Adolesc Health. 2013;52:419–26.
55. Khan R, Maharaj R, Seerattan N, Babwah F. Effectiveness of personalized writtenasthma action plans in the management of children with partly controlledasthma in Trinidad: a randomized controlled trial. J Trop Pediatr. 2014;60:17–26.
56. Rhee H, Belyea MJ, Hunt JF, Brasch J. Effects of a peer-led asthma self-management program for adolescents. Arch Pediatr Adolesc Med.2011;165:513–9.
57. Rikkers‐Mutsaerts ER, Winters AE, Bakker MJ, van Stel HF, van der Meer V, deJongste JC, Smashing Study Group. Internet-based self-management
compared with usual care in adolescents with asthma: a randomizedcontrolled trial. Pediatr Pulmonol. 2012;47:1170–9.
58. Shah S, Sawyer SM, Toelle BG, Mellis CM, Peat JK, Lagleva M, et al.Improving paediatric asthma outcomes in primary health care: arandomised controlled trial. Med J Aust. 2011;195:405–9.
59. Van Gaalen JL, Beerthuizen T, van der Meer V, van Reisen P, RedelijkheidGW, Snoeck-Stroband JB, Smashing Study Group. Long-term outcomes ofinternet-based self-management support in adults with asthma: randomizedcontrolled trial. J Med Internet Res. 2013;15, e188.
60. Wong SS, Nathan AM, de Bruyne J, Zaki R, Tahir SZM. Does a writtenasthma action plan reduce unscheduled doctor visits in children? Indian JPediatr. 2011;80:590–5.
61. Coelho AC, Cardoso LS, Machado CS, Machado AS. The impacts ofeducational asthma interventions in schools: a systematic review of theliterature. Can Respir J. 2016;2016:8476206.
62. McLean G, Murray E, Band R, Moffat KR, Hanlon P, Bruton A, et al. Interactivedigital interventions to promote self-management in adults with asthma:systematic review and meta-analysis. BMC Pulm Med. 2016;16:83.
63. Morawska A, Mitchell AE, Burgess S, Fraser J. Effects of Triple P parentingintervention on child health outcomes for childhood asthma and eczema:randomised controlled trial. Behav Res Ther. 2016;83:35–44.
64. Yeh HY, Ma WF, Huang JL, Hsueh KC, Chiang LC. Evaluating theeffectiveness of a family empowerment program on family function andpulmonary function of children with asthma: a randomized control trial. IntJ Nurs Stud. 2016;60:133–44.
65. Hoskins G, Williams B, Abhyankar P, Donnan P, Duncan E, Pinnock H, et al.Achieving Good Outcomes for Asthma Living (GOAL): mixed methodsfeasibility and pilot cluster randomised controlled trial of a practicalintervention for eliciting, setting and achieving goals for adults with asthma.Trials. 2016;17:584.
66. Rice JL, Matlack KM, Simmons MD, Steinfeld J, Laws MA, Dovey ME, et al.LEAP: A randomized-controlled trial of a lay-educator inpatient asthmaeducation program. Pat Ed Counsel. 2015;98:1585–91.
67. Plaza V, Peiró M, Torrejón M, Fletcher M, López-Viña A, Ignacio JM, et al. Arepeated short educational intervention improves asthma control andquality of life. Eur Respir J. 2015;46:1298–307.
68. Zairina E, Abramson MJ, McDonald CF, Li J, Dharmasiri T, Stewart K, et al.Telehealth to improve asthma control in pregnancy: a randomizedcontrolled trial. Respirology. 2016;21:867–74.
69. Castro M, Zimmermann NA, Crocker S, Bradley J, Leven C, Schechtman KB.Asthma intervention program prevents readmissions in high healthcareusers. Am J Respir Crit Care Med. 2003;168:1095–9.
70. Clark NM, Gong ZM, Wang SJ, Lin X, Bria WF, Johnson TR. A randomizedtrial of a self-regulation intervention for women with asthma. Chest.2007;132:88–97.
71. de Oliveira M, Faresin S, Bruno V, de Bittencourt A, Fernandes A. Evaluationof an educational programme for socially deprived asthma patients. EurRespir J. 1999;14:908–14.
72. Gallefoss F, Bakke PS. Cost-effectiveness of self-management in asthmatics: a1-yr follow-up randomized, controlled trial. Eur Respir J. 2001;17:206–13.
73. Gruffydd-Jones K, Hollinghurst S, Ward S, Taylor G. Targeted routine asthmacare in general practice using telephone triage. Br J Gen Pract. 2005;55:918–23.
74. Honkoop PJ, Loijmans RJ, Termeer EH, Snoeck-Stroband JB, van den HoutWB, Bakker MJ, et al. Symptom-and fraction of exhaled nitric oxide–drivenstrategies for asthma control: a cluster-randomized trial in primary care.J Allergy Clin Immunol. 2015;135:682–8.
75. Kauppinen R, Sintonen H, Tukiainen H. One-year economic evaluation ofintensive vs conventional patient education and supervision for self-management of new asthmatic patients. Respir Med. 1998;92:300–7.
76. Krieger J, Song L, Philby M. Community health worker home visits for adultswith uncontrolled asthma: the HomeBASE Trial randomized clinical trial.JAMA Intern Med. 2015;175:109–17.
77. Lahdensuo A, Haahtela T, Herrala J, Kava T, Kiviranta K, Kuusisto P, et al.Randomised comparison of guided self management and traditionaltreatment of asthma over one year. BMJ. 1996;312:748–52.
78. Levy ML, Robb M, Allen J, Doherty C, Bland JM, Winter RJ. A randomizedcontrolled evaluation of specialist nurse education following accidentand emergency department attendance for acute asthma. Respir Med.2000;94:900–8.
79. Mancuso CA, Peterson MG, Gaeta TJ, Fernández JL, Birkhahn RH, MelnikerLA, et al. A randomized controlled trial of self-management education for
Pinnock et al. BMC Medicine (2017) 15:64 Page 31 of 32
asthma patients in the emergency department. Ann Emerg Med.2011;57:603–12.
80. McLean W, Gillis J, Waller R. The BC Community Pharmacy Asthma Study: astudy of clinical, economic and holistic outcomes influenced by an asthmacare protocol provided by specially trained community pharmacists inBritish Columbia. Canadian Respir J. 2003;10:195–202.
81. Moudgil H, Marshall T, Honeybourne D. Asthma education and quality oflife in the community: a randomised controlled study to evaluate the impacton white European and Indian subcontinent ethnic groups fromsocioeconomically deprived areas in Birmingham. UK Thorax. 2000;55:177–83.
82. Pilotto LS, Smith BJ, Heard AR, McElroy HJ, Weekley J, Bennett P. Trial ofnurse-run asthma clinics based in general practice versus usual medicalcare. Respirology. 2004;9:356–62.
83. Pinnock H, Bawden R, Proctor S, Wolfe S, Scullion J, Price D, et al.Accessibility, acceptability and effectiveness of telephone reviews forasthma in primary care: randomised controlled trial. BMJ. 2003;326:477–9.
84. Price D, Haughney J, Lloyd A, Hutchinson J, Plumb J. An economicevaluation of adjustable and fixed dosing with budesonide/formoterol via asingle inhaler in asthma patients: the ASSURE study. Curr Med Res Opin.2004;20:1671–9.
85. Ryan D, Price D, Musgrave SD, Malhotra S, Lee AJ, Ayansina D, et al. Clinicaland cost effectiveness of mobile phone supported self monitoring ofasthma: multicentre randomised controlled trial. BMJ. 2012;344, e1756.
86. Schermer TR, Thoonen BP, van den Boom G, Akkermans RP, Grol RP,Folgering HT, et al. Randomized controlled economic evaluation of asthmaself-management in primary health care. Am J Respir Crit Care Med.2002;166:1062–72.
87. Shelledy DC, Legrand TS, Gardner DD, Peters JI. A randomized, controlledstudy to evaluate the role of an in-home asthma disease managementprogram provided by respiratory therapists in improving outcomes andreducing the cost of care. J Asthma. 2009;46:194–201.
88. Sundberg R, Tunsater A, Palmqvist M, Ellbjar S, Lowhagen O, Toren K. Arandomized controlled study of a computerized limited education programamong young adults with asthma. Respir Med. 2005;99:321–8.
89. van der Meer V, van den Hout WB, Bakker MJ, Rabe KF, Sterk PJ, AssendelftWJ, et al. Cost-effectiveness of Internet-based self-management comparedwith usual care in asthma. PLoS One. 2011;6, e27108.
90. Yilmaz A, Akkaya E. Evaluation of long-term efficacy of an asthma educationprogramme in an out-patient clinic. Respir Med. 2002;96:519–24.
91. Yoon R, McKenzie DK, Bauman A, Miles DA. Controlled trial evaluation of anasthma education programme for adults. Thorax. 1993;48:1110–6.
92. Bodenheimer T, Wagner E, Grumbach K. Improving primary care for patientswith chronic illness: the chronic care model. JAMA. 2002;288:1775–9.
93. Ring N, Jepson R, Hoskins G, Wilson C, Pinnock H, Sheikh A, et al.Understanding what helps or hinders asthma action plan use: a systematicreview and synthesis of the qualitative literature. Pat Ed Counsel.2011;85:e131–43.
94. Fleming E, Gillibrand W. An exploration of culture, diabetes, and nursing inthe South Asian community: a metasynthesis of qualitative studies.J Transcult Nurs. 2009;20:146.
95. Nam S, Janson SL, Stotts NA, Chesla C, Kroon L. Effect of culturally tailoreddiabetes education in ethnic minorities with type 2 diabetes: a meta-analysis. J Cardiovasc Nurs. 2012;27:505–18.
96. Marshall IJ, Wolfe CDA, McKevitt C. Lay perspectives on hypertension anddrug adherence: systematic review of qualitative research. BMJ.2012;344, e3953.
97. Pérez-Escamilla R, Hromi-Fiedler A, Vega-López S, Bermúdez-Millán A,Segura-Pérez S. Impact of peer nutrition education on dietary behaviors andhealth outcomes among Latinos: a systematic literature review. J Nutr EducBehav. 2008;40:208–26.
98. GRADE Working Group. Grading quality of evidence and strength ofrecommendations. BMJ. 2004;328:1490–4.
99. Ayling K, Brierley S, Johnson B, Heller S, Eiser C. How standard is standardcare? Exploring control group outcomes in behaviour change interventionsfor young people with type 1 diabetes. Psychol Health. 2015;3:85–103.
100. de Bruin M, Viechtbauer W, Schaalma HP, Kok G, Abraham C, Hospers HJ.Standard care impact on effects of highly active antiretroviral therapyadherence interventions: a meta-analysis of randomized controlled trials.Arch Intern Med. 2010;170:240–50.
• We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript atwww.biomedcentral.com/submit
Submit your next manuscript to BioMed Central and we will help you at every step:
Pinnock et al. BMC Medicine (2017) 15:64 Page 32 of 32