Developing Behavior Change Interventions for Self-Management in Chronic Illness: An Integrative Overview
Full text
Special Issue: Adjus men o Ch onic Illness
O iginal A icles and Re iews
De eloping Beha io Change
In e en ions o Sel -Managemen
in Ch onic Illness
An In eg a i e O e iew
Ve a A aújo-Soa es,
1,2
Nelli Hankonen,
3
Jus in P esseau,
4,5,6
Angela Rod igues,
1,7
and Falko F. Snieho a
1,7
1
Ins i u e o Heal h & Socie y, Facul y o Medical Sciences, Newcas le Uni e si y, Newcas le upon Tyne, UK
2
School o Psychology, Facul y o Medical Sciences, Newcas le Uni e si y, Newcas le upon Tyne, UK
3
Facul y o Social Sciences, Uni e si y o Tampe e, Finland
4
Clinical Epidemiology P og am, O awa Hospi al Resea ch Ins i u e, O awa, Canada
5
School o Epidemiology, Public Heal h and P e en i e Medicine, Uni e si y o O awa, Canada
6
School o Psychology, Uni e si y o O awa, Canada
7
Fuse. The UK Clinical Resea ch Collabo a ion Cen e o T ansla ional Resea ch in Public Heal h
Abs ac : Mo e people han e e a e li ing longe wi h ch onic condi ions such as obesi y, ype 2 diabe es, and hea disease. Beha io change
o e ec i e sel -managemen can imp o e heal h ou comes and quali y o li e in people li ing wi h such ch onic illnesses. The science o
de eloping beha io change in e en ions wi h impac o pa ien s aims o op imize he each, e ec i eness, adop ion, implemen a ion, and
main enance o in e en ions and igo ous e alua ion o ou comes and p ocesses o beha io change. The de elopmen o new se ices and
echnologies o e s oppo uni ies o enhance he scope o deli e y o in e en ions o suppo beha io change and sel -managemen a scale.
He ein, we e iew key con empo a y app oaches o in e en ion de elopmen , p o ide a c i ical o e iew, and in eg a e hese app oaches in o
a p agma ic, use - iendly amewo k o igo ously guide decision-making in beha io change in e en ion de elopmen . Mo eo e , we highligh
no el eme ging me hods o apid and agile in e en ion de elopmen . On-going p og ess in he science o in e en ion de elopmen is needed
o emain in s ep wi h such new de elopmen s and o con inue o le e age beha io al science’s capaci y o con ibu e o op imizing
in e en ions, modi y beha io , and acili a e sel -managemen in indi iduals li ing wi h ch onic illness.
Keywo ds: Beha io change, in e en ion de elopmen , complex in e en ions
Li e expec ancy con inues o inc ease wo ldwide, wi h he
global a e age li e expec ancy ha ing inc eased by 5yea s
be ween 2000 and 2015 (Wo ld Heal h O ganiza ion,
2014a). Howe e , non-communicable condi ions such as
ca dio ascula disease, espi a o y disease, cance , and dia-
be es ha e also inc eased since 2000 in e e y egion o he
wo ld and a e now he mos p e alen causes o mo ali y
and mo bidi y (Wo ld Heal h O ganiza ion, 2014a,
2014b). Ch onic non-communicable condi ions sha e
beha io al isk ac o s such as obacco smoking, poo die ,
and physical inac i i y (Lim e al., 2012). These condi ions
a e also associa ed wi h an inc eased isk o unde mining
men al heal h (Moussa i e al., 2007). Mul imo bidi y is
also p e alen and heal h beha io s can bene i pa ien s
by posi i ely impac ing on mo e han one condi ion
(Ba ne e al., 2012). Sel -managemen is hus a complex
endea o , in ol ing adhe ence o ea men , change o
mul iple heal h beha io s, and egula con ac wi h heal h-
ca e p o ide s (Depa men o Heal h, 2012; Schulman-
G een e al., 2012).
In e en ions add essing isk ac o s and suppo ing
beha io change o he e ec i e sel -managemen o
ch onic condi ions can make a conside able di e ence o
heal h and well-being and educe he cos s o deli e ing
heal h ca e o an aging popula ion li ing longe wi h
ch onic condi ions (OECD/EU, 2016). In he US, 157 mil-
lion people a e p edic ed o li e wi h ch onic condi ions
by 2020. Popula ion aging aises capaci y conce ns o
heal hca e sys ems, in hei cu en con igu a ions, o
cope wi h he inc easing bu den o ch onic condi ions
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)] h ps://doi.o g/10.1027/1016-9040/a000330
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
(Bodenheime , Chen, & Benne , 2009; NHS England,
2016). The e is consensus o he need o in e en ions
o suppo indi iduals and popula ions by a ge ing he p e-
en ion and sel -managemen o ch onic disease (Boon
e al., 2014) and o he key ole o beha io change in e -
en ions in his p ocess (Ha deman, Su on, Michie, & Kin-
mon h, 2004).
Wha Is a Heal h Beha io Change
In e en ion?
In e en ions a e coo dina ed se s o ac i i ies and ech-
niques in oduced a a gi en ime and place o change
he beha io o indi iduals, communi ies, and/o popula-
ions h ough a hypo hesized o known mechanism (NICE,
2007,2014). The heal h o popula ions and he indi iduals
wi hin hem is in luenced by a complex sys em o de e mi-
nan s, om indi idual li es yle ac o o communi y in lu-
ences, h ough li ing, wo king, and social condi ions
(Dahlg en & Whi ehead, 2006). Heal h beha io change
in e en ions can be a ge ed a a combina ion o le els:
policy (e.g., laws and egula ion), communi y (e.g., neigh-
bo hoods), mac o-en i onmen s (e.g., oo ou le s o ans-
po links), mic o-en i onmen al (e.g., choice a chi ec u e
in shops), ins i u ional (e.g., schools and employe s), in e -
pe sonal ( amilies and social ne wo ks), and/o in ape -
sonal (e.g., weigh loss p og am o he apy) le el (A aújo-
Soa es & Snieho a, 2017; Hollands e al., 2017;McLe oy,
Bibeau, S eckle , & Glanz, 1988).
Heal h beha io change in e en ions a e usually com-
plex (C aig e al., 2008). Wha makes an in e en ion com-
plex is he numbe and complexi y o i s in e ac ing
componen s, he beha io s in ol ed, he o ganiza ional
g oup, and indi idual le els a ge ed and he ou comes as
well as he deg ee o lexibili y o ailo ing pe mi ed. The
TIDieR checklis (Ho mann e al., 2014) was de eloped
o imp o e he comple eness o epo ing, and ul ima ely
he eplicabili y, o in e en ions by desc ibing: (a) a a io-
nale o heo y desc ibing he goals o he in e en ion ele-
men s, (b) he con en in e ms o beha io change
me hods (Adams, Giles, McColl, & Snieho a, 2014;Hol-
lands e al., 2017; Kok e al., 2016; Michie, Richa dson,
Johns on, Ab aham, F ancis, Ha deman, e al., 2013), ma e-
ials, and p ocedu es, (c) p o ide (s) (including quali ica ion
and aining needed), (d) modes o deli e y (e.g., p o ided
ace- o- ace o h ough a digi al pla o m) o indi iduals
o g oups (Domb owski, O’Ca oll, & Williams, 2016),
(e) loca ion and equi ed in as uc u e, ( ) iming and dose,
and (g) any planned mechanisms o ailo ing o adap a ion
o he in e en ion o needs/ ea u es o he ecipien (s). An
ex ension o he TIDieR guideline o epo ing popula ion
heal h and policy in e en ions has ecen ly been published
(Campbell e al., 2018). In e en ions also o en include
addi ional componen s o build and sus ain appo and
engagemen h ough in e pe sonal s yles (Hagge & Ha d-
cas le, 2014) o ea u es such as gami ica ion in digi al
in e en ions (Cugelman, 2013). Heal h beha io change
in e en ion de elopmen is he p ocess o deciding he
op imal combina ion o hese ea u es and he anspa en
epo ing o hese decisions.
Wha Makes a Good Heal h Beha io
Change In e en ion?
“P imum non noce e”(eng. “ i s , do no ha m”). The p in-
ciple o non-male icence is he single mos impo an c i e-
ion o any heal h in e en ion (C aig e al., 2008;Michie,
A kins, & Wes , 2014). In addi ion, a good in e en ion
should be designed o impac , should be e aluable, should
no inc ease social inequali ies, and should ha e a demon-
s able bene i o e exis ing in e en ions and se ices.
The impac o in e en ions on he heal h o he a ge
audience can be illus a ed h ough he RE-AIM (Reach,
E ec i eness, Adop ion, Implemen a ion, Main enance)
model (Glasgow, Vog , & Boles, 1999). Reach e e s o
he p opo ion o he in ended a ge popula ion ha can
ac ually be and is ul ima ely eached wi h an in e en ion;
E ec i eness e e s o he bene icial and unin ended e ec
he in e en ion achie es on key ou comes unde eal-
wo ld condi ions, including cos -e ec i eness; Adop ion
e e s o he up ake o he in e en ion by he s a , se ings,
and o ganiza ions; Implemen a ion e e s o he deg ee o
which he in e en ion can/will be deli e ed consis en ly
and wi h ideli y o e ime and se ing; and Main enance
e e s o he sus ainabili y o in e en ion e ec i eness in
indi iduals and se ings o e ime. To achie e his, in e -
en ions should be based on he bes a ailable e idence-
based heo y and di ec e idence o op imize impac and
o model whe he and how he in e en ion is likely o c e-
a e bene i (Ba holomew Eld edge e al., 2016;C aige al.,
2008;Wigh ,Wimbush,Jepson,&Doi,2016). Op imizing
RE-AIM is aided by maximizing he accep abili y and easi-
bili y o in e en ion p ocedu es and ma e ials (Lancas e ,
2015). This is bes achie ed h ough he ac i e in ol emen
o key s akeholde s in all s ages, om de elopmen h ough
o e alua ion o accep abili y and easibili y in ini ial pilo /
easibili y s udies as well as subsequen e icacy/e ec i e-
ness, implemen a ion and main enance e alua ions (C aig
e al., 2008;O’B ien e al., 2016).
A p e equisi e o a good in e en ion is i s “e aluabili y,”
ha is, whe he i s e ec can be obus ly e alua ed. In e -
en ions wi h a clea de ini ion, elabo a ed logic model,
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
8 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
and de ined p ima y and in e media e a ge s a e easie o
e alua e, which in u n acili a e unde s anding i , how and
o whom an in e en ion wo ks, acili a ing op imiza ion
and he eby con ibu ing o he accumula ion o knowledge
(Le i on, Khan, Rog, Dawkins, & Co on, 2010; Ogil ie
e al., 2011; Windso , 2015).
Good in e en ions should no inc ease social inequali-
ies in heal h (Lo enc, Pe ic ew, Welch, & Tugwell,
2013). Heal h and heal hy li e expec ancy a e s ongly
ela ed o socioeconomic s a us (OECD/EU, 2016). To
a oid in e en ion-gene a ed inequali ies, in e en ion
design should be sensi i e o PROGRESS indica o s (Place
o esidence, Race/e hnici y/cul u e/language, Occupa ion,
Gende /sex, Religion, Educa ion, Socioeconomic s a us,
and Social capi al (T. E ans & B own, 2003;O’Neill
e al., 2014). In e en ion de elope s need o conside
up ake, usage, and le el o indi idual agency equi ed o
minimize he po en ial o gene a ing inequali ies (Adams,
My on, Whi e, & Monsi ais, 2016).
Finally, good in e en ions should c ea e inc emen al
bene i o e al eady exis ing in e en ions and se ices.
In e en ions ha e high u ili y i hey add ess gaps in p o i-
sion, inc ease he po en ial o be implemen ed and sus-
ained, educe cos s and/o add ess ba ie s compa ed
wi h p e ious and exis ing in e en ions. In pa icula , scal-
able in e en ions, ha is, e ec i e in e en ions which
ha e a a each and modes cos s, add ess he need o
solu ions which ha e ew esou ce and geog aphic ba ie s
and can be p o ided o la ge numbe s o indi iduals and
communi ies (Mila , King, Bauman, & Redman, 2013).
The heal h esea ch landscape is no sho o beha io al
in e en ions. In ligh o his, a ho ough en i onmen al
scan analysis is needed o iden i y gaps in p o ision o
ensu e ha new in e en ions ha e a ai chance o make
a posi i e con ibu ion o heal h and well-being. Unde -
s anding usual ca e and compe ing in e en ions in a gi en
se ing enables s a egic decision-making abou po en ial
inc emen al bene i o a new in e en ion. Inc easingly,
he bounda ies o usual ca e a e no longe physical o geo-
g aphical. As in e en ions can ake yea s o be de eloped
and ully e alua ed, his analysis o he heal h in e en ion
ma ke should also conside pilo s udies and e alua ion
s udies unde way, o example, by analyzing ial egis ies
and g ey li e a u e (Adams, Hillie -B own, e al., 2016).
The P ocess o In e en ion
De elopmen
The e is a ange o amewo ks ha can in o m he de el-
opmen o heal h beha io change in e en ions such as he
MRC guidance o he de elopmen and e alua ion o
complex in e en ions (C aig e al., 2008), In e en ion
mapping (IM; Ba holomew Eld edge e al., 2016), Theo y
In o med Implemen a ion In e en ion (S. D. F ench
e al., 2012), PRECEDE-PROCEDE (G een & K eu e ,
2005), he Pe son-Based App oach (Ya dley, Mo ison,
B adbu y, & Mulle , 2015), he 6SQuID app oach in quali y
in e en ionde elopmen (Wigh e al.,2016), e idence-
guided co-design (O’B ien e al., 2016), he Knowledge-
o-Ac ion (KTA) cycle (G aham e al., 2006), he ORBIT
model (Czajkowski e al., 2015), he Expe imen al Medicine
Model (Shee an, Klein, & Ro hman, 2017), Mul iphase op i-
miza ion s a egy (MOST; Collins, Mu phy, & S eche ,
2007), and he Beha io Change Wheel (Michie, an S a-
len, & Wes , 2011; see Appendix A o a summa y o ame-
wo ks and hei pu pose). While each has a di e en ocus
and app oach, hey con e ge on a co e se o key s eps ha
include: analyzing he p oblem and de eloping an in e en-
ion objec i e, causal modeling, de ining in e en ion ea-
u es, de eloping a logic model o change, de eloping
ma e ials and in e ace, and empi ical op imiza ion ol-
lowed by ou come and p ocess e alua ion and implemen a-
ion. In e en ion de elopmen is i e a i e, ecu si e, and
cyclical a he han linea . De elope s may need o go back
and o h be ween s eps o achie e he op imal in e en ion
de ini ion pai ed wi h mos app op ia e logic model o
change wi hin a ailable esou ces.
In e en ion de elopmen should ideally be led by an
in e disciplina y Planning and De elopmen G oup ep e-
sen ing ele an expe ise (e.g., clinical ca e, psychology,
policy, sociology, heal h economics, epidemiology, se ice
design) and key s akeholde s (e.g., ci izens, pa ien s, ca e s,
heal hca e p o essionals, deli e e s, commissione s, policy-
make s, unde s) o unde s and he con ex o in e ening
and o make s a egic decisions ha e lec scien i ic e i-
dence and he p e e ences and iews o hose o whom
he in e en ion is de eloped and hose whose inpu is
needed o adop and implemen he in e en ion (Ba holo-
mew Eld edge e al., 2016;Wi emane al.,2017). To docu-
men he sequence o decisions in ol ed in in e en ion
de elopmen , wo kbooks can help o eco d in e en ion
de elopmen s eps, c ucial decisions, and he p ocess and
in o ma ion in o ming hese decisions (Ba holomew
Eld edge e al., 2016); Appendix B con ains a comp ehen-
si e lis o Key Conside a ions o he Repo ing o In e -
en ion De elopmen ). Nex , we add ess each key s ep in
de ail:
A. Analyzing he P oblem and De eloping
an In e en ion Objec i e
The de elopmen o a beha io change in e en ion es s
on a ounda ion o a ho ough analysis o he p oblem ha
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 9
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
he in e en ion de elope s aim o sol e and a clea de ini-
ion o in e en ion objec i es. PRECEDE/PROCEED was
concei edin he1970s o guide policymake s and in e en-
ion planne s in analyzing he likely cos s and bene i s o
heal h p og ams. I consis s o wo main pa s: PRECEDE
desc ibes an “educa ional diagnosis”and is an ac onym
o P edisposing, Rein o cing and Enabling Cons uc s in
Educa ional Diagnosis and E alua ion. PROCEED e e s
o an “ecological diagnosis”and s ands o Policy, Regula-
o y, and O ganiza ional Cons uc s in Educa ional and
En i onmen al De elopmen (G een & K eu e , 2005). I
p o ides he i s amewo k o analyzing how heal h and
quali y o li e ela e o beha io , physiology, and en i on-
men al ac o s and o he iden i ica ion o p edisposing,
ein o cing, and enabling ac o s o beha io s, which can
be ackled wi h in e en ions.
Many in e en ion de elopmen amewo ks include a
Needs Assessmen , which in ol es assessing he heal h
p oblem and i s likely beha io al, social, and en i onmen al
causes. This ini ial s age in ol es he iden i ica ion and de -
ini ion o he sequence o beha io s needed o modi y
heal h ou comes he eby iden i ying in e media e ou comes
ele an o he hypo hesized mechanisms o he in e en-
ion (Ba holomew Eld edge e al., 2016), ha is, “who
needs o do wha di e en ly, when, whe e, how?”(S. D.
F ench e al., 2012). The pe son-based app oach o in e -
en ion de elopmen (Ya dley e al., 2015)aims og ound
he de elopmen o beha io change in e en ions in an
unde s anding o he pe spec i e and psychosocial con ex
o he people who will use hem. Beha io s a ge ed o
change a e embedded in a ne wo k o mul iple beha io s,
some o which may acili a e o con lic wi h each o he
(P esseau, Tai , Johns on, F ancis, & Snieho a, 2013).
Unde s anding how a a ge heal h beha io i s alongside
o he beha io s, and he essen ial p epa a o y beha io s
equi ed, can help o iden i y he mos iable beha io al a -
ge s o an in e en ion ha may ex end beyond he single
beha io al ou come o he in e en ion. Ta ge beha io s
need o be de ined in con ex and in e y speci ic e ms,
ideally in e ms o Ta ge (s), Ac ion, Con ex (s), Time(s)
and ac o s (Fishbein, 1967; F ancis & P esseau, 2018),
including he in e - ela ionships be ween beha io s and
ac o s. Conside a ions abou changeabili y guide he p io -
i iza ion and selec ion o a ge beha io s and a ge ed
an eceden s o beha io , o example, which changes a e
achie able based on cu en e idence and heo y, and
how much impac would such changes ha e o he key ou -
comes (Ba holomew Eld edge e al., 2016;Czajkowski
e al., 2015; Shee an e al., 2017; Wigh e al., 2016).
Key s akeholde s should con ibu e om he beginning
o de ining he ini ial p oblem, a he han he in e en ion
de elopmen being a esea che -d i en op-down design
ask. S akeholde in ol emen helps o b idge be ween
he e idence and he local con ex and ensu es owne ship,
accep abili y, and widesp ead suppo o he in e en ion
essen ial o implemen a ion (O’B ien e al., 2016). In some
ins ances, in e en ion p io i ies a e d i en by use s o
pa ien o ganiza ion. Such p io i ies can be obus ly su -
aced, o example, in ol ing James Lind Alliance (2017)
me hods ha b ing clinicians, pa ien s, and ca e s oge he
o use a o mal me hodological app oach o gene a e
esea ch p io i ies ha a e impo an o pa ien s ac oss a
ange o se ings.
B. De ining he Scien i ic Co e o he
In e en ion
Heal h beha io change in e en ions a e guided by a logic
model o a heo y o change ha combines he in e en ion
echniques used o a ge causal mechanisms in o a com-
p ehensi e and es able se o assump ions (Moo e e al.,
2015). Th ee s eps go hand in hand and a e bes desc ibed
as one i e a i e p ocess:(i) causal modeling o he p oblem,
(ii) de ining in e en ion ea u es, and (iii) o mula ing a
logic model o change o he in e en ion (Ba holomew
Eld edge e al., 2016; Moo e e al., 2015; Wigh e al., 2016).
Decisions need o be made on me hod(s) and mode(s) o
deli e y, beha io change echnique(s), p o ide (s), loca ion
(s), iming, dose, pe sonaliza ion and hypo hesized causal
mechanisms o op imize each, (cos -) e ec i eness, adop-
ion, implemen a ion, and main enance. These design deci-
sions should be eco ded and made explici o cla i y he
con ibu ion ha all new in e en ions make o p e ious
e idence. The p ocess should be led by a pa icipa o y plan-
ning g oup ep esen ing s akeholde s such as use s and
commissione s o he in e en ion and he esea ch eam
o i e a i ely build a hypo hesis o change and make design
decisions based on scien i ic e idence and he needs o he
a ge audience. This ensu es he ele ance o he de el-
oped solu ion and c ea es co-owne ship as a esul o
cop oduc ion.
(i) Causal Modeling
The iden i ica ion o causal and con ex ual ac o s a ec ing
sel -managemen beha io s is a key s ep in in e en ion
de elopmen . Beha io is he esul o a complex ecologic
sys em o in luences which ange om p oximal indi idual,
cogni i e, and emo ional ac o s o social and communi y
in luence up o mo e dis al ac o s such as ca e deli e y sys-
ems (e.g., access o specialis medical ca e), li ing and
wo king condi ions (employmen , en i onmen , educa ion,
and housing), and socioeconomic, cul u al, and en i on-
men al condi ions (e.g., legisla ion; Dahlg en & Whi ehead,
2006). Modi iable ac o s ha ha e a s ong ela ionship o
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
10 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
he a ge beha io a e po en ial a ge s o in e en ions
(Michie, an S alen, e al., 2011; Wigh e al., 2016).
Beha io change app oaches end o ope a e on he
assump ion ha in e en ions a ec beha io by modi ying
social, en i onmen al, and/o cogni i e p edic o s o he
a ge beha io . In e en ions a e hen hough o ope a e
h ough a sequen ial causal model beginning om p edic-
o s o beha io , o beha io , o physiological changes and
e en ually leading o heal h ou come(s) (Ha deman e al.,
2005). IM (Ba holomew Eld edge e al., 2016) p oposes
o wo k backwa d om he a ge ed heal h p oblems (and
ha impac on quali y o li e), o he beha io and en i on-
men al ac o s ha shape hese heal h p oblems, and inally
o he p edic o s o he causal beha io al and en i onmen-
al isk ac o s. P edic o s a e a ed by ele ance and
changeabili y o de e mine hei p io i y o inclusion in
he in e en ion (Ba holomew Eld edge e al., 2016;Ya d-
ley e al., 2015).
Li e a u e e iews a e ecommended o syn hesize e i-
dence o he causes and p edic o s o he a ge beha io
(Ba holomew Eld edge e al., 2016;C aige al.,2008), ide-
ally, wi h sys ema ic sea ches (C aig e al 2008). In e iew-
ing exis ing e idence, ensions be ween s eng h and igo
and applicabili y o e idence can occu . Decisions abou
e idence e iews should be s a egically d i en o add ess
key unce ain ies. While usually sys ema ic e iews o s ud-
ies wi h low isk o bias a e p e e able, he mos ele an
e idence in o ming an in e en ion migh be supplemen ed
by g ey li e a u e such as local go e nmen epo s o hos-
pi al eco ds (Adams, Hillie -B own, e al., 2016;O’B ien
e al., 2016;Rod igues,Snieho a,Bi ch-Machin,Oli ie ,
&A aujo-Soa es,2017). Re iews may highligh he deg ee
o which esul s a e likely o be ans e able o he p esen
con ex bu o en addi ional empi ical esea ch is needed o
iden i y he mos impo an p edic o s and o es hei sen-
si i i y o con ex ual ea u es o communi ies, se ices, o
geog aphies.
Theo y has a cen al ole in his p ocess. In e en-
ion de elopmen is o en based on ope a ionalizing he
p inciples om a single heo y and selec ing in e en ion
echniques wi h he po en ial o modi y he heo e ical p e-
dic o s o beha io . This app oach can be use ul when
he e is insu icien esou ce o conside collec ing u he
empi ical da a and gi en he inhe en ly e idenced-based
na u e o a heo y, in ha i has been success ully
applied o di e en beha io s and/o in di e en con ex s
(D. P. F ench, Da ke , E es, & Snieho a, 2013). Howe e ,
his app oach is limi ed when he obse ed p ospec i e
ela ionships conside ed o he selec ion o in e medi-
a e in e en ion a ge s a e no s ong enough o in e -
en ions changing beha io al p edic o s o achie e
changes in beha io (Snieho a, P esseau, & A aújo-Soa es,
2014).
When no app op ia e heo y can be iden i ied, o when
mo e han one may seem ele an , in e en ion de elope s
can use he Theo e ical Domains F amewo k (TDF) o
o ganize e idence abou key ba ie s and enable s and link
back o ele an heo ies (F ancis, O’Conno , & Cu an,
2012; Heslehu s e al., 2014).TheTDFisasimple ool
de eloped h ough e iew and consensus me hods o
desc ibe he mos common explana o y cons uc s in
beha io al heo ies o ganized in o 14 domains: knowledge,
skills, social in luences, memo y, a en ion and decision
p ocesses, social/p o essional ole and iden i y, ein o ce-
men , belie s abou capabili ies, belie s abou conse-
quences, op imism, in en ion, goals, beha io al egula ion,
emo ion, en i onmen al con ex and esou ces (Cane,
O’Conno , & Michie, 2012; Michie e al., 2005). The TDF
can be used o in o m bo h quali a i e and quan i a i e
s udies wi h he aim o unde s and key p edic o s o beha -
io and o iden i y he mos ele an heo e ical app oach
(Beens ock e al., 2012; Laine, A aújo-Soa es, Haukkala, &
Hankonen, 2017; P esseau, Schwalm, e al., 2017).
Addi ional empi ical s udies can inc ease unde s anding
o he key in luences o he beha io in he a ge g oup.
Fo example, a su ey iden i ying he mos impo an co e-
la es o physical ac i i y beha io and in en ion could help
in selec ing he key ba ie s and enable s o a ge wi h
an in e en ion (Hankonen, Heino, Kujala, e al., 2017;
P esseau,Schwalm,e al.,2017; Snieho a, Schwa ze ,
Scholz, & Schüz, 2005). Quali a i e in e iews o n-o -1
s udies can p o ide an indi idualized assessmen o ba ie s
and needs (McDonald e al., 2017; Rod igues, Snieho a,
Bi ch-Machin, & A aujo-Soa es, 2017; Ya dley e al.,
2015). A key weakness o app oaches based on co ela ion
is he lack o causa ion and he p oblem o a enua ion, ha
is, la ge changes in p edic o s a e needed o achie e mod-
es changes in beha io (Snieho a, 2009).
Whe e mul iple beha io s a e a ge ed, a p ocess o es -
ing mul iple heo ies ac oss mul iple beha io s can be used
o iden i y he mos consis en ly p edic i e cons uc s
wi hin hei heo ies ac oss beha io s, hen heo ize and
es how such heo ies and hei cons uc s can be com-
bined, o example, in o a dual p ocess model (P esseau,
Johns on, e al., 2014) o in o m a logic model (P esseau,
Haw ho ne, e al., 2014). This app oach combines he
s eng h o p eexis ing heo y (and i s es ed media ing
and mode a ing mechanisms) wi h he empi ical compa -
ison o heo y ac oss beha io s o acili a e he selec ion
o beha io (s) and heo y upon which o u he de elop
he in e en ion. Theo y is used o add ess unce ain ies
and may include heo e ical ideas ha a e no di ec ly
ela ed o beha io , o example, heo ies o pe suasion
(Pe y & Cacioppo, 1986) o o symp om ecogni ion (Pe e -
sen, an den Be g, Janssens, & an den Be gh, 2011). Fig-
u e 1p o ides wo examples o in e en ion de elopmen .
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 11
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
(ii) De ining In e en ion Fea u es
In e en ion echniques (e.g., o change beha io , cogni-
ions, pe cep ions, o en i onmen al a iables) a e selec ed
based on e idence o hei e ec i eness in changing he
iden i ied causal and con ex ual ac o s in luencing he a -
ge beha io . In e en ion de elopmen app oaches di e
in how hey app oach he analysis o causal ac o s ocuss-
ing on in e en ion a ge s o echniques (Michie e al.,
2014; Shee an e al., 2017;Webb,Michie,&Snieho a,
2010). Ta ge -based app oaches iden i y modi iable p edic-
o s o beha io , whe eas echnique-based app oaches
ocus on in e en ion echniques hemsel es and con ex ual
modi ica ions which di ec ly in luence beha io (Webb
e al., 2010).
As highligh ed in he knowledge c ea ion unnel wi hin
he KTA cycle (G aham e al., 2006), use o e iew e i-
dence se s he ounda ion and p e en s epea ing p e i-
ously unsuccess ul beha io change echniques o
wi hholding in e en ion s a egies wi h demons a ed
e ec i eness in changing beha io . In some cases, e idence
syn hesis may iden i y ha a sui able in e en ion al eady
exis s ha could be e o i ed (i.e., ans o med o use in
a no el con ex and o in a no el popula ion) a he han
e-in en ed. Bu sys ema ic e iews o andomised con-
olled ials (RCTs) o in e en ions wi h simila aims do
no always p o ide su icien answe s. Fo example, in he
de elopmen o he “Le ’sMo eI ”in e en ion o change
physical ac i i y and seden a y beha io s in oca ional
school, a sys ema ic e iew (Hynynen e al., 2016)in o med
he designe s abou wha wo ks in ge ing olde adolescen s
mo e ac i e, bu i was no su icien . A ange o o he el-
e an sou ces o e idence con ibu ed o i s de elopmen
including exis ing e idence ega ding he se ing (school-
based heal h p omo ion), e idence abou he a ge beha -
io using a ange o me hods and esea ch on simila in e -
en ions in o he age g oups and popula ions con ibu ed o
in o m he in e en ion design.
Di e en le els o e idence answe di e en ques ions.
While sys ema ic e iews o RCTs o beha io change in e -
en ions p o ide he s onges e idence o e ec i eness,
hey o en say li le abou each, adop ion, and implemen a-
ion ou side o a esea ch s udy o abou longe - e m main-
enance (Domb owski e al., 2012). Likewise e idence om
igo ous s udies conduc ed in e y di e en se ings o in
communi ies wi h di e en ea u es may be applicable o
he local needs when e o i ed. E idence syn hesis should
be s a egic and sequen ial, de eloping an i e a i e unde -
s anding o how o op imize he in e en ion (Michie
e al., 2014). Whe e p e ious heal h beha io change in e -
en ions had he e ogeneous e ec s, i is o en possible o
code beha io change echniques and o he in e en ion
ea u es such as modes o deli e y (Ab aham & Michie,
2008; Adams e al., 2014;Koke al.,2016;Michie,Ash o d,
e al., 2011; Michie, Richa dson, Johns on, Ab aham, F an-
cis, & Ha deman, 2013) and o explo e whe he such ea-
u es a e associa ed wi h in e en ion e ec i eness
(Domb owski e al., 2012). Such an in e en ion ea u es
e iew-based app oach begins by iden i ying in e en ion
echniques and o he TIDIER ea u es (Ho mann e al.,
2014) o in e en ions o a gi en heal h beha io in a sys-
ema ic e iew o ials. TIDIER ea u es, including beha -
io change echniques and o he in e en ion echniques
can hen be coded wi hin in e en ions in he e iew o es
which echniques and combina ions o hese a e associa ed
wi h g ea e e ec i eness in o he se ings. E en hough i-
als o in e en ions make causal s a emen s o e ec i e-
ness, he e alua ion o in e en ion echniques wi hin he
e iew is co ela ional and should be ea ed wi h due ca e.
Ne e heless, his app oach can help o combine e idence
o in e en ion s a egies ha ha e been ound o be e ec-
i e in o he se ings and/o using heo y o in o m he
selec ion o in e en ion echniques.
In addi ion o e iew-based iden i ica ion o e ec i e
in e en ion ea u es, some app oaches p omo e an expe i-
men al me hod o in e en ion de elopmen o es ablish
causal e idence o he hypo hesized change by iden i ying
he po en ial modi iable causal ac o s and assessing
whe he changes in he a ge beha io occu as a esul
o manipula ing he p edic i e ac o (s) (Shee an e al.,
2017). The emphasis is on unde s anding he mechanisms
o change and using expe imen al designs o obus ly cla i y
how o change hese and in eg a ing his knowledge in o
applied esea ch. En i onmen al in e en ions a ge ing
poin -o -choice decisions such as s ai s e sus escala o
use(Ryan,Lyon,Webb,E es,&Ryan,2011) and on- he-
spo oppo uni ies o egis e o o gan dona ion (Li e al.,
2017), nudges (Hollands e al., 2013; Ma eau, Ogil ie,
Roland, Suh cke, & Kelly, 2011) o poin o sale decisions
(Dolan e al., 2012) a e mo e likely o be in o med by expe -
imen al han by co ela ional conside a ions.
Some in e en ion echniques may be e ec i e when
es ed in an RCT bu no widely accep able by acili a o s
o a ge audience alike, while o he in e en ion ech-
niques migh be highly accep able bu show smalle e ec
sizes. Accep abili y can be de ined as a “mul i- ace ed con-
s uc ha e lec s he ex en o which people deli e ing o
ecei ing a heal hca e in e en ion conside i o be app o-
p ia e, based on an icipa ed o expe ienced cogni i e and
emo ional esponses o he in e en ion”(Sekhon, Ca -
w igh , & F ancis, 2017,p.4). Engaging s akeholde s in
he de elopmen p ocess om ea ly on will inc ease he
po en ial o accep abili y. In e en ion p inciples ha a e
heo e ically sound and in line wi h good e idence, migh
s ill no be seen as accep able wi hou adap a ion o con ex
and audience. Fo example, some migh no be willing o
engage in planning in e en ions unless key modi ica ions
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
12 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
a e implemen ed o inc ease accep abili y and easibili y
(Wi eman e al., 2017). An icipa ed accep abili y o candi-
da e ea u es can be empi ically examined o in o m deci-
sions, o example, eache s’ iews on po en ial s a egies
o educe s uden si ing in schools was examined using a
mixed-me hods app oach (Laine e al., 2017). This example
also illus a es ha in addi ion o he main a ge g oup (s u-
den s), he en i onmen al agen s o “p o ide s”( eache s)
The In e en ions Suppo ing Long- e m Adhe ence and Dec easing ca dio ascula e en s (ISLAND) mul i -
cen e ial and heo y-based p ocess e alua ion in ol ed an in e en ion o suppo on-going medica ion
adhe ence and a endance o ca diac ehabili a ion ollowing a myoca dial in a c ion (MI) (I e s e al., 2017).
In e en ion
de elopmen conside ed exis ing
Coch ane
e iew
e idence
o
bo h hese
beha io s
(Ka mali
e al., 2014; Nieuwlaa e al., 2014), key ials o beha io change in e en ions (Snieho a, Scholz, & Schwa ze ,
2006), cos -e ec i eness da a (I o e al., 2012) and pilo ial da a (Schwalm e al., 2015). De elopmen also
in ol ed conduc ing quali a i e in e iews based on he Theo e ical Domains F amewo k wi h pa ien s pos -MI
o iden i y po en ial ba ie s and enable s, as well as quan i a i e analyses based on he Heal h Ac ion P ocess
App oach o iden i y mo i a ional and oli ional co ela es o beha io (P esseau, Schwalm, e al., 2017). These
sou ces o e idence in o med he basis o de eloping a logic model, beha io change echniques and modes o
deli e y o he in e en ion (I e s e al., 2017). An in e disciplina y eam was assembled in ol ing pa ne ing
wi h a design i m, pa ien s, a pa ien s akeholde o ganiza ion, p ima y and seconda y heal hca e p o ide s, and
esea che s (heal h psychologis s, s a is icians, heal h se ices esea che s, heal h economis s, implemen a ion
scien is s, and human ac o s enginee s) om he s a and h oughou o ensu e ha i could be implemen ed a
scale wi hin heal hca e sys ems. An explici use -cen ed design p ocess was used o i e a i ely de elop
ma e ials including de eloping pe sonas, p o o ype ma e ials, wo design cycles, pilo ing ma e ials using hink
aloud and semi-s uc u ed in e iews (Wi eman e al., 2017).
The Le ’s Mo e I (LMI) in e en ion aimed o inc ease physical ac i i y and dec ease excessi e seden a y
beha io among adolescen s – especially hose wi h insu icien PA le els (Hankonen e al., 2016), physical
ac i i y can p e en o delay onse o se e al li es yle- ela ed ch onic diseses such as ype 2 diabe es o hea
disease. The aim o he in e en ion de elopmen was o c ea e a easible, accep able, e ec i e and cos -
e ec i e school-based in e en ion ha could la e be scaled up. In e en ion de elopmen conside ed exis ing
e iew e idence o hese beha io s and school-based heal h p omo ion in e en ions, bu also ca ied ou a
sys ema ic e iew o he a ge g oup, beha io s, and con ex (Hynynen e al., 2016). De elopmen also
in ol ed conduc ing quali a i e analysis o in e iews o be e unde s and he ole o PA in daily li e o Finnish
oca ional s uden s, as well as analysis o pe sonal s o ies on key inci den s ela ed o PA change o e childhood
and adolescence. Fu he , quan i a i e analyses in o med by he The ec ical Domain F amewo k (F ancis e al.,
2012) aimed o iden i y he key co ela es o hese beha io s (Hankonen, Heino, Kujala, e al., 2017). As
some pa s o he in e en ion we e o be deli e ed by eache s, a mixed-me hods s udy o examine accep-
abili y o po en ial in e en ion s a egies was conduc ed among eache s (Laine e al., 2017). We conduc ed
e.g., scena io wo k wi h a g oup o expe s and s akeholde s, and wi h a s uden panel, did p ac ical small ials
o e.g., discussion exe cises wi h s uden s in o de o ge apid eedback o al e na i e p ac ical s a egies
wi hin he s uden p og am. This esul ed in he i s e sion o he in e en ion, he accep abili y). and easibili y
was in es iga ed in a andomised easibili y ial (Hankonen, Heino, Kujala, e al., 2017). An enhanced e sion
o he
in e en ion was hen
de eloped based on his eedback (Hankonen, Heino, Kujala, e al., 2017). An ad e -
isemen agency designed he ma e ials and he isual look o he in e en ion, in close collabo a ion wi h he
esea ch eam, including es ing wi h end-use s and a close linkage wi h heo y. An in e disciplina y eam
in ol ing esea che s (disciplines including social and heal h psychology, s a is ics, exe cise physiology and
measu e
men , spo s science, implemen a ion science, sociology), heal h p omo ion o ganisa ions, eache s,
s uden s, school heal h specialis s, e c. was assembled om he s a and hey con ened egula ly h oughou
he in e en ion de elopmen p ocess.
Figu e 1. In e en ion de elopmen examples.
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 13
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
ha deli e he in e en ion a e also he a ge o a “sec-
onda y”in e en ion, hence, hei iews and beha io s
should also be unde s ood. In implemen a ion science he
en i onmen al agen s a e he a ge o he in e en ion.
(iii) De eloping a Logic Model o Change
The MRC amewo k o he de elopmen and e alua ion
o complex in e en ions highligh s ha in e en ions
should be heo y-based (C aig e al., 2008). A common
misconcep ion is equa ing “ heo y”wi h “hypo hesis.”A
scien i ic heo y has been empi ically demons a ed o
explain beha io . I , while designing an in e en ion, he
eam concludes ha he e is a need o a ge a combina ion
o cons uc s om di e en heo ies ha ha e ne e been
es ed oge he , wha will ac ually happen is ha a speci ic
scien i ic hypo hesis ( ha can lead o a new heo y i suc-
cess ul) is being es ed, no a heo y.
I is use ul o c ea e a p og am’sscien i ichypo hesisin
e ms o he e idence-based mechanisms associa ed wi h
beha io and beha io change. In con as o o mal scien-
i ic heo ies, p og am heo ies a e p ac ical, conc e e wo k-
ing models and hypo heses o in e en ions, and a e
speci ic o each p og am o in e en ion. They (1)speci y
he in e en ion componen s, he in e en ion’sexpec ed
ou comes, and he me hods o assessing hose ou comes,
o en in he o m o a logic model, and (2) o e an in e en-
ion’s“hypo heses”( he a ionale and assump ions abou
mechanisms ha link p ocesses and inpu s o (bo h
in ended and unin ended) ou comes, as well as condi-
ions/con ex necessa y o e ec i eness; Da ido ,
Dixon-Woods, Le i on, & Michie, 2015).
This hypo hesis o change may be based on o in o med
by scien i ic heo ies, bu he main equi emen is o o -
malize he hypo hesized causal assump ions, de ail he
planned implemen a ion and heo ized mechanisms o
impac wi hin a se o ele an con ex s (C aig e al.,
2008). Theo y can also iden i y speci ic issues ha c ea e
ba ie s o in e en ion success (e.g., compe ing goals in
ime-limi ed GP sessions; P esseau, Snieho a, F ancis, &
Campbell, 2009). Ra he han using a single heo y o guide
in e en ion de elopmen , i is o en sensible o use heo y
o add ess he unce ain ies in he p ocess and o c ea e a
map o assump ions/hypo hesis linking heo ies and
e idence.
Acco ding o UK MRC Guidance, modeling an in e en-
ion be o e e alua ion p o ides he insigh s ha a e key o
in o ming he design o bo h he in e en ion and i s e al-
ua ion. Modeling may ake he o m o a p e ial economic
e alua ion es ing i he se o assump ions used o de elop
he in e en ions a e su icien o p o ide a good chance o
success ul impac . Mapping links be ween ou comes, de e -
minan s, change objec i es, and in e en ion echniques
e lec his p ocess o c ea ing he logic o in e en ion
(Ba holomew Eld edge e al., 2016). Fo example, in a
school-based in e en ion o p e en obesi y, pe o mance
objec i es (e.g., Communica e heal hy beha io messages
o pa en s and seek hei suppo ) a e mapped agains pe -
sonal (e.g., sel -e icacy) and ex e nal, en i onmen al p e-
dic o s (e.g., amily suppo ), and hus c ea ed ac ionable
change objec i es (e.g., con idence o seek pa en al suppo
and social ein o cemen om pa en s/ amily o in e es in
heal hy li es yles. These change objec i es become he a -
ge o in e en ion echniques (Lloyd, Logan, G ea es, &
Wya , 2011).
This p ocess should also in ol e he explici elabo a ion
o a “da k”logic model, ha is, a ca e ul elabo a ion o
po en ial pa hways h ough which he in e en ion may
lead o nega i e o ha m ul consequences (Bonell,
Jamal, Melendez-To es, & Cummins, 2014). This ex ends
beyond iden i ying po en ial ha ms by clea ly ou lin-
ing he mechanisms h ough which such ha ms may ake
place.
The Beha io Change Wheel (Michie, an S alen, e al.,
2011) is a pa icula ly use ul ecen ool o in eg a e heo y
and e idence and o b ing oge he s akeholde s in making
in e en ion design decisions. I is a me a-model o he
in e en ion de elopmen p ocess based on a comp ehen-
si e e iew and syn hesis o exis ing me hodological and
heo e ical app oaches om a ious disciplines. The Beha -
io Change Wheel links policy ca ego ies (guidelines, en i-
onmen al/social planning, communica ion/ma ke ing,
iscal measu es, egula ion, se ice p o ision and legisla-
ion) wi h in e en ion unc ions ( es ic ions, educa ion,
pe suasion, incen i iza ion, coe cion, aining, enablemen ,
modeling, and en i onmen al es uc u ing) and commonly
heo ized sou ces o beha io ; Capabili y (physical and psy-
chological), Oppo uni y (social and physical) and Mo i a-
ion (au oma ic and e lec i e), known as he COM-B
model (Michie, an S alen, e al., 2011).
C. De elopmen o Ma e ial and In e ace
Design decisions abou he look and eel o an in e en ion
can p omo e hei sus ained use and a e hus highly depen-
den on he mode o deli e , a ge audience and beha io .
In a digi al in e en ion, he g aphics used, decisions abou
gami ica ion and de ices used o deploy he in e en ion
in luence he o e all success o a beha io change in e en-
ion. This calls o mul idisciplina y wo k o inco po a e he-
o ies and me hods om o he disciplines. Heal h beha io
change heo ies a e no su icien o in o ming all deci-
sions abou he design o an in e en ion, and o he disci-
plines ha e a key ole in op imizing design decisions. The
use o communi y-based pa icipa o y esea ch (Teu el-
Shone, Siyuja, Wa ahomigie, & I win, 2006)suchas
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
14 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
consensus con e ences (Be y, Chan, Bell, & Walke , 2012)
o co-design wo kshops (O’B ien e al., 2016) and use -cen-
e eddesign(Ca azzo,Casselman,Hamming,Ka zman,&
Palme , 2012) help o make he in e en ion a ac i e,
clea and ele an o he use .
P oducing inal p og am ma e ials such as pos e s and
ideos may in ol e c ea i e consul an s, a is s o g aphic
designe s. IM sugges s w i ing design documen s o guide
he c ea ion and e iewing o he ma e ials: They can help
in ensu ing ha beha io al science insigh s and in e en-
ion s a egies a e adequa ely ans e ed in o ac ual ma e-
ial p oduc ion.
D. Empi ical Op imiza ion
Once he in e en ion p og am is designed and ma e ials
de eloped in o a ‘be a’ e sion, he e is he need o e ine-
men and op imiza ion. Building in ime o his ex a s ep
will inc ease u u e accep abili y and easibili y o he in e -
en ion. The e a e igo ous me hods ha can be used o ge
ex a in o ma ion o p oceed wi h empi ical op imiza ion/
e inemen o he in e en ion p io o wide scale e alua-
ion, such as he Mul iphase Op imiza ion S a egy (MOST).
Quali a i e and/o quan i a i e me hods can acili a e op i-
miza ion/ e inemen .
MOST is a amewo k o obus empi ical op imiza ion
and e alua ion o beha io change in e en ions (Collins
e al., 2007; Collins, Nahum-Shani, & Almi all, 2014).
MOST p oposes h ee phases: p epa a ion (i.e., de elop he-
o e ical model and highligh unce ain ies abou mos e ec-
i e in e en ion ea u es), op imiza ion (i.e., componen
selec ion using empi ical es ing), and e alua ion (i.e.,
de ini i e RCT). A he op imiza ion phase in e en ion
de elope s ga he empi ical in o ma ion on each in e en-
ion ea u e by conduc ing a andomized expe imen (e.g.,
ac o ial design, ac ional ac o ial design, SMART
designs). The esul s om his o mal es ing in o m deci-
sion-making p ocess in e ms o ea u e selec ion and o -
ma ion o he op imized in e en ion. The amewo k
p oposes an i e a i e p ocess s a ing ha i an op imized
in e en ion is shown o be e ec i e h ough a o mal es ,
i can be made a ailable o he public. The key elemen in
MOST is he p ocesses by which a mul icomponen beha -
io change in e en ion and i s componen s a e op imized
be o e a de ini i e ial o po en ially while he in e en ion
is in use (e.g., op imiza ion o an exis ing app).
Quali a i e me hods p o ide a complemen a y app oach
o suppo he de elopmen and e inemen o an ini ially
d a ed in e en ion. De elope s should aim o unde s and
and inco po a e he pe spec i es o hose who will use he
in e en ion by unde aking i e a i e quali a i e esea ch.
This is impo an o digi al in e en ions (Baek, Cagil ay,
Boling, & F ick, 2008) bu also o adi ional me hods o
deli e y. An example on how his can be ansla ed in p ac-
ice is by elici ing and analyzing se ice use s’ eac ions o
he in e en ion and i s elemen s. I migh also be impo -
an o conduc consul a ion wi h opic expe s (e.g., com-
pu e scien is s) and o he s akeholde s (e.g., heal hca e
p ac i ione s) o he in e en ion o accommoda e hei
iews and expe ise (P esseau, Mu sae s, e al., 2017;Rod i-
gues, Snieho a, Bi ch-Machin, Oli ie , e al., 2017). This
can be achie ed using esea ch me hods such as ocus
g oups, indi idual semi-s uc u ed in e iews coupled wi h
a hink-aloud p ocess. Mixed me hods can also be used o
e ine an in e en ion coupling bo h quali a i e wi h quan-
i a i e o ms o collec ing in o ma ion ha can in o m
e inemen .
E. E alua ing he In e en ion
De eloping in e en ions ha es explici hypo heses could
allow o syne gy be ween knowledge gene a ed ia he
implemen a ion and e alua ion o in e en ions and heo-
ies, allowing o hei es and e olu ion. In he pilo and
easibili y s age he easibili y and accep abili y o he in e -
en ion and e alua ion p ocedu es is es ed and i needed
op imized and addi ional in o ma ion needed o design
he e alua ion is collec ed (Eld idge e al., 2016;Lancas e ,
2015). Once a iable in e en ion and e alua ion p o ocol
has been achie ed, a ull-scale e alua ion o whe he he
in e en ion has i s in ended e ec s on he main ou come
should ake place assuming esou ces a e a ailable o do so.
The s udy design should be chosen based on wha is i
o pu pose –based on ques ion, ci cums ances, and speci-
ic cha ac e is ics o he s udy (e.g., expec ed e ec size and
likelihood o biases). Conside ing he ange o expe imen al
and non-expe imen al app oaches should lead o mo e
app op ia e me hodological choices (Shadish, Cook, &
Campbell, 2002). UK MRC guidance s ongly encou ages
conside ing andomiza ion, due o i being he mos obus
me hod o p e en ing selec ion bias (i.e., in e en ion ecip-
ien s sys ema ically di e ing om hose who do no ). In
case a con en ional indi idually- andomized pa allel g oup
design is no app op ia e, e alua o s should conside o he
expe imen al designs, o example, clus e - andomized i-
als, s epped wedge designs (Li e al., 2017), p e e ence ials
and andomized consen designs, o n-o -1designs (C aig
e al., 2008; Shadish e al., 2002). E en when an expe i-
men al app oach may no be easible, o example, he
in e en ion is i e e sible, obus nonexpe imen al al e na-
i es should be conside ed. In any case, in e en ion e alu-
a o s should be conscious o he need o a oid
unde powe ed ials o p e en p oducing esea ch was e
(Ioannidis e al., 2014).
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 15
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
synd ome. JMIR Human Fac o s, 4, e6. h ps://doi.o g/10.2196/
human ac o s.6502
Wo ld Heal h O ganiza ion. (2014a). Global Heal h Es ima es:
Dea hs by cause, age, sex and coun y, 2000–2012. Re ie ed
om h p://www.who.in /heal hin o/global_bu den_disease/
en/
Wo ld Heal h O ganiza ion. (2014b). Global s a us epo on
noncommunicable diseases 2014. In WHO.. Gene a, Swi ze -
land: WHO.
Ya dley, L., Mo ison, L., B adbu y, K., & Mulle , I. (2015). The
pe son-based app oach o in e en ion de elopmen : Applica-
ion o Digi al heal h- ela ed beha io change in e en ions
(Eds.), Jou nal o Medical In e ne Resea ch, 17, e30. h ps://
doi.o g/10.2196/jmi .4055
Recei ed May 20, 2017
Re ision ecei ed Feb ua y 21, 2018
Accep ed Ap il 19, 2018
Published online Augus 16, 2018
Ve a A aújo-Soa es
Ins i u e o Heal h & Socie y
Facul y o Medical Sciences
Newcas le Uni e si y
Baddiley-Cla ke Building
Richa dson Road
Newcas le upon Tyne NE2 4AX
UK
e a.a aujo-soa es@newcas le.ac.uk
Ve a A aújo-Soa es, PhD, is a Senio
Lec u e in Heal h Psychology in he
Facul y o Medical Sciences, Ins i-
u e o Heal h & Socie y and he
School o Psychology, Newcas le
Uni e si y, UK. He esea ch ocuses
on he de elopmen and assessmen
o e idence-based in e en ions o
he p omo ion o heal h beha io s,
p e en ion and sel -managemen o
ch onic condi ions. She is P esiden
Elec o he Eu opean Heal h Psy-
chology Socie y.
Nelli Hankonen, PhD, is Assis an
P o esso (Social Psychology) in he
Facul y o Social Sciences a he
Uni e si y o Helsinki. He esea ch
ocuses on changing mo i a ion and
beha io in he a ea o well-being
and heal h and in mechanisms o
change in complex in e en ions.
Jus in P esseau, PhD, is a Scien is
and Heal h Psychologis a he O -
awa Hospi al Resea ch Ins i u e and
Assis an P o esso in he School o
Epidemiology, Public Heal h and
P e en i e Medicine a he Uni e si y
o O awa. His esea ch d aws upon
heo ies and app oaches om heal h
psychology and beha io al medicine
o de elop and e alua e in e en-
ions ocused on changing heal h-
ca e p o essional beha io s and
heal h beha io s o pa ien s and he
public.
Angela M. Rod igues, PhD, is Re-
sea ch Associa e in he Facul y o
Medical Sciences, Ins i u e o Heal h
& Socie y, a Newcas le Uni e si y
and in Fuse, he UK Cen e o
Excellence o T ansla ional Re-
sea ch in Public Heal h.
Falko F. Snieho a, PhD, is Di ec o
o he NIHR Policy Resea ch Uni
Beha iou al Science and P o esso
o Beha io al Medicine and Heal h
Psychology in he Facul y o Medical
Sciences, Ins i u e o Heal h &
Socie y, a Newcas le Uni e si y and
in Fuse, he UK Cen e o Excel-
lence o T ansla ional Resea ch in
Public Heal h. His esea ch ocuses
on he de elopmen and e alua ion
o complex beha io al in e en ions
o indi iduals and popula ions. He
is pas p esiden o he Eu opean
Heal h Psychology Socie y.
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
22 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
Appendix A
In e en ion De elopmen and E alua ion F amewo ks and Pu pose
F amewo ks Pu pose
MRC F amewo k o he De elopmen o Complex In e en ions
(C aig e al., 2008)
To p o ide guidance on he p ocess o de elopmen ,
e alua ion and implemen a ion o a a ge in e en ion.
In e en ion Mapping (Ba holomew Eld edge e al., 2016) To desc ibe he i e a i e pa h (six phases) o designing,
implemen ing and e alua ing an in e en ion.
S eps o de eloping a heo y-in o med implemen a ion
in e en ion (S. D. F ench e al., 2012)
To suppo he de elopmen o an in e en ion designed
o change clinical beha iou based on a heo e ical
amewo k.
PRECEDE-PROCEEDE (G een & K eu e , 2005) The model aims o explain heal h- ela ed beha iou s and
en i onmen s, and o design and e alua e he
in e en ion.
The Beha iou Change Wheel (Michie, A kins, & Wes , 2014) This ool de ails how o design and selec in e en ions
acco ding o a beha iou analysis, mechanisms o
ac ion, and he in e en ions equi ed o change hose
mechanisms. This ool is also used o link in luences on
beha iou o po en ial in e en ion unc ions and policy
ca ego ies.
The Pe son-Based App oach o In e en ion De elopmen
(Ya dley, Mo ison, B adbu y & Mulle , 2015)
To design in e en ions based on igo ous, in-dep h
unde s anding o he psychosocial con ex o use s, and
de i ed om i e a i e in-dep h quali a i e esea ch.
6SQuID: 6 s eps in quali y in e en ion de elopmen (Wigh ,
Wimbush, Jepson, & Doi, 2016)
To p o ide a p agma ic and sys ema ic six-s ep guide o
in e en ion de elopmen , maximising i s likely
e ec i eness.
E idence-guided co-design (O’B ien e al., 2016) To desc ibe a sys ema ic, sequen ial app oach o
in eg a e scien i ic e idence, expe knowledge, and
s akeholde in ol emen in he co-design and
de elopmen o an in e en ion.
Knowledge- o-Ac ion (KTA) cycle (G aham e al., 2006) A concep ual amewo k o in eg a e he oles o
knowledge c ea ion and knowledge applica ion,
con ibu ing o sus ainable, e idence-based
in e en ions.
ORBIT model (Czajkowski e al., 2015) To p o ide guidance on he p ocess o ea men
de elopmen by sugges ing he use o a p og essi e,
ansdisciplina y amewo k o acili a e he ansla ion
o basic beha iou al science indings o clinical
applica ion.
EM Model (Shee an, Klein, & Ro hman, 2017) To de ail he p ocess in ol ed in designing in e en ions
o gain mo e cumula i e science o heal h beha iou
change.
Mul iphase op imiza ion s a egy (MOST; Collins, Mu phy, &
S eche , 2007)
To p o ide a guide o he op imiza ion and e alua ion o
mul icomponen beha iou al in e en ions.
Social Ma ke ing (e.g., Le eb e, 2011) The sys ema ic applica ion o ma ke ing concep s and
echniques o achie e beha iou change.
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 23
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
Appendix B
Key Conside a ions o he Repo ing o
In e en ion De elopmen
P epa a o y wo k: Desc ibe he eam and planned de elop-
men p ocess
a. Desc ibe he expe ise o he co e eam and ad iso y s ake-
holde eam in ol ed in de elopmen /design p ocess (in
di e en phases): mul i-disciplina i y, p io expe ience
b. Desc ibe ime used (and a ailable) o in e en ion de el-
opmen p ocess (e.g. leng h o design pe iod, equency
o design mee ings, e c.)
c. Desc ibe o he esou ces a ailable
d. Desc ibe possible unde /commissione demands/limi a-
ions/ eques s o he in e en ion o he de elopmen
p ocess (e.g. u u e use, use o echnology, limi ed inan-
cial esou ces, quick imeline o de elopmen )
e. Desc ibe o iginal gene al aims and in ended use/scalabil-
i y o he u u e in e en ion
S ep 1: Analyse he p oblem and de elop an in e en ion
objec i e
a. Desc ibe how he planning g oup wo ked o de ine he
heal h p oblem, heal h beha io s, a ge heal h beha io s
b. Desc ibe po en ial ma ke analysis, segmen a ion, and
possible subsequen esul ing decisions
c. Desc ibe he decision p ocess leading o p io i isa ion
and selec ion o a ge g oup(s) and beha io change
a ge s
d. Desc ibe how p epa a o y beha io s and ne wo ks o
o he beha io s we e iden i ied and p io i ised
S ep 2: De ine he scien i ic co e o he in e en ion
(i) Unde s and causal/con ex ual ac o s (Causal
Modelling)
a. Desc ibe o mal (beha io al) heo ies used in
unde s anding he p edic o s o he a ge heal h
beha io
b. Desc ibe how key unce ain ies we e iden i ied o
selec aim o e idence syn hesis
c. Desc ibe li e a u e sea ch and e iew p ocess
d. Desc ibe he a ionale/aims and he p ocess o
(possible) o iginal empi ical esea ch
e. Desc ibe a ing o in luencing ac o s (psychologi-
cal, social, p edic o s/mechanisms) o changeabil-
i y and ele ance
(ii) De elop a logic/ heo e ical model
a. Desc ibe he p ocess o de eloping he logic model
(i possible, include ea ly and la e e sions o he
logic model)
b. Desc ibe key explici c i e ia (e.g. accep abili y,
cos -e ec i eness) in making decisions o logic
model
c. Desc ibe whe he and which o he simila exis ing
in e en ions we e used in de eloping he logic
model, o whe he an exis ing in e en ion was
used as co e basis and e o i ed o accoun o
new con ex
d. Desc ibe key unce ain ies le in he causal chain
o logic model and he possible “weak links” he
de elopmen eam hinks he e may emain
e. Assess e aluabili y po en ial o such an in e en-
ion
. De elop a da k logic model ha desc ibes conside -
a ions made a ound po en ial unin ended conse-
quences and s eps made o a oid i
(iii) De ine in e en ion ea u es
a. Desc ibe decision p ocesses (including conside ed
al e na i e op ions) leading o decisions abou
i. p og am componen s/ac i i ies
ii. in e media e a ge s
iii. beha io change echniques o me hods o a -
ge p edic o s/mechanisms e.g. o wha ex en
a ious combina ions o BCTs we e explici ly
conside ed and le ou
i . dose/in ensi y/ equency/du a ion o in e en-
ion
. deli e y channel(s)
i. p o ide s (expe ise/backg ound/ aining)
ii. loca ion/in as uc u e
b. Desc ibe whe he and how an icipa ed accep abil-
i y o in e en ion among a ge pa icipan s and/
o p o ide s and/o commissione s was
in es iga ed
c. Desc ibe he decision p ocesses ela ed o oom o
local adap a ion and necessi y o ideli y o a ious
componen s
S ep 3: Design/De elop in e en ion ma e ials
a. Desc ibe how p o ocol was w i en
b. Desc ibe key p inciples in designing ma e ials (e.g.
design documen s)
c. Desc ibe how s akeholde inpu was ob ained o key
decisions (e.g., scena io-based wo k)
d. Desc ibe whe he and how small-scale p e- es ing o
in e en ion componen s (e.g. g oup exe cises, key
messages) was conduc ed, o make decisions abou
p og am con en
e. Desc ibe decisions leading o pe sonaliza ion and ai-
lo ing (how and why)
. Desc ibe he p ocess o de eloping p ocedu es o
ensu e ideli y
Eu opean Psychologis (2019), 24(1), 7–25 Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
24 V. A aújo-Soa es e al., Me hods o In e en ion De elopmen
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17
S ep 4: Conduc an empi ical op imiza ion
a. Desc ibe key ( esea ch) ques ions o empi ical
op imisa ion
b. Desc ibe empi ical design used in es ing he in e en-
ion (o i s componen s), including da a collec ion
me hods, sample, e c.
c. Desc ibe da a analysis me hods
d. Desc ibe whe he and how quali a i e and quan i a i e
me hods we e mixed
e. Desc ibe how judgmen s and op imiza ion decisions
we e made based on empi ical es ing
S ep 5: Design and unde ake in e en ion e alua ion
a. Desc ibe he plan o e alua ion o e ec i eness
b. Desc ibe a ionale (e.g. esou ces a ailable, unde
in e es s) leading o decisions ega ding e alua ion
c. Desc ibe he plan o e alua ing p ocesses
d. Desc ibe he in ended use o in o ma ion gained (e.g.
o po en ial adap a ions)
S ep 6: Design implemen a ion and unde ake implemen a-
ion e alua ion
a. Desc ibe how decisions ela ed o implemen a ion
(speci ic plans on how he in e en ion will be used
in ou ine p ac ice) we e done, e.g., was he imple-
men a ion in o med by a heo e ical amewo k o a
model
b. Desc ibe he implemen a ion in e en ion de elop-
men p ocess
c. Desc ibe each and allowed adap a ions
d. Desc ibe he plan o e alua ion o implemen a ion
e. Desc ibe a ionale (e.g. esou ces a ailable, unde
in e es s) leading o decisions ega ding e alua ion
. Desc ibe he plan o e alua ing p ocesses o
implemen a ion
g. Desc ibe he in ended use o in o ma ion gained (e.g.
o po en ial adap a ions)
Ó2018 Hog e e Publishing. Dis ibu ed as a Hog e e OpenMind a icle Eu opean Psychologis (2019), 24(1), 7–25
unde he license [CC BY 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/)]
V. A aújo-Soa es e al., Me hods o In e en ion De elopmen 25
${p o ocol}://econ en .hog e e.com/doi/pd /10.1027/1016-9040/a000330 - F iday, Feb ua y 15, 2019 12:57:09 AM - IP Add ess:153.1.49.17