In 2006 in north Somerset we launched the Postnatal pathway for maternal mental health and early attachment.
Ref: Milford R. Community practitioner Aug 2006
Hospital Anxiety Scale (HAD)
This is divided into two aspects depression and anxiety.
Depression
A score of 7 or lower is low concern
A score of 8-9 is moderate concern
A score of 10 or more is high concern
Anxiety
A score of 7 or lower is low concern
A score of 8-9 is moderate concern
A score of 10 or more is high concern
Mothers Object Relations Scales (MORS) Warmth and Invasion scores Warmth scale A score of 11 and lower is high concern A score of 15 - 10 is moderate concern A score of 16 and above is low concern Invasion scale A score of 17 and above is of high concern A score of 12 -16 is of moderate concern A score of 11 and below is of low concern
No concern Moderate concern High concern
Routine Health Visiting service
Up to 6 listening visits (review)
GP & Mental Health specialists
(NICE Guidance)
Refer to Mental Health Specialist for consultation
Additional appropriate services offered from a menu
of complementary care packages including
BABY MASSAGE FOR ALL
* GP
(NICE Guidelines)
Counselling IAPT
Referral to Adult Mental Health Team
Wait, watch and wonder/ Theraplay
based
Attachment group
No further action
We no longer use the HAD and now use
PQ9
GAD 7
We have commenced a pilot for the complimentary antenealt care pathway which aims to pick up women with mental health concerns at the first booking clinic
A secure attachment relationship between mother and infant is recognised as a protective factor and a key component in increasing the likelihood of children developing positive mental health in adulthood (Oates et al., 2007).
Recent research suggests 23% of non-clinical mothers have insecure-dismissive attachment, 19% insecure-preoccupied attachment and 58% secure attachment (Bakermans-Kranenburg and Van IJzendoorn, 2009).
By considering a possible links between maternal attachment style and the
dyadic relationship with the infant, this study looked at the effectiveness of infant massage in promoting improved mother-infant attachment and dyadic attunement.
cure attachment relationship between mother and infant is recognised as a protective factor and key component in children developing positive mental health in adulthood (Oates et al., 2007).
Recent research suggests 23% of non-clinical mothers have insecure-dismissive attachment, 19% insecure-preoccupied attachment and 58% secure attachment (Bakermans-Kranenburg and Van IJzendoorn, 2009).
Acknowledging these links between maternal attachment style and the dyadic relationship with the infant, this study looked at the effectiveness of infant massage in promoting improved mother-infant attachment and dyadic attunement.
NICE guidance for planning and commissioning children’s services suggests that health visitors and midwives should consider evidence-based interventions, such as baby massage, as part of provision for the social and emotional well-being of new mothers and the under 5’s.
Robust evidence of the effectiveness of baby massage courses would give confidence to decision makers about funding these resources within perinatal services.
Cut in local offer- lack of research used as a reason.
A Systematic review was conducted using 4 databases, including publications from January 1980 to May 2016.
The criteria included massage or touch, when used as an intervention on more than one occasion.
Two reviewers screened the extracted data for eligibility and quality, using standardised forms.
STUDY M ethod Part icipants Intervent ion Outcome M easures Data Collect ion Process Data ItemsAuthor's Judgment of
Bias/Limitat ionsSummary of M easures
232
Fujiki
279
Hart
392
Lee
520
O’Higgins
49
Beyer
115
Clarke
685
Watanabe
662
Underdown
523
Onozawa
264
Gurol
Very small sample size. High
drop out rate. Dif fering lengths
of sessions. Short period of
intervent ion.
data was analysed by non
parametric methods due to
sample size and M ann-Whitney
U-test (two tailed) or Fisher's
prbability test as appropriate
Quasi-
experimental
design.
Randomised
Controled
117 (3 excluded)
Breastfeeding singleton
M others of healthy babies
living in a specif ic city and
with a minimum of
highschool level of
educat ion.
Baby massage taught
at 5-7 days by IAIM
instructor, CD given
and technique checked
at 15 and 38 days). 15
minute massage given
daily for 48 days
Implementat ion of
M other-Infant Ident if ier
Poll.
M aternal Attachment
Inventory
Test between 24-48 hours after
birth in hospital and on the last day
of the study at home.
M aternal Affect ionate
Attachment
M AI conducted in dif ferent
environments. All mothers well
educated.
Analysis used SPSS. Percent iles
and means. Chi-square test. One-
way analysis of varience. Paired
sample t-test
Randomised
Controled
25 (9 dropped out)
M others scoring >13 on
the EPDS at 4 weeks
postpartum, with healthy
babies. 12 took part in a
massage group and 13
took part in a support
group
5 weekly sessions (1
hour for massage
group and 1 hour for
support group).
Edingburgh Postnatal
Depression Scale,
Video's coded using
Global Rat ings for
mother-infant
Interact ions at 2
months by Fiori-
Cowley and M urray
First and Last Sessions mother's
given EPDS and videoed playing
with their child for 5 mins.
Depression status. M aternal
contribut ion to interact ion.
Infant contribut ion to
interact ion. Quality of
interact ion.
No author comment but we could
consider limitat ions regarding
number, shortness of the interval
between outcome measures (no
follow up). Unknown care to
control group.
T-test
Crit ical Realist
research design
(mixed method -
quantat ive and
qualatat ive)
39 mothers at tending a
six week infant massage
programme in a group (six
did not complete). 72%
were white-Brit ish and
28% were other ethnic
minorit ies.
A f ive week massage
course (group) led by
the Internat ional
Associat ion of Infant
M assage (IAIM )
Working model of the
child interview (WM CI),
the Edinbrugh Post-
Natal Depression scale
(EPDS) and the care
index.
Interviews (unknown environment).
Videos. Pre and post intervent ion
Parental response classif ied
into one of three broad
attachment classif icat ions:
balanced, disengaged and
distorted. EPDS scores for
levels of depression and Care
Index (CI)measuring three
aspects of maternal behaviour
(sensit ivity, covert and overt
host ility and
unresponsiveness). Four
aspects of infant behaviour
(cooperat iveness, compulsive
compliance, dif f icultness and
passivity).
Diverse range of infant pract ices
observed between parents and
infants. They had a high
proporiton of white Brit ish
which may impact on the
contextual generalisat ion.
Possible bias in recruitment e.g
1/4 of part icipants had degrees,
meaning they may have been
more motivated to attend having
heard of the offer of high quality
infant massage programmes.
High level of low risk women
meaning that no signif icant
change was likely.
Quantat ive data was entered
direct ly into SPSS as well as
demographic details, WM CI
classif icat ions and EPDS and CI
scores. Appropriate stat ist ical
t -test were performed to
compare the means for
quantat ive data collected
before and after the
programme. Qualatat ive data
was also entered into NVIVO
whcih is a qualatat ive data
analaysis computer-software
packaged and analysed
thematically.
Intervent ion
research - Non-
equivalent
Control
First t ime mums, three
months post natal who
attended infant massage
classes. Full term infants.
20
10-15 minutes massage
per day over four
weeks. Control group
received normal care.
The Prof ile of M ood
States and the
Postpartum Bonding
Quest ionnaire. Pre and
post intervent ion.
Unknown
Tension-anxiety, depression,
anger-host ility, vigour, fat igue
and confusion.
Very small sample size. No
control group. Not randomised
in design. All part icipants were
married, proffessional and had
exist ing social support
networks. There were no high
levels of risk for stress or
relat ionship dif f iculty.
Recruitment via poster.
Visual comparison of Pre and
Post stress scores.
Post
Intervent ion
Evaluat ion
94 (52 dropped out)
Parents who had
completed IAIM baby
massage courses. 60 (140
dropped out) Parents who
had not taken part in the
massage courses
Unspecified 5
week programme
of baby massage.
Control group did
not attend this
programme.
Programme Evaluat ion
Quest ionairre, Infant
Social Interact ion
Quest ionairre, Focus
group Interviews.
Post interact ion quest ionairre by
mail, requiring mailing back and
Focus Group Interview.
Items about the baby, Parent 's
Competancy Scale, Items about
the parents, breastfeeding,
reading, chatt ing. Impact of
Social Interact ion and
Parent ing.
Post intervent ion data only.
M ost of the data was from the
focus group and just used one
trainer. Short length of t ime so
unknown longer term effects.
Sensit ivity of tools was limited.
Non intervent ion group was not
well matched to intervent ion
group so comparisons are
limited.
Qualitat ive data entered into
SPSS and analysed using
standard non parametric non
stat ist ical tests. Tapes were
transcribed and analysed using
Open Coding and Thematic
Analysis.
Single Pretest-
Posttest Design
4 (3 dropped out)
M others of healthy 1-3
months old infants
responding to a poster
advert .
One off t raining
session in non specif ic
massage and parental
sensit ivity. Applied 5 x
weekly for 15-30 mins
per day. Journal
complet ion. Contacted
at midpoint to discuss
progress and concerns.
Parent ing Stress Index -
Short Form
Pre and post quest ionairres.
Narrat ive analysis was done on the
journals.
Total Stress Score. Parent 's
rat ings of Infant 's temperament,
parental competancy,
at tachment, social support and
role habit disturbance.
Sensit ivity to infant cues and
other themes (eg. Time of day).
Not fully randomised in design.
Limited collect ion of
physiological parameters. All
babies lived with their families.
t -test and chi-square test
Prospect ive
Block -
Controled
Randomised
design
62 M others scoring 13+
on the EPDS at 4 weeks
post partum. Control
Group - 34 (56 dropped
out) scoring less than 9 on
the EPDS at 4 weeks post
partum. Both groups were
split equally between 2
intervent ions. M other's
attending 4 or more
sessions were included.
6 Sessions of either
IAIM Infant M assage
or a support group
Edinburgh Post Natal
Depression Scale,
Spielberg State Anxiety
Inventory, Infant
Characterist ics
Quest ionairre, Video
interact ions using the
Global Rat ings for
M other-Infant
interact ions. Social
economic quest ions.
9 to 12 weeks Quest ionairres and
Videos in Clinic. Repeated at 19
weeks in clinic. Social ecomomic
prof ile quest ions asked at the
beginning.
Depression and Anxiety State
M aternal Sensit ivity in
Interact ion, Infant Performance
in Interact ion, Overall
Interact ion, Fussiness/Dif f iculty
Scale.
The group were well educated
with no socio economic stress.
Part icipants were predominant ly
white and married. No treatment
for depression group was not
included for ethical reasons.
Init ial fall in EPDS scores may be
attributed to ant icipated
support (Appleby, et al. 1997).
ANOVA with post hoc
Bonferroni Tests.
Non-equivalent
Control Group
Pretest-
Posttest Design
129 (13 dropped out)
M others of Full term,
healthy 2-6 month old
babies, recruited at
M assage programme or
Well Baby Clinic. 26
M others agreed to the
massage programme and
103 took part as the
control group.
Non specif ic Baby
M assage course and
writ ten checklist .
Applied 4 x weekly at
home for 4 weeks.
M other's Percept ion of
Infant 's Temperament
Scale (modif ied),
Weight, Height, M other-
Infant Interact ion
(M IPIS Walker and
Thompson 1982)
Quest ionairre - not specif ied how
given. Video recordings of M other-
Infant interact ion in the health
centre, Weight -Cas electronic
scales, Height - collected in
triplicate and averaged. Social
demographics and feeding
information - quest ions were also
collected at t ime of select ion.
M other's Percept ion of Infant 's
Temperament (def ined as the
individual's emotional react ivity
and behavioural style in
interact ing with the
environment), M other-Infant
Interact ions, Height, Weight.
Wilcoxon rank sum test,
Student t-test. The level of
signif icance was set at 0.05.
Cort isol Levels determined by
enzyme-linked immnuosorbent
assay.
Non-equivalent
Control Group
Pretest-
Posttest Design
94 (52 didn't respond)
Parents usually present ing
with a perceived problem
(eg. Crying). 60 (140
didn't respond) Parents
with no perceived
problems. No select ion -
voluntary part icipat ion.
IAIM 5 week course. 15
courses in a variety of
sett ings. 12 parents at
each course.
Infant Social Interact ion
Quest ionairre,
Edingburgh Postnatal
Depression Scale,
Programme evaluat ion
Quest ionairre. Focus
group interviews
Quest ionairres mailed to
part icipants before and after
intervent ion. Interviews conducted
at focus group.
Parent ing sense of competancy,
Self Esteem, Breast Feeding,
Reading to child, Chatt ing,
General health of parent and
child, Depression status. Health
act ion zone funding
Not fully randomised in design.
Study focuses on one trainer.
Short period of intervent ion
requiring part icipants to mail
back quest ionairres. Long term
benef its can only be conjecture.
Indicit ive of future research
Percentages. Signif icance not
specif ied. Detailed results from
the report are reported
elsewhere in a further study.
Clarke, C et al. Social
Interact ion in Parent ing study:
Project Report , Newcast le
upon Tyne: university of
Northumbria @ Newcast le
2000.
STUDY DESIGN
Randomised
controlled
39 (18 dropped out) full
term, singleton M others @
6 weeks postnatal with no
complicat ions
Field's M ethod of
Baby M assage (did
not at tend a course).
Applied 10 mins per
day unt il 3 months
Cort isol in saliva,
Prof ile of M ood States
Cort isol taken between 10am and 3
pm at the start and end of the study
(in hospital and at home),
Quest ionairres were completed at
home before and after intervent ion.
All mothers were mailed the
quest ionairres.
Tension, Depression, Anger,
Vigour, Fat igue,
Confusion,Concentrat ion of
Cort isol Levels determined by
enzyme-linked immnuosorbent
assay
Short period of intervent ion,
requiring part icipants to mail
back responses, The second
Cort isol samples were delayed
unt il 6 weeks after the
intervent ion ended and taken in
dif ferent environments. Indicat ive
of further research.
Of 732 papers identified, 72 were included for consideration of the full paper.
10 papers met the criteria for inclusion in the systematic review.
Synthesis of the papers showed some evidence that baby massage has a positive effect on the dyadic relationship.
There was moderate evidence for short term improvement in infant attachment, no long term conclusions could be made from the review. Longitudinal research would be needed to support the current findings and careful consideration given to the use of appropriate outcome measures. Two papers suggested that further research should focus on using infant massage programmes to improve the mental wellbeing of mothers with previously recognised moderate mental health concerns.
We propose to complete further research by following up on the findings from Underdown et al.
The review they carried out suggested of possible link between women with mild to moderate mental health concerns have improved outcomes following the completion of a structured baby massage programme.
Focus will be on the elements that may or may not be important in this improvement.
Systematic Review Completed by
Rebecca Balakrishna
Melanie Teixeira
Roxanne Hart
Contact for full paper and references.