The Impact of ParentChild Interaction Therapy on Maternal Sensitivity Rae Thomas and Melanie J' Zimm

1
The Impact of Parent-Child Interaction Therapy
on Maternal SensitivityRae Thomas and Melanie J.
Zimmer-GembeckSchool of Psychology, Gold
CoastAustralia
Abstract   The aim of the current randomised
controlled trial was to increase maternal
sensitivity and decrease child abuse potential in
a sample of families at risk of or engaged in
child maltreatment. The families were referred to
Parent-Child Interaction Therapy (PCIT), a
parenting program that is designed to improve
parent-child relationships and assist with using
appropriate methods for managing child behaviour
problems. Outcomes of treatment were compared to
an attention only control group. PCIT is a
manualised treatment program founded in social
learning theory and attachment theory. PCIT was
designed to increase positive interactions and
play between parents and children, reduce
coercive parent-child interactions, improve
parenting skills, and decrease child behaviour
problems in families with young children (age 3
to 7). Maternal sensitivity levels were assessed
using an observational technique (Emotional
Availability Scales, Biringen, Robinson Emde,
1998) and child abuse potential was assessed
using a self-report measure, the Child Abuse
Potential Inventory (Milner 1986). In the current
study, families at risk of or engaged in child
maltreatment were referred to PCIT and randomly
assigned at a ratio of 21 to a Treatment PCIT
group or an Attention Only Waitlist group.   No
change in maternal sensitivity was found from the
pre- to the 12-week assessments in either the
PCIT treatment group (N36) or the Attention Only
Waitlist group (N22). Yet, among the PCIT
treatment group and when compared to the
pre-treatment assessment, maternal sensitivity
had improved by completion of the program
(approximately 30 weeks) and this pattern of
improvement was also found when mothers reports
of child abuse potential were examined. Maternal
sensitivity and mother-report of child abuse
potential were positively correlated, r -.20, p
lt .05 suggesting both measures assessed distinct
behaviours and cognitions. The importance of
including observational measures in data
collection is highlighted.
Child Maltreatment, Attachment and Maternal
Sensitivity   Child maltreatment constitutes a
serious disruption to a stable and nurturing
caregiving environment. Consequences of child
maltreatment and exposure to violence have
considerable impact on child development
extending to diagnosable disorders and impacting
on developmental pathways (Cicchetti Rizley,
1981 Margolin Gordis, 2000 Rogosch,
Cicchetti, Shields, Toth, 1995). Maladaptive
emotional and behavioural outcomes can include
cognitive distortions (Milner, 2001),
externalising behaviour (Kolko 1002), maladaptive
peer relationships (Bolger Patterson, 2001
Dodge, Pettit, Bates, 1997), impaired academic
functioning (Margolin Gordis, 2000), increased
risk of psychopathology (Higgins McCabe, 2000)
and insecure attachment relationships (Cicchetti
Toth, 1995 Crittenden Ainsworth, 1989).
  The effect of maltreatment on parent-child
attachment relationships is well documented
(Cicchetti Toth, 1995, Crittenden, 1988).
Several studies have reported that approximately
90 of maltreated children are classified as
insecure in their attachment pattern (i.e.,
either insecure-avoidant, insecure-ambivalent or
disorganised/disoriented Cicchetti Toth,
1995). Indeed Egeland, Sroufe and Erickson (1983)
reported that 100 of the children of
psychologically unavailable mothers were
classified as insecurely attached at 18 months of
age. Bowlby (1951) emphasised that psychological
unavailability of parents had a more pervasive
and detrimental effect on attachment of an infant
to a caregiver than social, financial or medical
difficulties. Psychological availability of the
caregiver has been postulated to distinguish the
attachment of a child and maltreating caregiver
from the attachment of a child and more
nurturing, nonmaltreating caregiver (Cicchetti
Toth, 1995 Egeland Erickson, 1987 Margolin
Gordis, 2000).   Related to psychological
unavailability, but perhaps a more subtle aspect
of parenting, is the concept of maternal
sensitivity. Maternal sensitivity has been
identified as a key aspect of parenting style
with young children and a predictor of attachment
status (Ainsworth Blehar, Waters, Wall, 1978).
Ainsworth et al. (1978) described maternal
sensitivity as a mothers appropriate responding
to the infants needs and the negotiation and
flexibility of her attention and behaviour.
Isabella, Belsky, and von Eye (1989) expanded the
construct of maternal sensitivity to an
interactive experience between the infant and the
mother. In a meta-analysis of attachment studies,
maternal sensitivity and attachment security had
a modest, but robust correlation (r.24, De Wolff
and van IJzendoorn, 1997) and recently maternal
sensitivity was identified as the strongest
predictor of attachment security
(Bakermans-Kranenburg, van IJzendoorn
Kroonenberg, 2004).   Therapeutic interventions
for families engaging in maltreatment,
underpinned with attachment theory concepts, such
as maternal sensitivity, may succeed where
behaviourally based, psychoeducational models
have been less effective (Pearce Pezzot-Pearce,
1994). The Family Interaction Program at Griffith
University built a focus on maternal sensitivity
into the implementation of PCIT to families at
risk of, or engaged in, child maltreatment in
attempts to increase maternal sensitivity,
decrease child behaviour problems and ultimately
decrease child abuse potential.
Parent-Child Interaction Therapy   PCIT is a
behavioural family skills program for families
with children aged between three and seven years
experiencing challenging, externalising
behaviours. There are two phases of PCIT,
Relationship Enhancement and Additional Skills
(also known as Child Directed Interaction and
Parent-Directed Interaction). Progression from
one phase to the next is predominantly dependent
on attaining prescribed levels of specific skills
known as Mastery Criteria (Hembree-Kigin
McNeil, 1995). Direct coaching sessions are
conducted with the parent and child in a play
therapy room with the psychologist in another
room behind a one-way mirror. The psychologist
and the parent communicate through a
bug-in-the-ear device. This device permits the
therapist to provide direct coaching of parental
communication and behavior management skills,
immediate feedback and social reinforcement of
the parent. Parents are expected to practice the
skills at home. Through PCIT parents are taught
strategies to increase positive interactions and
reduce the potential for child maltreatment.

• Results
•  
• Maternal Sensitivity
• From the pre- to the 12-week assessment, there
  were no significant differences and no
  significant changes in maternal sensitivity among
  the PCIT treatment group and the waitlist
  comparison group.
• From pre-treatment to treatment completion,
  significant improvement in maternal sensitivity
  was found among the treatment group, d.39.
• Correlation between Maternal Sensitivity and
  Child Abuse Potential
• At pre-treatment (N 112, maternal sensitivity
  was negatively correlated with child abuse
  potential scores, r -.20, p lt .05. This
  indicates that mothers who were less sensitivity
  were also relatively higher in child abuse
  potential.
• Child Abuse Potential
• As was found with maternal sensitivity, from the
  pre- to the 12-week assessment, there were no
  significant differences and no significant
  changes in child abuse potential among the PCIT
  treatment group and the waitlist comparison
  group.
• From pre-treatment to treatment completion,
  significant improvement in child abuse potential
  was found among the treatment group, d.36.

• Method
•  
• Participants were clinic-referred families at
  risk of, or with a history of, child
  maltreatment. Referral sources included the
  Queensland, Australia Department of Child Safety,
  Queensland Health, Education Queensland and
  non-government organisations.
• 121 families participating in an ongoing PCIT
  intervention were included in this analysis.
• Families were randomly allocated to either the
  PCIT treatment group or the attention only
  waitlist comparison group at a ratio of 21.
• N 87 treatment participants
• N 33 attention only waitlist participants
• An attrition rate of 52 occurred between
  pre-treatment and 12-week assessment resulting in
  a final sample size for pre-treatment to 12-week
  assessment of
• N 36 treatment participants
• N 22 attention only waitlist comparison group
  participants.
• 70 of the children were males with a mean age of
  4 years, 11 months.
• 94 of parents identified themselves as White
  Australians while 3.5 identified themselves as
  either Aboriginal or Torres Strait Islander.
• Weeks to treatment completion ranged from 15 to
  56 (M 30.4, SD 9.5).
• Maternal sensitivity was measured using the
  Emotional Availability Scales Middle Childhood
  Version 3rd Edition (Biringen, Robinson Emde,
  1998)
• Maternal sensitivity was assessed through
  video-taped observations of mother-child dyads
  during a 10-minute free play.
• Maternal sensitivity was rated by independent
  coders blind to the treatment status of the
  dyads. Intraclass correlation was 0.83 for
  maternal sensitivity.
• Child abuse potential was measured using a parent
  self-report measure, the Child Abuse Potential
  Inventory (Milner, 1986).

• Clinical Implications and Future Directions
•  
• Observational measures may complement more
  traditional data collection assessments such as
  self-report measures. The inclusion of a maternal
  sensitivity observational measure contributed to
  an understanding of the effectiveness of PCIT for
  maltreating families. The data suggests that
  observational measures add important findings for
  interventions.
• Therapeutic services for families engaged in or
  at risk of child maltreatment may require longer
  interventions to produce significant effects on
  entrenched cognitions and behaviours.
• Alternatively, families who continued in PCIT
  until treatment conclusion, may differ to those
  families who dropped out (such as more commitment
  to treatment, less stress, less child behaviour
  problems etc).
• Interventions to improve parent-child
  interactions among families at risk of child
  maltreatment often focus on improving
  parent-child interactions as a way to reduce
  parenting stress, coercive parenting behaviours,
  and improve child behaviour problems. Studies
  have shown that many parenting interventions do
  have these effects (Thomas Zimmer-Gembeck,
  under review). Just as likely, these programs may
  be improving other parenting behaviours that may
  indicate the potential for improvement of
  parent-child attachment relationships, such as
  maternal sensitivity. Yet, few intervention
  studies have examined this as a potential
  positive program outcome.