What do we know about effectiveness of parenting

What do we know about effectiveness of parenting interventions for preventing child maltreatment? Translating research evidence into the real world FRANCE S GARDNER PROFESSOR OF CHILD AND FAMILY PSYCHOLOGY DEPT SO CIAL PO LICY & SOCIAL WORK UNIVERSI TY OF OXFORD

Parenting interventions: We can think about these as potentially applicable to preventing maltreatment on 3 levels: Level 1: Population level prevention: Improving parenting in the general population Level 2: Indicated prevention Improving parenting in families identified as ‘at risk’ or having some difficulties Level 3: Secondary prevention ‘Treatment’ for maltreating parents to prevent re-abuse

Parenting interventions for preventing child maltreatment: outline Why parenting? What interventions? Early-middle childhood focus Basic evidence on effectiveness: prevention, treatment, how do they work, & for whom? How effective are they at preventing physical abuse? How well do parenting interventions transport and adapt? to varying cultures and countries? can they work in real-world, overstretched services? can they be taken to scale? Using examples from our RCTs (randomised controlled trials) in Wales & England

Why parenting interventions? Parenting skill strong predictor of child maltreatment & child problem behaviour Predictive of other outcomes (eg school failure; drug use; poor health /mental health; partner violence) We know a great deal about how to help parents to parent more effectively (much evidence)

Evidence-based parenting interventions: What are they?

Essential components of effective parenting programs (early and middle childhood) Specific factors - theory of change for your program e. g. • Social learning theory principles • Attachment - building warm relations • Cognitive behavioral principles - to deal with parent stress/ anger General factors: Effective practitioners • work in a way compatible with clients’ beliefs and values. • encourage clients to focus on present & future possibilities instead of past problems • are empathic, accepting, warm Effective programs share general & specific principles eg - Triple P - Family Check Up - Incredible Years - Parent Management Training (PMTO) - Parent-Child Interaction Therapy (PCIT)

Specific components of effective parenting interventions - emphasise principles rather than prescribe techniques - active problem solving by parents around own family needs & goals for kids - using social learning principles: eg importance of parent child interaction, reinforcement, clarity, consistency, small steps. - relationship building, praise and affection reduce conflict, & harsh, physical techniques. sanctions (non-violent) for problem behaviour apply principles to own situation - using discussion, role play, modelling, problem solving, practice, homework etc

One example: the Incredible Years parenting intervention (Webster Stratton) Parent, teacher, child group-based programs Strong evidence base Effective as both clinical and preventive programs Evidence for applicability across cultures and countries Identified and adopted by a number of governments as effective program.

The Incredible Years Basic programprinciples and components • 12 -14 two-hour group sessions, weekly • Emphasis on shared problem solving • Collaborative flexible leader style, start with parents needs and values • Emphasis on parents learning principles not rigid techniques; on changing behaviour…. • Discuss video clips of parents • Role-play & home practice in the group

Components of Incredible Years parenting program • Content pyramid - stresses first relationship building, then discipline strategies • Nice video examples on website <incredibleyears. com>

Parenting interventions: how do we know they work? - in general, for improving parenting & child outcomes

What constitutes good evidence? Randomised controlled trial (RCT): Considered ‘gold standard’ for minimising bias in evaluation Well-worked out methods for complex and communitybased interventions; parents’ views central to evaluation Good research is expensive, but so is ineffective or untested practice - even more so Well-meaning interventions may do no good - and worse, may do harm (ethical obligation). Many egs of well-liked interventions doing harm Systematic Review: Summarise findings of trials - rigorous, transparent, replicable, up to date

Effective parenting interventions: What is the evidence base? Over 100 randomised trials (RCTs) demonstrating effectiveness of structured, cognitive-behaviorally based interventions for: Reducing child problem behavior Improving parenting skill & confidence; mental health and a smaller number for: Reducing maltreatment -same programs may work: IY, PCIT, TP Not just young children early; similar findings with teens Many good systematic reviews (Barlow- Cochrane/ Campbell) Why use interventions without an evidence base? (at least in settings similar to the trials. . ? )

In what service contexts have parenting interventions been tested in RCTs? Settings for these randomised controlled trials: Many countries; wide range of targets: Universal and targeted prevention, vs ‘treatment’ studies. Range of service settings: health services - primary care and specialist, schools, child protection services, community groups, voluntary sector, ‘Head Start’ nurseries, day care, workplace, WIC nutritional centres (FCU), ‘Sure Start’ Range of delivery methods: Individual vs group; at home vs not; telephone, via primary care staff, media-based- TV, video (Sanders trials, TV) Can be flexible according to family need, preference, context (FCU maximally). Many more settings in practice- these are just the trials

Data from RCTs suggests effective for a range of families: l l Low income and average families Families with complex, multiple needs including maternal mental health problems, child abuse, history of maternal abuse, high poverty, those with incarcerated sibs (eg Scott; Webster-Stratton; Sanders; Gardner). Remove barriers. Child problems range from severe conduct disorder to very mild problems or none; children with multiple problems (eg ADHD, LD) Example - moderator analyses in our UK trials- test if intervention more or less effective for these subgroups

Transporting a US program to UK realworld services: 2 randomised trials Incredible Years parenting program Training, fidelity, adaptation to local family needs Oxford trial: Children with severe behaviour problems, referred for treatment, in community settings, in voluntary sector, age 2 -9 (Gardner et al. , 2006) North Wales trial: Targeted prevention in multiple ‘Sure Start’ services, age 3 -5. (Hutchings et al 2007; Gardner et al, 2010)

Overcoming community barriers to engagement – practical & process Local community venues (respond to parent views) Build relations with families, referrers, staff in community Food, child care Daytime and evening groups- offer choices Active recruiting of fathers Home assessment visits before group starts Collaborative not didactic process; builds on parents’ strengths and wishes; Individual goals. Phone calls between sessions if needed North Wales: offer Welsh-language groups & materials

Oxford trial in voluntary sector: Family Nurturing Network (Gardner et al. 2006, Jnl Child Psychology & Psychiatry) 76 families, randomised to intervention vs wait-list. Children 2 -9, referred for severe conduct problems. ‘Incredible Years’ parenting groups. Clear intervention effects on child problem behaviour, harsh and positive parenting skill - by both parent report and direct observation in the home. Good effect sizes, in Intention-to-Treat analyses, ES. 5 -. 8 Effects maintained to 18 month follow up. High consumer satisfaction. Very much a ‘real world’ service

North Wales ‘Sure Start’ trial. Hutchings, Bywater, Daley, Gardner et al. , BMJ 2007 Children 3 -5 yrs in 11 low income ‘Sure Start’ areas, screened for risk for conduct problems 133 families, randomised to intervention vs wait-list ‘Incredible Years’ parenting groups. Clear intervention effects on child problem behaviour (ES. 9), positive and negative parenting skill (ES. 6 -. 9), by both parent report and direct observation in home. Effects on parent depression (ES. 5). Effects maintained for 1 year Effect Sizes impressive – nb are from Intention-to-Treat analyses Conclude: similar effects+, in more complex multi agency ‘real world’ setting- first RCT of its kind. Low income, bilingual areas, high unemployment

1. Observed Positive parenting: (sum of praise, affectionate, positive affect, problem-solving) Evidence-Based Intervention Group, Department of Social Policy & Social Work University of Oxford

Oxford and Wales trials Moderators of change - for whom do they work? - equally well or better with the most troubled parents Mediators of change - how & why do they work, or via what mechanisms? - positive parenting skill as active ingredient in both Important to know for all interventions, and especially when trialling interventions in ‘real world’

SO WHAT? So, parenting interventions work to reduce harsh parenting, parent child conflict, and child problem behaviour…. . That should help prevent abuse…. . BUT…. . Is there evidence they actually prevent abuse? - level 1 23 How relevant are these interventions to the ‘high end’ families (level 3)?

Level 1: Population level prevention: Improving parenting in low or medium risk groups eg Triple P Parenting program (Australia) implemented widely in child & family services in 9 counties of South Carolina: did training, media campaign,

Can you change child outcomes that matter across a whole population? Taking Triple P to scale in South Carolina: Only RCT of widespread dissemination of an evidence-based parenting • intervention Aimed to reduce maltreatment in whole population of children 0 -8, by implementing Triple-P parenting system, county-wide. In a cluster randomised trial design* 9 counties randomised to triple P 9 comparison counties got ‘services as usual’ Bold outcome measures – child abuse indicators- after 2 yrs Counties not self-selected Counties first matched in pairs on demographic indicators: county rate of poverty, child abuse & population size. 85, 000 families in the 18 counties; N=649 service providers participated; often in disconnected services

Results for primary outcomes substantiated child maltreatment Increased much more in Control counties p<. 03 child out-of home placements Reduced in Triple P, increased in Control counties, p<. 01 child maltreatment injuries Reduced in Triple P counties, increased in Control p<. 02 effect sizes large: 1. 1 - 1. 2

Conclusions from S Carolina trial Unique trial of implementation of preventive evidence based parenting program Cluster randomised by county to implementation Tall order to change abuse indicators - but they did Strengths - well matched at outset; deliberately, as small N of 18 counties; public awareness up Cost and cost benefit studies (Foster et al, 2008)

LEVEL 2 Level 2: Indicated prevention Improving parenting in families identified as ‘at risk’ or having some difficulties For example, David Olds’ highly structured, home visiting program

LEVEL 3 Level 3: Secondary prevention ‘Treatment’ for maltreating parents to prevent re-abuse Example of adaptation of parenting intervention to child protective services - Webster Stratton paper http: //www. incredibleyears. com/library/items/adapting-incredible-yearsfamilies-welfare-system_10. pdf Systematic review of effects on maltreated children - Montgomery et al; Barlow et al.

Families in the child protection system Webster Stratton (2010). “Adapting Incredible Years (with fidelity) for families involved in child welfare system” Analysis of Head Start RCT data (n=630) 20% of kids in child protection system their problems were worse but intervention just as effective for these families - ie improved parenting & chid problem behaviour

Families in the child protection system Webster Stratton 2010: adaptations with fidelity for child welfare: flexibility of application of program seen as part of training & accreditation process- eg: spend longer on: basic relationship building thru play; parent attributions, developmental expectations; anger management and hitting use home visits; collaborate with case workers Case series of 136 families in CPS

Effectiveness of interventions for children who have been physically maltreated 3 systematic reviews Montgomery, Gardner et al, for UK govt DCSF: parenting; family-therapy, child focussed. Found: 7 RCT’s of Parenting Interventions- compared to ‘treatment as usual’ Poor quality: small, few had maltreatment outcomes Some evidence for reducing maltreatment Better evidence for improving parenting skills and child mental health outcomes Found: 2 RCTs of Family Therapy- both showed less effect on harsh parenting than comparison interventions: CBT

Conclusions from systematic reviews of parenting interventions for abuse • Several trials; small, quality not great • Some evidence structured parenting interventions can reduce harsh parenting • In one study, reduced maltreatment • Promising - especially as consistent with: i) evidence from trials of same interventions with other high risk groups, for reducing harsh parenting / parent child conflict; and ii) with findings of population level prevention trial

Can parenting interventions translate well to other countries and cultures? Evidence-based programs developed in Australia & US have been used & tested in many countries: e. g: Randomised trials in: UK, US, Ireland, Canada, Australia, Germany, Norway, Hong Kong, New Zealand…. Implemented in many more: Iran, Russia, Thailand, Denmark, Germany, Israel, Holland, Portugal, France… ( - many initial misgivings in UK)

Can parenting interventions translate well to other countries and cultures? Culture: l Diverse families in many studies (eg Scott, Gross, Miller) - program can be seen as inherently flexible to family / cultural needs l l Bilinguality - how dealt with, especially in group interventions? - approaches in Wales, Seattle Large Webster-Stratton (n=650) study found no ethnic differences in any child & parent outcomes, nor in engagement, attendance, parent satisfaction

Effective parenting interventions for preventing child maltreatment What do we know? Strikingly strong evidence base for parenting interventions changing child & parent behaviour, reducing harsh parenting. Much evidence that transferable across diverse real world service settings and families, in several countries Appear to be able to engage and help very marginalised families - - and to do as well with the highest risk families as the others But do they actually prevent child abuse? Probably -evidence promising at each level, from general population prevention to treatment for abusive parents

Thank you!

What next? As we know a lot, we need to implement more- both in prevention and response mode In accessible settings, non stigmatising - where do families go? (WIC); at what stages are parents receptive? ‘sure start’, neighborhood nurseries…. Need careful training, supervision and support of staff Need to test effectiveness in new settings that are very different, or when taken to scale. Need more work on cultural adaptation, and testing of new programs, new versions Keen to hear your views on your services & situation

Key refs supplied: Montgomery et al systematic review (summary) Barlow Cochrane review parenting for maltreating parents Webster Stratton on adaptation of parenting programme to i) child protection services ii) culturally diverse families Eg of two ‘transported’ parenting interventions tested in UK RCTs for high risk families, Gardner et al 2006, Hutchings et al 2007

Next- some spare slides:

Rolling out in UK: National Academy of Parenting Practitioners: Rationale: Many effective parenting programs But also much untested or ineffective practice (home start eg, much enthusiasm…. no effects) Many staff working with parents had little training for parenting work Govt DCSF set up academy in Nov 2007 to address these issues, via training and research. Partnership between voluntary sector, govt and KCL

Components of implementation fidelity • is the program delivered as designed? • are all the core components present? • with all of the core components? • to the right population? • with appropriately trained & supported staff? • using the right protocols, techniques and materials? • in the right context? • are participants engaged in the program?