Search tips
Search criteria 


Logo of archdischfnArchives of Disease in Childhood - Fetal & NeonatalVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Arch Dis Child Fetal Neonatal Ed. 2007 November; 92(6): F427–F428.
PMCID: PMC2675382

A good idea that doesnt work the Parent Baby Interaction Programme

Short abstract

Perspective on the paper by Glazebrook et alsee page 438

Keywords: interaction, parental stress, preterm infant, very premature birth

It is impossible to work with parents and children for any length of time without coming across situations where mothers, fathers, or both seem to need help with parenting. It has long been known that there are associations between the quality of parenting and childrens outcomes. Additional difficulties with establishing appropriate parenting styles are imposed on families as a result of their baby needing intensive care. It is therefore important to find out which interventions, provided in the setting of a neonatal intensive care unit NICU, might improve parenting and whether this in turn could mediate better outcomes for babies, and their parents, in families to which such interventions are given.

Over the past 20 years a great deal of work has evaluated interventions to improve parenting, and it is fortunate that many of the published studies have been randomised controlled trials. This in turn has allowed for two related Cochrane reviews, one in 20031 on child outcomes, and one in 20042 for maternal outcomes. For children, such programmes seem to lead to improvements in attachment and behaviour and for mothers, indices of mental health, such as depression, show improvements. These effects seem to be mediated by increased maternal sensitivity towards the baby that in turn allows more secure attachment for the child, and perhaps a more rewarding relationship for the mother. So there seems to be reasonable evidence that, in general, some of these interventions work but which, and for whom And what is the best way to evaluate the efficacy of such interventions

The paper by Glazebrook et al, evaluating the Parent Baby Interaction Programme PBIP in a NICU setting, is another randomised controlled trial that further contributes to the literature on interventional parenting programmes. It is important for two reasons its demonstration that this particular intervention programme was ineffective, and the rigour of the methodology with which that finding was demonstrated. The authors discuss the possible reasons for the ineffectiveness of the PBIP, and I will not repeat them here. But is worth emphasising just how similar the measured outcomes were in the subjects and controls, given that the study was strongly powered to detect clinically relevant group differences. This demonstration of complete ineffectiveness undermines the suggestion that more intervention would have thrown up an appreciable advantage for the index babies over the controls, and strongly suggests that PBIP simply does not work.

Finding that things do not work is good for healthcare because it prevents the enthusiastic adoption, on the basis of theoretical plausibility or extrapolation from other work, of programmes that have no benefit. Indeed a programme that is plausibly beneficial may even turn out to be harmful when properly studied. New programmes will either carry a new direct cost for instance, the establishment of a new post to deliver or coordinate the programme or impose opportunity costs for instance, if nurses deliver a new kind of care they may have to stop doing something else so if they do not work, we should not waste resources on them. But it can be an uphill struggle to do rigorous evaluations of interventions that relate to parenting because they carry a veneer of being obviously good. We all want to be better parents, and we all want parenting to be betterfor paediatricians, it is almost an article of faith. However, even though there is good evidence that some parenting programmes seem to work does not mean that every parenting programme will be effective. Each has to be evaluated on its merits.

So to the methodology. Glazebrook et al used a clusterrandomisation technique, a crossover design, pairing of units to account for socioeconomic variation, and two demographically different regions of the UK Trent and the South West. This allowed them to maximise power, minimise crosscontamination between groups, allow for deprivation and generalise the findings. Trials of this kind require a huge investment of time and money, but the payoff for the National Health Service is even bigger, given that no unit could now plausibly make a bid to introduce PBIP in its hospital. Indeed it should also have the effect of increasing the care with which any other proposed parenting programme should be scrutinised before introducing it. We have to wonder whether the other trials in the Cochrane review were as well designed as this. If they had been, would the metaanalyses have come up with such unequivocal positive findings

When a study that is as well designed and powerful as this one fails to show any effect of the intervention, it is important to look critically again at other studies that purport to show benefits of analogous interventions. Above all, what this study adds to the literature is that every parenting intervention programme should be evaluated with the same rigour. Glazebrook et al have set the standard to which others must aspire.


Competing interests Martin P Ward Platt is Deputy Editor of the Archives of Disease in Childhood.


1. Barlow J, Parsons J. Groupbased parenttraining programmes for improving emotional and behavioural adjustment in 03 year old children. Cochrane Database Syst Rev 2003. 1CD003680. [PubMed]
2. Barlow J, Coren E. Parenttraining programmes for improving maternal psychosocial health. Cochrane Database Syst Rev 2004. 1CD002020. [PubMed]

Articles from Archives of Disease in Childhood. Fetal and Neonatal Edition are provided here courtesy of BMJ Group