Comparing apples and pears: misleading conclusions about the population mental health impact of a parenting programme, a commentary on Marryat, Thompson and Wilson (2017).

backgroundThe article by Marryat, Thompson and Wilson (2017) in BMC Pediatrics presents an evaluation of the implementation of the Triple P system as a public health intervention conducted by the Glasgow City Council and NHS Greater Glasgow and Clyde. DISCUSSION: Unfortunately, the conclusions drawn are questionable for multiple reasons. The lack of a controlled design precludes defensible conclusions about intervention effects free from routine threats to internal validity. There was a substantial mismatch between the intervention sample and the population sample assessed. The article's title and abstract leave readers with the mistaken impression that the children assessed for outcome were suitably representative of intervention families, when in fact many of the children in the intervention families were missing from the teacher-report outcome assessment (a single questionnaire), and many or most of the children in the teacher-report outcome assessment belonged to families who had never received the intervention. Although Triple P targets parent-child relations and child behavioural and emotional problems at home, Marryat et al. narrowly defined mental health impact as child difficulties in nursery or preschool, while not reporting data from practitioners and parents in the same evaluation that did not support the authors' conclusion. The paper was further diminished by a number of misleading statements and factual errors related for example to other research on Triple P. Studying the extent to which child mental health functioning at home can generalise to school settings is an important topic of inquiry in relation to parenting support interventions, but unfortunately the Marryat et al. article did not move this area forward.