Why consistent completion criterion are required in childhood weight management programmes

- Inconsistent criterion for programme completion is problematic.
- Inconsistent criterion greatly limits the generalisability of study findings.
- The mean change in standardised body mass index was strongly associated with the completion criterion.
- The proportion of completers was strongly associated with the completion criterion.
- The predictors of completion vary substantially under alternative completion criterion.

ObjectivesCurrent research in the field of childhood weight management (WM) effectiveness is hampered by inconsistent terminology and criterion for WM programme completion, alongside other engagement-related concepts (e.g. adherence, dropout and attrition). Evidence reviews are not able to determine conclusive intervention effectiveness because of this issue. This study aims to quantify how various completion criterion impacts upon on: 1) the percentage of WM completers; 2) the standardised body mass index (BMI SDS) reduction; and 3) the predictors of WM completion.Study designA methodological, sensitivity analysis to examine how differential completion criterion affect programme outcomes and predictors.MethodsSecondary data of 2948 children were used. All children attended a MoreLife WM programme between 2009 and 2014. The completion criterion was incrementally adjusted by 10% (i.e. completer attends 10%, 20%, 30%... of sessions) for research aims 1-2, with the percentage of completers and change in BMI SDS calculated at each increment. For aim 3, the stability (strength, direction and significance) of the predictors were examined when using the completion criterion of four alternative studies against our previous study (completion ≥70% attendance).ResultsThe volume of programme completers decreased in a linear manner as the completion criterion became more stringent (i.e. 70-100% attendance). The change in BMI SDS conversely became incrementally greater. The strength, direction and significance of the predictors was highly dependent on the completion criterion; the odds ratio varied by 24.2% across a single predictor variable (delivery period). The degree of change is evidenced in the paper.ConclusionsInconsistent completion criterion greatly limits the synthesis of programme effectiveness and explains some of the inconsistency in the predictors of engagement. Standardised criterion for engagement-related terminology are called for.