ESPE Abstracts

Background: Growth delay is a common cause of consultation in pediatric practice and can be an early indicator of an underlying pathological process. As aetiological treatments (hormone therapy, special diets, etc.) have been routinely available for decades, it has become ethically challenging to conduct prospective randomised trials to test the efficacy of updated clinical approaches.

Objectives: To assess the influence of etiological treatment on catch-up growth using Inverse Probability of Treatment Weighting (IPTW) and investigate factors associated with good catch-up growth in children treated with etiological treatment (ET) and a wait-and-see approach (WS).

Study design: Retrospective cohort of 198 children referred for short stature as sole reason for consultation to a specialized endocrinology consultation between 2000 and 2019.

Intervention: etiological treatment of growth delay or wait-and-see approach

Primary Outcome: Catch-up growth evaluated one year after receiving either ET or a WS.

Secondary Outcome: factors associated with good catch-up growth independently of the specific intervention received.

Statistic: We used IPTW, a causal inference method, to analyse the effect of an etiological treatment on catch-up growth. To deal with missing data, we applied a variable selection procedure using backward selection by Akaike Information Criterion. Then we used weighted Poisson regression with robust variance to estimate relative risk. The effect of the treatment was reported through the average treatment effect in the treated.

Results: Etiologies were secondary growth failure (42%), idiopathic short stature (32%) and primary growth failure (22%). Etiological treatment were growth hormone, alone or combined with thyroid hormone (n = 96), thyroid hormone replacement therapy (n = 6), sodium bicarbonate therapy (n = 1) or gluten free diet (n = 2). Good catch-up growth was more frequent in the ET group (31.31%) than in the WS group (3.53%), P <0.0001, with patients in the ET group having a 10.3 timer higher chance of achieving good catch-up growth afer one year. Younger age and lower growth velocity at start of intervention were significantly associated with better catch-up growth, independently of the approach (ET or WS).

Conclusion: We demonstrated, using IPTW, that etiological treatment is effective in growth catch-up. Younger age and lower growth velocity are predictors of good catch-up growth in children referred for short stature, whether there is a treatment or not. While the retrospective design has some limitations, the use of causal inference methods helps strengthen the validity of the findings in a real-world setting.