Large sample size in child and adolescent psychiatric research: the way of salvation?

Abstract

As in other medical fields, the shift toward scientific-based clinical practice in both diagnostics and treatment is ongoing in child and adolescent psychiatry [9]. However, numerous characteristics more or less specific for our field impede the generation of evidence and therefore in the end the implementation of evidence-based guidelines. Three examples serve to illustrate this impediment. (1) The low level of introspection of young subjects reduces diagnostic accuracy. (2) The high rates of comorbidity limit specificity and generalization of findings. (3) The requirement of caregivers' consent for study participation hampers recruitment efforts. Such issues entail the frequently posed question: ''How large is the impact and cost-benefit ratio of the different sampling methodologies, i.e., comparing register-based versus clinically selected versus single case studies [8] ? '' Interestingly, in the current issue of European Child and Adolescent Psychiatry there is only one randomized controlled trial (RCT). Van den Hoofdakker et al. [12] explored the influence of paternal variables on the outcome of behavioral parent training (BPT) in n = 83 children with attention-deficit/hyperactivity disorder (ADHD) by comparing a group receiving BPT plus ongoing routine clinical care (RCC) versus an RCC alone group. This is all the more noteworthy as across the medical sciences RCTs are often considered as the most important source of evidence. But RCTs face important ethical and logistical constraints, particularly in children and adolescents with the consequence of, e.g., smaller sample sizes and shorter observation periods. In addition, RCTs have been criticized for focusing on highly selected populations and outcomes.All the other findings in the current issue are based on much larger sample sizes. Four studies used questionnaires in large samples of children and adolescents [1,2,5,7]. One population-based study rendered possible via linkage of records of three health data sets compared disordered (ADHD; n = 11,902) and non-disordered children (healthy controls; n = 27,304) under 18 years [11]. Two reviews, one on studies which reported white matter/gray matter changes in pediatric and adolescent bipolar disorder/unipolar depression, as detected by diffusion tensor imaging and voxel-based analysis [10], and the other providing an overview of ASD screening studies and ongoing programs across Europe [4], also report findings on large cumulated total samples of clinically diagnosed patients.But do studies with larger sample sizes result regularly in findings of higher quality or relevance?First, the advantages of a large sample size include a more precise estimate of the effect size and an easier assessment of the representativeness of the sample and the generalizability of the achieved results. However, a small effect size may not prove to be of clinical relevance. In addition, both selection bias and the negative impact of confounders need to be considered with care. The reduced number of variables and/or quality of ...

Internet Addiction Amongst University Students Under COVID-19: Prevalence and Correlates

“…Second, this study employed a large sample size which is uncommon in other existing studies. With a large sample size, the estimation would be more precise, and the generalizability of the results would increase (Roessner, 2014). Third, this study involved multiple risks and protective factors to IA which were considered important during the pandemic, while existing studies normally focused on one or two predicting factors.…”

Incidence of gestational diabetes mellitus in the United Arab Emirates; comparison of six diagnostic criteria: The Mutaba’ah Study

“…The main strength of our study was the large representative population that increased the study power and generalizability of the findings and minimized estimate errors ( 57 ). The Mutaba’ah study is the largest prospective mother and child cohort study in the UAE, which provides data on maternal and child health from conception to adolescence.…”

Gestational Diabetes Mellitus: A Cross-Sectional Survey of Its Knowledge and Associated Factors among United Arab Emirates University Students

“…One of the strengths of our study is the large sample size which increases the precision of estimates by reducing margins of error [ 41 ]. To our knowledge, this is the first study to report GDM knowledge among university students in the MENA region.…”