The impact of bias of underlying literature in guidelines on its recommendations: assessment of the German fluoride guideline.

Purpose: The significance of the underlying literature in clinical guidelines can be weakened by the risk of bias, which could negatively affect the recommendations. Especially in controversial matters, such as fluoride use for caries prevention in children, biased results may be not reliable and lead to incorrect conclusions. This study was performed to detect bias in underlying literature of the German guideline for caries prevention using fluoride in children, where no consensus was reached between paediatricians and paediatric dentists.

Methods: Three tools used for risk of bias assessments of different study designs were RoB 2 for RCTs, ROBINS-I for non-randomized studies, and ROBIS for systematic reviews. For each study cited in the guideline two independent risk of bias assessments were performed. Disagreements were resolved by consensus.

Results: Out of 58 papers, 48.3% (n = 28) showed high risk of bias, with the majority in sections regarding fluoride tablets, fluoridated toothpaste, and paediatricians' recommendations. 9 out of 20 recommendations and statements were based on studies with high risk of bias, all of which were in these three controversial sections. 13 out of 29 RCTs showed high risk of bias (44.8%), as all 13 non-randomized trials did, while only 2 of 16 (12.5%) systematic reviews had high risk of bias.

Conclusion: Considering risk of bias of cited studies in clinical guidelines may result in substantial changes in its recommendations and aid in reaching consensus. Efforts should be made to assess risk of bias of underlying literature in future clinical guidelines.