Graduate Program in Dentistry, Federal University of Pelotas, Pelotas, Brazil.
Faculty of Health and Behavioural Sciences, School of Dentistry, The University of Queensland, Queensland, Brisbane, Australia.
Community Dent Oral Epidemiol. 2023 Feb;51(1):62-66. doi: 10.1111/cdoe.12802. Epub 2023 Feb 7.
Oral conditions represent a critical public health challenge, and together with descriptive and predictive epidemiology, causal inference has a crucial role in developing and testing preventive oral health interventions. By identifying not just correlations but actual causes of disease, causal inference may quantify the average effect of interventions and guide policies. Although authors are not usually explicit about it, most oral health studies are guided by causal questions. However, methodological deficiencies limit their interpretability and the implementation of their findings. This manuscript is a call to action on the use of causal inference in oral research. Its application starts with asking theoretically sound questions and being explicit about causal relationships, defining the estimates to evaluate, and measuring them properly. Beyond promoting causal analytical approaches, we emphasize the need for more causal thinking to promote thoughtful research questions and the use of appropriate methods to answer them. Causal inference relies on the plausibility of assumptions underlying the data analysis and the quality of the data, and we argue that high-quality observational studies can be used to estimate average causal effects. Although individual efforts to embrace causal inference in dentistry are essential, they will not yield substantial results if not led by a systematic and structural change in the field. We urge scientific societies, funding bodies, dental schools, and journals to promote transparency in research, causal thinking, and causal inference projects to move the field toward more meaningful studies. It is also time for researchers to move forward and connect with the community, co-produce investigations and translate their findings, and engage in interventions that impact public health. We conclude by highlighting the importance of triangulating results from different data sources and methods to support causal inference and inform decision-making on interventions to effectively improve population oral health.
口腔状况是一个严峻的公共卫生挑战，描述性和预测性流行病学以及因果推断在开发和测试预防口腔健康干预措施方面发挥着至关重要的作用。通过识别不仅是相关性，而且是疾病的实际原因，因果推断可以量化干预措施的平均效果并指导政策。尽管作者通常没有明确说明，但大多数口腔健康研究都受到因果问题的指导。然而，方法学上的缺陷限制了它们的可解释性和实施。本文呼吁在口腔研究中使用因果推断。其应用始于提出理论上合理的问题，并明确因果关系，定义要评估的估计值，并正确测量它们。除了促进因果分析方法外，我们还强调需要更多的因果思维，以促进有思想的研究问题，并使用适当的方法来回答这些问题。因果推断依赖于数据分析和数据质量的假设的合理性，我们认为高质量的观察性研究可以用于估计平均因果效应。虽然在牙科中个别努力接受因果推断是必要的，但如果不在该领域进行系统和结构性的变革，这些努力将不会产生实质性的结果。我们敦促科学学会、资助机构、牙科学院和期刊提高研究、因果思维和因果推断项目的透明度，以使该领域朝着更有意义的研究方向发展。现在也是研究人员向前迈进并与社区联系、共同开展调查并翻译他们的发现以及参与影响公共卫生的干预措施的时候了。最后，我们强调了从不同数据源和方法汇总结果以支持因果推断并为干预措施提供决策信息以有效改善人口口腔健康的重要性。
