Pharmacoepidemiology in pregnancy: analysis protocol for an observational cohort study in the UK Clinical Practice Research Datalink.
Pharmacoepidemiology in pregnancy: analysis protocol for an observational cohort study in the UK Clinical Practice Research Datalink.
复制标题
妊娠的药物ePIDEMIOGY：英国临床实践研究数据链接的观察队列研究的分析方案。
DOI：
10.12688/wellcomeopenres.17523.1
复制
发表时间：
2022
影响因子：
--
通讯作者：
中科院分区：
文献类型：
--
作者：
大量妇女在怀孕期间服用处方药和非处方药。然而，关于这些药物对母亲和后代的潜在影响的确切证据很少。我们将研究在整个妊娠期间继续使用处方药治疗慢性既存母体疾病（如糖尿病、高血压和甲状腺相关疾病）的风险和获益。如果不及时治疗，这些疾病是新生儿和产妇不良结局的既定风险因素。然而，这些疾病的一些治疗与不良新生儿结局相关。 我们的主要目标有两个。首先，我们的目的是评估在怀孕期间继续治疗对母亲的有益影响。其次，我们的目的是确定是否有一个相关的不利影响，对新生儿的母亲在怀孕期间继续治疗。为了建立这一证据，我们将调查母体药物处方与不利和有益的后代结果之间的关系，以提供证据来指导临床决策。 我们将使用英国临床实践研究数据链（CPRD）的数据进行假设检验观察性代际队列研究。我们将使用四种统计方法：多变量调整回归、倾向评分回归、工具变量分析和阴性对照分析。在估计药物暴露与母体或新生儿结局之间的相关性时，这些方法应考虑潜在的混杂因素。 在本方案中，我们描述了我们研究的目的、动机、研究设计、队列和统计分析，以帮助研究的重现性和透明度。
Large numbers of women take prescription and over-the-counter medications during pregnancy. However, there is very little definitive evidence about the potential effects of these drugs on the mothers and offspring. We will investigate the risks and benefits of continuing prescriptive drug use for chronic pre-existing maternal conditions such as diabetes, hypertension and thyroid related conditions throughout pregnancy. If left untreated, these conditions are established risk factors for adverse neonatal and maternal outcomes. However, some treatments for these conditions are associated with adverse neonatal outcomes. Our primary aims are twofold. Firstly, we aim to estimate the beneficial effect on the mother of continuing treatment during pregnancy. Second, we aim to determine whether there is an associated detrimental impact on the neonate of continuation of maternal treatment during pregnancy. To establish this evidence, we will investigate the relationship between maternal drug prescriptions and adverse and beneficial offspring outcomes to provide evidence to guide clinical decisions. We will conduct a hypothesis testing observational intergenerational cohort study using data from the UK Clinical Practice Research Datalink (CPRD). We will apply four statistical methods: multivariable adjusted regression, propensity score regression, instrumental variables analysis and negative control analysis. These methods should account for potential confounding when estimating the association between the drug exposure and maternal or neonatal outcome. In this protocol we describe the aims, motivation, study design, cohort and statistical analyses of our study to aid reproducibility and transparency within research.