背景:家族性高胆固醇血症(FH)是一种严重未被诊断和可治疗的遗传性脂质紊乱,它显著增加了早发心血管疾病的风险。单基因FH的患病率被认为是250-350中的1。NHS长期计划旨在与初级保健合作,在5年内将FH检测提高到至少25%,由NHS基因组学计划支持。
目的:本系统评价了初级保健中≥18岁成年人FH的系统筛查方法。
方法:七个数据库[Cochrane,PubMed,奥维德,CINAHL,ProQuest,WebofScience,Scopus],四个临床试验登记处[ISRCTN,ANZCTR,Clinicaltrials.gov,检索了2020年3月至2023年5月的WHO-ICTRP和相关灰色文献[OpenGrey]。只有包括成年人在内的研究才有资格。使用ROBINS-I评估偏倚风险。
结果:筛选了831条记录。没有随机化,确定了对照研究。从全文回顾来看,在57项(6.90%)中,有5项符合条件的非随机研究被确定.纳入的研究均使用电子病历(EMR)中的自动FH病例识别,并且是具有中等偏倚风险的高质量研究。叙事综合报告的结果包括三项算法研究,合并检出率,DR14.4%(95CI11.67-16.62),一项有监督的机器学习[合奏]研究,DR15.5%(95CI15.47-15.53)和一项使用混合诊断EMR模型和/或FH基因型的研究确认DR25.0%(95CI16.30-35.8)。在这些研究中没有报告不良反应。
结论:将EMR的自动病例发现与初级保健的临床随访相结合可以增强FH识别。结合基因分型的途径表现出最好的检出率。
BACKGROUND: Familial Hypercholesterolaemia (FH) is a greatly underdiagnosed and treatable genetic lipid disorder which significantly increases risk of premature cardiovascular disease. The prevalence of monogenic FH is thought to be 1 in 250-350. The NHS Long Term Plan aims to increase FH detection to at least 25% over 5 years in collaboration with primary care, supported by the NHS genomics programme.
OBJECTIVE: This systematic review evaluates systematic screening methods for FH in adults aged ≥18 years in primary care.
METHODS: Seven databases [Cochrane, PubMed, Ovid, CINAHL, ProQuest, Web of Science, Scopus], four clinical trial registries [ISRCTN, ANZCTR, Clinicaltrials.gov, WHO-ICTRP] and relevant grey literature [OpenGrey] from March 2020 to May 2023 were searched. Only studies including adults were eligible. Risk of bias was assessed using ROBINS-I.
RESULTS: 831 records were screened. No randomised, controlled studies were identified. From full-text review, five eligible non-randomised studies out of 57 (6.90%) were identified. The included studies all used automated FH case-identification from electronic medical records (EMR) and were high quality studies with a moderate risk of bias. Narrative synthesis reported outcomes which included three algorithmic studies, with a pooled detection rate, DR 14.4% (95%CI 11.67-16.62), one supervised Machine Learning [Ensemble] study, DR 15.5% (95%CI 15.47-15.53) and one study utilising a hybrid diagnostic EMR model and/or FH genotype confirmation DR 25.0% (95%CI 16.30-35.8). No adverse effects were reported in these studies.
CONCLUSIONS: Incorporating automated case-finding from EMR with clinical follow-up in primary care can enhance FH identification. Pathways incorporating genotyping showed the best detection rate.