小狗阅读会员会员
医学顶刊SCI精读工具

扫码登录小狗阅读

阅读SCI医学文献
Document
订阅泛读方向 订阅泛读期刊
  • 我的关注
  • 我的关注
  • {{item.title}}

    按需关注领域/方向,精准获取前沿热点

  • {{item.title}}

    {{item.follow}}人关注

  • {{item.subscribe_count}}人订阅

    IF:{{item.impact_factor}}

    {{item.title}}

Prediction of gestational diabetes based on nationwide electronic health records

基于全国电子健康档案的妊娠期糖尿病预测

  • 影响因子:19.14
  • DOI:10.1038/s41591-019-0724-8
  • 作者列表:"Artzi, Nitzan Shalom","Shilo, Smadar","Hadar, Eran","Rossman, Hagai","Barbash-Hazan, Shiri","Ben-Haroush, Avi","Balicer, Ran D.","Feldman, Becca","Wiznitzer, Arnon","Segal, Eran
  • 发表时间:2020-01-01
Abstract

Leveraging the availability of nationwide electronic health records from over 500,000 pregnancies in Israel, a machine-learning approach offers an alternative means of predicting gestational diabetes at high accuracy in the early stages of pregnancy. Gestational diabetes mellitus (GDM) poses increased risk of short- and long-term complications for mother and offspring^ 1 – 4 . GDM is typically diagnosed at 24–28 weeks of gestation, but earlier detection is desirable as this may prevent or considerably reduce the risk of adverse pregnancy outcomes^ 5 , 6 . Here we used a machine-learning approach to predict GDM on retrospective data of 588,622 pregnancies in Israel for which comprehensive electronic health records were available. Our models predict GDM with high accuracy even at pregnancy initiation (area under the receiver operating curve (auROC) = 0.85), substantially outperforming a baseline risk score (auROC = 0.68). We validated our results on both a future validation set and a geographical validation set from the most populated city in Israel, Jerusalem, thereby emulating real-world performance. Interrogating our model, we uncovered previously unreported risk factors, including results of previous pregnancy glucose challenge tests. Finally, we devised a simpler model based on just nine questions that a patient could answer, with only a modest reduction in accuracy (auROC = 0.80). Overall, our models may allow early-stage intervention in high-risk women, as well as a cost-effective screening approach that could avoid the need for glucose tolerance tests by identifying low-risk women. Future prospective studies and studies on additional populations are needed to assess the real-world clinical utility of the model.

摘要

利用以色列超过 500,000 例妊娠的全国电子健康记录的可用性,机器学习方法提供了一种在妊娠早期高精度预测妊娠糖尿病的替代方法。妊娠期糖尿病 (GDM) 增加了母亲和后代的短期和长期并发症的风险 ^ 1-4。GDM 通常在妊娠 24-28 周诊断,但应尽早发现,因为这可能会预防或大大降低不良妊娠结局的风险 ^ 5,6。在这里,我们使用机器学习方法对以色列 588,622 例妊娠的回顾性数据进行 GDM 预测,这些妊娠具有全面的电子健康记录。我们的模型即使在妊娠开始时也以较高的准确性预测 GDM (受试者工作曲线下面积 (auROC) = 0.85),大大优于基线风险评分 (aurocc = 0.68)。我们在以色列耶路撒冷人口最多的城市的未来验证集和地理验证集上验证了我们的结果,从而模拟了现实世界的表现。询问我们的模型,我们发现了以前未报告的风险因素,包括既往妊娠葡萄糖激发试验的结果。最后,我们设计了一个更简单的模型,该模型仅基于患者可以回答的 9 个问题,准确率仅略有降低 (auroc = 0.80)。总的来说,我们的模型可能允许对高风险女性进行早期干预, 以及一种具有成本效益的筛查方法,可以通过确定低风险女性来避免糖耐量试验的需要。未来的前瞻性研究和额外人群的研究需要评估模型的真实世界临床效用。

关键词:
阅读人数:7人
下载该文献
小狗阅读

帮助医生、学生、科研工作者解决SCI文献找不到、看不懂、阅读效率低的问题。提供领域精准的SCI文献,通过多角度解析提高文献阅读效率,从而使用户获得有价值研究思路。

相关文献
影响因子:2.86
发表时间:2020-01-08
来源期刊:Acta Diabetologica
DOI:10.1007/s00592-019-01469-5
作者列表:["Benhalima, Katrien","Crombrugge, Paul","Moyson, Carolien","Verhaeghe, Johan","Vandeginste, Sofie","Verlaenen, Hilde","Vercammen, Chris","Maes, Toon","Dufraimont, Els","Block, Christophe","Jacquemyn, Yves","Mekahli, Farah","Clippel, Katrien","Den Bruel, Annick","Loccufier, Anne","Laenen, Annouschka","Minschart, Caro","Devlieger, Roland","Mathieu, Chantal"]

METHODS:Aims We aimed to develop a prediction model based on clinical and biochemical variables for gestational diabetes mellitus (GDM) based on the 2013 World Health Organization (WHO) criteria. Methods A total of 1843 women from a Belgian multi-centric prospective cohort study underwent universal screening for GDM. Using multivariable logistic regression analysis, a model to predict GDM was developed based on variables from early pregnancy. The performance of the model was assessed by receiver-operating characteristic (AUC) analysis. To account for over-optimism, an eightfold cross-validation was performed. The accuracy was compared with two validated models (van Leeuwen and Teede). Results A history with a first degree relative with diabetes, a history of smoking before pregnancy, a history of GDM, Asian origin, age, height and BMI were independent predictors for GDM with an AUC of 0.72 [95% confidence interval (CI) 0.69–0.76)]; after cross-validation, the AUC was 0.68 (95% CI 0.64–0.72). Adding biochemical variables, a history of a first degree relative with diabetes, a history of GDM, non-Caucasian origin, age, height, weight, fasting plasma glucose, triglycerides and HbA_1c were independent predictors for GDM, with an AUC of the model of 0.76 (95% CI 0.72–0.79); after cross-validation, the AUC was 0.72 (95% CI 0.66–0.78), compared to an AUC of 0.67 (95% CI 0.63–0.71) using the van Leeuwen model and an AUC of 0.66 (95% CI 0.62–0.70) using the Teede model. Conclusions A model based on easy to use variables in early pregnancy has a moderate accuracy to predict GDM based on the 2013 WHO criteria.

关键词: 暂无
翻译标题与摘要 下载文献
影响因子:19.14
发表时间:2020-01-01
来源期刊:Nature Medicine
DOI:10.1038/s41591-019-0724-8
作者列表:["Artzi, Nitzan Shalom","Shilo, Smadar","Hadar, Eran","Rossman, Hagai","Barbash-Hazan, Shiri","Ben-Haroush, Avi","Balicer, Ran D.","Feldman, Becca","Wiznitzer, Arnon","Segal, Eran"]

METHODS:Leveraging the availability of nationwide electronic health records from over 500,000 pregnancies in Israel, a machine-learning approach offers an alternative means of predicting gestational diabetes at high accuracy in the early stages of pregnancy. Gestational diabetes mellitus (GDM) poses increased risk of short- and long-term complications for mother and offspring^ 1 – 4 . GDM is typically diagnosed at 24–28 weeks of gestation, but earlier detection is desirable as this may prevent or considerably reduce the risk of adverse pregnancy outcomes^ 5 , 6 . Here we used a machine-learning approach to predict GDM on retrospective data of 588,622 pregnancies in Israel for which comprehensive electronic health records were available. Our models predict GDM with high accuracy even at pregnancy initiation (area under the receiver operating curve (auROC) = 0.85), substantially outperforming a baseline risk score (auROC = 0.68). We validated our results on both a future validation set and a geographical validation set from the most populated city in Israel, Jerusalem, thereby emulating real-world performance. Interrogating our model, we uncovered previously unreported risk factors, including results of previous pregnancy glucose challenge tests. Finally, we devised a simpler model based on just nine questions that a patient could answer, with only a modest reduction in accuracy (auROC = 0.80). Overall, our models may allow early-stage intervention in high-risk women, as well as a cost-effective screening approach that could avoid the need for glucose tolerance tests by identifying low-risk women. Future prospective studies and studies on additional populations are needed to assess the real-world clinical utility of the model.

关键词: 暂无
翻译标题与摘要 下载文献
影响因子:4.34
发表时间:2020-01-27
DOI:10.1128/AAC.01777-19
作者列表:["Lowes DJ","Hevener KE","Peters BM"]

METHODS::Repurposing of currently approved medications is an attractive option for the development of novel treatment strategies against physiological and infectious diseases. The antidiabetic sulfonylurea glyburide has demonstrated off-target capacity to inhibit activation of the NLRP3 inflammasome in a variety of disease models, including vaginal candidiasis, caused primarily by the fungal pathogen Candida albicans Therefore, we sought to determine which of the currently approved sulfonylurea drugs prevent the release of interleukin 1β (IL-1β), a major inflammasome effector, during C. albicans challenge of the human macrophage-like THP1 cell line. Findings revealed that the second-generation antidiabetics (glyburide, glisoxepide, gliquidone, and glimepiride), which exhibit greater antidiabetic efficacy than prior iterations, demonstrated anti-inflammatory effects with various degrees of potency as determined by calculation of 50% inhibitory concentrations (IC50s). These same compounds were also effective in reducing IL-1β release during noninfectious inflammasome activation (e.g., induced by lipopolysaccharide [LPS] plus ATP), suggesting that their anti-inflammatory activity is not specific to C. albicans challenge. Moreover, treatment with sulfonylurea drugs did not impact C. albicans growth and filamentation or THP1 viability. Finally, the use of ECE1 and Candidalysin deletion mutants, along with isogenic NLRP3-/- cells, demonstrated that both Candidalysin and NLRP3 are required for IL-1β secretion, further confirming that sulfonylureas suppress inflammasome signaling. Moreover, challenge of THP1 cells with synthetic Candidalysin peptide demonstrated that this toxin is sufficient to activate the inflammasome. Treatment with the experimental inflammasome inhibitor MCC950 led to similar blockade of IL-1β release, suggesting that Candidalysin-mediated inflammasome activation can be inhibited independently of potassium efflux. Together, these results demonstrate that the second-generation antidiabetic sulfonylureas retain anti-inflammatory activity and may be considered for repurposing against immunopathological diseases, including vaginal candidiasis.

方向

复制标题
发送后即可在该邮箱或我的下载查看该文献
发送
该文献默认存储到我的下载

科研福利

临床科研之家订阅号

报名咨询

建议反馈
问题标题:
联系方式:
电子邮件:
您的需求: