Enhancing the Safe and Effective Use of Medicines in Hospitalized Patients: An AI Led Population Health Study

加强住院患者药物的安全有效使用：人工智能主导的人口健康研究

• 批准号: 2265644
• 金额: --
• 依托单位: Swansea University
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2019
• 资助国家: 英国
• 起止时间: 2019 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2265644
• 关键词: Enhancing; Safe; Effective; Use; Medicines

Theme 2 - Biological, Health and Clinical ScienceCancer patients are at an increased risk of harm from medication errors as chemotherapy can have significant toxicity, even at therapeutic doses, and cancer patients can have co-morbidities which affect their tolerance and increase likelihood of drug interactions through polypharmacy. The aim of this project is to develop advanced machine learning and computational methods with electronic health records and reports to enable the safe and effective use of medicines in cancer treatments to achieve the best possible outcomes for hospitalized patients. This will enable the Specialist Cancer Pharmacist to be released to undertake more patient facing roles, and have additional time to focus on supporting the management and usage of high risk drugs.The objectives of this study are- To develop scalable machine learning methods for big data analytics with high dimensional data.- To extract phenotypic factors from clinical reports associated with drug safety and patient safety.- To identify the predictors of medication errors including adverse drug reactions, while considering co-morbidities, inter-current illness, co-prescribing factors, patient-related factors, antineoplastic drugs and hospital organisation.- To integrate these factors to identify which group of patients are more effective in certain drug treatments.

主题2 -生物学、健康和临床科学癌症患者因用药错误而受到伤害的风险增加，因为化疗可能具有显著的毒性，即使在治疗剂量下，癌症患者也可能患有并发症，这会影响他们的耐受性，并通过多种药物增加药物相互作用的可能性。该项目的目的是开发先进的机器学习和计算方法，并提供电子健康记录和报告，以确保在癌症治疗中安全有效地使用药物，为住院患者实现最佳效果。这将使专科癌症药剂师能够承担更多面向患者的角色，并有更多的时间专注于支持高风险药物的管理和使用。本研究的目标是-开发可扩展的机器学习方法，用于高维数据的大数据分析。从临床报告中提取与药物安全性和患者安全性相关的表型因素。确定用药错误的预测因素，包括药物不良反应，同时考虑合并症，并发症，共同处方因素，患者相关因素，药物和医院组织。整合这些因素，以确定哪组患者在某些药物治疗中更有效。