PDF(Pubmed)
Abstract：
Frailty, neurodegeneration and geriatric syndromes cause a significant impact at the clinical, social, and economic level, mainly in the context of the aging world. Recently, Information and Communication Technologies (ICTs), virtual reality tools, and machine learning models have been increasingly applied to the care of older patients to improve diagnosis, prognosis, and interventions. However, so far, the methodological limitations of studies in this field have prevented to generalize data to real-word. This review systematically overviews the research designs used by studies applying technologies for the assessment and treatment of aging-related syndromes in older people.
Following the PRISMA guidelines, records from PubMed, EMBASE, and Web of Science were systematically screened to select original articles in which interventional or observational designs were used to study technologies\' applications in samples of frail, comorbid, or multimorbid patients.
Thirty-four articles met the inclusion criteria. Most of the studies used diagnostic accuracy designs to test assessment procedures or retrospective cohort designs to build predictive models. A minority were randomized or non-randomized interventional studies. Quality evaluation revealed a high risk of bias for observational studies, while a low risk of bias for interventional studies.
The majority of the reviewed articles use an observational design mainly to study diagnostic procedures and suffer from a high risk of bias. The scarce presence of methodologically robust interventional studies may suggest that the field is in its infancy. Methodological considerations will be presented on how to standardize procedures and research quality in this field.
Following the PRISMA guidelines, records from PubMed, EMBASE, and Web of Science were systematically screened to select original articles in which interventional or observational designs were used to study technologies\' applications in samples of frail, comorbid, or multimorbid patients.
Thirty-four articles met the inclusion criteria. Most of the studies used diagnostic accuracy designs to test assessment procedures or retrospective cohort designs to build predictive models. A minority were randomized or non-randomized interventional studies. Quality evaluation revealed a high risk of bias for observational studies, while a low risk of bias for interventional studies.
The majority of the reviewed articles use an observational design mainly to study diagnostic procedures and suffer from a high risk of bias. The scarce presence of methodologically robust interventional studies may suggest that the field is in its infancy. Methodological considerations will be presented on how to standardize procedures and research quality in this field.
摘要：
背景：脆弱，神经变性和老年综合征在临床上造成重大影响，社会,和经济水平，主要是在老龄化世界的背景下。最近,信息和通信技术（ICT），虚拟现实工具，机器学习模型越来越多地应用于老年患者的护理，以改善诊断，预后,和干预。然而,到目前为止,该领域研究的方法论局限性阻止了将数据推广到真实单词。这篇综述系统地概述了应用技术评估和治疗老年人衰老相关综合征的研究设计。
方法：遵循PRISMA指南，PubMed的记录,EMBASE,和WebofScience进行了系统筛选，以选择原始文章，其中使用了介入或观察性设计来研究技术在脆弱样本中的应用，共病，或者多病人。
结果：34篇符合纳入标准。大多数研究使用诊断准确性设计来测试评估程序或回顾性队列设计来构建预测模型。少数为随机或非随机介入研究。质量评估显示，观察性研究存在很高的偏倚风险，而介入研究的偏倚风险较低。
结论：大多数审查的文章主要使用观察性设计来研究诊断程序，并且存在较高的偏倚风险。方法学上健壮的介入研究很少，这可能表明该领域处于起步阶段。将介绍如何标准化该领域的程序和研究质量的方法。
方法：遵循PRISMA指南，PubMed的记录,EMBASE,和WebofScience进行了系统筛选，以选择原始文章，其中使用了介入或观察性设计来研究技术在脆弱样本中的应用，共病，或者多病人。
结果：34篇符合纳入标准。大多数研究使用诊断准确性设计来测试评估程序或回顾性队列设计来构建预测模型。少数为随机或非随机介入研究。质量评估显示，观察性研究存在很高的偏倚风险，而介入研究的偏倚风险较低。
结论：大多数审查的文章主要使用观察性设计来研究诊断程序，并且存在较高的偏倚风险。方法学上健壮的介入研究很少，这可能表明该领域处于起步阶段。将介绍如何标准化该领域的程序和研究质量的方法。
