SAS Blog Post - 17 Feb, 2020
Risk Management of Hospital-acquired Infections During Hospitalization 住院感染的風險管理
In the case of a public health crisis, hospitals are one of the potential channels that lead to massive disease outbreak caused by cross-infection. With advanced analytics, hospitals can streamline and optimize the use of patients’ data to forecast which patients have a higher potential risk of developing a hospital-acquired infection during hospitalization, and as a result, reduce the number of cross-infections.
Hospitals in the Region of Southern Denmark have implemented the first complete system for monitoring hospital-acquired infections with the support of SAS Analytics. Within SAS Visual Data Mining and Machine Learning, the system’s risk models have been developed based on 284,000 previous patient cases in the region. The infection-monitoring overview is displayed through the Region’s management information portal, where doctors and departmental managers are accustomed to retrieving other information. Case study: https://www.sas.com/en_hk/customers/the-region-of-southern-denmark-global.html
當公共衛生問題發生時，醫院是其中一個經交叉感染而導致大規模疫症爆發的潛在途徑。醫院可利用進階分析，精簡及優化患者數據使用，以預測哪些患者屬住院感染的高風險人士，從而減少交叉感染個案。以南丹麥大區的醫院為例，它們利用SAS Analytics設置首個全面住院感染監察系統，並透過SAS Visual Data Mining and Machine Learning分析區內28萬4千個過往病人病例，構建風險模型。醫生與部門經理則能透過區內常用的管理資訊渠道獲取住院感染監測概況。案例參考: https://www.sas.com/en_hk/customers/the-region-of-southern-denmark-global.html
Prediction of Potential Epidemic Outbreaks 預測流行病爆發情況
Predictive analytics uses big data to predict the potential onset and spread of an epidemic outbreak. By extracting information from existing data sets and using AI-based analytics, medical professionals can allocate adequate resources not only to prepare for an imminent health crisis but also to determine future medical trends and patterns.
SAS Viya VTA, the big data analytics platform, can compare disease fatality rates in different affected areas through real-time monitoring of the number of confirmed diagnoses and determine the developing status of the disease. It also allows users to monitor and collect health data from global platforms and further analyze them by filtering hot topics, keywords, symptoms, behaviors, and complicated languages. With the filtered data, it facilitates visual investigation and visual text analytics via SAS Viya Demo, which provides innovative forecasting models and new approaches to analyze disease prediction with the aim of pinpointing key barriers and facilitators to the delivery of proven effective interventions.
With the public concern regarding the recent coronavirus (COVID-19), SAS also used its Viya VTA to consolidate and analyze the relevant data of the outbreak situation in China for near real-time reference and use.
預測性分析利用大數據來預測疫症爆發的可能性及潛在危機。透過從現有數據庫中擷取相關資訊，並進行AI分析，讓專業醫療人員可有效分配足夠資源，不但能夠為迫在眉睫的健康問題作好準備，同時亦可預測未來的醫療趨勢與模式。例如，大數據分析平台SAS Viya VTA可以透過實時監控疾病確診個案，對比不同受影響地區的死亡率，並斷定疾病的發展情況。平台還允許用戶監測及收集全球資訊平台上的公共衛生數據，通過過濾熱門話題、關鍵字、症狀、行為及複雜語言等，再利用SAS Viya Demo就過濾後的數據進行可視化調查（Visual Investigation）及可視化文本分析（Visual Text Analytics）。SAS Viya Demo能夠提供創新的預測模型和全新途徑來分析疾病預測方法，旨在準確指出關鍵障礙與催化原因，並提供可靠且有效的干預措施。
最近的新型冠狀病毒（COVID-19）引起公眾關注， SAS亦利用Viya VTA平台整合及分析中國疫情的相關數據，以提供近乎實時的數據參考及使用。
In the upcoming decade, the adaptation of AI and advanced analytics will be further enhanced to support preventative, predictive and personalized health services.