Bala Hota, vice president and chief analytics officer of Rush Medical College, was in disbelief. The Chicago-based institution received only three stars in the Medicare’s Hospital Compare rating system back in 2015. Together with his colleague, chief medical officer Omar Leteef, they were determined to find out the reasons behind the low-quality scores. 

Both Hota and Leteef were practicing physicians at Rush. They believe the quality care and excellent services the medical centre are providing for the patients, were not duly reflected in the ratings. After a series of systematic revisions, Hota realized some clinicians might not have fully accounted for the severity of patients. This has caused Rush to be constantly at the lowest rank when it comes to documenting patients’ conditions. 

The loophole became prominent when Hota noticed transferred patients were marked less acute than when they were at the previous hospital. The bizarre urged Hota to venture into technology for help. Hota and his team tailor-made algorithms for Rush to better report and document patients’ severity. Eventually, the institution returned to its top rating. 

Benefits of Cloud platform

Hota believes data analytics plays an important role in healthcare. Right now, Rush’s dashboard has more than 200 key performance indicators (KPIs). It gives management staff, real time data in the areas of patients’ length of stay, safety, care variation, mortality and cost etc. The dashboard not only raised any abnormalities, but also became the milestone for Rush to adopt a predictive model. 

However, Hota feels that this will not happen if not for Cloud. Speaking at the recent Healthcare Information and Management Systems Society (HIMSS) 2019 conference, Hota admitted Rush was struggling with the growing amount of data prior their migration to Cloud. The institution receives over 300,000 patients annually as electronic medical records (EMRs) boost to about 10 terabyte. “Our leaders believe prior adopting machine learning and AI (artificial intelligence), they have to first manage the exponential growth in data” said Hota. Indeed, infrastructures supporting AI development have to be agile and elastic. 

In general, Cloud platforms enable healthcare providers to get hold onto patients’ information gathered at different care points. This can translate into a smaller team required to set up and safeguard the network. By having health records stored in the Cloud also means scalability is no longer an issue, as it enables grow in scale and capacity whenever the institution is ready. 

Details to note 

Specifically, Rush adopted a mixed of IaaS (infrastructure as a service) and PaaS (Product as a service). The former acts as a “data centre”: it provides infrastructures such as servers, network and storage to Rush. Under IaaS, Rush has a complete control over the dashboards. Its flexibility also signifies that Rush does not have to commit to a certain form of hardware or software as it scales up. The latter provides a platform for software creation. This means Rush is free to create their in-house application without having to worry about their infrastructure. 

Hota recognized the challenge deriving from cost, as Cloud may not necessarily be cheaper than traditional data center. The more control an institution has over the infrastructure, the higher they have to pay. That was the reason Hota pushed for a hybrid Cloud, adopting a mixed of IaaS, PaaS, Hadoop distribution and cloud data lake. 

In 2017, the Cloud Standards Customer Council (CSCC) has updated its guidance on cloud computing in healthcare. Under the Health Insurance Portability and Accountability Act (HIPAA), both healthcare and Cloud service providers have to come into agreement with where protected health information and sensitive data are stored, handled and who has the access to them. 

Besides, CSCC requires healthcare organizations to constantly revise their KPIs to be certain of their reliability. Healthcare providers who are interested in Cloud are also encouraged to assess the standards of Cloud service providers in terms of data integration, portability and interoperability. 

Author Bio
synthetic gene empathy chinese artificial intelligence data medicine healthcare ai

Hazel Tang

A science writer with data background and an interest in current affair, culture and arts; a no-med from an (almost) all-med family. Follow on Twitter.