Data-Driven Healthcare - Fusion PPT

Data-Driven Healthcare

In 2011, HHS launched Healthdata.gov as part of President Barack Obama’s open government initiative. Six years later, the site now has more than 3130 datasets and hosts several apps. HHS has, over the years, made more and more data from CMS, CDC, FDA and NIH “easily available and accessible to the public and to innovators across the country.” This information includes everything from clinical care provider quality information and nationwide health service provider directories to community health performance information.

In addition to large data repositories like Healthdata.gov, a number of new sensors for individuals exist to collect health data, not on generalized groups, but on themselves.  Health analytics can provide insights into what has happened (reporting), what will happen (predictive analytics) and what should be done right now (prescriptive analytics).

Large Volumes of Data Available

Total patient data available to healthcare organizations for such analytics could reach approximately 25,000 petabytes by 2020. Conventional healthcare IT already generates considerable amounts of structured data in the form of medical and financial records. Unstructured data sources including medical images of all sorts, audio and video records, email conversations and exchanges over social media add considerably more.

Areas for the application of health data analytics include:

  • Clinical analytics using data from electronic health records (EHRs) for example, and pharmaceutical and research and development (R&D) data analytics.
  • Financial analysis, using claims and cost data, also extending to fraud detection.
  • Supply chain operations to identify trends and patterns in retail and over the counter (OTC) purchases and hospital and practitioner needs for supplies, distribution, and stocks.
  • Human resources, caregiver training, skillset coverage and patient behavior, sentiment and preferences.

Yet, despite the huge amount of patient and patient-related data available, relatively little of it is being used to generate actionable insights for better healthcare. A 2015 survey of 270 healthcare professionals showed only 10% of respondents were using advanced tools to collect data and run predictive analytics. Fewer than 20% used data warehouses for key performance indicator (KPI) tracking, while just 16% used data to make strategic decisions.

Changes Encouraging the Use of Health Analytics

Three factors could change these statistics, making data-driven healthcare a bigger part of all healthcare activity:

  1. A general trend toward evidence-based medicine, with the systematic review of clinical data as part of treatment decisions. This trend contributes to improving caregiver judgment.
  2. The U.S. federal government and other public entities have now made large volumes of healthcare information and knowledge available, from records of clinical trials and public insurance programs. Only suitable health analytics software tools can unlock the value and insights from such massive amounts of data.
  3. New generations of data analytics technology are making it easier for non-IT staff to manage their data, visualize it to spot immediate trends and ask ad hoc questions as health markets and patient needs change.

Practical and Advisory Roles for IT Teams

IT teams can advise the rest of their organization on the selection of health analytics tools, whether these tools remain in the IT department or are used by others.

  • IT teams often have a head start in understanding the requirements for acquiring and preparing data properly, to give valid results from business intelligence and data analytics applications. Data cleansing and harmonizing, through the removal of duplicate and erroneous information and conversion to a standard format, are essential operations that non-IT clinical, financial and healthcare supply chain staff may not know how to perform.
  • By working with end-users to understand their requirements, IT teams can advise on suitable software applications for “self-service,” allowing end-users to manage data and access health analytics by themselves, faster and more flexibly than if they were systematically obliged to go through the IT department.
  • IT teams can also put robust IT security procedures in place to protect data, prevent breaches, and raise end-user awareness of the importance of good IT security practices.

A coordinated approach to health analytics by IT teams will also reduce the risk of isolated short-term solutions, and allow a foundational approach to health analytics that will carry healthcare organizations and caregivers into tomorrow’s healthcare environments, as well as today’s.