What is Prescriptive Analytics
Prescriptive analytics is a business intelligence field that helps find the best course of action to take in a situation taking into account the objectives, requirements and constraints. It is the phase of Analytics where data is analyzed for optimization and simulation. Descriptive Analytics, the first phase, looks at data to quantify the different relationships to help organizations categorize and plan for the future. Predictive analytics is a branch of data mining that focuses on processes that provide clear and actionable initiatives to predict future probabilities, trends, patterns, deviations, anomalies, risks and opportunities. Prescriptive analytics takes into account the answers derived from Predictive Analytics (What, why and when it will happen) and helps answer questions such as – How can we change our course of action using the real time data?
Prescriptive Analytics synergistically combines data, business rules and mathematical models. Data can be structured or unstructured and can come from multiple sources. In healthcare, prescriptive analytics can be used for strategic planning by leveraging operational and usage data combined with data from external sources such as economic data, population demographic trends and population health trends. This knowledge is essential for making decision such as building new facilities and equipment utilization and to better understand the trade-offs between adding additional beds and expanding an existing facility versus building a new one (1).
Prescriptive analytics in healthcare
Organizations that are already using predictive analytics to meet the latest challenges in healthcare will benefit from this more sophisticated approach to keep themselves ahead in the ever-changing competitive environment. Healthcare Organizations are increasingly using analytics to drive clinical and operational improvements to meet business challenges. Healthcare Organizations that are interested to form Accountable Care Organizations and provide proactive or preventive healthcare instead of reactive care need predictive and prescriptive analytics embedded in their day-to-day operations to reduce costs and improve care. These organizations have to be able to merge large amounts of clinical care, process and claims data by using data warehouses and business intelligence software packages (2).
At present, most healthcare organizations have the essential foundation for analytics – data warehousing and are using descriptive analytics quite extensively to report and to create financial and operational dashboards by understanding the historical data. Some of these healthcare organizations are also using predictive analytic techniques to support enterprise analytics, evidence-based medicine and clinical outcome predictions. Very few healthcare organizations are currently taking advantage of predictive analytics to enable more personalized healthcare, dynamic fraud detection and to predict patient behavior and assist in behavior modification leading to healthier lifestyle choices (3). Organizations that embrace these approaches and build on the current use of prescriptive analytics at the point of care, with self-service analytics at the care provider level will be best positioned for the coming trend of sequencing data analysis and will facilitate discovery and dissemination of diagnostic and therapeutic breakthroughs (4).
Barriers to adoption of analytics by organizations
Healthcare enterprises need to have an open, flexible architecture and display platforms in order to make the data relevant and accessible for clinicians, researchers and regulators and a learning culture has to be developed around the use of clinical data (4). Along with the Information Technology infrastructure essential for gaining the benefits of analytics, organizations also have to take into account the organizational behavior issues as demonstrated in the report by a 2010 report by MIT Sloan Management Review and the IBM institute for Business Value. The authors identified the following barriers that organizations face and most are related to management and culture rather than being related to data and technology -
Lack of understanding of how to use analytics to improve the business
Lack of management bandwidth due to competing priorities
Lack of skills internally in the line of business
Ability to get the data
Existing culture does not encourage sharing information
Ownership of data is unclear or governance is ineffective
Lack of executive sponsorship
Concerns with the data quality
Perceived costs outweigh projected benefits
No case for change
Don’t know where to start (5).
References: 1. Walker M. Data Science Central [Internet]. 2012 Sep; [cited 2012 Nov 20]; Available from: http://www.datasciencecentral.com/profiles/blogs/predictive-descriptive-prescriptive-analytics
2. Loonsk, J. The HIT of ACOs, Part 1: Analytic Data; [Internet]. 2011 Aug; [cited 2012 Nov 20]; Available from: http://www.govhealthit.com/news/hit-acos-part-i-analytic-data-julyaugust-2011
3. Cortada, J, Gordon D, Lenihan B. The value of analytics in healthcare: From insights to outcomes[Internet]. [cited 2012 Nov 20]; Available from: www.ibm.com/.../se__sv_se__healthcare__the_value_of_analytics_in_healthcare.pdf
4. Taeger, J. OTB Solutions [Internet]. The route to personalized medicine solutions begins with BI strategy; 2012 Aug; [cited 2012 Nov 20]; Available from: http://www.otbsolutions.com/perspectives/Pages/Article.aspx?Article=38
5. MIT Sloan Management Review and the IBM Institute for Business Value. Analytics: The new path to value[Internet]. 2010 Fall [cited 2012 Nov 20]; Available from: cci.uncc.edu/.../MIT-SMR-IBM-Analytics-The-New-Path-to-Value-Fall-2010.pdf
Submitted by Sandhya Yanamadala