Data Analytics

From Clinfowiki
Jump to: navigation, search


Data analytics is the scientific process of finding, understanding, and communicating meaningful patterns and information within structured and unstructured data that informs fact based decision making. It aids decision makers by reducing and presenting the essential information in a clear and understandable form (1). Analytics refers to qualitative and quantitative techniques and processes used to enhance productivity and gauge economic impact. It varies according to organizational requirements (2).

Types of Analytics (3), (6):

• Descriptive: “What has happened?”

-Analyzes data to look for patterns where no targe variable or class exists.

• Diagnostic: “Why did it happen?”

-Uses data mining and correlations.

• Predictive: “What will happen?”

-Creates a model based on a target variable, using multiple independent variables

-Modeling approaches include:

i) Regression model: A linear model where data is numerical. It consists of two types: Simple linear regression, and multiple linear regression.

ii) Classification model: A nonlinear model where the target variable is categorical.

• Prescriptive: “What to do in given situation.”

-Uses machine learning, signal processing, and natural language processing

Approaches to analyzing data (3):

• Statistical modeling

• Machine learning

• Programming languages


Increased data accessibility is improving healthcare analytics. Uses of analytics include (4), (7):

• Identify at risk patients

• Identify patients requiring care coordination

• Deliver the most effective intervention

• Track and improve clinical outcomes

• Performance measurement

• Optimize clinical workflow

• Program adjustments

• Lower health care costs and spending

• Clinical decision support

Challenges (5), (6):

• Capturing data that is consistent, correct, and formatted appropriately for use.

• Data cleaning to ensure that datasets are accurate, consistent, relevant, and not corrupted in any way.

• Proper storage of growing data which is inexpensive to scale, easy to maintain, and not prone to producing data siloes across different departments.

• Security which protects data from high profile breaches, hackings, and ransomware episodes.

• Maintaining and understanding the data so that it may be reused or reexamined for alternative, future purposes.

• The ability to easily query data for reporting and analytics, so that organizations may engage in meaningful analysis of their big data assets.

• Generation of reports which are understandable, reliable and accessible to the target audience.

• Utilization of good practices and presentation techniques for data visualization.

• Maintenance and regular updates of data so that it remains current and relevant.

• Appropriate governance models which oversee and provide data analytic services for the organization.

• Require that the organization follow a consistent methodology for problem solving.

• Data interoperability


1. Shortliffe, E. H., & Cimino, J. J. (2006). Biomedical informatics: Computer Applications in Health Care and Biomedicine(3rd ed.). London: Springer Science Business Media, LLC.

2. Yang, H., & Lee, E. K. (2016). Healthcare analytics: From Data to Knowledge to Healthcare Improvement. Hoboken, NJ: John Wiley & Sons.

3. Hoyt, R. E. (2016). Health informatics: Practical guide for healthcare and information technology professionals(6th ed.). Informatics Education.

4. Berg, G. (2015, July 17). 3 ways big data is improving healthcare analytics. Retrieved April 28, 2018, from

5. New challenges, trends in data analytics: 5 survey findings. (2016, July 13). Retrieved April 29, 2018, from

6. Bresnick, J. (2018, April 29). Top 10 Challenges of Big Data Analytics in Healthcare. Retrieved June 12, 2017, from Top 10 Challenges of Big Data Analytics in Healthcare

7. “Health Policy Brief: Reducing Waste in Health Care,” Health Affairs, December 13, 2012.

Submitted by Amber Sidhu