Insight
Healthcare Analytics: Choosing the right data analytics tools, for the right role, and the right purpose

>>Healthcare Analytics: Choosing the right data analytics tools, for the right role, and the right purpose

When it comes to selecting the right data analytics tool, every organization is on their own evolutionary path. Some can meet their needs with spreadsheets and answering questions like “What happened?” Many organizations, however, want more savvy capabilities. They want to be able to gain foresight that will help them predict and plan for future changes in sales, delivery, product, and technology.

This article will explore how your analytics journey evolves in two dimensions — from descriptive-to-prescriptive and discovery-to-operationalization. Also, we’ll explore how your data analytics toolkit changes alongside your business purpose and what stakeholder roles your organization contains.

Why is this important?

Data is extensively consumed as it flows from, let’s say, established leadership through to an associate programmer who may have recently joined the organization. The advent of Big Data has caused the divisive line between each group of data consumers in the organization to become thin and overlap. These fuzzy borders can create confusion, as well as double-work and gaps in the analysis because stakeholders have intersecting goals and analytics capabilities. Yet, their means to accomplish the work – their specific set of data analytics tools — vastly differ, as well as the ease with which those tools can be applied.

Improving understanding of how roles and tools vary with purpose and analytics cycles will harmonize your teams’ efforts — resolving communication challenges and enabling your organization to spend more brainpower on the task at hand.

The Right Purpose: What types of insights do you need?

The top half of our guiding infographic (Figure 1) depicts the traditional three-level journey from descriptive to predictive to prescriptive analytics. This journey takes your organization from hindsight and fundamental insights to foresight and, ultimately, having analytics that helps you make excellent choices.

Within each level, there is a repeating cycle of discovery, codification, and operationalization. This cycle involves exploring the data to determine relevance and relationships, testing hypotheses, and establishing usable reports/algorithms/models for ad-hoc usage, then finally implementing for widespread usage via automation, alerts, applications, and so forth.

The journey is cumulative, with each level building on the previous; your business objectives dictating how far through the cycles and levels you need to progress. Table 1 lists examples of what a healthcare organization may want to know at each level.

Traditional Data Analytics Journey Chart
Analysis Journey Levels and Questions Answered
Table 1: Analysis Journey Levels and Questions Answered

The Right Role: Which role can best uncover the insights at each level and through each cycle?

There are many data stakeholders engaged with analytics within your organization. The bottom half of Figure 1 breaks down these stakeholders into three categories, depicting their shifting levels of engagement within the activity cycles in each level.

Business Analysts: 

  • Subject Matter Experts involved throughout all three cycles at each level, conducting analysis in the discovery phase, and leading or guiding teams during codification and operationalization
  • Developing business cases, identifying gaps, and problems; creating narratives with data; creating hypotheses and interpreting data analysis patterns, relationships; presenting analytical findings through data visualizations; implementing advanced strategies for gathering; reviewing and analyzing data requirements

Statisticians: 

  • Often equally involved with business analysts during discovery, depending on the situation. During codification and operationalize cycles, their role becomes more distinct from the business analyst and focuses on statistical “programming”
  • Spotting patterns and relationships among datasets; creating modeling techniques and methods; finding trends, insights and making recommendations; building analytical models with strong expertise in data management; designing algorithms to collect and analyze data

Programmer:

  • Mostly involved at the operationalize phase, creating production versions of the design created by business users/statisticians/modelers
  • Excelling in the latest tools and technologies; handling and extracting large amounts of data from Big Data Analytics tools like Teradata/Hadoop; proficient in writing code for the algorithms used to find data trends and for predictive analytics

You may notice that the term data scientist is missing from this list. The data scientist stakeholder could span all three roles and activity sets when one person can wear all three hats. Often in practice, the data scientist role spans the technology skillset of the statistician and the programmer.

The Right Tool: Which data analytics tool is best at each level and cycle?

Every organization has data, but how they use their data depends on the business objectives and analytical needs. Excel, or other transactional reporting tools, is often the starting point for organizations. It helps them see what happened.

As they “slice and dice” their data to find out why an outcome occurred, they progress to statistical packages and business intelligence (BI) tools and techniques. When organizations begin to address challenges that require predictive and prescriptive capabilities — what’s next and what action should we take — they move to more advanced functionality of statistical packages and begin using statistical programming tools, as well as machine learning tools.

Table 1 illustrates the broad user categories and the tools frequently used at each cycle. There are analytics tools for all jobs and all sizes of organizations. Some software tools may be used extensively by programmers while others are preferred by business analysts that require minimal coding.

Data Science Technology for Each Stakeholder
Table 2: Data Science Technology for Each Stakeholder

The promises of big payoffs are driving organizations to engage more deeply, and invest more heavily, in Big Data analytics tools.  With “can do” attitudes, teams jump in and quickly find themselves tripping over each other around objectives, accountabilities, and tools. This creates churn that can undermine speed and the entire endeavor as frustration wins.

This article gave you our insights and guidance on how to choose the right data analytics tool for the right role based on the right purpose. As you take a fresh look at your data analytics journey level and what your current business objectives may require, we hope that you refer to this article often and reach out when you need assistance with assessing your data analysis needs and toolset for delivering the right answers for better decisions.

Jeannine Siviy

Insight Contributor

Jeannine Siviy, Director of Healthcare Solutions

Insight Author

Deepan Mahadevan,Senior Consultant

Insight Contributor

Naveed Ismail, Senior Consultant