How Your Healthcare Company Can Use Big Data

Big Data and You

You won’t escape big data in 2015. Chances are, this is not the first article you've read about big data and if it’s not already at the top of your company’s 2015 list of strategic initiatives, it will appear soon enough.

But, what does it actually mean? How can you use it? Where do you start?

What does it actually mean?

“Big Data” describes initiatives, and the support systems of those initiatives, that utilize data to formulate quantitative cause and effect relationships regarding business processes.

Many of these cause and effect relationships could be intuitive when applied to small scale processes, but big data seeks to confirm larger strategic decisions with numerical guarantees and to facilitate scalability as your company grows.

Big data is not a company requirement but it can provide opportunities to leverage institutional knowledge for future gain – giving you a competitive advantage. Let’s look at a simple theoretical example in the Healthcare industry.

Big Data Healthcare Example:

In 2014, a hospital tracked every patient discharged. It discovered 25% of patients were readmitted within 30 days of discharge. Striving to reduce this number and to improve patient quality in 2015, the hospital analyzed the discharged and readmitted patient populations for patterns.

The hospital found 80% of readmitted patients never filled their prescribed post-discharge medication. In the first quarter of 2015, the hospital designated an employee to follow-up with every discharged patient and to reiterate the importance of the prescribed medication for a healthy/expedient recovery. By the end of Q1, 2015, hospital remittance rates dropped to 7%.

Breaking Down Big Data Components

This example contains three characteristics of big data. Three big data aspects, among others, that companies should consider for successful application:

• Company data: The hospital tracked, stored, and reported data pertaining to its discharged/readmitted patient populations.
• Statistical analysis: The hospital’s reported data was in a form that allowed for data analysis.
• Insight application: After data analysis hypothesized a causation between grouped patients and the observed outcome, the hospital created a plan to remove the identified variable contributing to increased hospital re-admissions.

Big Data Stumbling Blocks

Easy, right? Sort of.

My example assumes a lot about the operational capabilities of our hypothetical hospital: that the patient data collected throughout 2014 was complete, accurate, and in a format conducive to data analysis, that the data analysis found a clear-cut relationship between patient populations and the observed outcomes, and that, once a quantitative relationship was found, the hospital had resources necessary to affect the outcome.

Companies that could derive value from big data implementation but do not, frequently fail to overcome these roadblocks.

So, how can I use it?

Big data has applications across your business in the same way innovation applies to both big and small business processes. The “how” then hinges on the evaluation of cost, impact, and ability.

Small companies can derive great value by exporting data into Microsoft Excel Pivot tables. Whereas, large companies can justify robust BI platforms like Tableau. Only you can judge whether the levels of investment would produce results worth the cost.

Related Post: 21 Medical Practice Questions Big Data Answers

Where do I start?

There are three big data building blocks that can put your company on the right path toward successful initiatives today: strategic metrics, data management, receptive organizational behavior.

1. Strategic Metrics – Set goals

What gets you paid? Identify your business’ core competency and set goals. Once this goal is set, announce that goal, hold people accountable for the achievement of that goal, and start measuring it.

When creating company-wide strategic goals/metrics focus only on bottom line influencing business processes. Setting goals, collecting data, and holding people accountable is time consuming, expensive, and potentially exhausting. If the business process doesn't directly contribute to your company’s short and long-term positive cash flow, there is little argument it requires measurement.

2. Data management – Collect relevant data

Data management pertains to the tracking, storage, and reporting of raw data necessary to measure strategic metric achievement. If your current operational software, data storage space, or operational process does not allow for good data management, you will have a difficult time realizing big data success.

Of the three big data start-up requirements, due to increasing data growth, data management will contribute most to cost. [SOURCE] However, digital infrastructure investment should not kill big data projects; it merely requires additional consideration during your ROI “how” evaluation.

3. Organizational behavior – Shift paradigms

Organizational behavior remains, even for big data, the difference between positive and negative outcomes.

Defined strategic metrics, accurate data, and flawless analysis unravels in the hands of an unwilling company culture. Initiatives that deviate from existing business process or where employees do not agree on the value of improvements have a low probability of success. Conversely, willing company culture greatly enhances success and profitability. [SOURCE]

Therefore, you must ensure your organization is ready for big data. Successful organizations embrace innovation, prioritize creativity, and accept failures as necessary learning opportunities. Furthermore, strong big data initiatives engage employees throughout the entire process, rely on employee collaboration regarding strategic metrics, and strengthen employee ownership by communicating initiative outcomes.

Additional thoughts and resources

For more practical tools and big data recruitment strategy I recommend:

• Reassessing your Big Data Analytics Strategy by Emmanuel Amamoo-Otchere
• Chasing the Data Science Unicorn by TDWI

The author, Michael Berger, is a University of Miami MBA and Lean Six Sigma Black Belt specializing in process optimization within the healthcare industry. Michael has experience creating and evaluating predictive algorithms as they apply to patient stratification.