Skip Navigation
White Papers

The Power of Prescriptive Analytics in Healthcare

Imagine two health plan sponsors that each employ ~10,000 mostly blue-collar workers. These plans decide to launch care management programs designed to reduce unnecessary emergency department (ED) visits and costs. The initiatives make sense, based on reams of data suggesting that a number of employees at both companies are receiving care in the ED that could be treated more appropriately — and less expensively — in an office or retail clinic setting, or possibly through a telemedicine visit.

Both programs include care managers and access to a triage service designed to divert patients from the ED to a primary care clinician. For those who end up in the ED despite these efforts, the member’s primary care clinician is notified of the patient’s status in the ED, increasing the likelihood of better care coordination. To reduce the chance of an ED revisit, the patient is sent home with educational materials about appropriate settings for health care.

One year after the programs were launched, ED visits and spending are down for both firms, but one realizes significantly larger savings. It turns out that when setting up the program, this employer had the foresight to act on an insight the other did not — the fact that a small number of its workers were habitual visitors to the ED. Armed with that knowledge, this employer rolled out a “right-sized” program that hit the bullseye. Meanwhile, the other company must scramble to play catch-up. In hindsight, the second company now sees that it could have succeeded with fewer care managers and a more streamlined telemedicine program focused mainly on habitual ED users.

Prescriptive analytics is the difference between having foresight and relying on hindsight. It is the difference between getting it right the first time, rather than doing so months or years — and potentially millions of dollars — later.

Prescriptive Analytics Tell You What to Do

The answer for employers like those profiled above is to invest in a platform that taps into the power of predictive and prescriptive analytics.

More often than not, reports that employers and payers rely on to make decisions are based on descriptive analytics — a summary of historical data that explains what happened. It’s great information, but provides only half the equation. Sometimes reports include predictive analytics, where information is extracted from existing data to explain what might happen in the future. This is the other half of the equation, which outlines multiple potential futures based on many possible actions. While predictive analytics is an improvement over descriptive analytics, using it alone can lead to confusion, given the many potential scenarios it produces.

What’s missing is a clear path to the best course of action. That is where prescriptive analytics comes in. It utilizes specialized software that pores over the many potential solutions and helps select the best one. The bottom line: Prescriptive analytics tell you what to do.

Payer Account Reporting “Must-Haves”

Payer account reporting is only as good as the data that comprises it. Insurers and employers whose prescriptive analyses are hitting the mark are likely using an analytics platform that offers the following:

  • Airtight security and data-integrity processes. Health care ranks near the bottom of major US industries in this regard. Legacy systems are particularly vulnerable in an era when attacks are more frequent and sophisticated. Best practices include:
    • Annual SOC II audits
    • Internal pen testing and static code scanning
    • Third-party pen testing and vulnerability scanning
    • Internal security auditing
    • Third-party daily review
    • Compliance with the National Institute of Standards and Technology’s security and     privacy controls
    • Role security
    • Secure passwords
  • Ability to link claims to biometric and other non-claims data. Allow plans and sponsors to go beyond diagnostic codes and adjudicated medical claims to identify illnesses earlier, in some cases before the recording of a diagnosis code on a claim. As an example, elevated BMI, cholesterol and blood pressure readings coupled with low current use of health services can unmask metabolic syndrome and potential high utilization in the future.
  • An online portal with key performance indicators and insights. 24/7 access to an easy-to-read dashboard of KPIs, along with an alert system that sends messages when a measure prompts a predetermined alert.
  • Proven methods to analyze trends. After spotting a troubling trend, use trusted tools from respected third parties to quickly get to the bottom of what’s causing it. For example, the Total Cost of Care model from HealthPartners of Minnesota can help determine whether cost, volume or intensity of service is the key driver. An analytics platform that offers a suite of such tools saves time and avoids aggravation.
  • Detailed insights into specialty pharmacy. This is the fastest growing component of medical spending, accounting for 36% of total drug spending in 2015. The keys are to:
    • have a common definition of specialty pharmacy and KPIs
    • design and refine cost containment strategies such as preauthorization programs, closed formularies and limited networks
  • Relevant benchmarks. Plan sponsors want to know how they compare to others. It’s important to have benchmarks that are up to date; large enough to be statistically credible and valid; and able to be broken out by region, plan design and industry type. They also need to be adjusted for age, gender and illness burden.
  • Network analysis. This is particularly important for employers who have moved to narrow or tiered networks. They will want to know why employees are going outside the network — specifically, what services are they seeking, from which providers and at what price? They’ll want to know whether to include a favored provider and at what price point. More importantly, they will want to know if providers in the network are the most efficient and/or high-quality in the area — or are there better choices?
  • Indicators of value-based care. As plans move from fee-for-service to prospective payment or value-based contracting, employers will want to know if their employees are actually receiving more high-value care and less low-value care. Just as important, are members receiving enough care? This is another instance where respected third-party tools come into play, making it easier for plan sponsors to monitor care and communicate findings with local providers.

Liberating the Power of Data with Storytelling

Finally, payer account reporting requires more than just a good analytics platform. Analytics can be little more than statistical noise unless they tell a story. Bridging the gap between data and a relevant story requires specialized skills from subject matter experts. They are the people who understand a plan sponsor’s goals and objectives, and they can tell the story to liberate the power of the data.


Prescriptive Analytics Whitepaper