The Power of Healthcare Data Analytics

The Power of Healthcare Data Analytics

The Power of Healthcare Data Analytics

The Power of Healthcare Data Analytics

By Dr. Abbie Leibowitz, MD, F.A.A.P., Chief Medical Officer, President Emeritus, Co-Founder

In today’s ever-changing consumer-driven healthcare landscape, engaging employees in their health is critical to controlling costs and improving clinical outcomes. According to the Kaiser Family Foundation’s 2017 Employer Health Benefits Survey, more than half of workers are covered by an insurance plan with a general annual deductible of at least $1,000 for individual coverage. It is certain that the number of people covered by high deductible plans will continue to increase. As these employees take on more financial responsibility for their healthcare, it is more important than ever that they are provided with resources to help them make better-informed healthcare decisions.

Employers struggle to develop effective strategies for getting employees to focus on managing their health and well-being. Healthcare data can be used to increase employee engagement in the medical management process. Data analytics applied to administrative information (medical and pharmacy claims, self-reported health information, laboratory and biometric screening results) can be used to identify gaps in recommended care, stratify risk and initiate outreach.

Organizations that adopt a data-driven approach in which interactions are personalized to each employee and their family members are more likely to be successful in motivating employees to proactively attend to medical needs and lifestyle issues.

Utilizing and Understanding Data Analytics

Healthcare data is complex. Identifying members across multiple data sets to create a picture of an individual’s health journey requires a level of analytic sophistication and computing power that has not been widely available at an affordable price to most employers. This is truly “Big Data!” The goal is to make the information actionable to both employers and employees. For every state and stage of health, there are applicable recommendations for care. These evidence-based practices apply to the management of chronic conditions as well as to preventive care. The challenge is to present them to individuals at a time and in a manner that makes the message more likely to get a response.

In a system in which most workers get health insurance through their employer, the employer is the logical conduit through which this kind of information flows. Regardless of their size or insurance structure, avoidable medical costs impact employers. While the threshold varies, most employers with 500 or more employees are self-insured. They function like small, or not so small, health insurance plans, and depending on their tolerance for financial risk and the limits of their stop-loss insurance, these companies directly pay for the medical care their employees receive. Even when employers are fully-insured, the claims experience of their group will be reflected in the insurance premiums they pay. This is to say nothing of the huge impact health issues have on the productivity and cohesiveness of the workforce.

Employers have historically attempted to address these issues by stringing together multiple interventional programs designed to target individuals with particular needs. However, without sophisticated data analytics and a comprehensive engagement strategy, these efforts are needlessly expensive and routinely ineffective. The programs tend to duplicate each other’s features, and the layering of multiple vendors and touch-points adds to the complexity of the health benefits program. The result ironically, although not surprisingly, is to discourage the very participation they were intended to promote.

Applying Data at the Point of Contact

It is unfortunate that it is so difficult to get people focused on their health, but it is true nevertheless. Using a Big Data approach, we can aggregate data from multiple sources, effectively creating an administrative version of a personal health record. We can identify an individual’s health status, and using this information, determine steps that could be taken to reduce future medical risks. We then can determine if gaps in recommended care exist on this road to better health, and if they do, initiate steps to get the employee to attend to these needs.

Using the individual’s preferred channel for communication is most effective, but even in the absence of this knowledge, communicating using multiple channels increases the likelihood of success. Our motto is that, “Everything works a little bit.” While the general bias is that people don’t respond to mailed reminders, in fact some people do. Just as “some” people respond to email, or text messages, or online notifications, or automated phone calls. It’s an advertising approach to push the message out in as many ways as possible. It is hard to predict which approach will work with which consumer for which message at which time, so the idea is to try them all, or as many of them as reasonable!

Better still is the ability to “tag” an unexpected message to a request for assistance. Psychology explains that people are more receptive to an additional suggestion when they have made an active effort to get information or assistance for a problem they face. In consumer marketing this is referred to as an “up-sell,” and we can take the same approach in healthcare, reminding the person who contacts the service center that there is something extra they need to do to address a gap in care. The fact is, up-sells are far more effective than “cold calls” (to borrow another marketing term) when trying to motivate individuals to address a health need.

Data analytics enables us to create this interface. Using information we know about a person, we can develop a scripted environment and provide the additional message at the point of contact. Engagement is driven by offering the assistance the employee requests, like finding a doctor, getting an appointment, arranging a second (or third) opinion, or discussing the cost of care. Help with benefit questions, claims errors, or denials of coverage also are frequent needs employees seek help for. Once engaged, the agent, whether in a live or a virtual environment, can not only provide assistance for the need, but also offer an additional suggestion tied to the participant’s medical situation. The exchange might go something like this:

“Oh, certainly, I can explain what your deductible is for physical therapy. While I’m bringing up that information, I see that you have not had your annual diabetic eye exam. I can help you schedule that visit if you’d like.”

And with that, we’ve taken advantage of the employee’s need for assistance to deliver a message about an unrelated medical need. Even if she was not interested in further assistance with the appointment at this time, using data analytics we can track whether over the next months she scheduled the eye exam. A large percentage of participants do. In fact, across large populations using a multi-channeled approach, we can expect to close nearly half of the gaps in care in the population over the course of the year.

Developing a Data-Driven Strategy

The additional benefit of applying a data-driven strategy to managing workforce health is the speed with which information can be made available to allow the employer to evaluate results and implement new strategies and programs to address continuing challenges. This calls for relationships with vendors flexible enough to modify approaches on the fly. A good data analytics partner should be consultative in this review, providing not only insights from the trends reflected in the data, but also adapting predictive modeling techniques that show where things are likely to go in the future. Not all data analytics teams are geared towards this anticipatory guidance, but without it, the employer is likely to be continuously “chasing its tail.”

The employee population is dynamic and ever-changing. Building programs and designing outreach based on continuous data analytics and predictive modeling can have a significant impact on engagement, outcomes and medical costs. However, it’s an ongoing process best served when the data analysts and the employer work together towards these common goals.

Next story: Data Target At-Risk Employees to Improve Outcomes and the Bottom Line