Big Data Strives for Individualized Coverage

Anne-Marie Meyer, PhD

Healthcare spending is one of the largest expenses for a company and, more and more, employers are using population-based Big Data to ensure their employees have coverage tailored to their needs at a reasonable cost. Employers are hiring experts in healthcare analytics to help obtain insurance policies and benefits that are aligned with historical claims data from their employees. These analytics companies are attempting to improve cancer coverage options via this method, but the complexity of this disease and its variable costs have so far proved to be a barrier.

“On the employee side, health plans are getting more complex,” said Aaron Huang, vice president of marketing at Lumity of San Mateo, CA, which does data analysis to develop a selection of ideal plan choices for employees. “There’s a whole variety of plans that employees have a hard time choosing from, and based on the fluctuating needs of an employer, the health plan recommendation could change from year to year.” There’s huge waste in cancer treatment, Huang says, and he believes Big Data will eventually help to streamline care and provide greater savings.

Other health data companies in this field include Benefits Science Technologies, Oscar, and Zenefits. Lumity serves as a broker for payers and does not charge employers. The company analyzes data from employees’ prescription claims to detect patterns in usage and estimate an employer’s optimal cost of healthcare coverage based on different coverage options. Much depends on the amount and quality of medical data available. For companies with 250 or more workers, insurance companies will provide access to employees’ medical claims data, which leads to a more precise picture of costs, Huang said. “With that level of information, you can actually do some pretty accurate forecasting of one-time and chronic conditions.” For smaller companies, such information is harder to come by.

Currently, Lumity’s ability to improve employer-specific coverage for cancer and other high-cost conditions is limited. Huang hopes that in the future Lumity and companies like it will be able to use data to identify patients with cancer and pair them with advocates who can help with navigation. Huang believes that unnecessary spending for chronic conditions such as cancer amounts to 20% to 30% of the medical tab, which can easily top $1 million for a patient. “That’s a lot of overbilling that could be preempted with the right type of advocates,” he said.

However, the complexity and high cost of cancer treatment makes it very difficult—even when armed with Big Data—to bring costs down, Huang said. “Cancer coverage is very much a case-by-case basis with carriers. With cancer, there are a lot of clinical trials that carriers will decide to cover on a one-off basis. Many of these things are more up to the insurance companies than they are up to the design of the bene ts programs themselves.” Huang said it would be possible to design a plan for a specific type of cancer, but it would raise the costs for everyone in the company. “There’s not an easy answer to providing better cancer coverage today. A lot of it comes down to how they can improve the clinical trial process and fast-track some of these experimental new treatments for people who have rare or hard-to-treat types of cancer.”

Direct costs of cancer screening and treatment, which often include advanced diagnostic testing, planned and unplanned hospital admissions, and multiple therapies that include specialty pharmaceuticals, can be difficult to predict because they are complex and vary considerably, agreed Anne-Marie Meyer, PhD, research assistant professor in cancer epidemiology at UNC-Chapel Hill, NC. In order to find better coverage options, companies such as Lumity would have to take into account the great variability among patients with cancer, she said. “Patients, especially very sick patients, are going through really intensive and diverse treatments, from surgery to chemotherapy to radia- tion. It’s a lot more complicated than other Big Data.”

Most employers are not experts in clinical oncology and struggle to keep up with the complexity and heterogeneity of cancer, along with the rapid innovations in treatment. In a 2015 survey by The Northeast Business Group on Health Solutions Center, employers said they perceived cancer care to be more complex than other conditions due to rapidly increasing costs of care, variable definitions of quality, the extensive services needed to support employees and caregivers, and the multiple roles of the benefits professionals who assist employees. “It’s important for employer groups to understand the delicate balance between developing an affordable health plan and ensuring that, when you get to some of those high-cost diseases, patients who need the insurance are able to use it,” said Heather Morel, chief operating officer for The US Oncology Network.

The data Lumity obtains is HIPAA-compliant and provided voluntarily by employees through business partners connected with pharmacy bene ts managers. In exchange, Lumity provides employees with forecasts of healthcare expenses to help them pick the most suitable insurance plans. For the employer, Lumity’s bene ts advisors help choose the best plan options and rates based on employ- ee health risk profiles.

Observers believe the potential in Big Data for tailoring healthcare coverage can develop in numerous ways. Health care analytics companies could consider using large-scale data to identify patterns in the availability and use of preventive services (such as mammography and colonoscopy), said Morel. This would enable the development of employer health plans that encourage screening for cancers, particularly those common in specific employee populations. Morel explained that such an initiative could lower costs for a company if more cancers are detected in the early stages when treatment is less expensive. “Providing some of these services with very low or no co-pay does incentivize people to not worry about the cost and worry more about their health,” said Morel.

Debra Patt, MD, vice president of Texas Oncology in Austin, TX, said treatment outcomes available in population-based data also can be useful for payers and employers to ensure they are making the best decisions for their employees, such as which providers to include in-network. She expects that more decisions will account for patient preferences in the future, although she believes that patient preferences may also include factors unrelated to good outcomes, such as patient experience preferences.

There are other roles for big data in bringing down the costs of care, said Jean-François Beaulé, executive vice president of design and innovation at UnitedHealth Group in Hartford, CT. He said companies that consult on healthcare analytics can identify specific reasons—a large number of unnecessary emergency department visits, for example—why an employer may be overspending. “You need to look beyond the average and determine who is performing below and what are some of the challenges that they may be facing.”

Patt and Meyer said that patient concerns about confidentiality can lead to refusals to share data. According to Meyer, educating employees on how their health data are used can help them make informed decisions about whether to share personal information. “If patients understood the value of that data to help guide their treatment, they would hopefully be in favor of that.” Huang estimated that 80% to 90% of employees do give consent for Lumity to use their data; in the case of non-consenting employees, Lumity uses a benchmark based on a database of 80 to 90 million self-insured individuals to estimate healthcare needs according to age, gender, and geographic location.

Even with patient consent, the need to comply with HIPAA and other health laws means that data are often removed for patients with rare diseases in sparsely populated geographic areas, which can compromise the usefulness of the data for guiding rare-disease coverage decisions, Morel said.

Lumity and its competitors hope to eventually use employer data to provide employees with better local services for their needs, as well as improve predictive models to reduce overbilling and help employers and employees make better decisions about healthcare. "There’s not a lot of visibility” into the link between pricing and what patients actually receive for the money, Huang said.

Patt said that Big Data will become an even more important tool in controlling healthcare costs. She has observed the drastic increase in the cost of employer-sponsored health insurance, caused in part by the aging of the workforce, the higher incidence of cancer, the increasingly longer survival of patients with cancer, and patients frequently requiring chronic treatment. “The transformation of cancer from an acute illness and rapid death to a chronic but manageable disease is amazing but sometimes coincides with expenses that may be as much as or more than $10K per month. The increase in healthcare costs has become a high risk for insurance companies to bear, and they’ve heightened their data assets in terms of human resources and analysts to help them better understand how to plan for the population they serve,” she said.

Northeast Business Group on Health. Cancer and the workplace: the employer perspective. http://nebgh.org/wp-content/uploads/2015/10/CancerWorkplace_ FINAL.pdf. Published October 2015. Accessed January 14, 2017.