Using analytics to rationalize healthcare spending
Employers typically have two key goals for their employee health benefits program. The first, of course, is to recruit and retain top talent.
The second is to improve productivity by maintaining and improving the health of their employees. Workers who are unable to work because they are sick, or who are distracted because they are worried about the health of a loved one, are less productive, and employers know that providing competitive health benefits to their workforce will mitigate these issues.
Yet as the cost of health benefits continues to rise, more employers are beginning to ask their insurance brokers if the money they’re spending is delivering the outcomes they’re seeking, e.g., a reduction in absenteeism and “presenteeism” (where employees are ill but show up to work anyway).
Given that health insurance continues to be one of the highest-cost line items (after wages and salary) in total employee compensation, employers are strongly motivated to work with their benefit advisers to improve their outcomes.
The Kaiser Family Foundation’s 2016 Employer Health Benefits Survey found the average employer contributes $5,306 for individual employee health plan coverage and $12,865 for families. So it comes as no surprise that employers are keenly interested in finding ways to reduce the amount that they are spending on healthcare without sacrificing the health of their workforce.
Advisers, however, have traditionally been hard-pressed to deliver answers about outcomes, because the data that could provide those answers is so fragmented. Part of it comes from claims systems; other portions are found in electronic health records (EHRs), population health management (PHM), finance and other siloed systems. Some of it simply consists of details about plan constructs such as eligibility, co-pays, co-insurance and overall plan design.
Even when they can gain access to all these different data sources, the sheer volume and variety is so massive and disparate that it’s exceedingly difficult to interpret. But—and here’s the kicker—even if an adviser can surmount all these hurdles, the results are not only retrospective but so far in the past that they are no longer useful as a guide for the future.
The good news is that a new generation of predictive and prescriptive analytics tools are making it possible to aggregate all of this disparate data, along with relevant demographic, sociographic and psychographic data cast a light not only what is happening now, but also areas of risk that are likely to increase going forward. This gives brokers the insight they need to better work with employers and health plans on designing plan options that better match their clients’ needs.
By using analytics to evaluate and measure the effectiveness of the programs they come up with, advisers can demonstrate their ROI to their clients and recommend adjustments as-needed.
Here’s an example. Suppose 100 of an employer’s 2,500 employees have a chronic condition such as diabetes. Health plan members with chronic conditions typically cost significantly more than those without them. Medications to manage their conditions can be expensive, and they typically will use the most expensive care settings - emergency department (ED) visits and hospital in-patient services – on a more frequent basis.
Extracting anonymous data from the health plan and other sources, next-generation predictive analytics can create risk scores for each of those diabetic patients based on personas that account for factors such as where they live (Zip+4 data), their income and education levels, co-morbid conditions, ethnicity, lifestyle considerations and other factors. This information is then used to optimize health plans based on “impactability” (the likelihood of a positive outcome from closing care gaps) and “intervenability” (the likelihood the employee will become engaged in and follow a care plan).
When put into action, more effective, personalized care plans lead to far greater employee engagement and superior outcomes. When a care manager can view a member profile that includes income, education and other social determinants, they can design realistic plans that members will actually utilize. Once these programs are implemented, brokers can share progress reports and use prescriptive analytics to devise additional improvements.
Another way brokers can help employers reduce their healthcare costs is by using next-generation analytics to determine how often employees go out-of-network for certain specialist services, such as ophthalmologist exams. If the incidence is high, it may reveal a shortcoming in the current provider network. Armed with that information, brokers and employers can take corrective action, improving the quality of the network while reducing costs.
Analytics can also be used to improve price transparency, one of the biggest hot-button issues in healthcare. Currently, employees who opt for a high-deductible health plan can be shocked to discover that a procedure they thought was covered under their $25 co-pay costs many times more because they were referred to an out-of-network provider. Or they may be unaware that having a procedure such as a colonoscopy in a hospital setting, rather than a clinic or an office, can cost thousands instead of hundreds of dollars.
All too often, however, the decision about which plans to offer, or which providers to use, is based on gut feel—not factual data. By employing next-generation prescriptive analytics, brokers can steer employers to the right plans for their specific employee population, reducing their costs and helping them achieve the improvements in productivity and workforce health they desire.