Effective Prevention and Management of Diabetes Requires Health Plans to Rethink Clinical Measures

By Bea Capistrant, Katherine Bobroske and Caitlyn Hall , Health Care Innovation - Data Science & Research at Morgan Health

Today, more than 37 million Americans have diabetes and roughly 96 million more are considered pre-diabetic. Among those with employer-sponsored insurance (ESI), 1 in 10 people have diabetes, of whom 1 in 5 are undiagnosed. The speed with which diabetes has evolved into a significant and widespread public health concern underscores the importance of early identification and management of the disease, particularly for employers and health plans navigating the country’s largest insurance market.

For health plans to monitor population health for plan members enrolled in ESI coverage, they first need reliable tools and clinical definitions to identify individuals most at risk for common chronic conditions.

Gold-standard definitions to identify #diabetes and other chronic conditions typically have come from the Medicare program – even for commercial plans within ESI. But there are challenges in applying such standards to a non-Medicare population, and in the case of diabetes, they may lead health plans to unintentionally overlook key populations most at risk for prediabetes and diabetes.

Most notably, Medicare definitions rely predominantly on #healthcare claims data, whereas employers and their health plans often capture novel health risk assessment (HRA) data that Medicare and other plans do not have. Most commonly, an HRA includes a biometric screening as well as survey component that asks about employee’s demographics and self-reported health history/status. To gain a more complete and accurate picture of plan member/patient health, especially for identifying populations at risk for diabetes, we encourage health plans to combine HRA biometric data with claims data.

Employers’ unique data insights can support a more robust, complete picture of patient health

Many employers, particularly large and jumbo employers, offer employee plan members and spouses online wellness assessments and annual wellness exams that produce timely and high-quality HRA data – both reported by patients and measured during visits. Plan members often are incentivized to participate, which drives increased participation and results in more complete data about population health. In many cases, plans may use this data to identify health and wellness programs and services for participants.

Because so many plan members participate, HRA biometric data could offset the gaps inherent in claims-based definitions of disease prevalence. HRA biometric data can identify plan members not yet diagnosed with diabetes (for whom claims have not yet been produced), and assess diabetes management among those who have already been diagnosed, between provider encounters.

Combining claims and HRA biometric data allowed us to identify more members with diabetes than claims alone

Morgan Health recently conducted an internal analysis using this method to determine how comprehensively claims data identified diabetes in the J.P. Morgan Chase population. For the purpose of conducting this analysis, we used a limited dataset that did not allow us to identify any individuals from the health plan. Combining annually collected HbA1C data from HRA biometric screenings with claims data, we identified 25% more adult health plan members with diabetes than those identified by claims data alone.

There were some notable insights from this combined approach. For example, we saw a distinct pattern of age among those identified as having diabetes from HRA data alone: younger plan members were more likely to have been identified only by HRA data, and have no related claims. Therefore, while HRA participation and biometric data are useful for plan members of all ages, they are especially useful for initial diagnoses in younger plan members who generally use less health care.

Enhancing population health research and clinical practice in ESI

HRA data – self-reported and biometrics – can be a valuable tool for plans? to identify chronic diseases among employees. We hope that other plans and ESI researchers with access to lab and claims data repeat this analysis in their own populations: greater reporting on ESI population health will help establish best practices and standard definitions for commercial plans.

At JPMorgan Chase, our investments in wellness, health care clinics, especially in value-based care models with our partner apree health , present a natural opportunity to connect patients with the care they need to manage their chronic conditions. Subsequent analyses may consider how many patients with an elevated HbA1C level during an HRA biometric screening visit had follow-up care with their primary care provider to evaluate how well HRA visits connect to ongoing condition management.

The analysis for diabetes was straightforward, given the reliability of a single measure of HbA1C as a basis for diagnosis. Other conditions, like high cholesterol and hypertension, require patients to fast before labs or return for multiple readings. At-home biometric screening tools such as those offered by our partner LetsGetChecked might offer an alternative to an in-person visit, either for follow-up or for remote employees to complete the HRA without being onsite. As plan sponsors and plans continue to evolve how they administer and incentivize HRA visits, they should consider the substantial potential value HRA data can add to the understanding and support of patients’ overall health and wellbeing.?