Collecting and analyzing non-clinical data to target interventions

Covid-19 invited much-needed interest among executives in tackling systemic care disparities so they can promote more equitable health outcomes. Health equity interventions must be targeted to a specific patient population or identified disparity likely to meaningfully impact outcomes.

Yet Covid-19 also exposed how much our traditional lack of data collection on health disparities masked our health equity problem. We now know the industry has far less data on social determinants of health than it needs to make process. The root cause problem is that few organizations universally screen for nonclinical needs or stratify their data along demographic lines.

General best practices exist but are blunt tools. Organizations have limited resources, meaning they must collect and analyze data to determine which disparities to target to have an outsized impact for their specific patient populations. Digital tools like electronic health records and virtual touchpoints can help organizations track non-clinical data that can support analyses to uncover disparities at specific organizations.

Using machine learning to make treatment pathways more precise and equitable

Progressive organizations are beginning to use their rich, non-clinical data to power machine learning models that identify and address care disparities. For example, a group of researchers used a machine learning algorithm to prove that traditional radiographic severity measures were missing certain knee pain features of Black patients—who experienced 75% worse knee pain than non-Black patients, at the median. The researchers trained their model to identify patients with severe pain based on radiographic imaging that was missed by traditional grading systems. The model helped prevent Black patients from being overlooked as candidates for surgery—22% of Black patients were eligible for surgery based on the researchers’ model, compared to 11% using traditional grading.

Adoption of machine learning models for health equity use cases are limited currently but can begin to generate outsized impact as their developers overcome legacy barriers like training data availability and representativeness.

Industry leaders have long been concerned about how sustainable our legacy senior care model—often viewed as costly and unprofitable—will be as the Baby Boomer generation ages into Medicare. With each passing year, population aging becomes a more acute problem—the 85+ age group will nearly double as a share of the population between 2025 and 2050, according to MedPAC. This growth will only exacerbate rising spending and the looming insolvency of the Medicare Trust Fund, which is projected to occur in 2026.

An opportunity is emerging for digital health tools to drive outsized impact on this challenge by bending this cost curve and helping seniors fulfill their growing desire to age at home. Two types of digital tools could have especially high leverage in helping industry leaders achieve these goals:

Tools that enable remote patient monitoring

Most simplistically, virtual visits can help physicians monitor their patients without them needing to travel to the office or hospital site. Wearable devices (e.g., smart blood pressure monitors), meanwhile, can impact chronic condition management by enabling providers to track their patients’ key health metrics on an ongoing basis. This allows for more targeted, and ideally, preventative, interventions.

Virtual visits and remote monitoring are especially impactful under hospital at home models, which have grown significantly in the last few years. For hospital at home care, telemedicine and remote monitoring give the care team insight into the patient’s condition without requiring their physical presence in the patient’s home.

Tools that enable aging in place

Seniors overwhelming prefer to age at home—90% of seniors prefer to remain in their homes instead of moving to an assisted living facility. Doing so safely and effectively currently requires significant caregiver support. Technology can help offload certain caregiver tasks to make it more accessible. A variety of technologies can play a role:

• Non-clinical tools: tools that ensure the home is safe for seniors, such as apps and devices that can turn off your stove or alert caregivers that help is needed.
• Clinical tools: tools that ensure adherence to medication use and care protocols, such as pill boxes that remind seniors to take their medicine.
• Social tools: tools that connect seniors to social services that can provide additional, non-clinical support.

Success in risk-based payment models can improve quality and reduce cost—but only if providers have access to the data and tools they need to identify and act upon opportunities for quality improvement and savings. Digitally enabled healthcare can support this effort by improving patient risk stratification and creating data clinicians can use to make more proactive care decisions for their patients.

Risk stratification

Risk stratification helps providers identify how to allocate resources across different subgroups of patients with similar complexity and needs, enabling rightsized care and lower overall cost. To stratify risk, provider must have access to longitudinal and comprehensive data on their patient populations. This process is most effective when providers standardize their data collection to ease analysis with tools like electronic health records and when they partner with other industry sectors to access new datasets (e.g., pharmacy claims) or combine their datasets (e.g., medical claims and clinical records).

Some progressive organizations are even bolstering risk stratification by incorporating non-clinical data—like credit reports and social determinants of health data—that add depth to patients’ future likelihood of requiring services. Using non-traditional data to achieve predictive risk stratification, in turn, enables providers to provide preventative care that prevents health challenges before they arise.

Digital health-enabled data for clinical decision making

Wearable devices can help clinicians make more proactive and accurate care decisions by providing an ongoing look at the patient’s condition. This can allow clinicians to adapt patients’ care plans in real time to meet their current needs, rather than needing to rely on data reflecting a snapshot of the patient’s condition at the time of a visit. In turn, this smarter care management can improve quality and reduce cost of treating chronic diseases under value-based payment.