Closing the Gap Between Lifespan and Health Span with Predictive Analytics

The U.S. is a graying nation. Americans 65 and older are expected to make up more than 20% of the population by 2030, up from 17% in 2022, and as much as 23% by 2050. We’re living longer, but not better; an increasing portion of our lifespan is spent in poor health. 

This demographic shift brings with it costly and complex healthcare needs, since older adults often suffer from chronic conditions and comorbidities that require more intensive medical care. With our current healthcare system plagued by workforce shortages, capacity issues, and rising costs that are already outpacing budgeted Medicare spending projections, we are woefully underprepared for the growing demands of our aging population. 

To effectively meet the challenges the “gray wave” presents, healthcare organizations and payers must collaborate and adopt a new approach that can help reduce the risks to older adults through earlier identification and intervention. Despite the growing number of technology resources available, we have not yet figured out a way to maximize the interaction of these resources with one another — and, more importantly, with payers, providers, patients and public health.  

Predictive modeling can help mitigate one of the biggest health span risks

One of the biggest dangers we face in growing older is falling, which is the leading cause of disability and death among patients over the age of 65. Falls don’t just represent physical stumbles; they are a profound threat to the health and independence of older adults. They pose an economic threat as well; approximately $50 billion is spent on medical costs related to older adult falls each year in the U.S., according to the Centers for Disease Control and Prevention (CDC).  Additionally, falls happen for a myriad of reasons and their convergence, be it from the consequences of polypharmacy, deconditioning and sarcopenia, cognitive changes or patterns of declining health

Predictive models provide a potential answer for detection and to ultimately tee up necessary prevention strategies. These models analyze disparate data points to predict health events, allowing for timely and targeted interventions that can extend health spans and improve quality of life. By leveraging vast amounts of patient data and artificial intelligence (AI), predictive models can generate precise mortality and morbidity predictions, such as identifying patients at risk of falling, months, and even years, in advance.

There are reams of data available to predictive models that could lead to accurate risk prediction with conditions like falls, but data capture at the point of care and data interoperabilsty remain inconsistent. Clinical data is fraught with variability from EHR to EHR, from implementation to implementation of the same EHR, between providers, and even between patients based on social and medical differences. These discrepancies in data capture, presentation, and quality lead to challenges in weaving a cohesive “story” that is requisite for not only predictive modeling, but even more critical, for care delivery. To overcome these hurdles, interoperability solutions and structured data frameworks are imperative.

Achieving a standards-based approach

The CDC’s STEADI (Stopping Elderly Accidents, Deaths, and Injuries) initiative is an example of a successful standards-based approach. Premised on the clinical fall prevention guidelines of the American and British Geriatrics Societies, STEADI provides healthcare providers with a structured process framework for screening fall risks, assessing modifiable risk factors, and implementing targeted interventions. However, as noted earlier, the impact of such initiatives like this can be limited by inconsistent and nonstandard data recording in EHRs, thereby blunting the impact of predictive models and other novel technologies reliant on optimized data for their effective use. Many patient-provider interactions are documented using questionnaires that are not appended to existent vocabularies and when they are, they are not integrated into EMR systems, leading to fragmented data capture that is difficult to analyze comprehensively. By adopting and enforcing standards for such data collection tools with vocabularies like Logical Observation Identifiers Names and Codes (LOINC), healthcare providers can ensure that data from the STEADI and other questionnaires are consistently recorded and tracked over time as data. This standardization would enable better longitudinal analysis, improving patient care by providing clinicians with a clear view of a patient’s health trajectory. By tailoring its models to incorporate STEADI criteria, a technology partner can ensure that the predictive analysis aligns with the established guidelines for fall risk assessment. 

The issue of falls among older adults is a major concern for public health, given that we are all living longer. There are many other health metrics and clinical conditions that are critical to address for more effectiveness of public health initiatives. By establishing a standard method of gathering data and utilizing programs like STEADI, we can significantly advance our frailty and fall prevention efforts and improve delivery models for other conditions that impact the most vulnerable members of our population. Standardization not only paves the way for compatible data across the system, but also arms healthcare providers with valued insights to make informed decisions. Incorporating advanced predictive technologies within fall prevention efforts will further improve the outcomes, relieving the burden on the healthcare system, and providing an opportunity for all players within the ecosystem to drive to value-based care and better serve an aging population.

Photo: Getty Images

This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.