by Scott Code, Senior Director, LeadingAge CAST
In the last 10 years, senior living providers have gone from primarily capturing handwritten notes about their residents and other information to digitalizing information across their organization. Now the THE BASICS: GETTING TO “WHAT” IS HAPPENING AND “WHY”
An organization may already be using basic descriptive analytics, such as graphic dashboards and reports, to visualize sudden changes, trends, and potential correlation relationships (but not causality), to help users understand and summarize what has happened in the past. An example may include a report of the monthly rehospitalization rate of residents. Descriptive analytics helps identify “what” has happened in the past, such as a stabilization, decrease or increase. However, we really want to know “why” something has happened. Diagnostic analytics, on the other hand, helps an organization narrow down on the root cause of “why” something is happening. In the case of rehospitalizations, that root cause may be accepting patients from a new hospital partner that is discharging patients too early or are not cooperating on medication reconciliations. Together, descriptive and diagnostic analytics provide the “what” has happened and “why” it is happening.
LEARNING FROM PAST EXPERIENCES: PREDICTIVE ANALYTICS AND DECISION MAKING/SUPPORT
With so much uncertainty these days, having the ability to predict future events or understand the probability of a future events is very powerful. That is where predictive analytics come into play. Predictive analytics can build a mathematical model that uses past “what” and “why” to help identify residents at higher risks of falling and/or rehospitalizations at certain times, and may need attention or care at night or during the weekend when staffing may be lower. Predicative analytics can allow a senior living provider to start providing a proactive approach to care rather than a reactive approach. Decision-making and decision support capability rely on predictive analytics and models, but go one step further in modeling decisions and actions that have helped improve the desired outcome in the past, or mitigate the adverse consequences based on best practices and learning historic data – for example, increasing staffing at nights and weekends and would make a recommendation to the user.
EXAMPLES FROM THE COVID-19 PANDEMIC
During COVID-19 pandemic, predictive analytics can help an organization identify which residents are at higher risk for contracting the virus as well as most likely to have poor health outcomes. During the early months of the pandemic, a journal article from JAMDA titled, “Predicting Coronavirus Disease 2019 Infection Risk and Related Risk Drivers in Nursing Homes: A Machine Learning Approach” highlighted how the use of predictive analytics can help describe and predict the risk of COVID-19 outbreak in nursing homes, providing data-driven support for nursing home infection control policies, strategies for the prioritization of resources to high-risk nursing homes, and the relaxation and restriction of nursing home visitation policies. The study concluded that “The prevalence of COVID-19 infections in a nursing home’s surrounding community and a nursing home’s size were identified as the primary risk factors associated with nursing home infection, suggesting that the introduction of infection from the outside community as a likely infection mechanism. Developing financially sustainable testing and screening approaches to identify pre-symptomatic and asymptomatic individuals entering a nursing home are critical to preventing and controlling COVID-19 outbreaks in these settings.”
The U.S. Coronavirus Map: What Do the Trends Mean for You?, is an interactive map included in Mayo Clinic’s COVID-19 online resource center. The map represents key data and trends in an easy-to-use format and displays figures like the total number of cases in each county and state, new cases per day, positive test rate and fatality rate, presented with trends over time. To create the map, data scientists developed content sources, validated information, and correlated expertise for the tracker, all of which will be enhanced with more real-time data and refinement of the predictive analytics models used.
If you are not utilizing the data you have, from your EHR, Financial Systems, Telehealth, Assessment Data, etc. to comprehensively understand what is happening in your organization, and/or your market, and why, as well as potentially leveraging the powers of predictive analytics and decision support to improve your quality and performance, which we know affect your financial bottom-line, you are at risk of falling behind! However, utilizing the right data analytics tools and capabilities, some of which are potentially already embedded in some of the technology systems you already invested in, can help prevent you from falling behind!
WHAT IS CAST DOING?
In an effort aimed at preventing you from falling behind, LeadingAge CAST will be releasing a Technology Selection Tool late this year focusing on Data Analytics to help you better understand this field, which is quickly gaining importance in the aging services sector, and assist you in sifting through commercially available solutions. The portfolio will include a whitepaper providing background on Data Analytic technologies, types of available systems, benefits and the most important planning steps that long-term and post-acute care (LTPAC) providers should consider before adopting and implementing these solutions. In addition, the portfolio includes a Product Selection Matrix and an Online Selection Tool that helps providers select the most appropriate solution for their setting/ business line, functional requirements and needs. Finally, vendors will nominate a provider user to contribute a Case Study.
Stay tuned to LeadingAge and LeadingAge CAST.
