Early Findings: Observed Behaviors by Time of Day

Behavioral Analysis to Detect Early Indicators of Changing Acuity of Seniors

In August of 2022, Fynn launched its operations platform that enables senior living operators to leverage behavioral analysis—a powerful and underutilized tool—to gain a deeper understanding of residents for earlier detection of potential changes in acuity. This report highlights early findings on how the platform is using behavioral analysis to proactively identify changes in physical, social, and emotional well-being sooner. Fynn will continue to update these ongoing findings as it learns about the impacts of continuous behavioral analysis on resident and community outcomes and the engagement and support for senior living care teams.


According to Cleveland Clinic, sundowning or late-day confusion is a phenomenon experienced by 20% of people with Alzheimer’s and dementia. Symptoms begin in the late afternoons through the evening and include changes in behavior such as increased feelings of agitation, confusion, etc., at times escalating to wandering, hallucinations, and in some cases, violence. To help care teams anticipate potential adverse events caused by sundowning, Fynn has developed a continual assessment of resident behaviors to detect early patterns of behaviors associated with sundowning. This platform has empowered care staff to work with medical professionals and families to deploy earlier interventions such as adjusting levels of care, environmental triggers, or medications for residents who may be experiencing sundowning.

To date, Fynn has been deployed in four senior living communities offering Independent Living through Memory Care. The goal is to address the challenge of improving the resident care experience by identifying early shifts in resident well-being. Fynn enables senior living operators to continually assess behavioral trends to provide an earlier indication of patterns that may be related to sundowning and enable staff to anticipate and intervene to improve residents’ well-being and their care experience.

The Big Idea

Fynn is committed to promoting Health Sustainability—a conceptual and actionable approach to transforming the availability, efficiency, and quality of senior living options and operations. While the Health Sustainability Movement seeks to bring together the larger senior living ecosystem, Fynn is contributing by helping care teams identify physical and emotional changes earlier for proactive, personalized care.

Through Fynn’s novel data collection and analysis of behavioral data recorded in the platform, Fynn is able to detect early patterns of behaviors associated with sundowning. These early findings are the first step of many toward improving the resident care experience and ultimately, toward creating Health Sustainability.

A Proactive Approach

Fynn leverages behavioral analysis to detect earlier indicators of potential shifts in acuity. Rather than waiting for quarterly wellness assessments to track residents’ well-being, Fynn provides a means to continually record resident behaviors as well as physical health data collected by staff upon completion of tasks and Activities of Daily Living (ADLs). By examining both behavioral and physiological data, Fynn provides a more complete picture of a resident’s health across frequently recorded data points rather than limiting data to a few formal assessments per year.

Working with data scientists and physicians, Fynn used the Mini-Mental State Examination (MMSE), a questionnaire that is used extensively in clinical and research settings to measure cognitive impairment, to identify a limited set of clinically-significant behaviors for staff to observe and record. Fynn transforms the data into baseline trends created for individuals. Deviations from expected trends drive proactive intervention based on individual baselines for more personalized care.

Recognizing that sundowning often has negative impacts on both resident and employee well-being, Fynn began analyzing behavioral trends among residents compared to the time of day to see if negative behavioral patterns increased in the afternoon and evening hours. The graphs below are examples of some anonymized behavioral trends from residents who experienced worsening behaviors as the day progressed. It is important to note that while Fynn records trends and informs care teams of deviations from trends, Fynn is not a diagnostic tool. 

How Fynn Works:

To consistently collect behavioral data on a daily basis, resident behaviors are collected by staff upon completion of tasks and Activities of Daily Living (ADLs). Fynn is device agnostic, meaning staff can record behaviors on tablets or mobile devices into Fynn’s secure app. The behaviors collected by staff are supplemented with passive sensor signals to develop insights into resident well-being.

In one example, a staff member shared that her team was able to use Fynn to identify an increase in negative behaviors for one Memory Care resident that has a diagnosis of dementia. She noted that his behaviors were time-of-day specific and identified as sundowning. The increase in negative behaviors triggered a change in caregivers, as the data helped identify that the resident responded better to male caregivers. It also prompted an appointment with the resident’s neurologist to discuss adjusting his medications as well as a shift in his daily routine. She credited Fynn with helping her to identify “residents at-risk” through reports on these residents that helped her implement interventions based on the findings.

Fynn’s novel approach of using behavioral data to better understand and proactively care for residents generates a positive feedback loop of action. By providing empirical evidence of a dramatic shift in resident behavior in correlation with the time of day, care staff can work with residents’ physicians and families to create action plans backed by data rather than anecdotal details. This improves care collaboration and experience while seeking to reduce adverse events that may result in hospitalization, decreased quality of life, and legal ramifications.

“Combined with objective physiological data, behavioral analysis is a powerful and underutilized tool for understanding the ‘wellness pulse’ of each individual resident and potential best practices or common failure points when compared across communities. As our data collection becomes more robust, we anticipate Fynn’s behavioral analysis will become even more precise in its identification of ‘residents at-risk’, which will be a major game-changer for resident well-being and cost-saver for communities.” –John Thomas, Chief Technology Officer at Fynn

Fynn Early Findings Data

The following charts show how Fynn enabled staff to identify behavioral pattern changes and make care adjustments that had a positive impact on residents.

Resident 1 (Memory Care)

  • OBSERVATION: Typical behavior pattern for a resident in Memory Care
  • ANALYSIS: Staff members knew Resident 1 in Memory Care was stable throughout most of the afternoon and sharply declined in the evening
  • RECOMMENDED ACTION: Maximize resident activity during the late morning through mid-afternoon and remove environmental triggers in the evening

Resident 2 (Memory Care)

  • OBSERVATION: Steep decline around 2:30 pm is likely an indication of early sundowning
  • ANALYSIS: By knowing when a resident experienced their most positive and negative behaviors, care staff could tailor a resident’s care plan and improve their care experience. For example, Resident 2’s higher friction ADLs such as showering could be scheduled for the morning when he/she is most agreeable and engaged
  • RECOMMENDED ACTION: Discuss medications and/or programming changes with doctor and family

Resident 3 (Assisted Living)

  • OBSERVATION: Behaviors peak mid-morning followed by a decline after lunch
  • ANALYSIS: Staff can more closely observe Resident 3 to investigate the causality of his/her declining behavior following lunch and work with the resident’s physicians and family to determine if the shift in behavior is associated with early signs of sundowning and dementia. From here, Resident 3’s doctors can provide the necessary treatment and work with the community staff and family to safely maximize the resident’s cognitive functioning for longer
  • RECOMMENDED ACTION: Investigate the causality of decline after lunch and partner with physician and family to discuss care level

To Date, Fynn has Demonstrated the Following:

1) Staff are able to determine when resident behavior would begin to worsen and anticipate those changes.

2) Staff are empowered to take preventative measures. 

Rather than relying on “gut feelings” or anecdotal and inconsistent observations, the care team could create an action plan based on each resident’s behavioral trends—which clearly validated a shift in behavior and the need for intervention.

3) Staff are celebrated for promoting resident well-being and community NOI.

By recognizing the changes in resident behavior and beginning to form action plans to ease the negative symptoms associated with sundowning and/or other behavioral declines, team members were rewarded for better sustaining resident and community well-being. The community believes Fynn’s behavioral analysis will help them to better understand when and how to intervene earlier, thus improving resident care and potentially reducing incidents. Furthermore, community ownership and managers hope to use the correlation between sundowners and adverse events to improve root-cause analysis of the major challenges affecting NOI.

Next Steps

As communities’ volume of data increases, Fynn’s intelligent system will expand from demonstrating trends to recognizing specific risks and offering recommended actions based on identified trends. Using behavioral data to generate automated alerts and system-directed tasks based on identified patterns is the next step as Fynn works to help communities provide proactive care.