Want to reduce Falls risk?
- chiarahaeussermann
- 42 minutes ago
- 3 min read
Falls risk is not always obvious
Falls are rarely caused by a single factor. The risk can build slowly over time due to changes in health, mobility, strength, balance, medication, pain, memory, mood, or daily activity.
These changes are not always easy for families, carers or medical professionals to spot. A person may appear 'fine' during a visit, yet small changes in their walking ability, confidence or daily routine can still increase their risk of falling. Short on resources, it can be challenging for care providers to prioritise clients based on their risk of falling, particularly when they are responsible for managing the needs of many individuals.
Current research is clear: falls prevention should be proactive, targeted and personalised. The 2022 World Guidelines for Falls Prevention and Management recommend identifying
different levels of risk and using multifactorial assessment. NICE's 2025 guidance also focuses on assessing risk, identifying contributing factors, and reducing the likelihood of injury, distress, loss of confidence and independence, and mortality.
What is the La Casa Care Predictive Falls Risk Assessment?
It is an AI-supported decision tool that helps to estimate a person’s risk of falling. It uses information about a person’s health, lifestyle and daily activities to provide a clearer indication of risk.
It considers factors such as age, long-term conditions, pain, mood, memory, medication, strength, mobility, history of falls and daily activities. It then converts this data into a risk score, which it presents in a simple traffic-light format.
Green indicates a lower estimated risk.
Yellow indicates a boundary group that may require closer monitoring or further investigation.
Red indicates a higher estimated risk, where a more detailed falls assessment or prevention plan may be required.
The tool is designed to support decision-making, not to make decisions on its own. Instead, it provides families and care teams with a clearer way of identifying those who may require attention sooner.

How the Falls Risk Assessment works
The La Casa Care Predictive Falls Risk Assessment has been trained on several major ageing studies, including the CHARLS (China) and HRS (USA) datasets, which provide thousands of records describing older adults’ mobility, strength, pain, mood, memory, medication use, chronic conditions and fall events across multiple waves.
Using this data, the system learns patterns through a specialised version of LightGBM, a gradient‑boosting model that builds many small decision trees and combines them to detect subtle risk signals. We use a variant called focal loss, which makes the model pay more attention to harder‑to‑classify individuals. Once trained, the model produces a probability score for each person, and a validation process selects two confidence thresholds: a lower threshold (below which the model is confident the person is at lower risk) and a higher threshold (above which the model is confident the person is at higher risk). Scores in between form a boundary zone where signals are mixed. This creates the final green/yellow/red output. To support interpretation, the system also computes SHAP‑style feature contributions, which highlight the factors that most influenced each prediction (such as mobility limitations, pain, medication effects or previous falls), helping families and care teams understand why the model sees elevated risk.
From risk signals to earlier action
Falls prevention is most effective when people can take action before risk turns into harm. La Casa Care’s Predictive Falls Risk Assessment makes this possible by turning multiple risk factors into a clear signal that supports earlier conversations, targeted check-ins, and better prioritisation.
For older people, this means that changes in mobility, confidence, or daily functioning can be recognised sooner. For families and informal carers, it provides a clearer indication of when extra support may be needed. For care providers, it helps them to focus their time and resources on those most likely to benefit from follow-up or a more detailed assessment.
Falls risk is also influenced by the home environment. When a person is identified as being at a higher risk of falling, La Casa Care’s Safety Scan can help to reveal visible hazards, such as poor lighting, clutter and trip hazards. If you would like to read more about it, visit our blog post: “How Smartphones Can Help Make Home Safer for Older People”. These two tools support a more proactive approach to prevention: identifying risk earlier, understanding the home context, and guiding practical next steps.
La Casa Care helps older people live more safely and independently at home, with AI-supported falls risk assessment built for earlier prevention.
Want to learn more about La Casa Care’s falls prevention solutions?
Contact us at info@lacasa.care.




Comments