Smart Care Technology: Fall Prevention or Fall Detection?

Can smart care technologies assist in preventing falls, or only in detecting them after the fact? The overarching goal of falls-prevention tech is not to restrain or limit a patient’s mobility, but rather to avert the negative consequences of a fall. 

---

In cardiology, a pacemaker revolutionised arrhythmia treatment by maintaining normal heart rhythm over the past 50 years. Is there an equally simple, robust “pacemaker-like” solution for falls? Unfortunately, falls are a complex, multifactorial problem, and no single device can guarantee prevention of falls at this time. Most smart care technologies today focus on detecting falls or alerting caregivers when risky movements occur, which can prevent injuries by enabling a quick response. Some emerging systems even attempt to predict falls by monitoring subtle changes in a person’s movement patterns (e.g. slowing gait speed) that precede a fall. 

However, these predictive solutions are still evolving and are not yet foolproof. Any fall-prevention product must be chosen in the context of an individual patient’s needs, and it must be easy for staff to use – technology that is cumbersome will see reduced effectiveness. Even if devices are easy-to-use, low-cost, and meet electrical and cybersecurity guidelines, their functional maturity (measured by technology readiness levels, TRL 1–9) remains moderate for most fall technologies – only certain location tracking systems are close to the highest maturity (TRL 9), while many other solutions are far from fully mature. 

Choosing the right technology requires understanding different sensor settings, knowing the top fall risk predictors, and reviewing the current range of fall-related technology options. In the sections below, we examine how device settings impact performance, identify the top 5 predictors of falls, and survey various fall-related technologies (with pros and cons) to see how they can aid in prevention.

Precision vs. Sensitivity in Fall Detection Settings

A key challenge in fall-detection sensors is finding the right balance between sensitivity and precision (often called specificity in medical contexts). Effective use of this technology means tuning it to detect as many true falls as possible (high sensitivity) without drowning staff in false alarms (high precision). There is usually a trade-off: increasing sensitivity tends to decrease precision (more false positives), whereas pushing for very high precision can miss real falls.

For example, one hospital study found that of 67 patient falls in a cardiac care unit over 2 years, only one patient who fell had a bed alarm active – and that alarm was set to its lowest sensitivity. Because it only triggered when the patient was completely out of bed, it failed to alert during the attempt to get up. The authors speculated that a more sensitive setting (triggering an alarm earlier, e.g. on shifting weight to the bed edge) might have prevented that fall. But is cranking up sensitivity always better? Not necessarily – an overly sensitive alarm could ring constantly with false alarms whenever a patient shifts normally, leading to staff alarm fatigue.

In fall detection terms:

• Sensitivity is the ability to correctly identify true positives (falls that really happen). A true positive means a fall occurs and the system correctly detects it. A false negative means a fall occurs but the system fails to detect it (missed fall).

• Precision (or positive predictive value) is the ability to avoid false alarms – effectively identifying true negatives. A true negative means no fall occurs and the system correctly stays quiet. A false positive means the system incorrectly alarms for a fall that did not occur.

Every 24 hours, a continuous monitoring system might analyze 17,280 discrete five-second periods for “fall vs. no fall” events. Even a precision of 99.994% – which sounds extremely high – would still result in about 1 false alarm per day (0.006% of 17,280 is ~1) due to the sheer number of opportunities for error. Real-world studies reflect this challenge. For instance, one wearable sensor system achieved 80% sensitivity but with a false alarm rate of ~0.05 per hour (around 1 false alert per day). 

On the other hand, if you adjust thresholds to eliminate virtually all false positives (aiming for ~100% precision), sensitivity can drop dramatically. One smartphone-based fall detection study attained >99.9% specificity (almost no false alarms, about one false alert every 46 days) but at the cost of only ~73% sensitivity (it missed about 1 in 4 genuine falls). In other words, to avoid “crying wolf” with false alarms, the system became far less sensitive to actual falls.

So what’s the better approach? A high sensitivity (“better safe than sorry”) will catch more falls but risk overwhelming staff with false alarms and causing alert fatigue. A high precision setting avoids false alarms but risks missing real falls – potentially missing 1 out of 5 falls, as in the example above. In practice, a balance must be struck based on the care setting and staff capacity. Many facilities opt for moderately high sensitivity with some tolerance for false alarms, combined with clear protocols so staff do not become desensitized. Ongoing adjustment and staff feedback are important – if too many false alarms occur, staff may start ignoring alarms, defeating the purpose. 

Conversely, missing falls is unacceptable from a safety standpoint. The optimal tuning often involves iterative trials to approach the sweet spot where the system is as sensitive as possible while still maintaining a high enough precision that false alarms are infrequent and credible. The goal is a near-100% precision and high sensitivity – a tough target – which is why multiple technologies and layers of detection are often used together.

(Every facility is different: a dementia unit with very high fall risk might accept more false alarms to catch every possible fall, whereas an acute care ward might set devices to alarm only on more obvious danger to avoid constant distraction. Either way, continual monitoring of the alarm data and outcomes is key to finding the right setting.)

Top 5 Predictors of Falls

When assessing fall risk, research shows there are many contributing factors – but a few predictors stand out above the rest. The top five predictors of falls in older adults (especially in care settings) are usually:

1. History of Previous Falls – Has the person fallen before?

2. Gait or Balance Impairment (Unsteady Gait) – Do they have difficulty walking or frequent loss of balance?

3. Use of Antipsychotic Medications – Are they on antipsychotic drugs?

4. Use of Antidepressant Medications – Are they on antidepressants?

5. ADL Impairment – Do they have trouble with Activities of Daily Living (e.g. needing assistance with basic tasks?)

Notably, the first two factors dominate fall risk by far. A history of falls is arguably the strongest single predictor of future falls. If someone has fallen in the recent past, it indicates underlying issues (e.g. mobility or balance problems, cognitive impairment, environmental hazards) that are likely still present. In fact, one meta-analysis found that a previous fall, along with functional impairments in ADLs, had the strongest association with fall risk in nursing home residents. Gait impairment – often evidenced by an unsteady or shuffling gait – is another leading indicator closely tied to fall likelihood. Simply put, people who walk with instability or muscle weakness are far more prone to falling.

Fall History: It’s easy to be wise after the event – anyone who has already fallen is clearly at risk of falling again. Technology can’t change the past, but it can help respond better in the future. Monitoring systems for those with a fall history can at least ensure faster response and provide data to tailor a prevention plan. For example, wearable fall detectors or room sensors can automatically log when and where a fall occurred, potentially informing caregivers about patterns (e.g. falls happen during nighttime bathroom trips or in a particular location). While knowing someone’s fall history flags them as high-risk, it doesn’t by itself prevent the next fall – proactive measures are needed. This might include closer supervision, exercise programs, medication review, or sensor technology that alerts staff when the person is attempting risky movements.

Gait Changes: Declines in gait speed or stride length are one of the best leading indicators of an impending fall. Remarkably, research using in-home motion sensors found that if an older adult’s usual walking speed slows down by just 5 cm/sec over the course of a week, there is an 86% probability of a fall in the next three weeks. That is four times the fall risk of someone whose walking speed hasn’t changed. Similarly, a decrease in average stride length by about 7.6 cm in a week was associated with a ~50% probability of falling within three weeks. These are huge predictive signals – essentially, a sudden decline in mobility is often a red flag. It might reflect growing muscle weakness, pain, dizziness, or other emerging health issues that directly increase fall risk. Unfortunately, continuously measuring a person’s gait in real life is not easy without technology. This is where smart care tech could actually help predict falls: by tracking a resident’s daily step counts, walking speed, or general activity level, subtle declines can trigger an early warning. For instance, if a usually active resident has been moving significantly less or slower this week, staff could be alerted to check on them proactively (perhaps the person is developing an illness, vertigo, or other issue that could precipitate a fall).

Incorrect Weight Shifting – The #1 Cause of Imbalance: Beyond individual risk factors, it helps to understand how most falls happen in order to predict and prevent them. One landmark study used video footage in long-term care to analyze circumstances of falls. The findings busted the myth that most falls are due to tripping or slipping. In fact, the single most common root cause was incorrect weight transfer or shifting of body weight, accounting for roughly 41% of falls. In other words, nearly half of all falls happened not because the person stepped on a hazard or was pushed, but because they lost balance due to their own movement – leaning too far, turning improperly, or mis-stepping such that their center of gravity moved outside their base of support. (By comparison, true trip or stumble events caused only about 20% of falls in that study, and slips on wet surfaces only ~3%.) This means imbalance during routine activities is the biggest culprit.

What activities were people doing when they fell due to incorrect weight shifting? The majority were on the move performing ordinary actions. The top three activities just before a fall were:

• Walking – e.g. walking forward or turning while walking (a significant portion of falls, ~24% in one analysis, and even higher if including initiating walking).

• Standing – e.g. standing in place or standing up from a seated position (around 13% of falls were from standing quietly, and many others during the act of standing up or transferring).

• Transferring – especially transitions like rising from a chair, bed, or toilet, or pivoting from one surface to another (sitting down or lowering oneself was involved in ~13% of falls in the video study, and other data shows a large share of falls occur during sit-to-stand attempts or similar transfers).

The key point is that most falls happen during motion, not when a person is just sitting safely. About 70% of falls in frail older adults occur during everyday activities – walking, getting up, reaching – rather than external hazards. This underscores why mobility monitoring is so important. Since incorrect weight shifting (a sign of impaired balance control) is such a prevalent mechanism, the best way to predict falls is to detect changes in how a person moves: slower gait, shorter steps, more imbalance or swaying, etc. Technology that can continuously observe these factors – for example, motion sensors that build a baseline of someone’s typical movement patterns and gait, or wearables that measure walking speed and stability – could potentially alert caregivers to deterioration before a fall happens. Knowing a person’s gait is slowing or they are stumbling more could prompt preventative interventions (medical review, physical therapy, assistive devices) to avert an imminent fall.

(In summary: Fall history tells you who is high-risk; gait decline tells you when risk is spiking; and weight-shift imbalance tells you what scenarios to guard against (e.g. be extra careful when the person is walking or standing up). The next section looks at what types of tech tools exist to address these issues.)

The Types of Fall-Related Technologies

The most common method for fall detection and prevention is the use of sensors monitoring the person’s activity. Broadly, these technologies fall into two main categories: wearable sensors and ambient (environmental) sensors. Each category includes several types of devices:

1. Wearable sensors: These are devices worn on the person’s body (on the wrist, hip, around the neck, etc.) that detect motion or position. The most prevalent are inertial sensors – typically accelerometers and gyroscopes that measure movement and orientation. A simple accelerometer can capture the sudden acceleration and impact of a fall, making it sufficient for basic fall detection algorithms. Another wearable approach is real-time locating systems (RTLS), where the person wears a tag (like a badge or pendant) that communicates with receivers installed around the facility. These can track the person’s location and movement in real time, alerting staff if, say, a high-risk patient is out of bed or wandering into unsafe areas. Location tracking doesn’t detect a fall per se, but by monitoring unusual movement patterns or lack of movement, it can infer that something may be wrong. RTLS is a relatively mature technology (widely used in hospitals for asset and patient tracking) and can support fall management by quickly pinpointing where a fall occurred or by triggering alerts when certain conditions are met.

   
   

2. Ambient “smart” sensors: These are placed in the person’s environment (room, bed, chair, etc.) and do not require the person to wear anything. There are a variety of ambient sensors used for falls:

   • Pressure sensors: Often integrated into mats or pads on beds, chairs, or the floor. For example, a bed pressure pad can detect when a patient is attempting to get up; if weight shifts toward the edge, it triggers an alarm so staff can assist before the patient fully stands (potentially preventing an unassisted fall). Chair pads and floor mats likewise alert when pressure is applied or removed unexpectedly. Some beds have built-in pressure sensors that can tell if a patient has left the bed.

   • Sound sensors: These can listen for loud noises or thuds that might indicate a fall, or changes in the usual sound pattern of activity. While not as commonly used alone, sound detection can complement other systems (for instance, detecting a crash or cry for help).

   • Infrared motion sensors: These detect movement using IR beams or motion detectors. They might be set up at doorways (to detect if a person exits a room or enters a bathroom) or in a room to monitor activity. If no motion is detected for a long period when there should be (or conversely, motion is detected at an unusual time), the system can send an alert. Simpler versions include IR beam devices that create an invisible “tripwire” – for example, an IR beam across a bed that triggers an alarm if the beam is broken when the patient gets up.

   • Vision sensors (cameras): Camera-based fall detection uses video to identify falls, either through computer vision algorithms or human monitoring. Modern AI can analyze video feed for patterns like a person collapsing or lying on the ground. Cameras can be very effective in detection and even allow remote assessment of the situation, but they raise privacy concerns, especially in personal spaces like bedrooms or bathrooms.

   • Radar sensors: This is a newer, cutting-edge approach. Radar-based fall sensors use radio waves (often high-frequency FMCW radar) to detect human movement and posture without cameras. They can sense falls by the characteristic way a body moves and even monitor breathing or slight movements. Radar can work in the dark, through steam (e.g. in a bathroom with a hot shower), and does not capture an image of the person (better for privacy). So radar offers a promising balance of visibility for the caregiver and privacy for the resident. Some radar devices can create a 4D imaging of movement and have been shown to detect falls even behind obstacles like shower curtains. The downside right now is cost and the fact that these are still emerging products (not yet standard in most facilities).

Barriers: Each type of sensor technology has challenges. Achieving accurate fall detection is inherently difficult – distinguishing a real fall from normal activities (or even from someone dropping an object) often leads to trade-offs as discussed earlier. Wearable sensors face issues of form factor (the device must be comfortable and acceptable to wear), battery life (frequent charging or battery changes can be a hassle), and user compliance. An older adult might forget or refuse to wear a gadget every day, or wear it incorrectly, reducing its effectiveness. 

The position of a wearable (waist vs. neck vs. wrist) also affects accuracy; for example, a wrist device (like a smart watch) might interpret sudden arm movements as falls. On the other hand, ambient sensors require that the fall occurs within their detection range – a camera or radar in the living room won’t help if the person falls in the bathroom, unless multiple sensors are installed throughout the home. This can be expensive and complex to manage. Ambient devices can also be triggered only when the person is in view or within the sensors’ area; if the person crawls out of the camera frame after falling, the system might lose track. Additionally, some older facilities have physical layouts or clutter that make sensor placement tricky (e.g., need to avoid too many false triggers from a pet moving around, etc.).

Resident Motivational Factors: The willingness of older adults to use fall prevention technology can depend on several perceived factors:

• Control: They want to feel in control of their lives, not “watched” or constrained by devices.

• Independence: A technology should enhance independence (by providing safety) rather than be seen as a sign of dependency.

• Privacy: Especially with cameras or microphones, privacy is a big concern. Many seniors (and their families) are uncomfortable with constant surveillance, so non-camera solutions or those that respect privacy tend to be preferred.

• Self-efficacy: If the person believes the device actually helps them remain safe, they are more likely to accept it. Education about how an alert pendant, for instance, empowers them to get help quickly may improve adoption.

• Perceived need for safety: Individuals who acknowledge their fall risk are more open to technology; those in denial about their risk may resist.

• Usability: The device must be easy to operate – simple interfaces, minimal steps. Anything confusing (multiple buttons, complex charging procedures) will discourage use.

• Cost: Many seniors live on fixed incomes, so the cost of devices or subscription monitoring services can be a barrier. Technologies that are affordable (or covered by insurance/programs) have better uptake.

  
  

Wearable vs. Ambient Sensors: Pros and Cons

There is an ongoing debate on wearables versus ambient sensor solutions for falls. Each has pros and cons, and in many cases a hybrid approach is best.

A WEARABLE sensor system (like a fall-detecting pendant or smartwatch) moves with the person. The big advantage is that it can protect the person anywhere they go, including outdoors or across different rooms – you’re not limited by fixed sensor coverage. Wearables are great for people who are active or move around a lot; the sensor is always “on them,” so if they fall in the garden or on the way to the mailbox, it can still trigger an alert. They also typically allow detection of not just the impact of a fall but some pre-impact clues (sudden changes in motion). 

However, wearables can be impractical for some: the person has to remember to wear it and not remove it. Compliance can be especially poor in patients with dementia who may remove or fiddle with devices. Comfort is key – a device that is heavy, ugly, or interferes with daily life (like needing frequent charging) will be quickly abandoned. Moreover, wearables depend on battery power; if a device dies or isn’t charged, the protection is lost. Accuracy can vary with placement – a pendant at the chest might detect falls better than a wrist accelerometer, for example, but each has blind spots for certain types of falls or movements. Finally, some situations (like bathing) are problematic: most electronics shouldn’t get wet, and many people don’t wear pendants or watches in the shower, yet the bathroom is a high-risk location for falls.

A non-contact AMBIENT sensor system avoids the compliance issue – the resident doesn’t have to do anything or wear anything. This can be good for forgetful individuals or those who dislike wearing devices. Ambient sensors (aside from cameras) can also be more discreet – e.g. a radar device mounted on the ceiling, or a pressure mat under the mattress – which the senior might not even notice. Privacy concerns are lower with non-camera ambient sensors; something like a bed sensor under a bed leg is invisible and doesn’t intrude on the resident. 

The downsides include limited range (each sensor usually covers one room or area), so you need multiple sensors for full coverage – which increases cost and complexity. Ambient systems also often cannot provide information outside the home or in transit between rooms. If a senior falls in an unmonitored spot (like a backyard without sensors), the system won’t know. Additionally, ambient sensors might generate more false alarms from environmental factors (pets triggering motion sensors, a heavy object dropping could trigger a vibration sensor, etc.). Acceptance can also be an issue if the person feels like they’re under surveillance – cameras being the prime example, but even motion sensors can be perceived as “Big Brother” by some.

In practice, combining wearable and ambient sensors can yield the best coverage. For instance, a senior might wear a pendant that has a manual call button and basic fall detection, while their room is also equipped with a radar sensor that can detect falls and even monitor gait speed changes. The wearable provides protection when the person is out of their room, and the ambient sensor provides an extra layer when they are in their private space (and maybe forgot to wear the pendant). Redundancy is valuable because if one method misses a fall, another might catch it. Many studies suggest that multi-sensor fusion – integrating data from various sources (wearables, floor sensors, vision, etc.) – can significantly improve both sensitivity and precision of fall detection.

Wired vs. Wireless Systems

Another distinction in fall detection devices is how they transmit alerts: wired or wireless. Originally, many alarm systems were wired into nurse call systems – for example, a bed pressure pad connected via a cable to a wall unit. Now, wireless technology is common, meaning the sensor (bed pad, chair pad, pendant) sends a signal via radio (Bluetooth, Wi-Fi, proprietary RF) to a central receiver or the nurse call system without a physical cable.

Wireless systems offer two big benefits over wired ones:

1. Reduced tripping hazards: There are no cords running across floors or dangling from beds, which is ironic but important – you don’t want your fall prevention device to create a fall hazard! Removing wires makes the environment safer for both residents and staff.

2. Ease and flexibility of installation and movement: With wireless pads or sensors, staff can more easily move the patient or rearrange furniture without unplugging cords. For instance, a wireless bed alarm can stay with a patient if you move them to a different room – just bring the sensor pad and transmitter – whereas a wired alarm might need replugging or might not be compatible elsewhere. Wireless sensors can also be paired to different receivers on the fly (some systems have tap-to-pair features for simplicity), making them more convenient when setting up in a new location. Essentially, wireless tech allows monitoring without being tethered to one spot.

However, wireless devices must be reliable and secure. They can suffer interference or range issues in some buildings. It’s important that they connect into the nurse call or alarm system in a robust way (no dropped signals). Many facilities choose wireless for the fall monitors but integrate them such that alerts still go directly to staff pagers/phones or central systems, ensuring quick response.

(Note: “Wireless” in this context refers to the communication method of the sensor. It is separate from whether the device is battery-powered or plugged in – see next section.)

Power Source: Mains vs. Battery

For any electronic fall prevention device, power loss means lost protection. Devices can be mains-powered (plugged into an electrical outlet) or battery-powered, or sometimes a combination (plugged in with a battery backup).

• Truly wireless devices (like a pendant or a sensor pad with a transmitter) are usually battery-powered. The advantage is they continue to work even if the power in the building goes out (as long as the batteries have charge). They also don’t require an outlet, which gives more freedom in placement (and again avoids cords).

• The downside of battery power is the need for constant monitoring and maintenance of battery levels. Staff or the managing company must routinely check batteries and replace them before they die. A dead battery in a fall detector essentially renders it useless, which can give a false sense of security. There’s also cost and environmental impact over time in replacing batteries (though many devices use long-life batteries that last months or years).

• Mains-powered devices (plugged in) ensure continuous power without battery worries, but of course only work where there’s an outlet and will not function in a power outage unless on a generator or UPS. Many modern systems use a hybrid: for example, a bed alarm unit that plugs into the wall but has a battery backup in case of power failure. Or a camera system that is plugged in but has a short-term battery if it loses power.

From a safety perspective, redundancy is key. If possible, having battery backup on important devices is recommended so that falls are still detected during power failures. Also, alarms should trigger when a battery is low – many wearable call buttons will send a low-battery warning to the base station so that staff know to change it.

Given the limitations of any single device, many experts recommend using a combination of fall technologies to cover all bases. No system is perfect: a wearable might miss a slow slide off a chair that an ambient sensor would catch, or a camera might miss something behind a privacy curtain that a wearable would detect via motion. By layering multiple approaches (wearable + bed alarm + bathroom sensor, for example), the chance of completely missing a fall (false negative) is greatly reduced. Of course, this must be balanced with practical considerations like cost and complexity – one cannot simply install every gadget on every person. But identifying the most appropriate mix for each scenario (maybe a high-risk individual gets more layers of monitoring) is a prudent strategy.

What Is the Best Fall Predictor?

Considering all the above, what can we do to actually prevent falls, not just detect them? The holy grail is to predict and intervene before a fall happens. As discussed, the single best predictor is a previous fall, but that only tells us who is high-risk, not when they will fall. The next best predictor – which we can measure in real time – is gait or motion reduction. Since the majority of falls happen during movement and often involve a decline in mobility or balance, keeping an eye on a person’s activity patterns is key.

Current fall-related technologies are starting to incorporate this insight. Instead of only reacting to a fall or an immediate attempt (like standing up from bed), advanced systems monitor trends in a person’s mobility over days and weeks. For instance, motion sensors throughout an elder’s apartment might track their walking speed in the hallway or their step count each day. If those metrics show a significant drop (e.g. the resident’s weekly average step count plummets, or their walking speed is noticeably slower than last week), the system could flag a fall risk alert. This could prompt a clinical assessment: maybe the person has a new medical issue (like an infection, medication side effect, or pain) that is impairing their gait, and addressing that could prevent a fall. Research like the sensor study by Phillips et al. (described above) gives credence to this approach – catching a gait slowdown early is like catching an arrhythmia before it causes a heart attack.

In summary, alerting on a reduction of motion or gait quality over time may be one of the best solutions for fall prediction given currently available technology. We might not have a perfect “pacemaker” for falls yet, but we do have the ability to continuously monitor certain vital signs of mobility. By leveraging wearables or ambient sensors to watch gait speed, stride, and activity levels, smart care systems can provide early warnings. This moves fall management from purely reactive (alarm after the person hits the floor) to more proactive (notification that John’s mobility has declined 15% this week – check on him before a fall happens).

Of course, human judgment is still crucial. Technology can present the data and alerts, but staff and clinicians must interpret and act on them appropriately. In the end, fall prevention is a combination of personalized care plans and smart use of technology. What works for one patient (e.g. a wearable alarm for an active individual) might not be right for another (who may need discreet room sensors due to dementia and device aversion). It’s important to evaluate each situation and choose the right mix of interventions – whether tech-based or otherwise – to keep our elders safe from falls while preserving their dignity and freedom of movement.

Conclusion and Further Information

While there is no single easy fix for preventing all falls, smart care technology can significantly mitigate fall risk and consequences by detecting falls quickly and, increasingly, by warning of elevated risk before a fall occurs. The key is understanding the capabilities and limits of each solution and applying them in a person-centered way. High sensitivity settings can catch more falls but may burden staff with alarms; high precision reduces false alarms but could miss incidents – so finding the right balance is essential. Remember that falls prevention technology should augment, not replace, good clinical care: regular risk assessments, exercise programs, medication reviews, and environmental safety improvements remain fundamental.

If you’re interested in implementing fall prevention technology, it’s wise to consult with experts and possibly pilot different systems. Consider the environment (hospital, nursing home, private home), the individual’s risk factors (do they wander? have cognitive impairment? etc.), and staff workflows. Proper staff training on any new device is also critical to ensure it’s used effectively.

For more information, you can contact us to help evaluate your fall prevention requirements. We can assist in ensuring you have the appropriate mix of systems in place to avoid falls and major injuries, improve your quality measures (and star ratings), and ultimately provide a safer environment for those in your care. Keeping our older adults safe from falls is a team effort – by combining the best technology with compassionate caregiving, we can make great strides in reducing falls and their impacts.

(Feel free to leave a comment below or reach out for a consultation on fall prevention solutions.)

Contact US

By Hubert van Dalen, Managing Director of eHomeCare, where he advises on the strategies and governance of smart care technologies across the health, aged care, and disability sectors.

---

Sources:

1. Hudson, B. “What You Need To Know about Falls Prevention Technology (Falls Prevention Tech Series, Part 1).” TIDI Products Blog. (2022) – Discusses the role of technology in falls prevention.

2. Hoke, L.M., & Zekany, R.T. “Two sides to every fall: Patient and nurse perspectives.” Crit Care Nurse 40(6):33-41 (2020) – Study of falls in a hospital unit; noted only 1 of 67 falls had an active bed alarm, which was set too low.

3. Oh-Park, M. et al. “Technology utilization in fall prevention.” Am J Phys Med Rehabil. 100:92-99 (2021) – Review of sensor technologies for fall prevention (pressure sensors, infrared, etc.).

4. Harari, Y. et al. “A smartphone-based online system for fall detection…” J NeuroEng Rehabil 18, 124 (2021) – Describes a phone accelerometer fall detector with 73% sensitivity and ~1 false alarm per 46 days (high specificity).

5. Schwickert, L. et al. “Sensitivity and False Alarm Rate of a Fall Sensor in Long-Term Care.” Gerontology 59(4):376-384 (2013) – Found an 80% sensitivity and 0.049 false alarms/hour (~1 per day) in real-world use.

6. Queensland Health. “Stay On Your Feet – Community Good Practice Guidelines.” (2011) – Notes that a previous fall is a strong predictor of future falls.

7. Shao, L. et al. “Incidence and Risk Factors of Falls in Nursing Homes: Systematic Review & Meta-analysis.” J Am Med Dir Assoc. 24(11):1708-17 (2023) – Identifies fall history and ADL impairment as top risk factors; also highlights medications (antidepressants, antipsychotics) and unsteady gait as significant contributors.

8. Phillips, L.J. et al. “Using embedded sensors in independent living to predict gait changes and falls.” West J Nurs Res. 39(1):78-94 (2017) – Found that a 5.1 cm/s decrease in gait speed over 1 week quadrupled fall odds (≈86% fall probability in 3 weeks).

9. Robinovitch, S.N. et al. “Video capture of the circumstances of falls in elderly people in long-term care.” Lancet 381(9860):47-54 (2013) – Video study showing ~41% of falls caused by incorrect weight shifting vs ~20% by trips; many falls occurred during walking, standing, transferring.

10. Global News (Canada). “Videotaping shows most falls among elderly due to incorrect weight-shifting.” (Jan 2013) – Coverage of Robinovitch’s study; emphasizes 70% of falls happened during routine activities (walking, standing, sitting) and most were due to weight-shift imbalance.

11. He, Y. et al. “Pre-impact and impact fall detection based on multimodal sensor.” Intelligent Automation & Soft Computing 36(3) (2023) – Discusses classification of fall detection sensors: vision-based, ambient (infrared, radar), and wearable, with respective pros/cons.

12. TIDI Products White Paper by Saver, C. “Role of Technology in Falls Prevention: A Patient-centered Approach.” (2022) – Provides insights on evaluating fall tech, importance of wireless systems (reducing cords), and balancing tech use with patient needs.