Fall detection for hospitals and care homes
Fall detection for hospitals and care homes is the visual detection of a person on the floor, on the corridor and common-area cameras the facility already has, raised to a member of staff within seconds of the fall. A fall in a corridor between rounds is found when someone walks past. Camzify watches the same cameras continuously, sends the frame and a clip to the person designated for that camera, and records who acknowledged it and when.
- A person on the floor, raised in seconds
- On the corridor cameras already installed
- Staff decide and attend, the record is kept
The fall is found at the next round, not when it happens
Hourly rounding is the standard on most wards and in most care homes, and it means a person who goes down in a corridor at ten past the hour can lie there until the next pass. In a memory-care unit or a night shift with two carers for thirty residents, the pass can be a long way off. The injury is often not the fall itself but the time on a hard floor afterwards.
The corridor camera saw the whole thing. It recorded it for the incident review, where it will establish how long the person was down, which is precisely the number the facility wanted to be small.
Nurse call buttons and bed alarms cover the bed. Wearables cover the people who wear them and keep them charged. The corridor, the dining room, the lounge and the garden are covered by cameras that nobody is watching.
“Dock door left unsecured after delivery — confirm and lock.”
Unacknowledged alerts escalate to the backup contact after the configured window.
The camera raises it, a person attends
Slip and fall detection runs on every enabled camera and evaluates each tracked person individually.
- A rapid, uncontrolled change of posture that ends with the person on the floor is the event.
- A person who stays on the floor rather than getting up raises the severity.
- The alert carries the frame and a clip, so the person receiving it sees what happened before they walk.
- The alert is acknowledged by a named person, and the time from detection to acknowledgement is in the record.
Detection runs at every hour. A notification window per camera limits when notifications go out, not when the detection runs, so the day room can notify the carer on the floor by day and the night station by night. The notification settings cover the channels and the escalation.
Between falls, the automated round checks the conditions that lead to them: a wet floor without a sign, a corridor blocked by a trolley, a handrail obstructed. Workplace safety monitoring covers the same detections for staff areas and loading bays.
Slip and fall detection
A tracked person down on the floor, and a person who stays down, raised with the frame and a clip.
DetectionMulti-object tracking
Each person in the corridor followed as one track, so a fall is evaluated on a person rather than on movement in the scene.
DetectionCamera tampering detection
A corridor camera that is covered, turned or failed, raised the same day rather than found after an incident.
DetectionNotifications and alerts
Channels and severity per camera, so the lounge notifies the carer on the floor and the nurse station at once.
- No person on the floorCompliant
- Corridor clear of trolleys and cablesNot compliant
- Wet-floor sign placed where floor is wetCompliant
- Camera view unobstructedCompliant
A failed item is resolved as Fixed or Pending before the round can close.
What a ward round checks
Fall detection is continuous and does not wait for a round. The round adds the conditions: at each stop, the corridor was clear, the floor was dry or signed, the camera could see. Repeated through the shift, it is a record that the environment was checked at each time, which is the record a facility is asked for after a fall.
A failed item messages the person designated for that camera and stays Pending in the report until it is marked Fixed. On an automated round, the AI also raises a critical notification for a person on the floor it sees at a stop, whether or not the checklist asked.
A check found Not Compliant captures a snapshot and messages the guard designated for that camera; on a manual round the operator chooses to send it, on an automated round it goes on its own. The item stays Pending until it is marked Fixed, which captures the after frame, and the report shows both.
Why falls in care settings matter
Published figures from the CDC, AHRQ and The Joint Commission, quoted as the source states them. None of them is a Camzify figure; we publish none.
- 1 in 4
- Over 14 million, or 1 in 4 older adults report falling every year, and the age-adjusted fall death rate rose 21%, from 64.7 per 100,000 older adults in 2018 to 78.4 per 100,000 in 2024.
- Source: CDC, Older Adult Falls Data
- 700,000 to 1 million
- Approximately 700,000 to 1 million patients fall in hospitals in the United States each year, and patient falls are the most common preventable adverse event within hospitals (December 2024).
- Source: AHRQ Patient Safety Network
- 776
- Patient falls were the most frequently reported sentinel event in 2024, with 776 events, a 15% increase from 2023. Of these, 51 falls (7%) resulted in patient death and 503 (65%) in severe harm.
- Source: The Joint Commission, Sentinel Event Data 2024 Annual Review
- 4.5 million
- Each year there are about 4.5 million emergency department visits due to older people falls, with 3.1 million treated and released and 1.4 million hospitalizations.
- Source: CDC, Facts About Older Adult Falls
What it will not do
The limits follow from what it is: a visual detection on the cameras that exist, verified by a person.
- It will not prevent a fall, and it will not assess who is at risk of one; those remain the care plan's job.
- It will not see a fall in a bedroom or a bathroom without a camera, and most facilities rightly have none there.
- It will not attend; it tells the person designated for that camera, and the response is theirs.
- It will produce some alerts that a person, looking at the clip, will dismiss, and the workflow is built for that.
- It will not identify anyone, and it does not link a clip to a patient record.
We do not publish detection rates, false-alert rates or response times for fall detection, because we cannot verify them for your corridors. The trust page sets out the policy.
Industries where this applies
Frequently asked questions
Ready to patrol your site 24/7?
Book a 15-minute demo and see a live patrol run on your own cameras.