Development of a GSM-Based Patient Fall Detection Alert System for Geriatric Wards in Nigerian Teaching Hospitals

Patient falls in geriatric wards represent a significant preventable safety incident in Nigerian teaching hospitals, and automated fall detection systems can reduce response time and injury severity in elderly inpatients. This study develops a GSM-based patient fall detection alert system suitable for geriatric ward deployment in Nigerian teaching hospitals. The system uses an MPU6050 accelerometer and gyroscope sensor module worn on the patient's waist, interfaced with an Arduino Nano microcontroller for real-time motion data acquisition. A threshold-based fall detection algorithm was programmed based on the combined acceleration magnitude exceeding 2g followed by a post-fall inactivity period of five seconds, discriminating falls from normal daily activities. Upon fall detection, the Arduino triggers a SIM800L GSM module to send an SMS alert containing the patient's identity and ward location to a designated nursing station phone number. The device is housed in a lightweight 3D-printed enclosure with a rechargeable 500 mAh lithium polymer battery. Algorithm performance was evaluated using a controlled fall simulation study with twenty adult volunteers on padded surfaces, testing seven fall scenario types and five activities of daily living. Fall detection sensitivity was 91.4 percent and specificity 94.0 percent across the test scenarios. False alarm rate was 6.0 percent during activities of daily living simulation. Battery life was sufficient for twelve hours of continuous monitoring. The study recommends clinical validation on actual geriatric inpatients, integration with nurse call systems, and ergonomic assessment of device wearability by elderly patients. Keywords: fall detection, accelerometer, geriatric ward, GSM alert, patient safety.