Background: One of the main reasons nurses leave the clinical field is the increasing workload. Researchers have estimated nursing workload by relying on self-reported nursing hours, tasks, and nurse-to-patient ratios. However, directly measuring nursing hours is essential for obtaining an accurate estimate. Objectives: The aims of this study were to measure nursing activities using real-time location information from beacons, smartwatches, and smartphones in the medical unit, and concurrently to collect physiological indicators through wearable devices. Methods: We used a descriptive, cross-sectional design to explore nursing workload in a medical ward based on a partial data from the SMART-NURSE project involving the collection of biometric and location data using smartwatches and Bluetooth beacons, with subsequent integration based on synchronized timestamps. Descriptive statistics, the Kruskal-Wallis test, and Spearman’s rank correlation were used to analyze physical activity indicators such as heart rate, step count, and energy expenditure. Results: A total of 49 work shifts were studied, with 15, 16, and 18 shifts for day, evening, and night, respectively. Nursing hours had significant differences depending on duties with night> day> evening. The average total energy expenditure during the duties was 906.21 kcal ( SD =205.99). The average heart rate was 92.55 ( SD =9.79), and it was significantly correlated with the number of patients’ procedures and the number of discharged patients. The average number of steps were 7343.65 ( SD =1376.62), which was positively correlated with direct nursing and negatively correlated with indirect nursing and nursing stress. Discussion: The nursing hours measured in this study were almost identical to the actual nursing hours, leading us to conclude that this method was successful in measuring nursing hours without any loss. As a result, physical activity data could be used as a fundamental basis for devising a measurement method to measure nurses’ workload and burden.
Kang et al. (2026) studied this question.