The constant hard work of providing nursing care, including handling emergencies, often causes medical accidents in hospitals. We have prototyped a wearable auto-event-recording system for medical nursing in order to capture the events that cannot be overlooked in analyzing such accidents. Our system consists of wearable sensors that record the footsteps and the posture tilt of the nurse in addition to the voice. We conducted experiments, and the results show that, without disturbing nurses' work, our sensors can record data and reconstruct nursing histories, including the important events that cannot be overlooked for analyzing medical accidents. In this paper, we show the results of sensor data analysis comparing a nurse's data taken in AM hours with those in the PM hours in order to study the effect of her fatigue on her motion. Also, we discuss on the possibility of detecting the nurse's fatigue by using our sensor system.