Objective To summarize and analyze the controller alarm data of 12 patients after implantation of left ventricular assist device（LVAD），so as to provide basis for formulating targeted measures to improve nurses’ response and processing ability to left ventricular assist controller alarm. Methods A total of 12 patients who received implantation of LVAD in the Department of Cardiovascular Surgery of Jiangsu Province Hospital from July 1，2 021 to August 9，2 023 were selected as investigation objects，and their controller alarm types，percentage，causes and nursing measures during ICU were classified and summarized，and then subgroup analysis of "suction alarm" was performed. Results A total of 1 080 alarms were generated during the ICU hospitalization among 12 patients. Single power supply alarm accounted for 76.9%，suction alarm accounted for 19.5%，the others included single battery low capacity and high power alarm. All of patients had a single power supply alarm with 831 times totally，5 people had suction alarm with 211 times totally，4 patients had single battery low capacity alarm with 36 times totally，and two persons had high power alarms with two times. There were no significant differences in preoperative body weight，body surface area，preoperative left ventricular end-diastolic diameter（LVEDD） and blood pump speed between suction alarm group and no suction alarm group，but there is a significant difference in the postoperative left ventricular end-diastolic diameter reduction（LVEDD-R）（P<0.05）. Conclusion The incidence of suction alarm was closely related to the short-term decrease of LVEDD after LVAD implantation. The analysis of alarm data is helpful to grasp the alarm status of LVAD and formulate targeted intervention measures.