Abstract: Objective: To discover the potential associations between the unplanned extubation and other factors in adult inpatients, and to provide reference for the prevention plans of unplanned extubation. Methods: Totally 42 024 items data from 618 patients with unplanned extubation in a tertiary grade A hospital in Hangzhou from January 2016 to December 2020 were collected. Apriori algorithm was used for data mining, chi-square test was used to evaluate the validity of the obtained potential association rules, and the strong association rules were formed combining with the opinions of experts and the results of data analysis. Results: A total of 13 pairs of strong association rules were formed. The high-risk predictors in unplanned extubation in adult inpatients include: male, conscious state, with smoking history, no occupation, no restraint, with low-risk catheter, non-venous catheter, single catheter fixation, night time, the working years of nurse on duty were 1-3 years. Conclusion: More attention should be paid to the male, those with smoking history, non-employment, and unconstrained, and the management of nighttime, low-risk catheters, and non-venous catheters, and training for low-age nurses also are important.

Key words: Unplanned Extubation; inpatients; data mining; association rules