Aim: Delayed recovery from general anesthesia is a well-known complication that requires predictive tools and approaches. This study aimed to determine significant factors associated with postanesthesia recovery and to develop an algorithm for estimating recovery time from general anesthesia.
Materials & methods: The genotypes of patients were determined by SNaPshot or ARMS-qPCR. The algorithm was developed via machine-learning and tested by the worm plot.
Results: Results showed that OPRM1 rs1799971 (p = 0.006) and ABCG2 rs2231142 (p = 0.041) were significantly associated with recovery time. Ten factors after random forest and stepwise selection were associated with recovery time. Ten factors after random forest and stepwise selection were associated with recovery time. Meanwhile, seven factors were associated with delayed recovery.
Conclusion: This study demonstrated that both clinical and pharmacogenetic data are significantly associated with recovery from general anesthesia and provide the basis for pre-emptive prediction tools.
Keywords: general anesthesia; machine learning; pharmacogenetics; postoperative recovery.