Risk Factors Analysis And Predictive Model Of Postextubation Swallowing Dysfunction In ICU Patients

Research Purpose:The purpose of this study was to identify the risk factors of swallowing disorders after extubation in ICU patients with endotracheal intubation and to establish a predictive model of dysphagia after extubation.According to the model,the patients with high risk of dysphagia were treated with swallowing rehabilitation training during endotracheal intubation to promote the recovery of swallowing function.In order to reduce the incidence of complications such as aspiration and aspiration pneumonia.Research Methods:During December 01,2018 to December 31,2019,316 patients who met the criteria for admission were recruited in intensive care unit of a top three hospital in Huzhou,70%of patients（n=222）were model cohorts.Group patients based on the screening results of swallowing disorders.General demographic data and clinical information of the patients were collected.General demographic data included age,gender,occupation,education,marital status,BMI,smoking history,drinking history and living conditions.Clinical data included diabetes history,history of hypertension,heart disease,lung disease,kidney disease.Duration of intubation,type of endotracheal intubation.Time to stop sedation,gastroscopy,APACHEⅡscore,B-type natriuretic peptide,whether for postoperative patient,CPOT score,days of the indwelling gastric tube,heart rate during extubation,systolic pressure during extubation,diastolic pressure during extubation,breathing during extubation,SPO₂ during extubation.EpiData 3.1 was used for data input and SPSS24.0 was used for statistical analysis.The normal measurement data were expressed by mean±standard deviation,and t-test was used for comparison between groups.Non-normal measurement data were expressed by median and quartile,and non-parametric test was used for inter-group comparison.The counting data was expressed as a percentage（%）and chi-square test was used for the comparison between groups.Binary Logistic regression analysis was used to determine the risk factors of dysphagia after extubation in patients with endotracheal intubation,and a regression prediction model was established to verify the effectiveness of the prediction model.Mini Programs,which is convenient for clinical use,was developed to predict the risk of PSD.Research Results:In the model cohort of this study,103 patients were positive screening results for dysphagia.Univariate analysis showed that there were 7 factors related to dysphagia,including age,living condition,lung disease,duration of intubation,CPOT score,days of indwelling gastric tube and heart rate during extubation.There was no statistical difference in other factors between the two groups.SPO₂ during extubation was also taken as a candidate factor when the P value was>0.05but<0.1,and the APACHEⅡscore which was considered important by clinical experience was added.The results of binary Logistic regression analysis showed that elderly patients（>70 years old）,long-time intubation（>48h）,high CPOT score（≥2 points）and elevated heart rate during extubation（>110bpm/min）were independent risk factors of PSD.The regression equation was:Logit（P）=-1.959+0.661*elderly patients+0.918*long-time intubation+0.398*CPOT（1）+1.921*CPOT（2）+1.201*elevated heart rate at extubation.According to this,the ROC curve was made,the area under the curve was 0.75,the sensitivity was 70.87%,the specificity was 68.06%,and the prediction accuracy was 73.36%.The Hosmer-Lemeshow test showed that the goodness of fit of the model was good（?²=1.863,P=0.932）.A prediction model Mini Program was developed according to the regression equation.Medical staff only need to enter the patient’s age,intubation time,CPOT score and heart rate during extubation,and then click"calculate"to directly obtain the probability value of PSD.When the probability value exceeds 48%,the patient is judged to be a patient with high risk of PSD.Research Conclusion:1.The independent risk factors of PSD were elderly patients（>70 years old）,long-time intubation（>48h）,high CPOT score（≥2points）and elevated heart rate during extubation（>110beats/min）.In clinical work,medical staff should pay attention to the patients with the above risk factors,evaluate their swallowing function in time,and give them appropriate nutrition programs.2.Logistic regression equation can predict the probability of PSD,quantify the risk of dysphagia,and develop Mini Program,which is convenient for clinical use.The area under the ROC curve is 0.75,which shows that the prediction efficiency of the model is good,the sensitivity is 70.87%,the specificity is 68.06%,and the prediction accuracy is 73.36%.Hosmer-Lemeshow test P=0.932,indicating that the model fits well.A small program developed based on the model is convenient for clinical doctors and nurses to predict the risk of PSD during intubation or after extubation,so as to take intervention measures as soon as possible.