Abstract
BACKGROUND: With the increased ageing of society, more and more elderly people are admitted to the intensive care unit, How to accurately predict whether elderly patients can successfully wean from the ventilator is more complicated. Diaphragmatic excursion (DE) and diaphragm thickening fraction (DTF) were measured by bedside ultrasound to assess diaphragm function. The lung ultrasound score (LUS) and the rapid shallow breathing index (RBSI) were used as indices of diaphragm function to predict the outcome of weaning from mechanical ventilation. The aim of this study was to examine the clinical utility of these parameters in predicting extubation success. METHODS: This prospective study included 101 consecutive elderly patients undergoing a trial of extubation in the ICU of Haidian Hospital between June 2017 and July 2020. Patients were divided into the successful weaning group (n = 69) and the failed weaning group (n = 32). Baseline characteristics, including RSBI, were recorded. Measurements of DE, DTF and LUS were made using ultrasound within 24 h before extubation. RESULTS: Median DE was greater in patients with extubation success than in those with extubation failure (1.64 cm vs. 0.78 cm, p = 0.001). Patients with extubation success had a greater DTF than those with extubation failure (49.48% vs. 27.85%, p = 0.001). The areas under the receiver operating curves for the RSBI, LUS, DE and DFT were 0.680, 0.764, 0.831 and 0.881, respectively. The best cut-off values for predicting successful weaning were DTF ≥ 30%, DE ≥ 1.3 cm, LUS ≤ 11, and RSBI ≤ 102. The specificity of DTF (84%) in predicting weaning outcome was higher than that of RBSI (53%), that of LUS (55%), and that of DE (62%). The sensitivity of DTF (94%) was greater than that of RBSI (85%), that of LUS (71%), and that of DE (65%). The combination of RSBI, LUS, DE, and DTF showed the highest AUC (AUC = 0.919), with a sensitivity of 96% and a specificity of 89%. CONCLUSIONS: DTF has higher sensitivity and specificity for the prediction of successful weaning in elderly patients than the other parameters examined. The combination of RSBI, LUS, DE and DFT performed well in predicting weaning outcome. This has potentially important clinical application and merits further evaluation.