Extubation success can be better predicted by diaphragmatic excursion using ultrasound compared to rapid shallow breathing index

Abstract

INTRODUCTION Mechanical ventilation is one of the cornerstones of ICU care and can change the outcome of critically ill patients.[1] Old age, prolonged ICU stay as well as critical illness myopathy and neuropathy due to the aforementioned factors compounded by the use of steroids or muscle relaxants can contribute to chances of extubation failure.[2] A rapid shallow breathing index (RSBI) of >105 is indicative of a high respiratory frequency and shallow breaths, or reduced tidal volumes, and maybe a predictor of unsuccessful extubation.[3] In a previous study, a change in diaphragmatic thickness of >30% had a positive predictive value of 91% for predicting successful extubation.[345] Our aim was to do a pilot study to compare diaphragmatic excursion by use of a point-of-care ultrasound by an ICU clinician who is proficient in the use of ultrasound to RSBI in predicting extubation success. METHODS Our study was a prospective, observational, comparative cohort study. Institutional Review Board approval was obtained for this study (for the use and recording of the diaphragmatic ultrasound) and informed consent was obtained from the family members and the patient post-extubation. A small predetermined sample size of 20 was used as a pilot to find a positive trend for use in a larger study thereafter. All patients received spontaneous breathing trials before extubation and were assessed for extubation by the ICU team prior to the study. An RSBI of >105 was predictive of poor chances of extubation success, and these patients were automatically excluded as they were not chosen for extubation by the ICU team at that point. Patients who had spinal, neurological, or diaphragmatic injury were excluded as this could hamper their diaphragmatic excursions. All patients could follow breathing commands on nil or minimal sedation in a semi-recumbent position. Extubation success was deemed to be as a lack of need for invasive or noninvasive ventilator support at 48 h. The zone of diaphragmatic apposition to the rib cage at the anterior axillary line in the right 8th intercostal space was used for identifying the diaphragm in all cases. It is technically difficult to visualise the left side due to the presence of the spleen. The patient was encouraged to take three or four vital capacity breaths. Since we were not timing the excursion to the breathing cycle in order for the clinician to complete the study successfully and easily, an average excursion of >1.5 cm at any point in the vital capacity breath was chosen as indicative of predictive of successful extubation. RESULTS Twenty observations on 20 patients were made. The ICU clinician practiced on three patients prior to the start of the study to achieve the learning curve of this strategy of identifying the diaphragm. The average age of the patients was 52 years and ranged from 21 to 89 years. 14 males and 6 female patients were recruited. The average length of stay was 5.5 days ranging from 1 to 14 days. The average RSBI was 40 (SD 17.0170), whilst the average excursion of the diaphragm was 1.98 cm (SD 0.47958). Univariate logistic regressions were performed to examine the effect of diaphragmatic excursion and RSBI on the likelihood of extubation failure. The odds of extubation success increased by 249.8 times (95% CI = {1.33, 46801}, P value = 0.039), and for an increase by 1 in RSBI, the odds of extubation failure increased by 1.008 times (95% CI = [0.94, 1.08], P value = 0.82). As shown in Figures 1 and 2, the modelled predictions of diaphragmatic excursion and RSBI are also compared using the area under the curve (AUC) from the ROC curves. A larger AUC indicates a higher predictive power of model. The diaphragmatic excursion model yields a larger AUC 0.92 (95% CI = [0.77, 1.0]) than the RSBI model with AUC 0.58 (95% CI = [0.27, 0.89]).Figure 1: ROC curve for diaphragmatic excursionFigure 2: ROC curve for RSBIDISCUSSION Evaluating diaphragmatic thickness and change with respiration is operator dependent and can lead to inaccurate results. Such studies have been done by experts in the field of ICU ultrasonography and require a fair deal of expertise. The odds ratio ranges for confidence intervals are wide probably due to the small sample size. From the results of this small pilot study, it seems that diaphragmatic excursion as evaluated by a point-of-care ultrasound is probably better in predicting extubation success than RSBI; however, bigger sample sizes are needed for verifying this result. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest. Acknowledgement We acknowledge the contributions of Ms. Jiexun Wang and Mr Robin Choo for helping us with the statistics.