Diagnostic accuracy of radiology (CT, X-ray, US) for predicting difficult intubation in adults: A meta-analysis.

OBJECTIVE: The aim of this study was to evaluate the overall accuracy of radiological measurements in prediction of difficult airway and compare the diagnostic value between the radiological measurements and the modified Mallampati score through a meta-analysis of published studies.

METHODS: A comprehensive electronic search of related literature was performed in PubMed, Embase, Cochrane Library and China National Knowledge Infrastructure. Meta-DiSc 1.4 and STATA 12.0 were selected for data analysis, and QUADAS-2 tool was used to assess the quality of included studies. Difficult airway was defined as Cormack-Lehane III-IV. Data from selected studies were pooled to yield summary sensitivity, specificity, positive and negative likelihood ratios, diagnostic odds ratio, as well as summary receiver operating characteristic curve.

RESULTS: A total of 17 studies dating up to November 2017 with 8779 individuals were enrolled in the present study. Heterogeneity existed in the non-threshold effect, but not in the threshold effect. Subgroup analyses based on radiological methods were conducted. The pooled diagnostic characteristics in the computed tomography subgroup were as follows: sensitivity 0.75 (95%CI, 0.64-0.84), specificity 0.75 (95%CI 0.68-0.81), PLR 3.19 (95%CI 1.91-5.32), NLR 0.38 (95%CI 0.23-0.64), DOR 11.74 (95% CI, 4.19-32.86) and AUC 0.8424 with Q* index 0.7741. In the X-ray subgroup, the sensitivity was 0.78 (95%CI, 0.73-0.82), the specificity was 0.88 (95%CI, 0.87-0.89), PLR was 5.03 (95%CI, 2.44-10.37), NLR was 0.27 (95%CI, 0.22-0.33), DOR was 23.18 (95%CI, 8.81-60.95) and AUC was 0.8970 with Q* index 0.8280. The corresponding values for the ultrasound subgroup were 0.69 (95%CI, 0.63-0.74) for sensitivity, 0.84 (95%CI, 0.82-0.85) for specificity, 6.25 (95%CI, 3.81-10.27) for PLR, 0.36 (95%CI, 0.27-0.47) for NLR, 22.26 (95%CI, 10.45-47.41) for DOR, 0.8942 for AUC with Q* index 0.8251. The pooled sensitivity, specificity and PLR of the modified Mallampati score were 0.61 (95%CI 0.56-0.66), 0.63 (95%CI 0.61-0.64) and 2.11 (95%CI 1.71-2.61) which were significantly lower than that of radiographic methods.

CONCLUSIONS: The results indicated that the diagnostic value of CT, X-ray and US was much better than that of modified Mallampati score. Ultrasound had diagnostic indices and the area under curve similar to those of CT and X-ray in predicting difficult airway. Considering being easy, readily availability, low cost, and free from radiological hazards, it can be considered as prior diagnostic strategy in this condition.