Do conventional predictors of “Difficult airway” truly predict difficult airway? Experience with cleft surgeries

Background: We designed this prospective observational study in order to evaluate whether conventional predictors of “Difficult airway” truly predict difficult airway in patients who present for cleft surgeries at Cleft Centre Peradeniya. Methods: One hundred and two babies between the ages of 2 months to 18 months who underwent cleft lip or palate repairs at Cleft centre, Peradeniya August 2015 to April 2016 were included in the study. For each of the patients, we gave a grade for the degree of mouth opening and the laryngoscopic view prior to surgery. Results: Out of the 102 participants, 9.8% of patients had difficult laryngoscopy. At least one out of seventeen studied “predictors of difficult airway,” was found in seven patients out of the 10 in whom laryngoscopy was difficult. There was a significant association of difficult laryngoscopy with reduced mouth opening. Furthermore, factors such as the presence of microstomia, micrognathia, retrognathia, and short neck were significantly associated with difficult laryngoscopy in Cleft Lip and Cleft palate patients who underwent surgical repair.


INTRODUCTION
Cleft deformities are considered the most common craniofacial abnormality, [1,2] with the incidence being 1 in 600 to 700 live births in the world [3,4] and cleft palate alone being 1in 2000 live births. [4] In Sri Lanka, cleft abnormalities occur at a rate of 0.83 per 1000 live births. [5] Generally, cleft lip repair and cleft palate repair are carried out at 4-5 months and 9-12 months respectively. [6] Establishing airway is always a challenge in paediatric anaesthetic practice. Unique differences in the anatomy of neonatal and paediatric airways make airway access more technically difficult than in an adult. [7] Difficult airway access leading to delays in establishing the airway together with a higher oxygen consumption rate and limited body oxygen reserves due to a lower functional residual capacity makes a paediatric patient vulnerable for hypoxia during induction of anaesthesia. The smaller the child the greater the differences in the airway compared to an adult. Small dimensions of the airway, anteriorly and higher placed larynx guarded by a long and intrinsically floppy epiglottis with a short neck, disproportionately large tongue with a small mandible and narrow cricoid makes pediatric airway management very difficult. Moreover, the presence of congenital airway abnormalities and craniofacial anomalies further complicate pediatric airway management. [7] Furthermore, multiple tracheal intubation attempts, with inadequate oxygenation in-between the attempts ending in a failed or delayed intubation have resulted in adverse consequences such as hypoxemic brain damage, cardiac arrests, and even deaths. [8,9,10,11]  in patients with retrognathia. They reported a significant positive association with the difficult laryngoscopic view when the age of the patient was less than 6 months. Tracheal intubation was reported as being successful in 99% of the sample and the difficulty with intubation was reported to have been overcome by the administration of muscle relaxant drugs. As this study was performed more than 25 years ago and the anaesthetic techniques, patient monitoring and drugs used have made a significant advancement since then, we felt that it was time to perform this study.
Therefore, we designed the present study in order to evaluate whether conventional predictors of "Difficult airway" truly predict difficult airway in patients who present for cleft surgeries at Cleft Centre Peradeniya. Anaesthesia was induced and maintained with Sevoflurane and oxygen. Laryngoscopy was performed by an anaesthetist using a laryngoscope with a size 2 mackintosh-curved blade after achieving muscle paralysis with intravenous Atracurium 0.5mg/kg body weight. The laryngeal view obtained was classified according to 'Cormack and Lehane' classification. Grade I is a full view of the glottis, grade II is a partial view of the glottis and full view of the epiglottis, grade III is no part of the glottis visible and only the epiglottis is visible and grade IV is a visualization of only the soft palate with no view of the glottis or the epiglottis. Results of Laryngoscopic views were classified as "easy", for laryngoscopic views of grade I and grade II and as "difficult", for laryngoscopic views of grade III and IV.

METHODOLOGY
Grade I -Full view of the glottis Grade II -Full view of the epiglottis and only partial view of the glottis Grade III -No part of the glottis is visible, only the epiglottis is visible Grade IV-neither glottis nor epiglottis is seen, only the soft palate is seen The laryngoscopic view was assessed just before intubation. An attempt was made, therefore, to assess the degree of correlation between the degree of mouth opening during the preoperative assessment and the laryngoscopic view grades observed during anaesthesia.
Using G-power statistical software, minimum total sample size of 60 was obtained (Effect size-0.2,  significance level-0.05 and the power of the study -80% were used). Assuming 15% of dropout rate for the study, the final sample size is adjusted to achieve the target statistical significance given by the hypothesis.
Data were analyzed using SPSS version 22 for windows. Numerical data were expressed as mean ± SD while categorical data was expressed as frequencies.   We studied 17 conventional predictors of difficult laryngoscopy in the sample of 102 patients who presented for surgery. Figure 2 indicates the factors which were tested and found to be present in patients with difficult laryngoscopy and intubation. They included microstomia, facial asymmetry, short neck, micrognathia, and retrognathia. one out of the five features mentioned above was found in 7 out of 10 patients in whom laryngoscopy was difficult (grades III or IV). It was also found that difficult laryngoscopy was significantly associated with poor mouth opening (Figure 3). When each of the difficult airway predictors were studied separately in the total sample of 102 patients, it was possible to do a Fishers exact test to see whether each of these predictors had a significant association with difficult laryngoscopy. Figure 6 shows that four out of the five predictors which were present in our sample had a significant association with difficult laryngoscopy. They were namely microstomia, micrognathia, retrognathia, and the presence of a short neck.

DISCUSSION
We attempted to find out whether the conventional predictors of difficult airway truly predicted the difficulty in intubation in patients who presented for clef repair surgeries. We studied 102 patients with 47 of them presenting for cleft lip repair surgeries while the rest were for cleft palate repairs. We found that at least one predictor of difficult airway occurred in seven patients out of 10 in whom laryngoscopic view was grade III or IV hence referred to as "Difficult Laryngoscopy". Evaluating some of the predictors of difficult airways in paediatric patients was extremely difficult, as the patients cannot follow instructions given by an investigator.eg Mallampati test or Delilkan's sign. Moreover, even simple mouth opening and tongue protrusion were very difficult with patients of this age group. In some patients, an observer had to wait for a long time till a child opened the mouth, or had to observe it when a child cried. However, these signs/parameters have not been validated properly for paediatric patients in previous studies.
The Mallampati test gives an idea about the degree of mouth opening and the ability to view the back of the pharynx. The mouth opening test we did, was also to get an idea about how easy or difficult it would be to perform laryngoscopy and intubation. The degree of mouth opening was put on a scale of grades 1-4, similar to that of Mallampati.
One drawback of the present study is that our sample is considerably smaller than the sample used by Gunawardane [13]. Therefore, we should be cautious in drawing definite conclusions regarding the importance of other predictive factors and difficulty in performing laryngoscopy/intubation in cleft lip and palate surgeries.

CONCLUSION
Among the 102 participants, 10 patients had 'Difficult Laryngoscopy (Laryngoscopic view grades 3 and 4), hence 9.8% of patients have had difficult laryngoscopy. At least one out of seventeen studied "predictors of difficult airway," was found in seven patients out of 10 in whom laryngoscopic view was grade 3 or 4 hence in expected Difficult Laryngoscopy. There was a significant association of the presence of microstomia, micrognathia, retrognathia, and short neck with difficult laryngoscopy in our sample. We also found that there was a significant association between the degree of mouth opening and the occurrence of difficult airway in Cleft Lip and Cleft palate patients.
Author declaration