Hui-Xian Li

Determining predictive value of preoperative tests for difficult intubation

In the recent article by Mahmoodpoor et al. [1] evaluating the predictive value of several preoperative tests for difficult intubation in a prospective descriptive study, they conclude that facial angle has a high sensitivity, positive and negative predictive value, and Youden index for prediction of difficult intubation, but the best result is achieved when facial angle is used in combination with either the modified Mallampati score or upper lip bit test. As difficult intubation prediction by preoperative tests is a crucial component of safe airway management algorithm [2], their findings have potential implications. However, several aspects of this study would have made interpretation of their findings questionable. First, the main aim of this study was to evaluate the predictive value of preoperative tests for difficult intubation, which was defined as the Cormack and Lehane grades 3 and 4 laryngoscopic views. According to the latest standards of difficult airways by the American Society of Anesthesiologists Task Force on Management of the Difficult Airway [2], however, the Cormack and Lehane grades 3 and 4 laryngoscopic views should actually be defined as difficult laryngoscopy, rather than difficult intubation. It must be emphasized that the laryngeal view obtained by direct laryngoscopy is often used as a primary variable for difficult or failed intubation, but they are not synonymous in the majority of patients [3]. Tracheal intubation depends more on the skill of the intubator than does laryngoscopy, and therefore the degrees of difficulty with direct laryngoscopy and tracheal intubation may be inconsistent. For example, some patients with a Cormack and Lehane grade 3 or 4 laryngoscopic view may be intubated by an experienced anesthesiologist at the first or second attempt if the distal end of the tracheal tube is appropriately curved by a malleable stylet or a introducer (e.g., a gum elastic bougie) is used [4]. As this study did not evaluate and report the occurrence of difficult intubation, we argue that their conclusions should be modified. Second, the laryngoscopic views using the Cormack-Lehane classification was stated as the primary endpoint. Besides the patients head was placed in the “sniffing” position, the readers were not provided with the experience of intubator and the use of external laryngeal manipulation during laryngoscopy. Thus, it was unclear whether an optimal laryngoscopy attempt was carried out in each patient. To obtain the optimal laryngoscopy attempt, a reasonably experienced intubator who has had at least 3

Current evidence for the use of C-MAC videolaryngoscope in adult airway management: a review of the literature

The C-MAC videolaryngoscope is the first Macintosh-typed videolaryngoscope. Since the advent of its original version video Macintosh system in 1999, this device has been modified several times. A unique feature of C-MAC device is its ability to provide the 2 options of direct and video laryngoscopy with the same device. The available evidence shows that in patients with normal airways, C-MAC videolaryngoscope compared with direct laryngoscopy can provide comparable or better laryngeal views and exerts less force on maxillary incisors, but does not offer conclusive benefits with regard to intubation time, intubation success, number of intubation attempts, the use of adjuncts, and hemodynamic responses to intubation. In patients with predicted or known difficult airways, C-MAC videolaryngoscope can achieve a better laryngeal view, a higher intubation success rate and a shorter intubation time than direct laryngoscopy. Furthermore, the option to perform direct and video laryngoscopy with the same device makes C-MAC videolaryngoscope exceptionally useful for emergency intubation. In addition, the C-MAC videolaryngoscope is a very good tool for tracheal intubation teaching. However, tracheal intubation with C-MAC videolaryngoscope may occasionally fail and introduction of C-MAC videolaryngoscope in clinical practice must be accompanied by formal training programs in normal and difficult airway managements.

Current Evidences for the Use of UEscope in Airway Management

Objective: UEscope is a new angulated videolaryngoscope (VL). This review aimed to describe the features of UEscope and provide clinical evidences regarding the efficacy and safety of this video device in adult tracheal intubation and its roles in airway management teaching. Data Sources: The Wan Fang Data, CNKI, PubMed, Embase, Cochrane Library, and Google Scholar were searched for relevant English and Chinese articles published up to January 15, 2017, using the following keywords: “HC video laryngoscope”, “UE videolaryngoscope”, “video laryngoscope”, and “videolaryngoscopy”. Study Selection: Human case reports, case series, observable studies, and randomized controlled clinical trials were included in our search. The results of these studies and their reference lists were cross-referenced to identify a common theme. Results: UEscope features the low-profile portable design, intermediate blade curvatures, all-angle adjustable monitor, effective anti-fog mechanisms, and built-in video recording function. During the past 5 years, there have been a number of clinical studies assessing the application and roles of UEscope in airway management and education. As compared with direct laryngoscope, UEscope improves laryngeal visualization, decreases intubation time (IT), and increases intubation success rate in adult patients with normal and difficult airways. These findings are somewhat different from the previous results regarding the other angulated VLs; they can provide an improved laryngeal view, but no conclusive benefits with regard to IT and intubation success rate. Furthermore, UEscope has extensively been used for intubation teaching and shown a number of advantages. Conclusions: UEscope can be used as a primary intubation tool and may provide more benefits than other VLs in patients with normal and difficult airways. However, more studies with large sample are still needed to address some open questions about clinical performance of this new VL.

Assessing Effects of Preoperative Anemia on Adverse Outcomes After Coronary Surgery

To the Editor, The recent article by Tauriainen et al [1] showed that after adjusting important baseline characteristics, operative factors, severity of perioperative bleeding and amount of transfused blood products by propensity score matching, regression and Cox proportional hazards analyses, preoperative anemia was not associated with an increased risk of short and late mortality. Other than the limitations described in discussion, we would like to make some comments on this study. First, in a retrospective study, the propensity score matching is indeed useful for adjusting the patients’ baseline characteristic and controlling selection biases. Nevertheless, an important limitation of propensity score matching is that the analysis is necessarily restricted to the matched sets [2]. In this study, to generate two propensityscore-matched groups, more than 70% of patients without preoperative anemia were excluded, while only about 14% of patient with preoperative anemia were excluded. We are concerned that loss of most non-anemic patients may shift the estimation target as the effect of non-anemia on the postoperative outcomes may no longer represent the effect for all non-anemic patients. Thus, findings of this study may only show the relation between preoperative anemia and postoperative short and late mortality in matched patients, but may not be representative of the full sample. Second, no matter what propensity score matching or regression analysis is used for risk adjustment, the statistical models carry the significant assumption of no unmeasured confounders; namely, all important known risk factors affecting the measured outcomes must be measured and taken into account within the model [2]. Apparently, this is an unrealistic assumption for study of Tauriainen et al, as many known perioperative risk factors affecting postoperative mortality of patients undergoing cardiac surgery were not considered into the models when propensity score matching and regression analysis were used for risk adjustments, for instance, preoperative hypoproteinemia, perioperative cardiac medications, intraoperative hemodynamic instability, postoperative anemia, delirium, respiratory failure, dialysis, sepsis, gastrointestinal complications [3, 4]. That is, even if after risk adjustments, many unmeasured confounders still exist and there may remain imbalance in unmeasured confounders between groups in this study. Thus, we argue that imbalance in unmeasured confounders would have biased the true effect of preoperative anemia on the adverse outcomes after CABG surgery or even resulted in a spurious association between preoperative anemia and postoperative mortality. Third, this study assessed the effects of preoperative anemia on short and late mortality. In fact, there are important differences in mortality risk factors between the early and late periods after CABG surgery, i.e., the majority of predictors of early mortality are cardiac-related variables, whereas the majority of predictors of late mortality are noncardiac-related variables and late mortality is mainly attributable to many causes, not necessarily related to patients’ cardiovascular and general health before surgery [5]. Thus, adjusting same preoperative and intraoperative confounders for short and late mortality in this study may be arbitrary and can confuse the interpretation of results. & Fu-Shan Xue xuefushan@aliyun.com

Association between intraoperative blood product transfusions and acute kidney injury following cardiac surgery

With great interest, we read the recent article by Kindzelski et al.1 assessing the association between intraoperative blood product transfusions and acute kidney injury (AKI) following cardiac surgery. By binary logistic regression analysis, they show that intraoperative blood product transfusions are independently associated with an increased risk of postoperative AKI. Furthermore, there is a stepwise increase in the probability of postoperative 30-day mortality with escalating AKI severity. In this study, the authors had used appropriate statistical methods to evaluate the influence of intraoperative blood transfusion on the occurrence of AKI and determine the association between postoperative AKI stages and mortality. However, this study is a retrospective analysis, which potentially introduces a number of confounders. Other than the limitations described in the discussion, we note that several issues of this study were not well addressed. First, when determining the association of intraoperative blood transfusion with postoperative AKI stages by logistic regression analysis, intraoperative variables used for risk adjustments only included type of surgery (CABG, valvular, both) and cross-clamp time, but not other important intraoperative variables associated with AKI, especially for hemoglobin levels during cardiopulmonary bypass. It has been shown that hemodilution anemia during cardiopulmonary bypass is independently associated with AKI after cardiac surgery, especially when prolonged cardiopulmonary bypass time and intraoperative transfusion are needed.2 Furthermore, the simultaneous occurrence of both hemodilution anemia and hypotension during cardiopulmonary bypass can synergistically act to increase the risk of AKI after cardiac surgery.3 We are concerned that the neglect of important intraoperative factors for risk adjustments would have biased the true contribution of intraoperative blood transfusion to the occurrence of postoperative AKI. Second, this study only focused on the influence of intraoperative blood transfusion on the development of postoperative AKI. The available evidence shows that blood transfusion after cardiac surgery is also an independent risk factor of postoperative AKI.4,5 Finally, when evaluating the association of escalating AKI severity with postoperative 30-day mortality by logistic regression analysis, postoperative anemia and blood transfusion were not included in the model for risk adjustment. Similarly, postoperative blood transfusion has also been associated independently with the increased risk of adverse events and mortality after cardiac surgery.6 Furthermore, postoperative anemia is common and frequently persists for months after cardiac surgery. When the postoperative hemoglobin level is considered as a continuous variable, every 1 mg/dl decrease in hemoglobin level is associated with a 13% increase in adverse cardiovascular events and a 22% increase in all-cause mortality.7 Thus, not taking postoperative anemia and blood transfusion into account would have tampered with the inference of the logistic regression model for adjusted impacts of AKI stages on postoperative 30-day mortality.

Effects of enteral different-dose levothyroxinesodium pretreatment on serum thyroid hormone levels and myocardial ischemia-reperfusion injury

Introduction: The available evidence shows that perioperative oral thyroid hormone can significantly attenuate the postoperative decline in the serum hormone level and improve postoperative hemodynamic and prognostic parameters. However, there has been no study assessing the effects of preoperative oral different-dose thyroid hormone on serum hormone levels and myocardial ischemia-reperfusion injury (IRI) after cardiac surgery. Methods: Forty-eight healthy Wistar rats, aged 35 days, were randomly allocated into six groups: Group BC, Group C and four pretreatment groups in which the rats were given levothyroxine-sodium of 10 μg, 20 μg, 40 μg and 80 μg/100 g. On the eighth day, the serum thyroid hormone levels were determined and then an isolated heart ischemia-reperfusion model was established with a Langendorff apparatus. Results: Compared with Groups BC and C, serum thyroid hormone levels on the eighth day did not significantly change in Group 10 μg, but were significantly increased in Groups 20 μg, 40 μg and 80 μg. The cardiac enzyme myocardial-bound creatine kinase levels in the coronary effluent during reperfusion were significantly lower in Groups 10 μg and 20 μg and 40 μg than in Group C. The recovery rates of +dp/dtmax and -dp/dtmax at 30 min during reperfusion were significantly lower in Groups 40 μg and 80 μg than in Groups 10 μg and 20 μg. Compared with Group C, myocardial expressions of heat shock protein 70 and myosin heavy chain α were increased in the four experiment groups and myocardial expression of thyroid hormone receptor α1 was significantly increased in Groups 20 μg, 40 μg and 80 μg. Conclusions: The pretreatment with enterally smaller doses levothyroxine-sodium does not significantly affect serum thyroid hormone levels and produces protection against myocardial IRI, whereas pretreatment with enterally larger doses of levothyroxine-sodium can only provide an attenuated or insignificant cardioprotection because of hyperthyroxinemia. Cardioprotection by levothyroxine-sodium pretreatment is probably attributable to increased myocardial expression of heat shock protein 70 and myosin heavy chain α.

Use of early postoperative serum creatinine elevation to identify acute kidney injury after cardiac surgery

To the Editor, We read with interest the article by Karkouti et al. on assessing the use of early postoperative creatinine elevation to identify acute kidney injury (AKI) after cardiac surgery. The AKI was defined based on the consensus recommendations of the AKI Network (AKIN) for serum creatinine (sCr) criteria. The early postoperative sCr change (DCreat) – denoted as the ratio of postoperative sCr (measured within 30 min of arrival in the intensive care unit)/preoperative sCr – was used as a primary endpoint. It was unclear, however, whether the sCr measured during the early postoperative period had been corrected based on the perioperative fluid balance. Available evidence suggests that not adjusting sCr for the fluid balance could underestimate the true incidence of AKI after cardiac surgery because a positive perioperative fluid balance could dilute sCr. Furthermore, in Karkouti et al.’s study, the threshold of DCreat [ 1.3 was used to identify postoperative AKI. In contrast, Ho et al. demonstrated that a smaller early postoperative sCr change (i.e., C 10% sCr increase) could significantly predict a higher AKI risk, and a C 10% decrease could significantly predict a lower AKI risk compared with the reference category. In a sensitivity analysis, Karkouti et al. evaluated the relation of the DCreat with AKI stage 1 in the development and validation sets but not the relations of the DCreat with AKI stages 2 and 3. It has been shown that AKI stages 2 and 3 incidences are lower than that of AKI stage 1 and are more significantly associated with shortand long-term adverse outcomes after cardiac surgery. It is thus generally believed that the need for a tool to predict the more severe categories of AKI is important for the patient’s prognosis and long-term outcomes after cardiac surgery. Thus, we argue that the predictives value of the DCreat for AKI stages 2 and 3 also require further validation in the development and validation sets. Finally, several models are available for predicting AKI following cardiac surgery that are based on readilyavailable clinical information before and during surgery. They can be easily applied at the bedside and provide a simple, interpretable estimation of risk. Furthermore, the best-validated models can effectively predict severe AKI that would require dialysis following cardiac surgery, such as the Cleveland Clinic model and the Mehta model. In Karkouti et al.’s study, the sensitivity, specificity, and positive and negative predictive values for using a threshold ratio of DCreat [ 1.3 for predicting postoperative AKI in the development set were 20% (95% confidence interval [CI], 16 to 24), 99% (95% CI, 99 to 99), 68% (95% CI, 59 to 76), and 93% (95% CI, 92 to 93), respectively. As Karkouti et al.’s study design did not include evaluation of the discrimination ability of DCreat compared with the best-validated clinical models to identify the patients who suffer AKI, an important question remains unanswered: Does the information gained from the DCreat facilitate early identification of AKI after cardiac surgery as well as, or better than the clinical models? This letter is accompanied by a reply. Please see Can J Anesth 2018; 65: this issue.

Evaluating LMA-ProSeal insertion techniques: A call for methodological clarification

In the recent article by Ishio et al. [1] evaluating the insertion efficacy of the LMA-ProSeal by novice doctors with or without a 90° bend created by an intubating stylet in anesthetized patients, they demonstrate that a 90° bend re-figuration facilitates LMA-ProSeal insertion, as shown by a higher successful insertion rate, a higher sealing pressure, and a lower subjective difficulty of insertion. To rightly compare effect of insertion techniques on performance of the studied device in a randomized controlled trial, however, standardization of equipment and rational design of study are crucial. In our opinion, there are several issues in their study that would have made interpretation of their findings questionable. First, in method, the authors did not specify the LMA-ProSeal insertion procedure, particularly at the control group. It must be emphasized that standardization of equipment is associated with improved performance. Given that all of 15 novice doctors participated in this study only had limited experience in the use of LMAProSeal, we would like to knowwhether they received formal teaching and training about the use of this device before this study. According to the manufacturers instructions, the LMA-ProSeal insertion can be performed with a digital method or a metal introducer technique [2]. When using a digital method, the doctors index finger is placed in the retaining strap of LMA-ProSeal: this is made easier by lateral compression of the body of the mask to bow the strap outward. When using a metal introducer, it distal end is located in the retaining strap of LMA-ProSeal and its proximal end is placed in the notch between airway tube and drain tube. Furthermore, what should be noted is that insertion of the LMA-ProSeal using a digital method is similar to the classic LMA, whereas the insertion with an introducer technique resembles the intubating LMA-Fastrach. We are concerned that no pre-study teaching and training would have resulted in a poor insertion technique of the LMA-ProSeal and then influence the final end-point of performance. Second, as to the Fig. 2 provided in this article, the shape of the LMA-ProSeal created by an intubating stylet was similar to that made by a metal introducer. Thus, a rational design of control group would be that the LMA-ProSeal was inserted with using a metal introducer, as performed in previous study [3]. Third, in this study, the cuff of the LMA-ProSeal was inflated to a pressure of 30 cm H2O with a pressure transducer. As to the manufacturers instructions, inflation to an intracuff pressure of 60 cm H2O is preferred for the LMA-ProSeal [2]. It has been shown that cuff inflation volume and intracuff pressure can significantly change airway sealing pressure of the LMA-ProSeal [4].

Use of dexmedetomidine to attenuate acute kidney injury after cardiac surgery

By a randomized clinical trial, Zhai and colleagues [1] showed that dexmedetomidine may attenuate postoperative acute kidney injury (AKI) and decrease the incidence of AKI after cardiac valve replacement under cardiopulmonary bypass (CPB). Many things of this study were done correctly and we had learned from their example. To differentiate the effects of one factor on primary study endpoint in a randomized clinical trial, however, all of other factors have to be standardized for avoidance of potential biases. In our view, there are several important issues in this study that are not well addressed. First, AKI was defined only by increases in serum creatinine (sCr) levels within postoperative 3 days according to the RIFLE criteria. To exclude effect of acute sCr fluctuations on diagnosis of postoperative AKI, the RIFLE criteria actually requires a 7-day time window, rather than a 3-day time window. The postoperative urine output was measured in this study, but it was not used for diagnosis of AKI. In fact, the RIFLE criteria require the urine output criteria. Furthermore, adding urine output to sCr criteria considerably increases the apparent incidence of AKI after major surgery and significantly changes the prognostic implications of AKI identification and staging [2]. In addition, we would like to know whether the sCr levels used for diagnosis of postoperative AKI in this study had been corrected based on perioperative fluid balance. The recent evidence indicates that not adjusting sCr levels for fluid balance may underestimate incidence of AKI after cardiac surgery as a positive perioperative fluid balance may dilute sCr [3]. We are concerned that these unknown factors would have confused the AKI identification in this study. Second, as a sensitive biomarker for early diagnosis of AKI, the serum NGAL levels at 12 h and 24 h after aortic unclamping were significantly increased in the two groups andwere significantly higher in the Placebo group compared to the Dex group. However, we noted that the measured values of serum NGAL at the two observed time points had the highly variable ranges. Given the fact that this study has a small sample size, only comparing median serum NGAL levels may have of limited clinical value. Most importantly, the readers were not provided with the cutoff value of serumNGALwith their ELISA for diagnosis of postoperative AKI. Furthermore, we were also very interested in knowing how many patients in each group had a higher serum NGAL level than the cutoff value and whether incidence of early postoperative AKI by NGAL measurement was significantly higher in the Placebo group than in the Dex group. Third, the readers were not provided with intragroup comparisons of some important intraoperative data associated with AKI after cardiac surgery. It has been shown that intraoperative systolic blood pressure decrease relative to baseline is independently associated with AKI

Is video laryngoscopy really superior to direct laryngoscopy for emergency intubation in prehospital trauma patients

In a randomized crossover manikin study comparing intubation performance of Macintosh laryngoscope (ML), McCoy laryngoscope (MCL) and C-MAC D-Blade video laryngoscope (CMDB) by prehospital emergency health workers, Yildirim et al. [1] show that the intubation success rate is significantly higher with the CMDB and MCL than with the ML, and intubation duration is significantly shorter with the CMDB than the ML and MCL in all scenarios including a normal airway, rigid cervical collar and a manual in-line cervical stabilization models. Given that airway management of prehospital trauma patients often presents an unique challenge to emergency medical technicians, their findings have potential implications. Other than the limitations described in the discussion, however, we note two issues of this study that might make interpretation of their findings questionable. First, this study was designed to evaluate the intubation performance of three devices in a setting simulating prehospital trauma scenarios, but it was conducted in a well controlled indoor environment under room light. Moreover, manikins were placed on a standard table in a training room. In an actual prehospital trauma situation, tracheal intubation is often required in patients lying on the ground in direct daylight. In the study using a manikin lying supine on the ground, Ueshima et al. [2] find that in daylight, tracheal intubation is always successful when using the ML, but only 1 of 15 participants could successfully intubate the trachea when using the Pentax AWS video laryngoscope. In a controlled, randomized clinical trial comparing the Glidescope video laryngoscope and ML for urgent intubation in the prehospital setting, Trimmel et al. [3] show that impaired monitor visibility of GlideScope video laryngoscope by ambient light occurred in 29 of 168 patients. The reason for the difficulty with video laryngoscopy under daylight is that it is impossible to see the glottis because of reflection of the sunlight on the monitor display. This is a common flaw of all video laryngoscopes requiring an LCD monitor to view the glottis [2]. Second, just like the authors had pointed out, this was a simulation using manikins rather than real humans. The manikin does not reflect normal human airway anatomy, [4] and can not well simulate the real intubation conditions of prehospital trauma patients, especially for oral contamination. In the prehospital trauma patients, presence of vomitus, secretions and blood in the upper airway is a common problem. Actually, smears of vomitus, or secretions and blood on the lens completely obliterate the view with the video laryngoscope. When the lens is obscured, the video laryngoscope must be removed from the mouth and cleaned, which requires extra time and possibly influences the emergent intubation time and success rate. In contrast, oral contamination rarely obstructs the laryngeal view during direct laryngoscopy, unless massive bleeding or vomiting occurs. In available literatures, prior randomized controlled clinical trials comparing direct and video laryngoscopes for urgent intubation in the prehospital patients do not demonstrate that video laryngoscopes are superior to direct laryngoscope with respect to the Re: Yildirim A, et al. (2016) Comparison of Macintosh, McCoy and C-MAC D-Blade video laryngoscope intubation by prehospital emergency health workers: a simulation study. Intern Emerg Med. Mar 21. In Press: doi:10.1007/s11739-016-1437-3.