Portal de Periódicos da CAPES

Comparison of the Reusable Standard GlideScope ® Video Laryngoscope and the Disposable Cobalt GlideScope ® Video Laryngoscope for Tracheal Intubation in an Academic Emergency Department: A Retrospective Review

2014; Wiley; Volume: 21; Issue: 4 Linguagem: Inglês

10.1111/acem.12351

1553-2712

John C. Sakles, Asad E. Patanwala, Jarrod Mosier, J. Dicken, Nathan Holman,

Tracheal and airway disorders

The objective was to compare the first-pass success and clinical performance characteristics of the reusable standard GlideScope® video laryngoscope (sGVL) and the disposable Cobalt GlideScope® video laryngoscope (cGVL). This was a retrospective analysis of prospectively collected data recorded into a continuous quality improvement database at an urban academic emergency department (ED). The intent of the database is to evaluate operator performance and to track practice patterns used for intubation in the ED. Between July 1, 2007, and June 30, 2013, operators recorded all consecutive intubations performed in the ED. The database included patient demographics and detailed information about each intubation, such as device(s) used, reason for device selection, method of intubation, difficult airway characteristics, number of intubation attempts, and outcome of each attempt. The operator also evaluated the presence of lens fogging and extent of lens contamination. The primary outcome measure was first-pass success. Secondary outcome measures were ultimate success, Cormack-Lehane (CL) view of the airway, presence of lens fogging, and extent of lens contamination. Only adult patients age 18 years or older intubated with the sGVL or cGVL using a stylet, and who had data forms completed at the time of intubation, were included in this study. A total of 583 intubations were included in the study, 504 with the sGVL and 79 with cGVL. First pass success was achieved in 81.0% (95% confidence interval [CI] = 77.3% to 84.3%) of patients in the sGVL group and in 58.2% (95% CI = 46.6% to 69.2%) of patients in the cGVL group. In a multivariate logistic regression analysis, the sGVL was associated with a higher first pass success than the cGVL (odds ratio [OR] = 3.3, 95% CI = 1.9 to 5.8). The ultimate success of the sGVL was 92.1% (95% CI = 89.4% to 94.3%) and the cGVL was 72.2% (95% CI = 60.9% to 81.7%). A CL grade I or II view was obtained in 93.2% (95% CI = 90.7% to 95.3%) in the sGVL group and 86.1% (95% CI = 76.5% to 92.8%) in the cGVL group. Lens fogging occurred in 33.3% (95% CI = 29.2% to 37.6%) of the cases in the sGVL group and 59.5% (95% CI = 47.9% to 70.4%) of the cases in the cGVL group. Significant lens contamination occurred in 5.0% (95% CI = 3.2% to 7.2%) of the sGVL group and 21.5% (95% CI = 13.1% to 32.2%) of the cGVL group. In this observational study, the sGVL had higher first pass and overall success than the disposable cGVL. The cGVL had significantly higher incidence of lens fogging and contamination, which may partially account for its lower success. A prospective randomized trial is needed to confirm these findings. Comparar el éxito del primer paso y las características de rendimiento clínico del laringoscopio GlideScope® Estándar Reutilizable (VLGE) con los del laringoscopio Cobalt GlideScope® Desechable (VLCG). Análisis retrospectivo de los datos recogidos de forma prospectiva en una base de datos de mejora continua de la calidad en un servicio de urgencias (SU) universitario. La intención de la base de datos es evaluar el rendimiento del operador y monitorizar los patrones de práctica utilizados para la intubación en el SU. Entre el 1 de julio de 2007 y el 30 de junio de 2013, los operadores grabaron todas las intubaciones consecutivas realizadas en el SU. La base de datos incluyó datos demográficos del paciente e información detallada de cada intubación, como el dispositivo usado, la razón de elección del dispositivo, el método de intubación, las características de la vía área difícil, el número de intentos de intubación y el resultado de cada intento. El operador también evaluó la presencia de empañamiento de las lentes y la extensión de la contaminación de las lentes. La medida de resultado principal fue el éxito de primer paso. Las medidas de resultado secundarias fueron el éxito final, la visión de la vía aérea de Cormack-Lehane, la presencia de empañamiento de las lentes y la extensión de la contaminación de las lentes. Se incluyeron en este estudio los pacientes adultos de 18 años o más intubados con VLGE o con VLCG usando un estilete y que tenían los formularios de los datos completados en el momento de la intubación. Se incluyeron 583 intubaciones, 504 con VLGE y 79 con VLCG. El éxito del primer paso se alcanzó en un 81,0% (IC 95% = 77,3% a 84,3%) de los pacientes en el grupo de VLGE y en un 58,2% (IC 95% = 46,6% a 69,2%) de los pacientes en el grupo de VLCG. En un análisis de regresión logística multivariable, el VLGE se asoció con un mayor éxito en el primer paso que VLCG (OR 3,3, IC 95% = 1,9 a 5,8). El éxito final de la VLGE fue de un 92,1% (IC 95% = 89,4% a 94,3%) yd el VLCG fue de un 72,2% (IC 95% = 60,9% a 81,7%). Se obtuvo una visión grado I o II de Cormack-Lehane en un 93,2% (IC 95% = 90,7% a 95,3%) en el grupo VLGE y en un 86,1% (IC 95% = 76,5% a 92,8%) en el grupo VLCG. El empañamiento de la lentes ocurrió en un 33,3% (IC 95% = 29,2% a 37,6%) de los casos en el grupo VLGE y en un 59,5% (IC 95% = 47,9% a 70,4%) de los casos en el grupo VLCG. Una contaminación significativa de la lentes ocurrió en un 5,0% (IC 95% = 3,2% a 7,2%) del grupo VLGE y en un 21,5% (IC 95% = 13,1% a 32,2%) del grupo VLCG. En este estudio observacional, la VLG estándar tuvo mayor éxito de primer paso y global que el VLCG desechable. El VLCG tuvo una incidencia significativamente mayor de empañamiento y contaminación de las lentes, que puede parcialmente contribuir a sus menores éxitos. Se necesita un ensayo clínico prospectivo aleatorizado para confirmar estos hallazgos. The GlideScope® video laryngoscope (sGVL) (Verathon Medical Inc., Bothell, WA) was first introduced into clinical practice in 2001 and its use has become increasingly more common for intubations performed in the emergency department (ED).1-5 The initial GlideScope® that was released into clinical practice used a multiple-use, reusable blade that requires high-level disinfection between each patient use. Due to concerns about infection transmission between patients, and other issues related to convenience, a single-use, disposable blade was developed and introduced into clinical practice in 2007 (Figure 1). The sGVL incorporates a blue reusable blade and must be disinfected in between each patient use. This typically takes the device out of clinical availability for several hours during the disinfection process. Additionally, the disinfection process can potentially result in accidental loss or damage of the blade due to improper technique, for example, autoclaving. The Cobalt GlideScope® video laryngoscope (cGVL), on the other hand, uses a reusable video baton, with a disposable clear plastic blade that is placed over it. Thus, after an intubation is performed, the disposable blade can be discarded and a new one placed on the video baton. This allows the cGVL to be almost immediately available for intubation on another patient. This also eliminates the need for disinfection of the cGVL blade and reduces the potential risk of cross-contamination of infectious material between patients.6, 7 The cGVL video baton must, however, undergo low-level disinfection after each patient use, and if it becomes visibly contaminated with gross material, high-level disinfection of the baton must be performed. To our knowledge, little research has been conducted comparing the efficacy of the sGVL and the cGVL for intubation in the clinical setting. A study performed by Jones et al.8 in the operating room demonstrated similar success and performance characteristics when comparing the sGVL to the cGVL for elective intubations. These results may not be applicable to the ED environment, where there are often other significant clinical factors that could affect the performance characteristics of the two GlideScope®s. For example, in the ED, patients often will have excessive secretions, blood, or vomit in the airway. These unique factors can potentially affect the efficacy of a video laryngoscope based on differences in their design. To our knowledge, there are no studies comparing these two versions of the GlideScope® in the ED. The purpose of this investigation was to compare the first pass success and clinical performance characteristics of the sGVL and cGVL when used for intubation of adult patients in the ED. This was an analysis of 583 intubations performed on adults in the ED with the sGVL or the cGVL, which were prospectively recorded in a continuous quality improvement database over the 6-year period from July 2007 to June 2013. This project was granted exemption from informed consent requirements by the university's institutional review board prior to conducting the study. This study was conducted at a 61-bed tertiary care academic ED with approximately 70,000 annual ED visits. This ED is a Level I trauma center and has a 3-year emergency medicine (EM) residency program and a 5-year combined pediatrics/EM (P/EM) residency program. The vast majority of intubations in the ED are performed by EM or P/EM residents under the direct supervision of attending emergency physicians who are ultimately responsible for all issues regarding emergency airway management. During the study period, there were three sGVL units and one cGVL unit available for ED intubations. All airways rolls in the ED contain both GlideRite rigid stylets (GlideRite) and standard malleable stylets. The decision regarding the method of intubation, initial device selection, and stylet selection are at the discretion of the operator and EM attending. Residents in this ED receive formal instruction on the use of multiple airway devices including the GlideScope®. This involves both didactic material as well as hands-on experience in the simulation laboratory. The technique taught for use of the sGVL and cGVL is the same. Residents are instructed to turn on the GlideScope® unit at least 2 minutes prior to intubation to allow the tip of the laryngoscope blade to warm up, thereby maximizing the effectiveness of the antifogging mechanism. They are instructed to insert the blade in the midline, keeping it flush against the tongue to avoid contamination from material in the posterior pharynx and to slowly navigate their way towards the glottic entrance. They are taught to advance the tip of the GlideScope® blade into the vallecula and are cautioned not to advance the GlideScope® too far, as this often makes tube passage more difficult. It is strongly recommended that the residents use the GlideRite stylet with all GlideScope® intubations. If they opt to use a standard stylet, it is recommended that the tube and stylet be bent to a 60° to 90° angle at the distal portion, to mimic the shape of the GlideScope® blade. When using the GlideRite stylet, it is suggested that once the tube passes beyond the vocal cords, the stylet be withdrawn 5 cm prior to advancement of the tube into the trachea. All patients requiring intubation in the ED were entered into the database. Only adult patients (age 18 years or older) who underwent initial intubation attempts using the sGVL (size 3, 4, or 5 blade) or the cGVL (size 3 or 4 blade) using a stylet, with data collection forms completed at the time of intubation, were included in this study (Figure 2). Following each intubation, the operator completed an airway data collection form that included a variety of clinically important information such as indication for intubation, method of intubation, device(s) used, reason for device selection, Cormack-Lehane (CL) view, number of intubation attempts, and outcome of each attempt. In addition, the operator evaluated the clinical performance characteristics of the video laryngoscopes by assessing the degree of lens fogging and gross lens contamination. Lens fogging was evaluated by the operator using a 10-cm visual analogue scale with "0" representing no fogging and "10" representing complete fogging. Lens contamination was evaluated in one of four categories: none (no contamination), mild (minimal contamination, cords easily visible), moderate (excessive contamination, cords still visible), and severe (complete contamination, cords not visible). For the purposes of this study, moderate or severe lens contamination was considered clinically significant contamination. These scales were developed by the authors for the purpose of this database, to allow evaluation of various optical devices used for intubation in the ED. The presence of difficult airway characteristics was also assessed by the operator and included the following factors: cervical immobilization, obesity, facial or neck trauma, short neck, large tongue, small mandible, airway edema, and blood or vomit in the airway. Intubation methods included rapid sequence intubation (RSI) in which a paralytic agent was used (usually in conjunction with a sedative), oral intubation in which only a sedative agent was used (SED), and oral intubation in which no medications at all were used (no meds). The operator had one of three options to choose for the reason of device selection: "standard" (routine airway, no suspected difficulty), "difficult" (difficult airway anticipated), or "education" (device selected for operator educational experience). The primary outcome was the first pass success of the sGVL and cGVL groups. An intubation attempt was defined as the insertion of the video laryngoscope blade into the mouth of the patient, regardless whether an attempt was made to insert a tracheal tube. Firspass success was defined as tracheal intubation on a single laryngoscopic insertion. Secondary outcomes were ultimate success, the CL view, the degree of lens fogging, and the degree of lens contamination. Ultimate success was defined as a successful tracheal intubation with the initial GlideScope® selected, regardless of the number of attempts required. The airway data forms were reviewed by the senior author for incomplete data and returned to the operator for completion if necessary. Data forms were cross-referenced to professional billing records, pharmacy records, and the ED admission log to identify any missing data forms. If an intubation was performed in the ED that was identified through cross-referencing, but was missing from the database, the operator was given a blank airway data collection form to complete. These cases had data forms completed usually less than 1 month after the intubation, but were eliminated from the data analysis to minimize the effect of recall bias. All statistical analyses were performed with STATA version 12 (StataCorp, College Station, Texas). Patient demographics were descriptively compared between the two groups. For categorical data, 95% confidence intervals (CIs) were calculated using the "exact" method and reported for each device. Multivariate logistic regression analysis was performed for the outcome variable of first pass success. The predictor variable of interest was intubation device (sGVL vs. cGVL). Other predictor variables added to the model to adjust for confounding included age, sex, trauma status, indication for intubation, method of intubation, reason for device selection, presence of at least one difficult airway characteristic, EM resident postgraduate year (PGY) level, and type of stylet used. These variables were selected a priori based on previous research and considered to be important based on clinical experience of the investigators. The intent was not to create a parsimonious model, but instead to adjust for all potential confounders. Thus automated variable selection methods were not used, and all variables were added directly to the model. Age was the only continuous predictor variable. This variable did not meet the assumption of linearity in the logit; thus it was dichotomized as age <65 or ≥65 years. Categorical predictor variables with more than two categories were dichotomized based on investigator consensus and clinical experience. This was done to avoid overfitting the model given the number of degrees of freedom available. The model was assessed for multicollinearity, and the Hosmer-Lemeshow goodness-of-fit test was used to evaluate overall model fit. To account for individual variability between operators, the model was clustered by operator. Influential observations were identified using Pregibon's delta-beta influence statistic. As part of a sensitivity analysis, highly influential observations were excluded to determine if it changed the results. Over the 6-year study period, a total of 2,972 intubations were performed in the ED. Of these, 814 were initially attempted using a GlideScope®. Of these, 118 were eliminated from analysis because they used non-sGVL or cGVL GlideScope®s (GVL Ranger, GVL Direct, GVL Groove), 59 were eliminated because they were pediatric patients, 52 were eliminated due to delayed completion of the data forms, and two were eliminated because no stylet was used (one used in conjunction with a flexible fiberoptic scope and one used a tube exchanger; Figure 2). This left 583 GlideScope® intubations, 504 attempted with the sGVL and 79 with the cGVL. The demographics for the two groups are presented in Table 1. The mean age of patients intubated in the sGVL group was 46.6 years (range = 18 to 96 years) versus 46.6 years (range = 18 to 89 years) in the cGVL group. A significant number of the patients in the study were trauma patients. RSI was performed in 85.1% in the sGVL group and 88.6% in the cGVL group. Fewer patients in the sGVL group had difficult airway characteristics than in the cGVL group. More operators in the sGVL group used the GlideRite stylet than in the cGVL group. First pass success was achieved in 81.0% of patients in the sGVL group and 58.2% of patients in the cGVL group (Table 2). When the sGVL was used with the GlideRite stylet, the first pass success rate was 83.7% (95% CI = 79.7% to 87.2%), and when used with a standard malleable stylet it was 71.2% (95% CI = 61.8% to 79.4%). When the cGVL was used with the GlideRite stylet, the first pass success rate was 66.7% (95% CI = 52.5% to 78.9%) and when used with a standard malleable stylet it was 40.0% (95% CI = 21.1% to 61.3%). After adjusting for potential confounders, the sGVL was associated with greater odds of first pass success than the cGVL (Table 3). The model fit the data well (Hosmer-Lemeshow goodness of fit, p = 0.752). After excluding influential observations, the sGVL was still associated with greater odds of first pass success (adjusted OR [aOR] = 3.4, 95% CI = 1.9 to 5.9) Ultimate success for the sGVL group was 92.1% and for the cGVL group was 72.2% (Table 2). In the sGVL group, a CL view of I or II was achieved in 93.2% (95% CI = 90.7% to 95.3%) of the patients, and in the cGVL group a CL view of I or II was achieved in 86.1% (95% CI = 76.5% to 92.8%) of the patients. Some degree of lens fogging occurred in 33.3% of the intubations in the sGVL group and 59.5% in the cGVL group. In the sGVL group the median fog score was 0 (interquartile range [IQR] = 0 to 0.5), and in the cGVL group the median fog score was 1.3 (IQR = 0 to 4.5). Significant lens contamination occurred in 5.0% (95% CI = 3.2% to 7.2%) and 21.5% (95% CI = 13.1% to 32.2%) of intubations in the sGVL and cGVL groups, respectively. After adjusting for confounders, the sGVL was less likely to be associated with fogging (aOR = 0.3, 95% CI = 0.2 to 0.5). Multivariate analyses were not conducted for other secondary outcomes such as ultimate success, CL view, and lens contamination, as this would result in overfitting of the models. In the sGVL group there were 40 intubation failures: esophageal intubation (n = 0), cannot see airway (n = 23), cannot direct tube (n = 17), and equipment failure (n = 0). In the majority of these cases the operator was subsequently able to successfully intubate the patient with a direct view laryngoscope device (n = 31). In the cGVL there were 22 intubation failures: esophageal intubation (n = 1), cannot see airway (n = 10), cannot direct tube (n = 9), and equipment failure (n = 2). A direct laryngoscope was also the most commonly used rescue device (n = 20; Figure 3). There were no cases where the sGVL was used to rescue the cGVL or where the cGVL was used to rescue the sGVL. In the sGVL group, there were 127 different operators, with a range of one to 16 uses per operator. In the cGVL group, there were 48 different operators, with a range of one to five uses per operator. The differences we found between the two groups are surprising as the two GlideScope®s are fairly similar in design. There are, however, some key differences in the design features of the sGVL and cGVL that may explain the findings in our study. For one, on the sGVL, the micro video camera lens is covered by a thin piece of glass and the antifog heating element is essentially in direct contact with this glass. On the cGVL, the heating element is on the video baton and thus must transfer the heat across the additional surface of the disposable plastic sheath. The transfer of heat across the plastic sheath might be less effective and this potentially could result in reduced capability of the cGVL to inhibit lens fogging. Another design difference is that on the sGVL, the micro video camera is located in a slightly recessed position on the blade compared to the cGVL, and this position may make it slightly more resistant to gross lens contamination from secretions, blood, or vomit in the airway. Another possible factor could be the difference in materials from which the two devices are constructed. The glass covering the micro video camera in the sGVL may be inherently more resistant to lens fogging and to the adherence of gross contaminates than the plastic material used in the cGVL sheath. Last, although the sGVL and cGVL are fairly similar in blade shape, the cGVL has a wide flange at its proximal portion that may interfere with operator's ability to maneuver the tracheal tube during the intubation. In this study, we found that the sGVL was used over six times more frequently than the cGVL, even though both devices were available in the ED over the entire study period. Part of the reason the sGVL was used more frequently may have been that the ED has three reusable sGVL units and only one disposable cGVL unit. In addition to this issue, there appears to be general operator preference of the sGVL over the cGVL. Even though there is a large discrepancy in the usage between the devices, we believe that the magnitude of difference in the success between the two devices is significant and represents a true difference in device performance. Our results are in direct contrast to the findings of a study comparing the sGVL to cGVL in the operating room.8 In this prospective, randomized study of 100 elective operating room intubations, the authors found no difference between the two GlideScope® models. In that study, the number of intubation attempts, the number of intubation failures, the CL views, the degree of fogging, the ease of intubation, and the times to intubation were all found to not be significantly different between the sGVL and the cGVL. It is important to note that in this study all patients were intubated for elective surgery, in a controlled environment, and had normal-appearing airways. Furthermore, any patients with anticipated difficult airways or those requiring RSI were excluded from the study. In the ED, clinical situations and patient characteristics are likely to be present that may significantly affect the performance of the GlideScope®. For example, patients in the ED requiring emergent intubation typically arrive unexpectedly, allowing the operator little time for preparation and assessment. ED patients are also frequently clinically more unstable with lower respiratory and hemodynamic reserves. Thus, operators often have less time to perform the intubation and are probably more likely to rapidly abandon one device if it is not immediately successful. Finally, it is not unusual for ED patients to present with copious secretions, blood, or vomit in the airway, which obviously can greatly affect the efficacy and performance of the GlideScope®. This was an observational study and thus lacked the benefits of randomization. Hence, there could be selection bias with regard to the device used because this was based on availability and operator preference. There was an obvious disparity in the sample group sizes, the sGVL group having over six times as many intubations as the cGVL group. Even though the cGVL group is much smaller, these intubations were attempted by a wide variety of operators, suggesting that it is not the operator having an effect on intubation failure, but rather a device issue. Another limitation of this study is that a larger percentage of operators in the sGVL group used the GlideRite stylet. We have shown in a previous study that operators in our ED have significantly higher success when using the GlideRite stylet when compared to the standard malleable stylet, and thus, this would bias the results against the cGVL. However, the type of stylet was accounted for in the multivariate regression model, and we still found a significant difference between the performance of the sGVL and cGVL. Also more patients in the cGVL had difficult airway characteristics. However, this difference was small and was furthermore incorporated in the multivariate regression analysis to adjust for this potential confounder. Although the operators completed the airway data collection forms shortly after the intubations, the data are still subject to self-report bias. Ideally, an unbiased observer would record data in real time to minimize the possibility of measurement bias; however, this is not practical in our clinical environment where intubations are typically performed emergently with little or no advanced warning. Also, this was a single-institution study where EM residents perform the majority of intubations and are very skilled using video laryngoscopes. Thus, these results may not be applicable to other clinical settings where emergency intubations are performed. However, we were able to evaluate these devices in a real-world practice setting, providing us with important information about the effectiveness of these commonly used devices. Finally, the scales used to evaluate lens fogging and contamination have not been validated, but may be a useful method for evaluating video laryngoscopy performance and should be considered for future studies. To our knowledge, this is the only study comparing the clinical effectiveness of the standard GlideScope® video laryngoscope to the Cobalt GlideScope® video laryngoscope for intubations performed in the ED. In this observational study, the standard GlideScope® video laryngoscope had higher first pass and overall success than the disposable Cobalt GlideScope® video laryngoscope. The Cobalt GlideScope® video laryngoscope had a significantly higher incidence of fogging and lens contamination, which may partially account for its lower success. A prospective randomized trial is needed to confirm these findings.