Video screen visualization patterns when using a video laryngoscope for tracheal intubation: A systematic review.

OBJECTIVE: Published studies of video laryngoscopes are often limited by the lack of a clear definition of video laryngoscopy (VL). We performed a systematic review to determine how often published studies of VL report on video screen visualization.
METHODS: We searched PubMed, EMBASE and Scopus for interventional and observational studies in which a video laryngoscope equipped with a standard geometry blade was used for tracheal intubation. We excluded simulation based studies. Our primary outcome was data on video laryngoscope screen visualization. Secondary outcomes were explicit methodology for screen visualization.
RESULTS: We screened 4838 unique studies and included 207 (120 interventional and 87 observational). Only 21 studies (10% of 207) included any data on video screen visualization by the proceduralist, 19 in a yes/no fashion only (ie, screened viewed or not) and 2 with detail beyond whether the screen was viewed or not. In 11 more studies, visualization patterns could be inferred based on screen availability and in 16 more studies, the methods section stated how screen visualization was expected to be performed without reporting data collection on how the proceduralist interacted with the video screen. Risk of bias was high in the majority of included studies.
CONCLUSIONS: Published studies of VL, including many clinical trials, rarely include data on video screen visualization. Given the nuances of using a video laryngoscope, this is a critical deficiency, which largely prevents us from knowing the treatment effect of using a video laryngoscope in clinical practice. Future studies of VL must address this deficiency.

INTRODUCTION

Background

Video laryngoscopes are increasingly the most common devices used for tracheal intubation in US emergency departments (EDs). Video laryngoscopy (VL) has consistently been reported to improve glottic visualization compared with direct laryngoscopy (DL). , , When incorporated into high‐volume training programs, there is some evidence that emergency medicine residents acquire airway skills better with a video laryngoscope.

Importance

Despite hundreds of published studies, however, including 79 total clinical trials across 3 Cochrane reviews, we still do not know whether a video laryngoscope improves procedural success. , , , , , A primary reason for this knowledge deficit is the lack of a consistent, explicit definition for VL. Especially with a standard geometry blade, using a video laryngoscope and VL are not equivalent. With a standard geometry video laryngoscope, proceduralists have the option to directly view the glottis by looking into the patient's oropharynx, or indirectly view the glottis by looking at the video screen. In addition to having the option to view the screen or not view the screen in a yes/no fashion, proceduralists can show significant variations in their patterns of screen visualization including variations in the percentage of time during an intubation attempt spent viewing the video screen, the number of times they look back and forth between the patient and screen during an attempt, when during an attempt they first view the screen, and whether or not they view the screen during key moments during an attempt. With such a range of possibilities, the lack of a standard definition of VL, especially based on accurate methods of data collection, is a critical and persistent deficiency in airway literature.

Goals of this investigation

We performed a systematic review of studies of tracheal intubation with a video laryngoscope and a standard geometry blade. We specifically sought to determine how often video screen visualization patterns were defined or measured, including the approach to data collection.

METHODS

Data sources and searches

We registered our systematic review methodology with PROSPERO before study commencement, including the search strategy, outcomes, and inclusion and exclusion criteria. We designed this systematic review to be consistent with Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) methodology.

We conducted a search of PubMed, EMBASE, and Scopus, from database inception through May 2021. We used each of the following search terms in each database: “VL,” “videolaryngoscopy,” “video laryngoscope,” and “videolaryngoscope.”

Study selection

From the initial database search, one author (PD) conducted a manual review of identified studies, starting with the title, then the abstract, and finally review of the full text. From this manual review, we included only original research studies available in English that involved tracheal intubation on a live person in the systematic review. Additionally, the study needed to include at least one group intubated using a video laryngoscope equipped with a standard geometry blade. We included observational and interventional studies, including clinical trials. We excluded studies of video laryngoscopes with only a non‐standard geometry or hyperangulated blade, as the proceduralist does not have the option for DL with these devices. We also excluded studies of nasotracheal intubation and of face‐to‐face intubation, given their fundamental procedural differences from traditional intubation techniques. We also excluded cadaveric studies, simulation‐based studies, case reports, and conference abstracts. We defined standard and non‐standard (hyperangulated) geometry blades as shown in Table 1.

TABLE 1

Standard geometry and non‐standard (hyperangulated) geometry blades

Device	Manufacturer	Location
Standard geometry blades
Macintosh and Miller versions of C‐MAC	Storz	Tuttlingen, Germany
Macintosh and Miller versions of Glidescope	Verathon	Bothell, WA
Macintosh and Miller versions of McGrath	Medtronic	Minneapolis, MN
Macintosh and Miller versions of Direct Coupled Interface	Storz	Tuttlingen, Germany
Macintosh and Miller versions of UE scope	UE Medical Devices	Newton, MA
Macintosh and Miller versions of Intubrite Video Laryngoscopes	Salter Labs	Vista, CA
CEL‐100	Connell Energy Technology Co.	Shanghai, China
Venner AP advance	Venner Medical	Singapore
VLP‐100	Daiken Medical	Osaka, Japan
Infantview	Vyaire Medical	Mettawa, IL
Non‐standard (hyperangulated) geometry blades
C‐MAC D‐Blade	Storz	Tuttlingen, Germany
Hyperangulated versions of Glidescope	Verathon	Bothell, WA
Hyperangulated versions of McGrath	Medtronic	Minneapolis, MN
Hyperangulated versions of UE scope	UE Medical Devices	Newton, MA
Airtraq	Prodol Mediteq	Getxo, Spain
King VL	Ambu	Ballerup, Denmark
Airway scope	Nihon Kohden	Tokyo, Japan
Truview	Truphatek	Natanya, Israel
Totaltrack	Medcom Flow	Barcelona, Spain
VL3	HugeMed	Shenzhen, China

Data extraction and quality assessment

After applying inclusion and exclusion criteria, one reviewer (PD) assessed included studies for any data reporting on the video screen visualization pattern by the proceduralist. We included both actual data collection of screen visualization (outcome data) as well as a description of approach to video screen visualization in the study protocol (ie, methodologic data). We also collected information on the approach to data collection for screen visualization, including self‐report, electronic record review, independent observer, and video. Video data sources included overhead or external cameras and the video laryngoscope itself.

For any study that reported either outcome or methodological data for screen visualization, both authors/reviewers (PD and BK) conducted an independent review. Both reviewers assessed the risk for bias using the Cochrane Collaboration Bias Appraisal Tool. We categorized the reporting of proceduralist visualization as high risk, low risk, or unclear risk of bias, based on the method of data collection (self‐report, video review, etc.). High risk studies included observational studies with self‐reported visualization patterns and interventional or observational studies in which the study protocol/manuscript methods section prescribe visualization patterns but both direct and indirect visualization were available to the proceduralist and visualization patterns were not objectively recorded. Low risk studies included those utilizing overhead video review and included a reliability assessment, studies in which the video screen was unavailable to proceduralists throughout the entire procedure and studies utilizing eye tracking devices. Unclear risk studies included those utilizing overhead video review but did not include a reliability assessment. Any disagreements between the two reviewers were resolved by a third party with content expertise. Because the goal of this review was to understand the video screen visualization patterns and not the outcome of an intervention, we did not perform measures of effect analyses or a meta‐analysis.

Outcomes

Our primary outcome was a study including outcome or methodologic data for video screen visualization patterns. To maximize sensitivity, we recorded any description of screen visualization, including the dichotomous “yes/no.” We also recorded more granular data, including duration of video screen viewing, the number of times a proceduralist switched their gaze between the patient and the video screen, which portion of intubation attempts (laryngoscopy vs tube delivery) the proceduralist viewed the video screen, and the total proportion of video screen viewing during intubation attempts. Secondary outcomes included the method of video screen data collection.

RESULTS

Enrollment

In our primary database search, we identified 4838 unique studies (Figure 1). We excluded 3615 studies after title review, 948 after abstract review, and 68 after full text review. We included 207 unique studies after all exclusions—120 were interventional (84 randomized controlled trials) and 87 were observational (Table 2). The included studies were published from 2001 to 2021.

FIGURE 1

Summary of the selection process for the systematic review

TABLE 2

Included studies

Study	Study design	Setting	Age (in years unless specified otherwise)	Device(s)	Visualization	Method of data collection (for studies in which visualization reported)	Risk of bias (for studies in which visualization reported)
Studies in which video screen visualization was reported
Boedeker et al., 2008 34	Interventional	OR	Does not specify	C‐Mac	Protocol prescribed direct and indirect glottic visualization with video laryngoscope. Video screen unavailable to proceduralist during direct laryngoscopy.	Self‐report	High risk
Brown et al., 2010 37	Observational	ED	≥15	C‐Mac	Protocol prescribed direct and indirect glottic visualization with video laryngoscope. Video screen unavailable to proceduralist during direct laryngoscopy.	Self‐report	High risk
Byhahn et al., 2010 42	Interventional	OR	Adults	C‐Mac	Protocol prescribed direct glottic visualization by proceduralist and indirect glottic visualization by independent observer, then protocol prescribed tube delivery under indirect visualization. Unclear if screen turned away from proceduralist.	Direct view based on self‐report, indirect view based on independent observer	High risk
Cavus et al., 2011 47	Interventional	Pre‐hospital	All	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Cavus et al., 2018 48	Randomized trial	Pre‐hospital	>18	C‐Mac, AP Advance	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Cengiz and Yilmaz, 2019 49	Randomized trial	OR	18‐75	C‐Mac	Protocol prescribed indirect laryngoscopy.	Independent observer recorded intubation success, not visualization patterns	High risk
De Jong et al., 2013 57	Interventional	ICU	≥18	McGrath Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
De Jong et al., 2021 58	QI	OR	≥18	C‐Mac, McGrath Mac, AP Venner	Protocol prescribed direct and indirect glottic visualization with video laryngoscope. Based on the protocol, it is assumed that most proceduralists likely intubated via indirect visualization, however the true visualization patterns of proceduralist were not specifically reported.	Self‐report	High risk
Dodd et al., 2019 62	Observational	ED	≥18	C‐Mac	Screen visualization reported in yes/no fashion.	Overhead video review	Low risk
Driver et al., 2016 63	Randomized trial	ED	Adults	C‐Mac	Proceduralists randomized to either direct or indirect visualization. When randomized to direct visualization, the video screen was made unavailable to proceduralist. Video screen visualization and protocol deviations in both directions (ie, were randomized to direct visualization and viewed video screen or were randomized to indirect visualization and did not view video screen or did not use video laryngoscope) were reported.	Self‐report	High risk
Driver et al., 2017 64	Observational	ED	≥18	C‐Mac	Screen visualization reported in yes/no fashion.	Overhead video review	Low risk
Driver et al., 2018 11	Randomized trial	ED	≥18	C‐Mac, Glidescope Mac	Screen visualization reported as “screen never used,” “screen viewed for entire attempt,” or “screen viewed during passage of tube or bougie into glottis.”	Self‐report	High risk
Elattar et al., 2020 69	Randomized trial	OR	≤2	C‐Mac	When the C‐Mac was used, a screenshot was taken using the C‐Mac device and additionally, a photograph was taken through the oropharynx simulating a direct view. Despite this, how the proceduralist actually visualized the glottis (direct vs indirect) was not reported.	Still‐image photographs	High risk
García‐Pintos et al., 2021 75	Observational	Pre‐hospital	≥18	C‐Mac	Protocol prescribed direct and indirect glottic visualization with video laryngoscope. Video screen “obscured or turned away from proceduralist's field of view” during direct laryngoscopy. Did not report screen visualization patterns during tube delivery.	Self‐report	High risk
Glasheen et al., 2020 79	Observational	Pre‐hospital	All	McGrath Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Guyette et al., 2013 10	Interventional	Pre‐hospital	≥18	C‐Mac	Screen visualization reported as only screen, mostly screen, mostly direct or only direct	Self‐report	High risk
Hackell et al., 2009 88	Observational	OR	Infants	Storz DCI	Case series of seven difficult infant intubations. Screen visualization patterns (direct vs indirect) described in six of the seven cases.	Self‐report	High risk
Hofstetter et al., 2006 91	Interventional	OR	≥18	Storz DCI	Protocol prescribed direct and indirect glottic visualization with video laryngoscope.	Self‐report	High risk
Hossfeld et al., 2015 92	Observational	Pre‐hospital	All	C‐Mac	Protocol prescribed direct and indirect glottic visualization with video laryngoscope.	Self‐report	High risk
Hossfeld et al., 2016 93	Observational	Pre‐hospital	All	C‐Mac	Protocol prescribed direct and indirect glottic visualization with video laryngoscope	Self‐report	High risk
Hossfeld et al., 2020 94	Observational	Pre‐hospital	All	C‐Mac	Direct and indirect laryngoscopic views reported in the majority of cases. Otherwise, did not report visualization patterns of proceduralist.	Self‐report	High risk
Howard‐Quijano et al., 2008 95	Interventional	OR	≥12	C‐Mac	Video screen unavailable to proceduralist	Intraoral video recordings	Low risk
Hwang et al., 2018 96	Observational	ED	≥18	C‐Mac	Screen unavailable to proceduralist on first attempt. If subsequent attempt(s) needed, screen visualization was permitted but not reported on.	Self‐report	High risk
Jungbauer et al., 2009 106	Randomized trial	OR	>18	Storz Berci‐Kaplan	Protocol prescribed direct or indirect glottic visualization with video laryngoscope.	Self‐report	High risk
Kaplan et al., 2006 107	Interventional	OR	≥18	C‐Mac	Protocol prescribed direct glottic visualization with video laryngoscope followed by indirect glottic visualization with video laryngoscope.	Self‐report	High risk
Knapp et al., 2021 116	Observational	Pre‐hospital	All	C‐Mac	Screen visualization reported in yes/no fashion.	Intraoral video and self‐report	High risk
Lascarrou et al., 2017 6	Randomized trial	ICU	≥18	McGrath Mac	Protocol prescribed indirect glottic visualization and tube delivery using video laryngoscope on initial attempt. On subsequent attempts, direct versus indirect visualization was self‐reported in a yes/no fashion.	Self‐report	High risk
Law et al., 2020 12	Observational	NICU	Neonates	C‐Mac	Only 2 cases in this observational study were performed with video laryngoscope. Although gaze switches between patient and equipment were reported on for the whole group, the 2 cases in which a video laryngoscope was used were not reported on individually.	Video recordings using eye tracking device (glasses)	Low risk
Macke et al., 2020 133	Randomized trial	Pre‐hospital	≥18	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Marrel et al., 2007 136	Randomized trial	OR	Adults	X‐Lite	For all patients, glottic visualization was performed both directly and indirectly using a video laryngoscope. Then for tube delivery, 1 group was allowed to use the video screen for indirect visualization and the other group had the video screen made unavailable to them and thus had to perform tube delivery with direct visualization.	Self‐report	High risk
Meininger et al., 2010 138	Interventional	OR	Adults	C‐Mac	Protocol prescribed direct glottic visualization by proceduralist and indirect glottic visualization by independent observer, then protocol prescribed tube delivery under indirect visualization. Unclear if screen turned away from proceduralist.	Direct view based on self‐report, indirect view based on independent observer	High risk
Mosier et al., 2013 144	Observational	ED	≥18	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Normand et al., 2018 155	Observational	OR	>18	McGrath Mac	Protocol prescribed direct glottic visualization performed with video laryngoscope screen covered. If direct view was CL IV (or attempt utilizing direct visualization was unsuccessful), the proceduralist was permitted to use indirect visualization, which was reported on.	Self‐report	High risk
O'Shea et al., 2015 157	Randomized trial	Delivery Room/NICU	Neonates	Modified Miller	Video screen unavailable to proceduralist	Intraoral video recordings	Low risk
O'Shea et al., 2018 158	Observational	NICU	Neonates	Modified Miller	Video screen unavailable to proceduralist	Intraoral video recordings	Low risk
Piepho et al., 2011 164	Interventional	OR	≥18	C‐Mac	Protocol prescribed indirect laryngoscopy only. No other visualization patterns reported.	Unclear	High risk
Pieters et al., 2018 166	Interventional	OR	≥18	C‐Mac	Protocol prescribed direct glottic visualization with video laryngoscope followed by indirect glottic visualization with video laryngoscope.	Photographs of glottis and self‐report	High risk
Raimann et al., 2017 170	Interventional	OR	Weight‐based: <10 kg	C‐Mac	Protocol prescribed direct visualization by proceduralist while second observer graded indirect view on video screen. Does not explicitly state that video screen was unavailable to proceduralist, only that direct visualization was performed.	Direct observation and self‐report	High risk
Raimann et al., 2019 171	Interventional	OR	18‐80	C‐Mac	Protocol prescribed glottic visualization by direct view with video laryngoscope with video screen unavailable to proceduralist. Then indirect visualization was “allowed” and screen was available to proceduralist. Reported view quality from direct and indirect visualization but otherwise did not report visualization patterns.	Self‐report	High risk
Sainsbury et al., 2017 175	Randomized trial	OR	Adults	Glidescope direct, C‐Mac	Video screen unavailable to proceduralist	Direct observation	Low risk
Sakles et al., 2012 176	Observational	ED	All	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Sakles et al., 2015 179	Observational	ED	All	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Sakles et al., 2016 181	Observational	ED	Adults	C‐Mac	Screen visualization reported in yes/no fashion.	Self‐report	High risk
Saran et al., 2019 183	Interventional	OR	<6 mo	C‐Mac	Attempts in which video screen unavailable to proceduralists or supervisors were compared to attempts in which video screen was available to proceduralists and supervisors. Did not report visualization patterns for attempts in which video screen was available.	Unclear	High risk
Sørensen and Holm‐Knudsen et al., 2012192	Interventional	OR	<2	Storz DCI	Protocol prescribed that proceduralists only perform indirect visualization when using Storz video laryngoscope. Visualization patterns were not described further.	Intraoral video recordings	High risk
Vadi et al., 2017 199	Randomized trial	OR	<2	Storz DCI	Reported hat in 2 of the patients randomized to Storz VL group, the proceduralist did not view the video screen. No additional visualization patterns were reported in cases in which screen was viewed.	Self‐report and independent observation	High risk
Weiss et al., 2001 205	Interventional	OR	<10	Modified Miller VL	Video screen unavailable to proceduralist	Self‐report	Low risk
Zhang et al., 2021 213	Randomized trial	OR	21‐80	C‐Mac	Protocol prescribed indirect laryngoscopy	Self‐report	High risk
Studies in which video screen visualization was not reported
Abid et al., 2020 15	Observational	Pre‐hospital	≤18	C‐Mac	Not reported
Aggarwal et al., 2019 16	Randomized trial	OR	25‐60	C‐Mac	Not reported
Ahmed et al., 2017 17	Randomized trial	OR	20‐60	C‐Mac	Not reported
Akbar et al., 2015 18	Randomized trial	OR	18‐60	C‐Mac	Not reported
Akbas et al., 2019 19	Randomized trial	OR	18‐65	McGrath MAC, C‐MAC	Not reported
Altun et al., 2018 20	Randomized trial	OR	18‐65	C‐Mac	Not reported
Alvis et al., 2015 21	Randomized trial	OR	Adults	McGrath Mac	Not reported
Amalric et al., 2020 22	Observational	ICU	Adults	McGrath Mac	Not reported
Amaniti et al., 2019 23	Observational	OR	≥18	C‐Mac	Not reported
Ångerman et al., 2018 24	Observational	Pre‐hospital	≥18	C‐Mac	Not reported
Arasu et al., 2020 25	Randomized trial	OR	18‐60	C‐Mac	Not reported
Aziz and Bambrink, 2011 26	Observational	OR	Adults	C‐Mac	Not reported
Aziz et al., 2012 27	Randomized trial	OR	≥18	C‐Mac	Not reported
Aziz et al., 2016 28	Observational	OR	>18	C‐Mac	Not reported
Bakshi et al., 2019 29	Randomized trial	OR	Adults	McGrath Mac	Not reported
Bensghir et al., 2013 30	Randomized trial	OR	>18	X‐Lite	Not reported
Bhat et al., 2013 31	Randomized trial	OR	≥18	C‐Mac	Not reported
Bhat et al., 2015 32	Randomized trial	OR	≥18	C‐Mac	Not reported
Blajic et al., 2019 33	Randomized trial	OR	All	C‐Mac	Not reported
Boehringer et al., 2015 35	Observational	Pre‐hospital	All	C‐Mac	Not reported
Breeman et al., 2020 36	Observational	Pre‐hospital	≥18	McGrath MAC	Not reported
Brown et al., 2015 38	Observational	ED	≥15	C‐Mac, Video Macintosh	Not reported
Brown et al., 2020 7	Observational	ED	>14	Multiple Standard Geometry VLs	Not reported
Brück et al., 2015 39	Randomized trial	OR	>18	C‐Mac	Not reported
Burjek et al., 2017 40	Observational	OR	<18	C‐Mac	Not reported
Burnett et al., 2014 41	Interventional	Pre‐hospital	≥18	C‐Mac	Not reported
Çaparlar et al., 2019 43	Randomized trial	OR	18‐65	C‐Mac	Not reported
Carlson et al., 2012 44	Observational	Pre‐hospital	All	C‐Mac	Not reported
Cavus et al., 2010 45	Interventional	OR	>18	C‐Mac	Not reported
Cavus et al., 2011 46	Randomized trial	OR	≥18	C‐Mac	Not reported
Chan et al., 2021 50	Observational	ED	≥21	C‐Mac	Not reported
Chandrashekaraiah et al., 2017 51	Randomized trial	OR	18‐65	C‐Mac	Not reported
Cheong et al., 2018 52	Observational	OR	Adults	C‐Mac	Not reported
Colak et al., 2019 53	Randomized trial	OR	≥65	McGrath Mac	Not reported
Corso et al., 2020 54	Observational	ED	Adults	I‐view	Not reported
Couto et al., 2020 55	Observational	ED	1‐18	McGrath Mac	Not reported
Couto et al., 2021 56	Interventional	ED	1‐18	McGrath Mac	Not reported
Deguchi et al., 2016 59	Randomized trial	OR	20‐85	McGrath Mac	Not reported
Desai et al., 2015 60	Observational	ED	≥18	C‐Mac	Not reported
Dey et al., 2020 61	Randomized trial	ICU	≥18	C‐Mac	Not reported
Driver et al., 2019 65	Observational	ED	All	C‐Mac	Not reported
Driver et al., 2020 66	Observational	ED	≥14	C‐Mac, Glidescope Mac blades	Not reported
Eberlein et al., 2019 67	Observational	Pre‐hospital	All	McGrath Mac	Not reported
Eisenberg et al., 2016 68	Observational	ED	0‐18	C‐Mac	Not reported
Ezhar et al., 2018 70	Randomized trial	OR	18‐60	C‐Mac	Not reported
Fiadjoe et al., 2016 71	Observational	OR	<18	C‐Mac	Not reported
Fogg et al., 2012 72	Observational	ED	Did not specify	C‐Mac	Not reported
Fogg et al., 2016 73	Observational	ED	All	C‐Mac	Not reported
Garcia‐Marcinkiewicz et al., 2020 74	Randomized trial	OR	<12 months	C‐Mac	Not reported
Gaszynski, 2017 76	Observational	OR	Adults	McGrath Mac	Not reported
Gaszynski, 2021 77	Randomized trial	OR	Did not specify	McGrath Mac, I‐view	Not reported
Giraudon et al., 2017 78	Randomized trial	OR	Weight‐based: 10‐20 kg	McGrath Mac	Not reported
Goksu et al., 2016 80	Randomized trial	ED	>16	C‐Mac	Not reported
Grant et al., 2021 81	Observational	ED	All	C‐Mac	Not reported
Green‐Hopkins et al., 2015 82	Observational	ED	<21	C‐Mac	Not reported
Grunwell et al., 2017 13	Observational	ICU	Children	C‐Mac	Not reported
Guerra‐Hernández et al., 2020 83	Interventional	OR	Adults	Hybrid 1.0 VDL	Not reported
Gümüş et al., 201484	Randomized trial	OR	18‐65	Storz DCI	Not reported
Gupta et al., 201385	Randomized trial	OR	18‐65	C‐Mac	Not reported
Gupta et al., 2015 86	Observational	OR	All	C‐Mac	Not reported
Gupta et al., 2020 87	Randomized trial	OR	18‐60	C‐Mac	Not reported
Hoşten et al., 2012 89	Randomized trial	OR	Adults	Storz DCI	Not reported
Hodgetts et al., 2011 90	Randomized trial	OR	≥18	C‐Mac	Not reported
Hypes et al., 2016 97	Observational	ICU	Adults	C‐Mac, McGrath Mac	Not reported
Hypes et al., 2017 98	Observational	ICU	Adults	C‐Mac, McGrath Mac	Not reported
Ing et al., 2017 99	Randomized trial	OR	18‐80	McGrath Mac	Not reported
Ives et al., 2021 100	Observational	ICU	Neonates	C‐Mac	Not reported
Jain et al., 2016 101	Randomized trial	OR	18‐60	C‐Mac	Not reported
Javaherforooshzadeh and Gharacheh, 2020 102	Randomized trial	OR	1‐5	Infant view	Not reported
Ji et al., 2018 103	Randomized trial	OR	>18	C‐Mac	Not reported
Jones et al., 2013 104	Observational	Non‐OR or ED	Adults	C‐Mac	Not reported
Joshi et al., 2017 105	Observational	ICU	Adults	C‐Mac, McGrath Mac	Not reported
Kaji et al., 2019 8	Observational	ED	<16	C‐Mac	Not reported
Kaur et al., 2020 108	Randomized trial	OR	20‐70	McGrath Mac	Not reported
Kerrey et al., 2015 109	QI	ED	Children	C‐Mac	Not reported
Kido et al., 2015 110	Randomized trial	OR	20‐85	McGrath Mac	Not reported
Kilicaslan et al., 2014 111	Observational	OR	>18	C‐Mac	Not reported
Kim et al., 2016 112	Randomized trial	OR	3‐7	McGrath Mac	Not reported
Kim et al., 2018 113	Randomized trial	OR	1‐10	McGrath MAC	Not reported
Kleine‐Brueggeney et al., 2016 114	Randomized trial	OR	≥18	McGrath Mac	Not reported
Kleine‐Brueggeney et al., 2017 115	Randomized trial	OR	Adults	AP advance	Not reported
Komasawa et al., 2017 117	Randomized trial	OR	20‐85	McGrath Mac	Not reported
Kontouli et al., 2013 118	Interventional	OR	≥18	C‐Mac	Not reported
Koylu Gencay et al., 2019 119	Randomized trial	OR	<2	C‐Mac	Not reported
Kreutziger et al., 2019 120	Randomized trial	Pre‐hospital	≥18	McGrath Mac	Not reported
Law et al., 2015 121	Observational	OR	Adults	C‐Mac	Not reported
Lee et al., 2016 122	Randomized trial	OR	18‐60	C‐Mac	Not reported
Lees et al., 2013 123	Interventional	OR	<17	Storz DCI	Not reported
Lim et al., 2020 124	Randomized trial	OR	19‐65	McGrath Mac	Not reported
Lin et al., 2012 125	Randomized trial	OR	≥18	CEL‐100	Not reported
Lin et al., 2012 126	Interventional	OR	Adults	CEL‐100	Not reported
Loughnan et al., 2012 127	Interventional	OR	>18	C‐Mac	Not reported
Loughnan et al., 2019 128	Randomized trial	OR	≥18	McGrath Mac	Not reported
Louka et al., 2018 129	Interventional	Pre‐hospital	Did not specify	C‐Mac	Not reported
Louro et al., 2020 130	Observational	ED	All	C‐Mac	Not reported
Maassen et al., 2009 131	Randomized trial	OR	≥18	Storz DCI	Not reported
Maassen et al., 2012 132	Interventional	OR	Adults	C‐Mac	Not reported
Mackie et al., 2020 134	Observational	ED	≥15	C‐Mac	Not reported
Macnair et al., 2009 135	Randomized trial	OR	2‐16	Berci‐Kaplan VL	Not reported
Marsaban et al., 2017 137	Randomized trial	OR	18‐65	C‐Mac	Not reported
Michailidou et al., 2015 139	Observational	ED	All	C‐Mac	Not reported
Miller et al., 2020 140	Observational	ED	Children	C‐Mac	Not reported
Min et al., 2019 141	Observational	ED	≥18	C‐Mac	Not reported
Modir et al., 2017 142	Randomized trial	OR	>15	C‐Mac	Not reported
Monette et al., 2019 143	Observational	ED	All	C‐Mac	Not reported
Mosier et al., 2013 145	Observational	ICU	Adults	C‐Mac	Not reported
Moussa et al., 2016 146	Randomized trial	ICU	Neonates	C‐Mac	Not reported
Mutlak et al., 2014 147	Observational	OR	Weight‐based: ≤10 kg	Storz C‐MAC	Not reported
Naito et al., 2016 148	Observational	Pre‐hospital	All	C‐Mac	Not reported
Nakanishi et al., 2018 149	Randomized trial	OR	20‐85	McGrath Mac	Not reported
Narayan et al., 2018 150	Interventional	OR	Did not specify	Modified Mac VL	Not reported
Nausheen et al., 2019 151	Observational	Pre‐hospital	All	C‐Mac	Not reported
Ng et al., 2012 152	Randomized trial	OR	>18	C‐Mac	Not reported
Ninan et al., 2016 153	Randomized trial	OR	≥18	C‐Mac	Not reported
Noppens et al., 2012 154	Interventional	ICU	Adults	C‐Mac	Not reported
O'Connell et al., 2019 156	Observational	ED	<21	C‐Mac	Not reported
Okamoto et al., 2019 159	Observational	ED	≥18	C‐Mac	Not reported
Pacheco et al., 2019 160	Observational	ED	<18	C‐Mac	Not reported
Pallin et al., 2016 161	Observational	ED	≤15	C‐Mac	Not reported
Paul Weng et al., 2020 162	Observational	ED	All	C‐Mac	Not reported
Peyton et al., 2021 163	Observational	OR	<18	C‐Mac, Storz DCI, Glidescope Direct, Mcgrath Mac	Not reported
Pieters et al., 2015 165	Randomized trial	OR	≥18	C‐Mac	Not reported
Pouppirt et al., 2018 167	Observational	NICU	Neonates	C‐Mac	Not reported
Purugganan et al., 2012 168	Observational	OR	>18	C‐Mac, McGrath Mac	Not reported
Puthenveettil et al., 2021 169	Randomized trial	OR	18‐60	C‐Mac	Not reported
Rajasekhar et al., 2020 172	Randomized trial	OR	18‐60	C‐Mac	Not reported
Rhode et al., 2016 172	QI	Pre‐hospital	≥15	McGrath Mac	Not reported
Rope et al., 2008 173	Observational	OR	Did not specify	X‐Lite	Not reported
Rowland et al., 2019 174	Observational	OR	All	C‐Mac	Not reported
Sakles et al., 2014 177	Observational	ED	≥18	C‐Mac	Not reported
Sakles et al., 2015 178	Observational	ED	≥18	C‐Mac	Not reported
Sakles et al., 2016 180	Observational	ED	≥18	C‐Mac	Not reported
Sakles et al., 2017 182	Observational	ED	All	C‐Mac	Not reported
Sarkılar et al., 2015 184	Randomized trial	OR	>18	C‐Mac	Not reported
Schalk et al., 2012 185	Interventional	ED	Adults	C‐Mac	Not reported
Serocki et al., 2010 186	Interventional	OR	≥18	Storz DCI	Not reported
Shravanalakshmi et al., 2017 187	Randomized trial	OR	18‐60	C‐Mac	Not reported
Singh et al., 2017 188	Randomized trial	OR	1‐6	C‐Mac	Not reported
Sinha et al., 2016 189	Randomized trial	OR	Weight‐based: 3‐15 kg	C‐Mac Miller	Not reported
Sinha et al., 2019 190	Interventional	OR	4‐14	C‐Mac	Not reported
Sinha et al., 2019 191	Observational	OR	≤60 weeks corrected gestational age	C‐Mac	Not reported
Steel et al., 2021 193	QI	Pre‐hospital	All	McGrath Mac	Not reported
Sulser et al., 2016 194	Randomized trial	ED	18‐99	C‐Mac	Not reported
Suzuki et al., 2019 195	Observational	ED	All	McGrath Mac	Not reported
Swain et al., 2020196	Interventional	OR	18‐65	C‐Mac	Not reported
Thion et al., 2018 197	Randomized trial	OR	18‐80	McGrath Mac	Not reported
Vadi et al., 2016 198	Randomized trial	OR	<2	Storz DCI	Not reported
Van Oeveren et al., 2017 200	Observational	ED	All	C‐Mac	Not reported
Vanderhal et al., 2009 201	Observational	Delivery Room/NICU	Neonates	Modified Miller VL	Not reported
Vassiliadis et al., 2015 202	Observational	ED	All ages	C‐Mac	Not reported
Vlatten et al., 2009 203	Randomized trial	OR	≤4	Storz DCI	Not reported
Wallace et al., 2015 204	Randomized trial	OR	>16	McGrath MAC	Not reported
Wong et al., 2017 206	Observational	OR	≥21	McGrath Mac	Not reported
Yatim et al., 2015 207	Randomized trial	OR	Did not specify, primarily adults	C‐Mac	Not reported
Yokose et al., 2016 208	Observational	OR	≥18	McGrath Mac	Not reported
Yoon et al., 2020 209	Randomized trial	OR	20‐80	McGrath Mac	Not reported
Yoon et al., 2020 210	Observational	OR	20‐80	McGrath Mac	Not reported
Yumul et al., 2016 211	Randomized trial	OR	18‐80	C‐Mac	Not reported
Yumul et al., 2016 212	Randomized trial	OR	18‐80	Video‐Mac	Not reported

Abbreviations: DCI, direct coupled interface; ED, emergency department; ICU, intensive care unit; NICU, neonatal intensive care unit; OR, operating room; QI, quality improvement; VL, video laryngoscope.

Main outcomes

Among the 207 studies included, only 21 studies (10%) reported outcome data for video screen visualization. Nineteen of these studies (9%) included only basic outcome data—that a screen visualization occurred or not (dichotomous), without further detail.

Only 2 studies (1%) included more detailed outcome data on video screen visualization. In the first, Guyette et al. used proceduralist self‐reporting to characterize screen visualization as only screen, mostly screen, mostly direct, or only direct. In the second study, Driver et al. used proceduralist self‐report to characterize screen visualization as screen never used, screen viewed for entire attempt, or screen viewed during passage of tube/bougie into glottis.

In addition to the 21 studies in which outcome data were explicitly reported, in 11 additional studies the approach to video screen visualization could be inferred based on screen availability to the proceduralist. In an additional 16 studies, methodologic data were available via a protocol in the methods section for how visualization was expected to be performed without reporting data collection on how the proceduralist actually interacted with the video screen. No identified study reported more specific patterns of video screen visualization, including duration of video screen viewing, proportion of attempts spent viewing the screen, and gaze switches between the video screen and the patient.

Methods of data collection

The method of data collection in the 207 included studies varied widely (Table 2). Of the 21 studies that reported outcome data for screen visualization, 19 were based on some version of self‐report, including the studies by Guyette et al. and Driver et al. Only 2 studies were based on video review, both overhead/external. Law et al. used eye‐tracking technology during intubation attempts. In this study, only 2 intubations were performed with a video laryngoscope, and although visualization between patient and equipment was reported, direct visualization versus indirect visualization was not characterized.

Risk of bias

In total, the 48 studies in which outcome or methodologic data were reported were assessed for risk of bias. The majority of assessed studies (n = 40) were categorized as high risk of bias. Eight studies were categorized as low risk and zero studies were categorized as unclear risk.

LIMITATIONS

Our study has several limitations. First, our original objective was describing the various approaches to video screen visualization in the published literature on VL, but the rarity of detailed information on the use of the video laryngoscope prevented us from accomplishing this objective. Second, although we attempted to maximize sensitivity in our search terms and definitions, it is likely that additional studies exist in which a video laryngoscope with a standard geometry blade was used that were not included. However, we feel confident that based on the nature of the search strategy, the vast majority of studies evaluating VL as an intervention (and not just a study in which a video laryngoscope happened to be used) were identified and included. Finally, because our study group has extensive experience with video‐based data collection across numerous studies focused on tracheal intubation, our interpretation of studies and the emphasis we placed on the need for video‐based data collection is subject to potential biases.

DISCUSSION

This systematic review demonstrates that in the majority of published research in which a video laryngoscope equipped with a standard geometry blade was used, video screen visualization patterns were not reported whatsoever. In the minority (10%) of studies in which video screen visualization patterns were reported or others (5%) in which video screen visualization patterns could be reasonably inferred based on screen availability, video screen visualization was almost always reported in a yes/no fashion.

Cochrane reviews present mixed although generally favorable results for VL in terms of procedural success. Studies that have reported on patient outcomes with VL generally have reported either no difference or a reduction in adverse events when compared to DL. , , , However, it is difficult to nearly impossible to interpret the true effects of an intervention when how the intervention was actually used is incompletely reported or not reported whatsoever. Knowing what investigators mean by VL is essential to interpreting the actual “treatment effect.” Product manufacturers have recommended a four‐step technique for performing VL, highlighted by first inserting the video laryngoscope blade while looking into the patient's oropharynx, then indirectly viewing the epiglottis by viewing the video screen, then looking back at the patient's oropharynx for endotracheal tube insertion into the mouth, then indirectly viewing the completion of tube delivery by viewing the video screen again. This technique is rarely described in the published literature, and it is often unclear how video laryngoscopes equipped with standard geometry blades were functionally used in studies. There are a wide range of possibilities of how these devices could have been used in studies, including using a video laryngoscope only as a teaching tool with supervisors viewing the video screen while the proceduralist functionally uses the device to directly view the glottis, performing primarily direct glottic visualization but using the video screen as a backup in the event of poor direct glottic visualization, performing primarily indirect glottic visualization throughout both laryngoscopy and tube delivery and a myriad of hybrid approaches combining these techniques. This variation in technique is likely greater in acute care settings (EDs and intensive care units), as attempts are more likely to be complicated by patient instability, oropharyngeal trauma, gastric contents in the oropharynx, and pulmonary edema, all of which can contribute to deviations from planned airway approaches and unplanned microbehaviors (ie, gaze switches) that are not necessarily intentional.

Valid data collection is essential to good research. Valid data collection during tracheal intubation, in particular when assessing the impact of VL, may require independent observers or video‐based data collection, especially outside of controlled settings. Because video‐based data collection is not always feasible (because of equipment availability, location of study, or size of study), valid data collection can be performed with self‐reporting/chart review. However, a detailed description of the approach to data collection is needed, and assessing the reliability of data collection is essential.

Our review suggests that the entirety of the available literature on VL, across all settings, has substantial and potentially significant flaws. Prominent among these are a lack of detailed description of data collection, a lack of confirmation that a prescribed airway approach was actually performed, the use of self‐reporting after emergency airway management, and a lack of specification of what VL actually means. We believe the results of our review should directly inform all future studies, in particular randomized trials, of video laryngoscopes, in all settings. We suggest the following be required components of all future video laryngoscope trials: detailed description of data collection, either assessment of the reliability of self‐reporting for airway outcomes or (preferably) the use of video‐based data collection, specification of what video/indirect/DL actually mean, and assessment during data collection of whether these pre‐defined approaches were actually followed.

In conclusion, video screen visualization patterns are often incompletely reported or not reported at all in the published literature. Because of this, it is often difficult to nearly impossible to assess the impact of video laryngoscopes as an intervention. Future studies focused on VL should include detailed assessments of video screen visualization patterns (including fine details such as duration of screening viewing, proportion of attempts spent viewing screen and number of gaze switches between the patient and video screen), detailed descriptions of data collection, and use of objective methods of data collection (ie, video‐based) when available.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.

AUTHOR CONTRIBUTIONS

PD and BK conceived and designed the study, performed literature search, data extraction, data analysis, and drafting of the manuscript. PD takes full responsibility for the paper as a whole.