A Quantitative Analysis of Smartphone-Based Endoscopy and Video Tower Endoscopy

Introduction

Laryngoscopy and nasal endoscopy are 2 key bedside procedures performed as a component of the complete otolaryngologic exam with 56% of inpatient consultations receiving laryngoscopy and 11% receiving nasal endoscopy. ¹ Laryngology, rhinology, and head and neck oncology patients, both adult and pediatric, regularly require endoscopic evaluation in an outpatient setting as well to guide management of their complaints, often with necessary comparison to previous endoscopic evaluations. These endoscopies may be recorded to allow review by the patient, otolaryngologist, and consulting teams. In academic centers, endoscopies performed by residents but not recorded may necessitate subsequent endoscopies for the purposes of accurate diagnosis, clinical decision making, and accurate billing by the attending physician.^{2 -8}

Traditional recording is done with a large, wheeled video tower. These videos then require transfer to a HIPAA-compliant storage or into the patient’s electronic medical record (EMR). Smartphone-based endoscopy (SBE) is an emerging technology which has been examined in adult populations as having equivocal quality in non-randomized, non-blinded laryngoscopic images ⁸ ; however rigorous image-quality studies and business-level analysis such as cost and time expenditure differences between these methods have not been described in the literature. We sought to explore these data and determine potential differences between these 2 methods.

Methods

This research was determined to be exempt from institutional review board approval due to not being human subject research. Comparison was performed with the use of a Karl Storz fiberoptic laryngoscope (El Segundo, CA). SBE recordings were made on an Apple iPhone (Cupertino, CA) with the use of the Save My Scope adapter (Chicago, IL). Tower-based endoscopy (TBE) was conducted with the use of the Karl Storz 4K HD portable endoscopic visualization and recording system (El Segundo, CA). Purchase price was extracted from purchase orders for either system. SBE total cost included both the cost of the Save My Scope product, along with the cost of a new dedicated iPhone Xs smartphone (as priced on initial public release), and a portable Karl Storz (El Segundo, CA) fiberoptic light source. HIPAA-compliant server storage cost was excluded from analysis, as costs were expected to be similar for both modalities.

Setup time was recorded among 10 otolaryngology care providers: 2 attending physicians, 2 fellows, 4 residents, 2 certified physician’s assistants (PA-C). Time was recorded from entering a simulated patient room to visual confirmation of an image on the screen (either TBE incorporated screen or smartphone screen) with the image focused. Time was recorded in seconds and repeated for each modality with each care provider. Providers were blinded as to the endoscopy recording format (SBE vs TBE) prior to entering the room and a random number generator determined the first setup conducted for each provider. The same laryngoscope was used for all trials.

Visual examination of both mechanisms was performed using a modified Snellen test ⁹ visual with fiberoptic laryngoscope set up (Figure 1). The end of the laryngoscope was suspended 50 mm above the paper and focused to maximize detail, verifying similar levels of discernment between the 2 writing samples. TBE was set up and cast on the 1080p HD monitor included with the standard set after placing Snellen test 1 (Figure 1a) in the area of visualization. An Apple iPhone Xs was then used to record a picture of the screen in a dark room with the auto-focus mechanism of the built-in camera application without the use of digital or optical zoom in order to simulate electronic transmission and storage of images. This modality for recording of the TBE image was specifically chosen to simulate rapid transmission of images over text message or upload to the clinical chart via a smartphone app. The TBE was then detached from the suspended fiberoptic laryngoscope and Snellen test 2 (Figure 1b) placed in the recording position. Photographs were obtained with the use of the same Apple iPhone Xs using the Save My Scope software. SBE and TBE photos were cropped for similar dimensions and sent via text message to 20 otolaryngology care providers (1 PA-C, 7 residents [PGY1-5], 2 fellows, 10 attendings varying in years of practice from 1 to 27).

OPEN IN VIEWER

Figure 1.

Snellen test 1 (a), and Snellen test 2 (b) along with photography recording setup (c).

These 20 providers were randomized to receive either TBE or SBE photos first via a random number generator, and instructed to respond via text message with the smallest line they were able to read in either photo. No indication was given as to which photo was from which modality. Providers were permitted to zoom, rotate, or change contrast and brightness on their phone as they desired to maximize their visualization, again in attempts to simulate the action taken with an actual consultation. Visualization of the text was determined to be adequate if 4/5 letters were able to be detected, congruent with standard Snellen scoring. If less than 4/5 were detected, providers were asked to attempt again with the line above written in the next larger font.

Statistical analysis for both setup time (quantified in seconds) and level of discernment (quantified by font size of discernment) was performed using SPSS (IBM-Armonk, NY) with paired-samples T tests. Kurtosis and skewness were calculated for discernment level visual testing to ensure relative conformality between providers.

Results

In bulk ordering at the primary institution, the TBE system was delivered at a cost of $41,667 per unit while the SBE system cost was $1750, a savings per unit of $39,917 (95.8%). This represents the cost of one system, although in our practice environment, a total of 6 systems are needed in order to cover both the outpatient clinic and inpatient consult service at the base hospital. Thus, at this particular institution total savings would be $239,502.

In time trials for setup of equipment, 100% (10/10) of providers completed the task successfully for both modalities. Median time for TBE setup was 61.5 seconds (interquartile range: 57.0-95.0) and median time for SBE setup was 23.5 (interquartile range: 22.0-27.0). Visual representation as a box-and-whiskers plot of both modalities is shown in Figure 2. Testing showed a statistically significant difference between the 2 methods with an estimated time savings of 46.7 seconds (95% CI: 30.3-63.1 seconds) per setup.

OPEN IN VIEWER

Figure 2.

Setup time comparison between smartphone-based endoscopy (SBE) and tower-based endoscopy (TBE).

Visual discernment testing was completed by 100% of providers (20/20) with the modified Snellen test. For SBE, the median font size of successful detection for all providers was 4.23 mm (interquartile value: 4.23 mm) and for TBE was 5.64 mm (interquartile value: 5.64 mm). Fisher-Pearson coefficient of skewness was measured for both samples and was 0.0 for SBE and 1.64 for TBE. Kurtosis was calculated for TBE as 0.70, but could not be calculated for SBE as all providers were unanimous in their level of visual detection. Histogram representation of these results is shown in Figure 3.

OPEN IN VIEWER

Figure 3.

Histogram comparison between size of smallest discernable letters with a modified Snellen’s test using images obtained from smartphone-based endoscopy system (SBE, right column) and tower-based endoscopy system (TBE, left column).

Discussion

This study examines the cost, time investment, and transmitted visual acuity of tower-based endoscopy as compared to smartphone-based endoscopy. Although non-inferiority in un-blinded video data has previously been published, ⁸ we sought to examine the data in a different manner. This blinded study is the first to our knowledge to examine the level of visual discernment between SBE and TBE, report on cost differences, and examine setup time variation between the modalities.

Our results regarding visual discernment outcomes between SBE and TBE are similar to those found in other studies external to otolaryngology. The use of various smartphone-based visualization systems has been evaluated and found to be equivalent or better to traditional visualization systems in nasolaryngoscopy, ¹⁰ uroscopy, ¹¹ colposcopy, ¹² and laparoscopy ¹³ when evaluating clinical photographs and visuals. Our data were collected to add to the current body of research, allow for a scaled rather than binary measure of visual perception, and to investigate a particular SBE system which had not previously been examined. Our data suggest there is a statistically significant increase in visual acuity of SBE when compared to TBE, congruent with these previous studies. However, due to the discordance in these studies between equivalency and superiority of SBE, this statistical significance may not always reach the level of clinical significance depending on specific images and anatomic area being observed. Many subjects within the study mentioned in follow up questioning they felt the ability of the SBE system to remove negative space from between fiberoptic cables (Figure 4) helped with discernment in smaller text on the modified Snellen test.

OPEN IN VIEWER

Figure 4.

Representative pictures from laryngoscopy performed with smartphone-based endoscopy and tower-based endoscopy systems.

There are important factors to consider when discerning between the 2 options of endoscopic recording systems. Although in an outpatient clinic setting the transport time with the equipment is not important (assuming the room is appropriately set up), an inpatient consult team may have other factors which influence their choice between the 2. Not only does TBE require an elevator to move between floors while an SBE system could fit in the pocket of a white coat, but maneuverability may be difficult with the heavier TBE system. While our study purposefully did not examine this feature in order to limit confounding, clinicians making a decision between these systems for an inpatient service may find it a valuable consideration. Our data do show that even within the limitations of time spent in the room, there is a significant time difference. Though this savings is statistically significant, the clinical significance of this change will vary between practitioners. A savings of 47 seconds of time per patient on in-room time alone was felt to be significant by our practice providers when queried, however this may vary between clinicians and practices. Although not examined in the context of this study, the ergonomics of use are different for TBE and SBE and should be a consideration for providers. SBE is not limited by the use of a cord (unless a light box is used) however the weight of the phone may modify the way in which the scope complex is held, which may change providers’ comfort with the use of this particular tool.

The pediatric patient presents a unique challenge to in-office laryngoscopy which is not isolated to patients younger than 18, but is more prominent than adults. Procedures can be highly anxiety-inducing in pediatric patients, and the frequent need to assess dynamic function of the larynx makes it difficult to permit any form of sedation during this procedure.^{5 -7,14} As such, it is especially pertinent in pediatric populations to reduce excessive procedures and fear.^15,16 Although the use of anxiolytics in an office setting for laryngoscopy has been described, ¹⁷ this could preclude the safety of certain procedures such as flexible endoscopic evaluation of swallowing (FEES). Pediatric anxiety can be reduced through the use of familiar objects, and introduction of an item prior to a procedure can significantly reduce anxiety and increase compliance. ¹⁸ With recent data showing 76% of households containing at least one smart phone and 96% of households having a cell phone, ¹⁹ children are frequently exposed to the technology involved with SBE, potentially leading to less anxiety and greater comfort on their part during procedural visits. This effect has been explicitly shown with the use of technology such as smartphones in adolescent and teenaged populations. ²⁰ Although there are no data on the effect of these 2 modalities on pediatric patient anxiety, it is possible this may reduce improve pediatric patient comfort, but further studies more directed toward this aim are needed.

Limitations to our study include our selection of both SBE and TBE equipment. It is possible that with different systems, these results may differ. Additionally, our study focused on fiberoptic evaluation, while some institutions have switched to distal-chip flexible laryngoscopes. Further studies will be needed to examine whether the outcomes of this study are maintained between a fiberoptic laryngoscope using an SBE format, and a TBE system which uses a distal chip laryngoscope. Within our study, we specifically chose to record the TBE images with the use of a smart phone in order to most closely imitate transmission of images over text, however if there is cross compatibility such that images can be directly transmitted from the TBE system, the imaging differences may not be the same. Our study does not account for certain features of TBE that are not readily available in SBE, such as video-stroboscopy. Clinicians should first discern whether their clinical needs could be met with SBE before seeking to compare it to their current systems. Due to the nature of testing for setup time, participants could not be fully blinded in the timing portion of our data collection, although participants were blinded as to the modality they would be using prior to entering the room. Finally, members of the clinical team participating in the timing portion of this study use TBE as their standard of practice. As such the effect size may have been dampened due to this familiarity with the TBE system.

Conclusion

Endoscopic procedures are a cornerstone of the otolaryngologic physical examination, and more than half of inpatient and emergency department consults necessitate endoscopic evaluation of the upper aerodigestive tract. Recording of the endoscopy increases patient satisfaction and allows review and inter-practitioner collaboration. We examined objective data on TBE and SBE systems in order to provide accurate information for otolaryngologists and otolaryngology practice managers to make appropriate purchasing decisions. Smartphone-based endoscopy offered decreased setup time, lower cost, and improved visual discernment in transmitted images via messaging when compared to traditional tower-based endoscopy.