An ultrashort video can teach residents to perform a fingertip injury repair

Abstract

Background

Acute fingertip injuries are common. Providers in rural and underserved areas often transfer these patients due to lack of comfort and skill with treating these injuries. Current learners prefer short and high‐density educational material. It is unknown if basic hand procedures can be taught using ultrashort training videos. This study investigates whether fingertip repair can be taught using a 60‐second educational video viewed immediately prior to performing the procedure.

Methods

A standardized cadaveric fingertip injury model was developed. Twenty‐three emergency medicine residents each having minimal experience with fingertip injury repair were randomized into one of three study arms: A) no video, B) standard‐length (8‐minute) video, and C) ultrashort (60‐second) video. Each subject was presented with an injured cadaveric finger and asked to prepare for and perform the repair within a 30‐minute time frame. The repair was graded on a 10‐point scale following a standard rubric. Time to completion, preparedness, and subjects’ confidence were also assessed. Results were analyzed by one‐way ANOVA and Kruskal‐Wallis tests.

Results

Mean repair scores for the standard‐length video group (9.5 ± 0.3) and the ultrashort video group (9.2 ± 0.3) were significantly higher than those of the no video group (4.0 ± 0.3, p < 0.05 for both comparisons). Mean time to completion of the exercise was significantly shorter in the ultrashort video group (19 ± 2 minutes) than in the standard‐length video group (26 ± 2 minutes). Subject‐reported outcomes (median preparedness, median post‐repair confidence, and median change in confidence following the procedure) were all significantly higher in the standard‐length video group and the ultrashort video group than in the no video group (p < 0.05 for all comparisons).

Conclusion

A 60‐second educational video viewed immediately prior to performing a fingertip injury repair can effectively teach an emergency medicine resident to correctly perform the procedure.

INTRODUCTION

Acute fingertip injuries commonly occur when fingers are subjected to various crush mechanisms. ¹ Most have a common injury pattern: an avulsed nail plate from the eponychial fold, a transverse nailbed laceration, and an open distal phalanx fracture. ² Providers in rural and underserved areas frequently encounter these injuries but often refer their patients to hand specialists due to minimal training and comfort with repairing these injuries. This results in long travel times for the patients, unnecessary consumption of medical resources, and increased medical costs. Repair of these acute fingertip injuries is a relatively simple procedure that requires only basic medical supplies: local anesthetic, laceration repair kit, suture, and dressings. These items are readily available in every emergency room and urgent care center, and thus the missing elements are education and confidence.

Using videos for surgical education is becoming increasingly common. Studies cite factors such as increased skill, ³ , ⁴ increased confidence, ⁵ and the changing educational expectations of the millennial generation ⁶ when explaining the use of video for education of trainees. When using video to prepare for surgeries, residents frequently use YouTube as their preferred source of videos. ⁷ Unfortunately, there is no regulation or quality control for videos intended for surgical education. When searching for YouTube videos, practitioners are bombarded with long videos of low quality. ⁸ While high‐quality educational videos on procedural techniques are available through other sources, many require logins or subscriptions. Most are still too long and are not practical to use at the point of care. Videos with longer total run times have been shown to reduce student engagement, with steep engagement drop‐offs just after 2 minutes. ⁹ According to a recent study on the development of effective educational videos in emergency medicine, videos should be made as short as possible, as they are used “just‐in‐time,” i.e., at the point of care as the patient is being actively treated. ¹⁰ Thus, there is demand for readily available, high‐quality educational videos that enable a practitioner to immediately perform a procedure correctly and with confidence.

It is not known whether basic hand procedures can be taught using ultrashort training videos. In this study, we tested the effectiveness of a 60‐second educational video on fingertip injury repair using a “just‐in‐time” training approach, where the trainee is educated immediately prior to performing the procedure. Just‐in‐time training has been shown to enhance trainee learning and confidence. ¹¹ In a controlled naturalistic setting, we hypothesize that a 60‐second “just‐in‐time” educational video can successfully train emergency medicine residents to correctly perform a fingertip injury repair with confidence.

METHODS

Injury pattern

A standardized cadaveric model for fingertip injuries was created to test the central hypothesis of this study. In contrast to prior studies that approximated fractures using a bone saw, ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ we chose to approximate the injury mechanistically. We created fingertip injuries using a novel device that simulates catching a finger in a closing door. The resulting fingertip injury has four components: injury to the pulp, avulsion of the nail plate from the eponychial fold, a transverse nailbed laceration, and an open distal phalanx fracture. All fingers underwent visual investigation and fluoroscopic imaging to confirm the desired injury. The severity of the fingertip injuries was considered similar across all fingers by the hand surgeons. While the device did not undergo any formal validation or piloting, it was deemed reliable by our hand surgeons since it consistently produced a fingertip injury pattern similar to what is seen in clinical practice. See Video S1 for a description of the device and the injury pattern.

Educational videos

The educational videos used in this study were created by our board‐certified hand surgeons using research‐proven video techniques. The videos outline the steps of the fingertip repair procedure using the principles outlined in the Green's Operative Hand Surgery textbook (same principles used to create the fingertip repair grading rubric). The videos displayed a “first‐person view,” which has been shown to be effective in training medical professionals on procedural skills. ¹⁷ An 8‐minute educational video describing the fingertip injury repair procedure was first created and designated as the “standard‐length video” (Video S2). A 60‐second video was then created by shortening the standard‐length video to focus on the key procedural steps; this video was designated the “ultrashort video” (Video S3). The longer video was modeled after common medical training videos found on various sites and used additional time to show and discuss the setup, supply list, and entirety of each step. Minimal editing was performed of the long video. The shorter video was created from the long video by aggressive editing, speeding up the video to 1.5x speed, deleting repeated steps, narrating off a script, and allowing some information to be implied rather than stated explicitly. The goal was to make the video as short as possible since recent studies have shown that shorter educational videos are more effective, particularly in the emergency medicine setting, and more likely to maintain a learner's attention. ¹⁰

Study design

Following institutional review board approval, the nonphysician program manager of our EM residency program (not an investigator of this study) sent an email to all residents seeking interest to participate in a one‐hour educational study during one of their research/skills half‐days. Twenty‐three of the 36 total EM residents (PGY1‐3) were available and verbally consented to participate in this study. The study was conducted on three different days (~8 subjects per day) given limitations in space. During each day, the subjects performed the simulation at the same time and, during the consent process, agreed to not discuss any information relating to the study with their co‐residents until completion of the entire study. Each resident enrolled in the study reported proficient suturing skills and minimal fingertip injury repair experience on the pre‐repair/screening questionnaire (5‐point Likert scale; see Appendix S1). Exclusion criteria included lack of proficiency in suturing and/or mastery in fingertip repair. The EM residents were then simply randomized into one of three study groups:

1. No video (n = 7)

2. Standard‐length video (n = 8)

3. Ultrashort video (n = 8)

The study was designed to be a simulation of an emergency room scenario. Subjects in all groups were presented with a clinical vignette, shown the standardized cadaveric fingertip injury and available medical supplies (Figure 1A). The fingertip injury models were randomly distributed to the subjects. The subjects were then given access to their assigned video. The no‐video group was notified they had no additional material available to them. Subjects in the standard‐length video and ultrashort video groups were permitted to watch their assigned videos via a YouTube link on a desktop computer that was provided. YouTube has the capability of skipping through the video, replaying the video, and slowing down or speeding up the playback speed. Subjects in all groups were given 30 minutes to prepare for and perform the fingertip injury repair, including any time spent watching videos. When ready, subjects proceeded to a simulated exam room to perform the procedure. This setup was designed to model a real‐life clinical scenario where emergency practitioners feel the constraints of time pressures but have free access to educational modalities. Video recordings were captured throughout the study to assess time management and score the repair (Figure 1B). Conduct of these research procedures was supported by the American Foundation for Surgery of the Hand Grant #2820.

FIGURE 1.

Testing setup. (A) Study subjects were provided with a complete laceration kit, absorbable and nonabsorbable sutures, finger tourniquet, and dressing material. (B) An overhead camera monitored activity throughout the study

Outcome measures

Each subject had already completed a pre‐repair questionnaire prior to randomization to determine their baseline confidence in performing the repair (Appendix S1). The fingertip injury repair was graded on a 10‐point scale using a novel unvalidated instrument created by our hand surgeon authors based on the principles outlined in pages 335–336 of Green's Operative Hand Surgery textbook (Table 1). ² Repairs were graded by using a combination of watching the repairs recorded on video and inspecting the final repaired fingertip. A single hand surgeon evaluated all final fingertip repairs and confirmed the scores in a blinded fashion. The medical student and research assistant knew which videos the subjects used, but the hand surgeon did not. Time to completion of the study was also recorded. Each subject then completed a post‐repair questionnaire (5‐point Likert scale) immediately following the repair to assess level of preparedness and confidence with performing a fingertip injury repair (Appendix S2).

TABLE 1.

Fingertip injury repair scoring rubric (10‐point scale)

Procedural step	Performed correctly (+2)	Performed incorrectly (+1)	Did not perform (+0)
1. Application of finger tourniquet	□	□	□
2. Removal of nail	□	□	□
3. Lavage of laceration & eponychial fold	□	□	□
4. Closure of lateral nail folds & volar pulp	□	□	□
5. Closure of nailbed laceration	□	□	□
6. Proper use of sutures (nonabsorbable for lateral nail folds/pulp, absorbable for nailbed)	□	□	□
7. Removal of finger tourniquet	□	□	□
8. Application of bacitracin	□	□	□
9. Placement of adaptic dressing correctly under the eponychial fold	□	□	□
10. Proper dressing application	□	□	□
Total score:

^*Raw score can be converted to the adjusted score by the following equation: $Adjustedscore=\left(\frac{rawscore}{20}\right)\times 10$.

Data analysis

An a priori power calculation was initially performed to inform sample size, although an interim decision was made to stop collecting data due to overpowering and large effect sizes observed in our desired outcome measures. One‐way ANOVA analyses were performed to ascertain differences in fingertip repair score and time to completion (continuous variables) among the three study groups. The Kruskal‐Wallis test was performed to assess for differences in pre‐ and post‐repair confidence levels and self‐assessed level of preparedness (ordinal variables). p < 0.05 was considered significant for both tests. When ANOVA and Kruskal‐Wallis tests were significant, post hoc analysis using the Tukey's honestly significant difference (HSD) and Wilcoxon rank sum tests was used for pairwise comparisons and to detect where the significant difference lies. Values reported in the text are results from these post hoc significance tests (summarized in Appendix S3 ). Continuous variables are reported as mean ± standard error, and ordinal variables are reported as median (interquartile range).

RESULTS

The median reported confidence levels before the repair was similar among the three groups (Table 2). Subjects in the standard‐length video group took on average 26 minutes (±2 minutes) to complete the exercise, an estimated 7 minutes longer (95% CI 1–13 minutes) than subjects in the ultrashort video group and subjects in the no video group, as demonstrated in Figure 2. Subjects in the standard‐length video group spent a mean 6 minutes, 10 seconds (±30 seconds) watching the standard‐length video; subjects in the ultrashort video group spent a mean 1 min, 50 seconds (±20 seconds) watching the ultrashort video.

TABLE 2.

Comparing education groups by variable

Variable	Overall sample (N = 23)	No video group (N = 7)	Standard‐length video group (N = 8)	Ultrashort video group (N = 8)	p‐value
Confidence before the repair (out of 5) Median (IQR)	3 (2.5–3)	3 (3–3.5)	3 (2.75–3.25)	3 (1.75–3)	0.22 b
Time taken to complete the repair (minutes) Mean (SE)	21 (1)	19 (2)	26 (2)	19 (2)	<0.05 a
Repair score (out of 10) Mean (SE)	7.7 (0.6)	4.0 (0.3)	9.5 (0.3)	9.2 (0.3)	<0.05 a
Self‐assessed level of preparedness (out of 5) Median (IQR)	4 (3.5–5)	2 (1.5–2.5)	4 (4–5)	4.5 (4–5)	<0.05 b
Confidence after the repair (out of 5) Median (IQR)	4 (3–5)	3 (2–3)	4 (3.75–4.25)	4.5 (4–5)	<0.05 b
Change in confidence (out of 5) Median (IQR)	1 (0–2)	0 (−0.5 to 0)	1 (0.75–1.25)	2 (2–2.25)	<0.05 b

All values reported to the first digit of uncertainty.

Abbreviation: IQR, Interquartile range; SE, standard error.

^a One‐way ANOVA test.

^b Kruskal‐Wallis rank sum test.

FIGURE 2.

Time to complete exercise by group. The standard‐length video group took significantly longer to complete the repair than the no video group or the ultrashort video group. Bar chart shows the estimated mean, while error bars show the 95% confidence interval

Mean repair score was significantly higher for subjects in the standard‐length video group (9.5 ± 0.3) and ultrashort video group (9.2 ± 0.3) than for the subjects in the no video group (4.0 ± 0.3, p < 0.05 for both comparisons), as demonstrated in Figure 3. There was no significant difference in repair scores between the two video groups. Median level of preparedness was higher in the standard‐length video group (4, IQR: 4–5) and the ultrashort video group (4.5, IQR: 4–5) than in the no video group (2, IQR: 1.5–2.5, p < 0.05 for both comparisons), as demonstrated in Figure 4.

FIGURE 3.

Repair score by group. The standard‐length video group and the ultrashort video group had significantly higher repair scores than the no video group. Bar chart shows the estimated mean, while error bars show the 95% confidence interval

FIGURE 4.

Self‐assessed level of preparedness by group. The standard‐length video group and the ultrashort video group had significantly higher self‐assessed level of preparedness than the no video group. Box‐and‐whisker plots show the median and quartiles

The median post‐repair confidence of the ultrashort video group (4.5, IQR: 4–5) and standard‐length video group (4, IQR: 3.75–4.25) were similar, while both scores were significantly higher than the median score of the no video group (3, IQR: 2–3, p < 0.05 for both). This is illustrated in Figure 5A. Finally, the change in confidence levels was determined, revealing significant differences between each pair of groups, as shown in Figure 5B. The increases in confidence in the standard‐length video group (+1, IQR: 0.75–1.25) and the ultrashort video group (+2, IQR: 2–2.25) were significantly different from zero, indicating a significant increase in confidence in these groups. The change in confidence in the no video group was not significantly different from zero.

FIGURE 5.

Confidence by group. (A) The standard‐length video group and the ultrashort video group had significantly higher confidence after the repair than the no video group. (B) The change in confidence was significantly larger in the ultrashort video group than in the standard‐length video group or the no video group. The change in confidence was significantly larger in the standard‐length video group than in the no video group. Box‐and‐whisker plots show the median, quartiles, and outliers

DISCUSSION

Medical professionals commonly use videos, among other learning tools, when seeking education on procedural techniques. However, the most frequently used video sources often provide videos of unknown educational quality or may be too long for a busy practitioner to watch during a clinical encounter. Most prior studies on the use of videos for surgical education have focused on long, comprehensive videos. However, a lengthy video is not necessary to improve procedural skills for minor procedures requiring only local anesthesia; Rowse et al. showed that a 90‐second instructional video viewed by general surgery residents significantly improved their performance of fine‐needle aspiration and showed that a short video can be used to increase comfort and confidence in performing a procedure for which the subjects already have basic skills (e.g., instrument handling). ⁶

In this study, the research subjects already had basic suturing proficiency and familiarity with the medical supplies. We were able to reduce the 8‐minute video down to 60 seconds by editing out content such as the supply list (supplies are seen as they are used), steps that were redundant (multiple passes of the suture needle) or showed skills they already had (tying sutures down). We also sped up the video to 1.5x speed and narrated faster. Thus, the 60‐second video contained only steps that were unfamiliar and needed to be learned.

We found that the standard‐length video group took significantly longer to complete the fingertip injury repair than either of the other two groups due to increased time spent watching the video. These subjects spent a mean 6 minutes, 14 seconds watching the video, despite a video length of 8 minutes. Many subjects in this group were observed to increase the playback speed and skip parts of the video. In contrast, subjects in the ultrashort video group spent a mean 1 minute, 50 seconds preparing with the 60‐second video, mostly due to re‐watching the video. We believe this behavior likely stems from the time constraints and quicker learning rates among current medical trainees.

Despite the increased time that the standard‐length video group spent in preparation, both the standard‐length and ultrashort video groups demonstrated similar competency in performing the procedure, as evidenced by nearly identical fingertip injury repair scores.

LIMITATIONS

One limitation of this study is the use of novel cadaveric injury models, which may call into question the relevance and generalizability of the findings for healthcare practitioners. However, use of a cadaveric model was the best available approach to answer our primary research question in a safe, ethical, and timely manner. Another limitation of this study is the variability in the subjects’ training levels and prior exposure to fingertip injury repairs given that this study included emergency medicine residents with experience levels ranging from PGY1‐3. However, subjects who reported mastery of the fingertip injury repair technique would have been excluded from the study. Additionally, we noticed that the senior and junior residents who participated reported similar experience levels with fingertip repairs and in fact scored similarly within each group (including the no‐video group, which assessed baseline experience level since this group did not receive any education). This may be due to infrequent exposure to these injuries and/or care often being referred to the hand specialists. While the PGY3s will have more experience in other areas, particularly suturing, we felt that the junior residents would still be proficient enough so that they would not be at a significant disadvantage. Despite this, the variability in the subjects’ suturing skills and their familiarity with the different suture materials is a limitation. Future educational videos should also specify both the suture type and size used. Additionally, while the no‐video group had access to a computer during their preparation period, we did not state whether they could or could not use it for preparation. In the future, we should specifically inform this group that they may access the computer during their allotted time to reflect “real” practice. Finally, the assessment instrument for scoring fingertip injury repairs was novel and needs to be incorporated in subsequent studies to build validity evidence. However, we believe this study supports internal validity for this instrument as there was responsiveness from the video groups compared to the noneducation groups. One additional limitation of this novel assessment instrument is that only one hand surgeon used it to evaluate the final fingertip repairs as opposed to multiple raters and, hence, does not allow for interrater reliability evaluation. Future studies should investigate the interrater reliability of multiple hand surgeons in multiple contexts when using this assessment instrument.

CONCLUSIONS

An increased supply of high‐quality, ultrashort educational videos on basic techniques and procedures requiring only local anesthesia could encourage providers to seek education and successfully acquire competence at the point of care. Future studies would need to investigate the effectiveness of ultrashort training videos in a live clinical setting and whether it could reduce unnecessary referrals, which is particularly costly and time consuming for patients coming from rural areas.

CONFLICT OF INTEREST

P.A. reports grant money from the American Foundation for Surgery of the Hand Award #2820 to Department of Orthopaedic Surgery, Carilion Clinic to conduct research conceived and written by Peter J. Apel from Department of Orthopaedic Surgery, Carilion Clinic. We have no intention of seeking financial gains from this educational study. Hence, all authors (P.A., Y.A., N.O., L.S., D.K., and H.H.) report no conflict of interest.

AUTHOR CONTRIBUTIONS

Y.A. – study concept and design, acquisition of the data, analysis and interpretation of data, drafting of manuscript. N.O. – administrative/technical support, study concept and design, obtained funding, analysis and interpretation of data, drafting of manuscript. L.S. – drafting of manuscript. D.K. – study concept and design, drafting of manuscript. H.H. – study concept and design, drafting of manuscript. P.A. – study supervision (principal investigator), critical revision of the manuscript for important intellectual content.

Supporting information

Appendix S1–S3

VideosS1–S3

Alshawkani YY, Orfield NJ, Samuel LT, Kuehl DR, Hagan HJ, Apel PJ. An ultrashort video can teach residents to perform a fingertip injury repair. AEM Educ Train. 2022;6:e10713. doi: 10.1002/aet2.10713