Quality and reliability Of YouTube videos on maxillofacial infections: a cross-sectional study

Abstract

Objective

This study aimed to evaluate the quality, reliability, and content level of maxillofacial infection–related videos on YouTube.

Materials and methods

A total of 101 videos meeting the predefined inclusion criteria were analyzed. The videos were independently evaluated by two researchers in terms of content scope, quality (Global Quality Scale, GQS), reliability (modified DISCERN and JAMA criteria), and user engagement metrics.

Results

The overall quality and reliability of the videos were low to moderate. The mean GQS score was 2.96 ± 0.71, reflecting an overall moderate level of video quality. The content predominantly focused on etiology, clinical findings, and diagnosis, whereas information regarding treatment was relatively limited. As the number of views, likes, and comments increased, the quality and reliability tended to decrease; conversely, longer video duration was associated with higher quality and more comprehensive content. Videos uploaded by healthcare professionals and institutions were found to be more reliable.

Conclusion

The quality of maxillofacial infection–related videos on YouTube is variable, with particularly insufficient coverage of treatment-related information. Popular videos are not necessarily reliable. Therefore, there is a clear need for more accurate and comprehensive content.

Introduction

Maxillofacial infections are among the most common emergency conditions. The majority are of odontogenic origin and may progress rapidly, spreading to the deep tissues of the head and neck region as well as adjacent anatomical spaces [1]. The dissemination of these infections into surrounding tissues poses a significant risk and, despite appropriate treatment, may lead to life-threatening complications such as mediastinitis, necrotizing fasciitis, sepsis, airway obstruction, cavernous sinus thrombosis, and cardiac and pulmonary complications [2]. Surgical drainage is considered the cornerstone of treatment for maxillofacial infections, while antibiotic therapy serves as an important adjunctive approach [3]. Non-odontogenic maxillofacial infections, on the other hand, generally arise from inflammatory processes involving lymph nodes, salivary glands, or soft tissues [4].

Today, the Internet has become one of the primary sources of health-related information, largely replacing traditional resources such as printed reference materials. Recent national data indicate that 58.5% of adults used the Internet to search for health or medical information, highlighting the growing role of digital platforms in healthcare information–seeking behaviors [5]. Among online platforms, YouTube stands out as one of the most widely used video-sharing websites worldwide. Founded in 2005, YouTube is an open-access platform that allows users to upload, view, and share videos, and it is currently the second most visited website globally after Google [6]. Due to its ease of access and provision of visual content, many individuals use this platform as a source of information on various fields of medicine.

However, there are significant concerns and ongoing debates regarding the accuracy, usefulness, and reliability of health-related content available on YouTube. Videos on the platform can be published without undergoing scientific peer review or content regulation and may be uploaded by both healthcare professionals and non-experts [7]. Consequently, the quality and reliability of information available to viewers may vary substantially.

Given that maxillofacial infections can progress rapidly and lead to serious complications, early diagnosis and appropriate treatment are of critical importance. Therefore, inaccurate or incomplete information on widely accessed platforms such as YouTube may adversely influence health-seeking behaviors, potentially contributing to delayed presentation, inappropriate antibiotic use, delayed surgical intervention or referral, and an increased risk of severe complications.

To date, no study in the literature has evaluated the reliability and content quality of YouTube videos related to maxillofacial infections. Therefore, in the present study, a total of 101 YouTube videos concerning maxillofacial infections were independently analyzed by two oral and maxillofacial surgery specialists using the Global Quality Scale (GQS), modified DISCERN, and JAMA evaluation tools. The aim of this study was to evaluate the reliability, accuracy, and educational quality of YouTube content related to maxillofacial infections, particularly videos intended for healthcare professionals.

Materials and methods

Study design

This study was designed as a cross-sectional content analysis to evaluate the content and quality characteristics of YouTube videos on maxillofacial infections. As the study was conducted using publicly available data, ethical approval was not required.

Video search strategy

The video search was conducted on YouTube, one of the most widely used video-sharing platforms worldwide. To minimize personalization bias and obtain unbiased results, all searches were performed on a single device using a stable internet connection, with browser history cleared and without logging into any user account.

Prior to the final search, pilot searches were performed using a broader set of keywords. In this context, symptom-based and alternative clinical terms such as “tooth abscess,” “jaw swelling,” “cellulitis,” and “deep neck infection” were also evaluated. However, in order to focus on educational content based on clinical and professional terminology, the keywords “maxillofacial infection,” “odontogenic infection,” and “facial space infection” were ultimately selected. This approach aimed to standardize the search strategy and target clinically relevant content. During the pilot phase, approximately 50 videos were screened. Based on this preliminary evaluation, symptom-based terms were found to predominantly retrieve patient-oriented and heterogeneous content, whereas professional terminology yielded more clinically focused educational videos. Therefore, the latter keywords were selected for the final analysis. Each keyword was searched independently, and the first 100 videos returned by YouTube’s default relevance algorithm were screened, yielding a total of 300 records. No additional YouTube filters (including upload date, duration, channel type, or alternative sorting options) were applied. All searches were conducted in incognito/private browsing mode from Türkiye to minimize personalization bias.

Of the 300 initially retrieved videos, 15 duplicate videos were identified and removed. Among the remaining videos, 34 were excluded because they were advertisements, 3 because they consisted of patient experience or vlog-style content, 7 because they were unrelated to the study topic, and 60 because of insufficient audio or visual quality. Consequently, 101 videos were included in the final analysis (Fig. 1). All searches were performed using YouTube’s default sorting option (“relevance”) on February 22, 2026, to minimize variability over time.

Fig. 1.

Flow diagram of video selection process

Inclusion and exclusion criteria

Videos were assessed according to predefined inclusion and exclusion criteria. Videos that provided information on maxillofacial infections, were in English, and had an educational or informative audiovisual purpose were included in the study. In contrast, duplicate videos, advertisements, patient experience or personal vlog-style videos, irrelevant content, and videos with insufficient audio or visual quality were excluded. In addition, YouTube Shorts were excluded because their limited duration may not allow comprehensive presentation of educational content and may reduce comparability with standard-length videos.

Data collection

For each included video, the duration (minutes), upload date, number of views, number of likes, number of comments, and source of the video (such as academic institution, specialist physician, healthcare organization, or individual user) were recorded. In addition, several derived metrics were calculated. The view ratio was calculated by dividing the total number of views by the number of days since upload. The interaction index was determined using the formula: (number of likes + number of comments) / number of views × 100.

Content evaluation

Videos were classified into four source categories according to predefined operational criteria. Academic institutions were defined as official channels affiliated with universities, dental schools, academic departments, or other recognized academic organizations. Affiliation with a university, use of academic titles, and the provision of education- or research-oriented content were considered indicative of this category.

Healthcare professionals included individual channels owned and managed by licensed healthcare providers, such as dentists, oral and maxillofacial surgeons, physicians, or other healthcare practitioners. Professional identity was verified through publicly available information, including professional titles, specialty credentials, and channel descriptions.

Educational media channels were defined as non-institutional channels primarily dedicated to the dissemination of educational content in healthcare or dentistry. These channels typically produced instructional videos, educational presentations, animations, or explanatory materials but were not directly affiliated with an academic institution or healthcare organization.

Healthcare organizations comprised official channels belonging to hospitals, clinics, healthcare centers, or other institutional healthcare service providers. In cases where source classification was unclear, channel descriptions and publicly available profile information were reviewed independently by both researchers, and final categorization was reached by consensus.

The scientific content of the videos was evaluated using a checklist covering the core clinical components of maxillofacial infections. The checklist was developed through consensus between two oral and maxillofacial surgeons based on current management guidelines and commonly reported educational domains in the literature.

This checklist included etiology, clinical findings, diagnostic methods, imaging, treatment principles, surgical drainage, antibiotic therapy, and complications. For each item, a score of 1 was assigned if the content was present and 0 if absent, and the sum of these values was calculated as the total content score.

Quality and reliability assessment

The overall quality and reliability of the videos were analyzed using three widely accepted tools for evaluating online health-related content. These included the GQS, the modified DISCERN reliability score, and the JAMA benchmarks (authorship, attribution, disclosure, and currency).

Inter-rater reliability

All videos were evaluated by two independent researchers. Inter-rater agreement was assessed using the Intraclass Correlation Coefficient (ICC) for continuous variables and Cohen’s Kappa coefficient for categorical variables. In cases of disagreement, consensus was reached with the involvement of a third researcher.

Statistical analysis

The normality of continuous variables was assessed using graphical methods and the Shapiro–Wilk test. Descriptive statistics were presented as mean ± standard deviation and median (minimum–maximum). The Kruskal–Wallis test was used for comparisons between groups and source types. Cross-tabulations were generated for content presence and analyzed using frequency (n), percentage (%), and the chi-square (χ²) test. Relationships between YouTube characteristics and quality scores were evaluated using Spearman correlation analysis. Interobserver agreement was assessed using the ICC and Cohen’s Kappa coefficients. All analyses were performed using IBM SPSS Statistics 26.0 and MS Excel 2024, and a p-value of < 0.05 was considered statistically significant.

Results

The mean video duration was 17.13 ± 25.41 min, and the mean number of views was 43,630.58 ± 144,416.81. The average number of likes was 621.72 ± 1,422.54, and the average number of comments was 60.92 ± 227.43. The mean number of days since upload was 1,621.26 ± 829.29. The mean view ratio was 28.94 ± 112.94, and the mean engagement index was 2.28 ± 1.64.

The substantial standard deviation relative to the mean for the view ratio indicates a highly skewed distribution, likely reflecting the influence of a limited number of highly viewed videos.

Evaluation of quality scores revealed that the mean GQS was 2.96 ± 0.71, indicating an overall moderate level of video quality. The mean modified DISCERN score was 2.74 ± 0.83, reflecting low-to-moderate reliability, while the mean JAMA score was 1.67 ± 0.83. The median values, which were generally within the range of 2–3, further suggest that most videos were of moderate quality, with only a limited proportion demonstrating high-quality educational content (Table 1).

Table 1.

Distribution of YouTube characteristics and quality scores

Variable	Mean ± SD	Median (Min–Max)
Duration (min)	17.13 ± 25.41	9.95 (1.27–204.0)
Views	43,630.58 ± 144,416.81	1,737.0 (3–1,024,957)
Likes	621.72 ± 1,422.54	31.0 (0–8,400)
Comments	60.92 ± 227.43	2.0 (0–1,645)
Days Since Upload	1,621.26 ± 829.29	1,703.0 (61–3,508)
View Ratio	28.94 ± 112.94	1.64 (0.05–923.39)
Engagement Index	2.28 ± 1.64	1.91 (0.0–13.33)
GQS	2.96 ± 0.71	3.0 (2–5)
Modified DISCERN	2.74 ± 0.83	3.0 (1–5)
JAMA	1.67 ± 0.83	2.0 (0–3)
Total Content Score	5.29 ± 1.82	5.0 (0–8)

Comparisons based on source type demonstrated significant differences particularly in visibility and engagement metrics. No significant difference was found in video duration (p = 0.257), indicating similar content lengths across groups. However, the number of views, likes, and comments differed significantly among groups (p < 0.001), with the highest values observed in videos produced by healthcare organizations and healthcare professionals. A significant difference was also found in the number of days since upload (p = 0.001), with individual content being available on the platform for a longer period. The view ratio was highest in videos by healthcare professionals and lowest in individual content (p < 0.001). In contrast, the engagement index was higher in individual content (p = 0.027). No significant difference was observed among groups in terms of overall quality (GQS) (p = 0.055). However, a significant difference was found in modified DISCERN scores (p = 0.020), with the highest values observed in healthcare organizations. While source type significantly influenced visibility and engagement, overall quality remained similar; however, information reliability was higher in institutional sources (Table 2).

Table 2.

Comparison of YouTube characteristics and quality scores by source type

	Educational Media Channel (n = 45)	Healthcare Professional (n = 45)	Individual (n = 6)	Healthcare Organization (n = 5)	Test statistic
	Mean±SDMedian (Min–Max)	Mean±SDMedian (Min–Max)	Mean±SDMedian (Min–Max)	Mean±SDMedian (Min–Max)	χ² / p
Duration (min)	12.35 ± 9.5110.45 (1.52–50.18)	18.14 ± 22.019.73 (1.27–121.88)	38.04 ± 81.354.16 (2.55–204.0)	26.00 ± 22.9012.53 (5.68–50.95)	χ²=4.043, p = 0.257
Views	22,075.82 ± 106,305.011,118.0 (3–711,089)	65,889.22 ± 176,220.2922,802.0 (179–1,024,957)	384.33 ± 289.62251.5 (128–916)	89,191.20 ± 198,259.32499.0 (224–443,848)	χ²=25.373, p < 0.001
Likes	226.09 ± 855.7418.0 (0–5,500)	1,047.31 ± 1,708.02579.0 (0–8,400)	10.33 ± 11.436.5 (2–33)	1,085.80 ± 2,411.716.0 (3–5,400)	χ²=26.114, p < 0.001
Comments	24.18 ± 130.041.0 (0–871)	89.91 ± 277.3724.0 (0–1,645)	0.83 ± 1.330.0 (0–3)	202.80 ± 453.470.0 (0–1,014)	χ²=19.788, p < 0.001
Days Since Upload	1,319.53 ± 781.69925.0 (61–3,326)	1,773.40 ± 742.112,065.0 (162–2,925)	2,716.83 ± 896.433,038.0 (1,378–3,508)	1,652.80 ± 568.591,649.0 (1,031–2,546)	χ²=16.856, p = 0.001
View Ratio	9.98 ± 41.441.28 (0.05–277.12)	51.06 ± 160.4312.59 (0.2–923.39)	0.14 ± 0.080.14 (0.05–0.26)	35.16 ± 77.790.30 (0.22–174.33)	χ²=28.612, p < 0.001
Engagement Index	2.27 ± 2.141.88 (0.0–13.33)	2.36 ± 1.122.32 (0.0–6.12)	2.58 ± 1.182.18 (1.56–4.05)	1.35 ± 0.161.34 (1.20–1.57)	χ²=9.142, p = 0.027
GQS	3.11 ± 0.573.0 (2–4)	2.78 ± 0.733.0 (2–5)	2.83 ± 0.753.0 (2–4)	3.40 ± 1.143.0 (2–5)	χ²=7.582, p = 0.055
Modified DISCERN	2.93 ± 0.653.0 (1–5)	2.51 ± 0.892.0 (1–5)	2.67 ± 1.212.5 (1–4)	3.20 ± 0.843.0 (2–4)	χ²=9.885, p = 0.020
JAMA Score	1.71 ± 0.662.0 (0–3)	1.60 ± 0.912.0 (0–3)	1.50 ± 1.221.0 (0–3)	2.20 ± 0.842.0 (1–3)	χ²=2.533, p = 0.469
Total Content Score	5.12 ± 1.606.0 (0–8)	4.80 ± 1.955.0 (0–8)	5.67 ± 1.515.0 (4–8)	6.40 ± 2.077.0 (3–8)	χ²=6.582, p = 0.087

χ²: Kruskal–Wallis test

Analysis of content distribution revealed that most videos focused on fundamental medical information. Etiology (98.0%), clinical findings (96.0%), and diagnosis (94.1%) were the most frequently covered topics, indicating that videos comprehensively addressed the basic aspects of disease understanding. Although treatment was included at a relatively high rate (72.3%), more specific treatment approaches such as surgical drainage (46.5%) and antibiotic use (42.6%) were addressed less frequently. Complications were mentioned in 58.4% of videos, while imaging was included in only 20.8% (Table 3).

Table 3.

Content Distribution

Content	ALL VIDEOS (n = 101)
	n (%)
Etiology	99 (98.0)
Clinical Features	97 (96.0)
Diagnosis	95 (94.1)
Radiological Imaging	21 (20.8)
Treatment	73 (72.3)
Drainage	47 (46.5)
Antibiotic Therapy	43 (42.6)
Complications	59 (58.4)

A statistically significant difference was found in terms of imaging presence according to source type (χ² = 12.138, p = 0.007). Imaging was present in 80.0% of videos from healthcare organizations, 50.0% of individual videos, 15.6% of those from healthcare professionals, and 15.6% of those from educational media channels.

Additionally, a statistically significant difference was observed in the presence of complications according to source type (χ² = 8.397, p = 0.038). Complications were reported in 73.3% of videos from educational media channels, 60.0% of those from healthcare organizations, 46.7% of individual videos, and 33.3% of those from healthcare professionals. No statistically significant differences were found for other content components based on source type (p > 0.05).

Correlation analysis revealed generally weak relationships between content and quality scores and video characteristics (Table 4). The total content score showed a weak positive correlation with video duration (r = 0.276, p = 0.005), whereas weak negative correlations were found with views, likes, days since upload, and view ratio (all p < 0.001). A moderate negative correlation was observed with the number of comments (r = − 0.429, p < 0.001).

Table 4.

Correlation between YouTube characteristics and quality scores

	Total Content Score		GQS		JAMA		Modified DISCERN
	r	p	r	p	r	p	r	p
Duration (min)	0.276	0.005	0.296	0.003	-0.041	0.684	0.247	0.013
Views	-0.384	< 0.001	-0.313	0.001	-0.236	0.017	-0.424	< 0.001
Likes	-0.362	< 0.001	-0.311	0.002	-0.268	0.007	-0.405	< 0.001
Comments	-0.429	< 0.001	-0.351	< 0.001	-0.248	0.007	-0.408	< 0.001
Days Since Upload	-0.345	< 0.001	-0.323	0.001	-0.321	0.001	-0.418	< 0.001
View Ratio	-0.346	< 0.001	-0.290	0.003	-0.184	0.065	-0.360	< 0.001
Engagement Index	-0.091	0.365	-0.199	0.046	-0.217	0.030	-0.086	0.393

r: Spearman’s correlation coefficient

The GQS demonstrated a weak positive correlation with duration (r = 0.296, p = 0.003) and weak negative correlations with views, likes, comments, days since upload, and view ratio. Its relationship with the engagement index was very weak and negative (r = − 0.190, p = 0.046). The JAMA score showed weak negative correlations with views, likes, comments, days since upload, and engagement index (all p < 0.05).

The modified DISCERN score was weakly positively correlated with duration and moderately negatively correlated with views, likes, comments, and days since upload, while showing a weak negative correlation with view ratio (all p < 0.001). Overall, as video popularity increased, content quality and reliability tended to decrease (Table 4).

A strong, positive, and statistically significant correlation was found between the total content score and both GQS and modified DISCERN scores (r = 0.722, p < 0.001; r = 0.615, p < 0.001). A moderate, positive, and statistically significant correlation was also observed between the total content score and the JAMA score (r = 0.411, p < 0.001).

Interobserver agreement between the two researchers demonstrated high consistency across all main scoring systems. Agreement was excellent for total content score (ICC = 0.994), GQS (ICC = 0.876), JAMA (ICC = 0.965), and modified DISCERN (ICC = 0.978) (p < 0.001).

Similarly, high agreement was observed for content components. Excellent agreement was found for clinical findings (κ = 0.884), diagnosis (κ = 0.847), imaging (κ = 0.970), treatment (κ = 0.952), drainage (κ = 1.000), antibiotics (κ = 1.000), and complications (κ = 0.959) (p < 0.001). In contrast, agreement for etiology was at a good level (κ = 0.658, p < 0.001). These findings indicate that the evaluation was highly reliable and reproducible.

Discussion

Social media platforms such as YouTube, which have become major sources of health-related information, have gained even greater prominence with the acceleration of digitalization during the COVID-19 pandemic [8, 9]. The YouTube algorithm prioritizes metrics such as views, likes, and engagement rather than content accuracy, which may lead to greater visibility of attention-grabbing but low-quality videos [9]. The findings of the present study support this notion, as significant negative correlations were identified between content quality and the number of views, likes, and comments.

According to analyses based on source type, significant differences were observed in terms of views, likes, and comments, with these metrics being higher in content produced by healthcare professionals and healthcare organizations. This suggests that professional sources may reach broader audiences. However, the higher interaction index observed in individual content should be interpreted with caution due to the relatively small sample size of this group. Sarıbaş et al. reported that 64.2% of videos were uploaded by healthcare professionals [10]. In the literature, Kovalski et al. reported this rate as 75%, while Yağız et al. reported 73.7% [11, 12]. Similarly, in the present study, the majority of content was produced by healthcare professionals and educational channels. It is frequently emphasized in the literature that videos created by healthcare professionals tend to have higher quality and lower bias. The findings of Buyuk and Alpaydın [13], who reported a significant association between reliability and quality in videos uploaded by dental professionals, support this view. Consistent with these findings, our study demonstrated that higher-quality videos were predominantly produced by healthcare professionals and institutions, with particularly higher reliability scores observed in these groups.

Although no significant differences were found between groups in terms of GQS and JAMA scores, modified DISCERN scores were significantly higher in favor of healthcare organizations. This indicates that institutional content may be superior, particularly in terms of presenting treatment options and ensuring information reliability. In contrast, a considerable proportion of low- and moderate-quality videos originated from individual sources, suggesting that differences in content quality may be related to the source. These findings contrast with those of Menziletoglu et al., [14] who reported no significant difference between video source and content usefulness, and instead suggest that source type may be a determining factor, especially in terms of reliability.

Similar to the findings of Spoyalo et al., [15] it appears that YouTube content may not fully meet the informational needs in complex clinical conditions such as cleft lip and palate. As also demonstrated by Srivastav et al., [16] a substantial proportion of content is of low quality, with notable deficiencies particularly in treatment and management aspects. Consistent with the literature, our content analysis revealed that most videos focused on fundamental topics such as etiology, clinical findings, and diagnosis. However, treatment-related content was more limited. Critical components of clinical management, such as surgical drainage and antibiotic therapy, were addressed in only approximately half of the videos.

Early drainage, appropriate antibiotic selection, and prevention of complications are vital in the management of maxillofacial infections. Therefore, the insufficient emphasis on treatment-related content represents a significant gap that may lead to incomplete or misleading information, particularly for non-professional audiences. The fact that surgical drainage was mentioned in only 46.5% of videos may lead viewers to perceive antibiotic therapy alone as sufficient. This misconception could result in delayed intervention, spread of infection to deep fascial spaces, airway obstruction, and even life-threatening complications such as mediastinitis. Inaccurate or incomplete treatment information may trigger inappropriate decision-making processes, particularly among individuals without medical training.

Current clinical approaches [17] emphasize appropriate surgical intervention and short-term, targeted antibiotic therapy rather than prolonged antibiotic use. The inadequate discussion of antibiotic selection, treatment duration, and antimicrobial resistance in the analyzed videos may contribute to inappropriate treatment approaches and the development of antimicrobial resistance.

Although statistically significant negative correlations were observed between popularity metrics and quality-related scores, the strength of these associations was generally weak to moderate. Therefore, the practical significance of these findings should be interpreted cautiously. Higher numbers of views, likes, and comments were associated with lower content, GQS, and modified DISCERN scores; however, these results indicate associations rather than causal relationships. Consequently, popularity metrics alone should not be considered reliable indicators of the educational quality or scientific reliability of YouTube videos.

Previous studies have reported an association between video duration and content quality, suggesting that it may reflect active viewer engagement. Lena et al. [18] and Gaş et al. [19] reported that longer video durations are associated with increased informational content. In line with these findings, our study identified a significant positive correlation between video duration and total content score, GQS, and modified DISCERN scores, indicating that longer videos provide more comprehensive and reliable information. Additionally, the higher quality and reliability scores observed in longer videos support the necessity of sufficient duration for effectively conveying complex medical information. Conversely, the negative relationship between content quality and engagement parameters suggests that shorter, easily consumable content may attract more attention but remain superficial, indicating that user engagement does not always correlate with informational depth.

In the analysis of content presence according to source type, the rate of addressing complications was found to be higher in educational media channels and healthcare organizations. This suggests that more institutional or education-oriented content provides a more comprehensive discussion of potential disease outcomes. In contrast, the lower emphasis on complications in individual content may indicate insufficient risk communication.

This study has several limitations that should be considered when interpreting the findings. Although all videos were independently evaluated by two researchers and excellent interobserver agreement was achieved, there is currently no universally validated standard tool for assessing the quality and educational value of YouTube videos. Moreover, the content checklist used in this study was developed by experienced oral and maxillofacial surgeons based on current clinical principles but was not formally validated. All content items were equally weighted to maintain methodological simplicity and objectivity; however, this approach may not fully reflect the relative clinical importance of individual components. In addition, the binary scoring system assessed only the presence or absence of information without evaluating the accuracy, comprehensiveness, depth, or clinical appropriateness of the presented content.

The findings may also have been influenced by the selected search strategy. Because the study focused on professional and clinically oriented terminology, symptom-based search terms commonly used by patients (e.g., “tooth abscess,” “jaw swelling,” and “facial swelling”) were not included. Consequently, the identified videos may preferentially represent educational content intended for healthcare professionals rather than the information most frequently encountered by the general public, thereby limiting the patient-centered applicability of the findings. Future studies incorporating symptom-based search strategies or sensitivity analyses may provide additional insight into the information most frequently encountered by patients.

The dynamic nature of YouTube and its continuously evolving recommendation algorithms represents another important limitation. As the search was performed on a single date, the results reflect only a snapshot of the available content at that specific time point. Repeating the search at different time periods may yield different video rankings and content distributions. Although measures were taken to reduce personalization effects, regional and algorithmic variations cannot be completely excluded.

The study design was based primarily on descriptive and correlational analyses. Therefore, potential confounding factors such as source type, video duration, video age, and engagement metrics were not controlled through multivariable modeling, and the observed associations should be interpreted with caution. Accordingly, the cross-sectional nature of the study does not allow conclusions regarding causality between video popularity and content quality.

In addition, some source categories contained relatively small numbers of videos, which may have limited the robustness of subgroup comparisons and reduced the stability of estimates within certain categories.

The inclusion of only English-language videos further limits the generalizability of the findings and may not reflect the information accessed by non-English-speaking populations. Furthermore, specific procedural and management-related aspects, including incision techniques, drainage methods, airway management, referral timing, and antimicrobial stewardship, were not evaluated separately. Future studies incorporating multilingual content, broader patient-oriented search strategies, repeated sampling at different time points, and more comprehensive assessment instruments may provide a more complete understanding of the quality and reliability of online information regarding maxillofacial infections.

Conclusion

This study demonstrated that YouTube videos on maxillofacial infections exhibit variability in both quality and content. While fundamental information is generally covered, treatment-related content remains limited. A negative association was identified between popularity indicators, such as views and likes, and content quality. In contrast, longer videos were found to have higher quality scores. Although videos produced by healthcare professionals and institutions tend to be more reliable, user engagement does not necessarily correlate with content quality. Therefore, there is a need to increase the availability of accurate, treatment-oriented content on YouTube.

Abbreviations

GQS

Global Quality Score

DISCERN

Modified DISCERN Reliability Score

JAMA

Journal of the American Medical Association Benchmark Criteria

ICC

Intraclass Correlation Coefficient

κ(Kappa)

Cohen’s Kappa Coefficient

SD

Standard Deviation

Authors’ contributions

MSS contributed to data collection, data analysis, and manuscript writing. HID and ÜS contributed to study design, supervision, and critical revision of the manuscript. EE contributed to data interpretation and manuscript review. All authors read and approved the final manuscript.

Funding

This research received no external funding.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study did not involve human participants, patient data, or biological material. All analyzed videos were publicly accessible. Therefore, ethics approval and consent to participate were not required.

Consent for publication

This study does not contain any identifiable personal data, images, or videos of individuals.

Competing interests

The authors declare no competing interests.