Digital platforms have become integral to contemporary life, and there is increasing consultation on their enormous and untapped potential (1). Among digital platforms, YouTube is the leading content community, boasting over one billion daily video views, 30 million active users daily, and 500 h of video submitted every minute. This platform has emerged as a potent instrument for rendering video content worldwide and making it available to a broad audience (2).

Today, YouTube plays an important role in access to information in the health field (3). It is widely used by parents, relatives of patients, and healthcare professionals to learn and raise awareness about serious diseases such as cancer (4). In this context, the quality of educational content provided, especially on the relationship between childhood cancer and nutrition, is very important. Nutrition is a critical factor affecting the disease’s course and patients’ quality of life during cancer treatment. While inadequate or unbalanced nutrition can negatively affect the treatment process, proper nutritional habits can support the fight against the disease (5). Therefore, parents and relatives of patients can access accurate information thanks to educational videos based on scientific foundations (6). Scientific content shared on digital platforms such as YouTube can be an important resource in providing families and health professionals with up-to-date and accurate information. However, choosing reliable and expert-prepared content to access accurate information is of great importance. Nonetheless, as with other digital platforms, YouTube lacks filters or systems to authenticate the veracity of material, presenting a danger of disinformation, particularly for less-informed users (2). In other words, individuals may post videos on YouTube without a reviewer’s approval. Patients could be misled by certain advertisements and posts intended for monetization specifically. Consequently, utilizing these platforms for medical advice or information may result in disseminating inaccurate information. Thus, it is crucial to evaluate the quality and trustworthiness of the data disseminated to millions of people on video-sharing sites such as YouTube (4).

Cancer is one of the leading causes of death around the world (7). Approximately 400,000 children and adolescents (0–19 years old) are diagnosed with cancer year worldwide (8). The rising incidence of pediatric cancer diagnoses continues to position it as a primary cause of non-accidental mortality (9). Investigating the connection between diet and cancer has been a topic for research. Several studies highlight the importance of diet in cancer patients, as well as finding foods that either prevent or promote cancer (10, 11).

Several studies examine the content of cancer-related videos on the digital platform YouTube (4, 12, 13). However, to the best of our knowledge, no study has yet analyzed the content of pediatric cancer and nutrition. This study aimed to identify the most viewed YouTube videos using keywords such as “nutrition in pediatric cancer” and “nutrition in childhood cancer,” evaluate them using internationally recognized scoring systems, and examine the videos’ scientific reliability, quality, and accuracy.

This study evaluated the reliability, quality, content, and scope of YouTube (www.youtube.com) videos related to pediatric cancer and nutrition. The terminology frequently used on the subject was analyzed, and nine keywords were determined as “Nutrition in pediatric oncology,” “Nutrition in pediatric cancer,” “Nutrition in childhood cancer, “Diet for pediatric oncology,” “Diet for pediatric cancer,” “Diet for childhood cancer,” “Food for pediatric oncology,” “Food for pediatric cancer”, and “Food for childhood cancer”.

Before scanning, YouTube search history was deleted, and the search filter was set to the most viewed videos, thus imitating the behavior of ordinary users. URL addresses of 100 videos for each keyword, totaling 900 videos, were recorded for future analysis. This entire scanning and recording process was carried out on May 14, 2024, using the methods used in previous studies, and the analysis process of the videos was started (4, 14). During the video analysis process, duplicate, irrelevant, non-English, audio-only, visual-only, YouTube Short videos (≤60 s), and videos with insufficient audio/visual quality were excluded from the study. After excluding videos that did not meet the criteria, the study was conducted with 25 videos (Figure 1).

Inter-observer agreement for the 25 included videos was assessed using Cohen’s kappa. Unweighted kappa was calculated for mDISCERN (κ = 0.728, p < 0.001) and NPCSS (κ = 0.738, p < 0.001), while quadratic weighted kappa was applied for GQS (κ = 0.950, p < 0.001).

When the video metrics analyzed, length of each video (second) median (IQR) values were found to be 868 (202–2,977), 120 (61–1,159) for the number of views, 2003 (275–3,884) for the number of subscribers of the accounts uploading the videos, 2 (0.5–9) for the number of likes, 0 (0–1) for the number of dislikes, and 1,095 days (730–3,285) for the duration that the videos were published on the platform. When interaction rates were examined, the like ratio was calculated as 100 (85–100), the view ratio (views/day) was 0.19 (0.07–0.58), and the VPI score was 0.2 (0.09–0.68). In terms of video quality, the GQS was found to be 4 (2–5), with 9 (36%) of the videos being rated as low quality, 2 (8%) as intermediate quality, and 14 (56%) as high quality. Regarding reliability, the mDISCERN score was calculated as 4 (2–5). 16 (64%) of the videos were classified as reliable, and 9 (36%) as non- reliable. Additionally, the NPCSS score was calculated as 10 (6–21) (Table 3).

When the distribution of the number of YouTube videos on pediatric cancer and nutrition by year is examined, it is seen that there were 3 videos in 2012, which constituted 12% of the total videos. The number of videos decreased to 2 (8%) between 2014 and 2015 and remained constant at this level. A gradual increase was observed in 2016–2020, reaching 3 videos (12%) in 2020. A remarkable increase was recorded in 2021, reaching 8 videos, constituting 32% of the total videos, and the peak point was recorded. In 2022, videos decreased to 5 (20%). In 2023–2024, the number of videos decreased significantly to 1 (4%) (Figure 2A).

In the distribution of videos examined by country, most videos originated from the United States, with 9 videos (36.0%) and the most content. There were 7 videos with no specified source, making up 28.0% of the total content. Among other countries, India, Kenya and Canada were represented with 2 videos (8.0%) each. France, Switzerland, and New Zealand were represented with 1 video (4.0%) each (Figure 2B).

Useful videos accounted for 60% of the sample, while 40% were classified as unuseful. Useful videos were significantly longer (p = 0.015, r = 0.49) and demonstrated markedly higher educational quality scores, including GQS (p < 0.001, r = 0.86), NPCSS (p < 0.001, r = 0.84), and mDISCERN (p < 0.001, r = 0.76), all indicating large effect sizes. Unuseful videos, on the other hand, were uploaded by channels with a higher number of subscribers (p = 0.040, r = 0.41), reflecting a moderate effect size. No significant differences were observed for other quantitative parameters (p > 0.05), which showed small to moderate effect sizes. Presenter type also did not differ significantly between useful and unuseful videos (p = 0.092), despite a large effect size (r = 0.51). Publisher affiliation showed no significant association with usefulness and had a moderate effect size (p = 0.217, r = 0.42) (Table 4).

The results of the correlation analysis indicated the presence of high, positive, statistically significant correlation between DISCERN and GQS and NPCS scores (r = 0.863 and r = 0.884; p < 0.001, respectively). A similarly very high positive correlation was found between GQS and NPCSS scores (r = 0.943; p < 0.001). A low negative but statistically significant correlation was found between the view ratio, mDISCERN, and GQS scores (r = −0.359 and r = −0.404; p < 0.05). No significant relationship was found between the like ratio and any parameter (p > 0.05) (Table 5).

ROC curve analyses were performed to identify high-quality videos based on the usefulness score. As a result of this analysis, mDISCERN (cut-off point 3.00, AUC 0.90), GQS (cut-off point 4, AUC 0.97) and NPCSS (cut-off point 10.00, AUC 0.94) scores were found to show significant results. All three scores showed above-perfect discrimination ability for high-quality videos (Table 6; Figure 3).

This study evaluated the reliability, quality, comprehensiveness, and accuracy of the most viewed YouTube videos related to pediatric cancer and nutrition. The fact that only 25 of the 900 initially identified videos met the inclusion criteria highlights the extent of information overload and misinformation on this topic. Furthermore, the finding that only 60% of the eligible videos were considered useful indicates that a substantial proportion of the already limited content available on YouTube does not provide meaningful benefit for patients and their families. In addition, the observed decrease in viewing rates as video quality and reliability increased suggests that patients and caregivers may show less interest in scientifically accurate content. Taken together, these findings underscore the challenges patients face in accessing reliable and useful health information on YouTube and emphasize the necessity of the present study.

YouTube is the world’s most visited video-sharing platform and is seen as an important source of information on various topics (4, 25). However, the lack of monitoring mechanisms and the ability of content producers to share videos freely leads to the spread of false and misleading information (25, 26). Today, with the increasing influence of the internet and social media, patients often use YouTube to seek information on health and nutrition (4, 14). YouTube videos on types of cancer and nutrition are also among the most searched content by patients and their relatives (4).

Our study is the first to evaluate pediatric cancer and nutrition videos on YouTube, but numerous studies in the literature examine the overall quality of cancer-related videos. To our knowledge, the only study examining the scientific reliability and quality of cancer and nutrition videos on YouTube is the study conducted by Sütçüoğlu et al. (4). Other studies have evaluated informative videos on specific types of cancer, such as prostate, testicular, lung, bladder, and breast (27–31). These studies show that cancer content on YouTube is highly variable in quality and that a significant portion of it either provides incomplete information or contains false/misleading information. However, the videos that receive the most interaction do not always provide the most accurate information, and sometimes, the opposite occurs, leading to information with a weak scientific basis reaching a 0broad audience (32, 33). This situation poses risks to both patient and general public health.

In the current study, the average length of the videos was approximately 15 min, which is consistent with the view that health-related videos longer than 10 min would increase comprehensiveness (26). However, the views of the videos vary widely (between 3 and 10,570). This shows that while some videos reach large audiences, others fail to attract enough viewers. This finding is similar to the study conducted by Sütçüoğlu and colleagues on cancer and nutrition-related content on YouTube (4). When evaluated in terms of interaction criteria, it is seen that although the like rate is relatively high, the viewing rate (daily views) is relatively low. This finding suggests that viewers generally rate the content positively, but such videos are not popular enough on the YouTube platform. This result is consistent with previous studies that educational and scientific videos generally receive lower interaction (25, 26). Additionally, considering that the average upload time of videos is approximately 3 years, some problems arise in terms of the up-to-dateness of the information. Especially in a rapidly developing field such as pediatric nutrition and cancer, where information must be constantly updated, it can be said that old content may contain false or misleading information. It may have harmful consequences for public health.

This research has examined the distribution of YouTube content on cancer and nutrition between 2012 and 2024. The findings demonstrate that the number of videos on this topic has followed a fluctuating course over the years, with a notable peak especially in 2021. The striking increase recorded in 2021 is particularly noteworthy. This period coincides with increased health consciousness and the search for information on strengthening the immune system due to the effects of the COVID-19 pandemic. Vanderpool et al. have emphasized that nutritional recommendations for cancer patients gained significant popularity during the pandemic period. Additionally, increased screen time and digital content consumption during pandemic restrictions may be among the factors explaining this increase.

According to the analysis of source countries of the examined videos, significant geographical differences have emerged in content production. The United States constitutes the country publishing the most content, accounting for 36% (n = 9) of all videos examined. This situation highlights the significant role played by US-based health institutions, researchers, and content creators in the dissemination of information about pediatric cancer nutrition through YouTube. Among other identified countries, three countries—India, Canada, and Kenya—each constituted 8% (n = 2) of the total content. The remaining content came from France, Switzerland, and New Zealand, each constituting 4% (n = 1) of the total sample. These findings indicate a clear predominance of Western, particularly North American, sources for pediatric cancer nutritional information on YouTube. This geographical distribution pattern may have significant implications for the cultural appropriateness, accessibility, and applicability of nutritional guidance for pediatric cancer patients in different health systems and cultural contexts worldwide.

In the study, when the videos were evaluated in terms of quality, it was found that 56% were of high quality and 36% were of low quality. However, 16 of the 25 videos evaluated (64%) were reliable. This result is often positive. However, many YouTube studies have reported that the videos evaluated are mostly low-quality and have low-reliability scores (26). In a study conducted by Garcia et al. examining YouTube content related to pancreatic cancer, 66.7% of the videos were found to have low reliability (31). Similarly, in a YouTube study by Sütçüoğlu et al. on cancer and nutrition, 61.25% of the videos were found to be unreliabl (4). In contrast, Kahlam et al.’s study of YouTube on colorectal cancer found that most videos were of good quality (34). Thus, YouTube does not provide homogeneous content for health education and that there are significant differences in quality and reliability. It can be said that this situation may cause the spread of false information on cancer issues and pose a potential risk to patient health.

In the current study, the methodologies of previous studies were taken as an example, and the videos examined with the participation of all researchers were divided into two groups useful and misleading. As a result, 60% (n:15) of the 25 videos were classified as useful, and 40% (n:10) were classified as misleading. When the metrics of useful and misleading videos were compared in terms of quality, reliability, comprehensiveness, and accuracy, remarkable results were obtained. First, the median duration of useful videos was approximately 43 min, while the median duration of unuseful videos was approximately 3 min. Sütçüoğlu et al. found in their study that quality videos have a longer duration (4). Second, although there was no significant difference in the number of views, it was observed that channels where unuseful videos were uploaded had more subscribers. This situation can be considered as evidence that YouTube channels that are popular with viewers do not always offer scientifically reliable content. It has been stated in the literature that popular channels can sometimes produce content that contains misleading or incomplete information in order to increase click and view ratio (35). This result highlights the need to carefully evaluate pediatric cancer and nutrition-related videos on YouTube for scientific accuracy.

As a result of evaluating the videos in terms of reliability, quality, and comprehensiveness according to whether they were useful or not, useful videos were found to be significantly higher in quality, more comprehensive, and more reliable. This result indicates that the classification of the videos according to their usefulness was carried out correctly. However, a video being evaluated as unuseful did not always stem from containing incorrect information; in many cases, it resulted from presenting short, superficial, and insufficiently comprehensive content despite including accurate information. Therefore, the absence of a significant difference between useful and unuseful videos in terms of presenter type and publisher affiliation should be interpreted by taking this situation into account. In the literature, it has also been shown that health professionals mostly produce high-quality content in cancer-related studies (4, 36, 37). These findings are consistent with the results of our study. Based on these results, it is important to encourage content produced by health professionals. Additionally, adjusting YouTube algorithms to highlight content approved by reliable institutions will facilitate access to higher-quality information for patients and their relatives.

Correlation analysis conducted in this study revealed strong positive associations among content quality indicators, including mDISCERN, GQS, and NPCSS. This finding indicates that videos with higher quality and reliability tend to provide more comprehensive information and is consistent with the existing literature (4, 30, 36). In addition, a negative relationship was observed between viewing rate and video quality and reliability, suggesting that higher-quality and more reliable videos are relatively less viewed. Similarly, Sütcüoğlu et al. reported that reliable videos related to cancer and nutrition had lower Video Power Index values, indicating reduced audience engagement (4). Previous studies have shown that videos with short and attention-grabbing titles tend to be more popular, whereas more scientific content generally receives lower viewing rates (26, 33). In this context, one possible explanation for the higher view counts of low-quality videos observed in our study is patients’ tendency to seek quick solutions. Additionally, YouTube’s algorithms may further reinforce this pattern by preferentially promoting content that generates higher engagement. This situation represents a significant public health concern, as inaccurate or incomplete information may reach wider audiences and increase the risk of misinformation. Therefore, health-related content on YouTube should be carefully evaluated in terms of scientific accuracy and reliability, and popular videos should not be assumed to be trustworthy. Alongside efforts to improve digital health literacy, it is also crucial that YouTube algorithms prioritize scientifically accurate and reliable content.

This study provides important contributions to pediatric cancer and nutrition but has some limitations. Firstly, due to the dynamic nature of YouTube, videos are constantly updated, new content is added, and popularity metrics change over time. This situation limits the generalizability of the study’s findings and is considered an inevitable limitation. Second, the study focused only on English language videos, which may introduce selection bias. In future studies, evaluating videos in different languages may be useful. Third, how viewers perceive the videos, whether the content is reliable, and how patients or their relatives interpret them have not been examined. Future research should be expanded to include analysis of comments made on videos and investigate how patients are affected by these contents. Fourth, as reported in previous studies, scoring systems used for video content evaluation, including the GQS, mDISCERN, and NPCSS, are inherently subjective and have not always undergone rigorous validation processes. Therefore, the development of more comprehensive and reliable assessment tools is warranted in future research. In addition, the small number of included videos limits the generalizability of the findings and the statistical power of the analyses. However, the fact that only a limited number of videos met the inclusion criteria despite the extensive content available on YouTube highlights both the rigor and the relevance of this study.

This finding indicates that pediatric cancer and nutrition is an insufficiently represented topic on the YouTube platform. Although many of the reviewed videos included high-quality and reliable information, only a portion of them met the criteria to be classified as useful. Furthermore, it was found that most of the useful videos were presented by dietitians. Another finding of the study was that viewer tendencies did not align with scientific accuracy and content quality. In sensitive topics such as pediatric cancer and nutrition, patients and their relatives should prioritize videos presented by healthcare professionals rather than relying on popular content. In this context, increasing the visibility of trustworthy health content on YouTube and supporting educational initiatives aimed at improving digital health literacy may be beneficial.