Educational Features of YouTube Videos Depicting Breastfeeding: Quality,Utility,and Reliability Analysis

Introduction

Breastfeeding is the cornerstone of child nutrition and development. In addition, breastfeeding has many positive effects for mothers, such as bonding with the baby, protecting from breast cancer, facilitating weight loss, and providing uterine involution. The World Health Organization recommends that infants be just breastfed for 6 months if there is no contraindication. ¹ However, despite this recommendation, only 39% of infants worldwide are exclusively breastfed for 6 months. ²

One of the important methods to increase these rates is to provide breastfeeding education. Prenatal breastfeeding education is recommended by The Academy of Pediatrics, Academy of Breastfeeding Medicine and Association of Women's Health.^3–5 Despite these recommendations, the number of women who receive prenatal breastfeeding education from a health care professional is very small. Studies have shown that the mother needs active support to initiate and maintain appropriate breastfeeding practices. ⁵ Social networking platforms and other channels of communication can play a key role in promoting breastfeeding.

Using training videos is both an effective method for patient education and a cost-effective strategy. ⁶ In the previous video-assisted breastfeeding study, the rate of early breastfeeding, knowledge about breastfeeding, breastfeeding motivation, and breastfeeding skills were found to be higher in the educated mothers. ⁷

YouTube is a free video hosting website that allows video content to be stored and served. YouTube website visitors can upload and share YouTube videos on a variety of platforms. In addition, the viewers can like or dislike them and post comments on those videos to express their thoughts. Although YouTube is considered to be a platform for providing medical information, videos are not considered peer reviewed. It is also ranked by popularity, views, user history, and comments, which are not normally eligible criteria for educational videos.^8,9 Although YouTube is a powerful educational tool that health care providers can take action to spread information and influence the public behavior, it can also be a source of misleading information if used improperly. In this study, we evaluated the educational features, reliability, and popularity of YouTube videos for breastfeeding education and sought answers to a few questions given hereunder.

Methods

We performed YouTube search on April 12, 2020, using the keyword “breastfeeding education.” In studies showing behavior patterns on the Internet, it shows that the vast majority of viewers only consider the first few pages of results to obtain information. ¹⁰ Considering this situation, we limited the number of videos to be scanned to 200 for the study. These 200 videos were saved in a playlist for further analysis because the search results in YouTube can change on a day-to-day basis.

Thirty-five videos were excluded due to irrelevant videos (n:11), the advertisements (n:9), duplicate videos (n:3), non-English language videos (n:4), length <1 minute videos (n:6), and videos with no sound (n:2).

Permission from YouTube was not needed to conduct this study as all the data used in our article were publicly available and no special access was required for collecting the data.

Videos were viewed by two experienced midwives responsible for breastfeeding education in our institution. Since English is the universal language approved by many countries around the world, only English videos were included in our study. Any discrepancies between the referees were resolved by discussing with the authors (S.Y.K. and D.O.).

Results

The first 200 videos related to breastfeeding education were analyzed. Thirty-five videos were excluded. Remaining 165 videos were evaluated. The most common video sources were classified as patients 57 (34.5%) and nonphysician health care providers 40 (24.2%). Of the videos, 36 (21.8%) were uploaded by physicians. Although the vast majority of videos were edited (82.4%), the majority of them were low resolution (80.6%). The descriptive information on the videos is presented in Table 2.

The characteristics of the scanned videos are given in Table 3. The mean number of views per video was 110.455 ± 210.477. The maximum number of views was 4.650.344, and the minimum number of views was 21.290. The mean length of duration for the videos was 274 seconds (between 125 and 4,421 seconds).

We divided the videos into two groups according to the GQS. Overall interobserver agreement was calculated as the weighted kappa score. Interobserver reliability for the GQS was 0.94. There were 134 (81.2%) videos in group 1 (poor and suboptimal quality) and 31 (18.8%) videos in group 2 (good and excellent quality (Table 4). Comparing both groups, the length of good–excellent videos was longer than that of poor–suboptimal videos. The most-liked and the most-viewed videos videos were in group 2. There were no differences between groups based on number of dislikes and number of comments. VPI values of group 2 videos were higher than those in group 1. In addition, the sources of 74% of the videos that make up group 2 were academic institutions.

Video sources were divided into academic institutions and private users (including physicians, health professionals other than doctors and patients). Table 5 gives the comparison of the descriptive features of the videos according to the sources uploading YouTube.

Discussion

Breastfeeding rates remain well below targeted values worldwide. The most important reason for this situation is that mothers consider themselves insufficient in breastfeeding and, therefore, do not trust themselves. Although health policies are aimed at increasing breastfeeding education, women view video sites where online health information can be accessed instantly when there is a lack of information provided by health care providers. In the near future, video-based resources will become the primary source of information for people. ¹⁵

The most important problem with video sources is that there is no control mechanism that checks whether it is reliable, correct, and appropriate. This can lead to irreversible consequences, especially in the field of health. The main purpose of our designing this study was to evaluate the quality of breastfeeding videos. In our study, the GQS was found to be 1.9 ± 0.7. Good and excellent quality videos made up only 18.8% of all videos and a significant number of the videos were uploaded by patients. This situation shows that YouTube videos have low accuracy in terms of breastfeeding, and their informative aspect is poor.

In our study, the majority of 165 videos were uploaded to the YouTube website by private users, followed by academic institutions. The video source may be related to the video reliability and quality. When we searched for high-resolution videos as YouTube filtering, almost all of the results came from academic institutions. In total, 93.7% of the videos originating from academic institutions were edited and 90.6% consisted of videos with high-definition resolution. In addition, the videos uploaded by academic institutions were more complete than the other sources. According to the global quality scoring, 74% of good and excellent quality videos (n:24/31) were uploaded to YouTube by academic institutions. Supporting our findings, it was thought that academic institutions provided higher quality and valuable videos in the study of Lee et al. ¹⁶ Educational videos on breastfeeding provided general information such as how breast milk is produced and how to continue breastfeeding, how to start breastfeeding, how the baby should be positioned, and how often and for how long the baby should be breastfed. In our study, we divided the YouTube videos into two according to the GQS classification. These groups were poor–suboptimal quality videos and good–excellent quality videos. It was observed that 81.2% of 165 videos about breastfeeding had low information quality. In the majority of poor videos, individuals share their own experiences and opinions to help the other individuals who want to learn breastfeeding.

Many people upload videos of high interest to gain recognition or profit on the Internet. When we analyzed the sources of uploaded videos in our study, the most striking feature was that private users except health professionals were the most video uploaders. When we analyzed the contents of these videos, we found that usually women who gave birth share their experiences with breastfeeding. In addition, it was understood that in most of these videos, detailed research on breastfeeding education was not done before uploading. In one of these videos, the mother was telling how she breastfed her baby while lying down. In another video, the patient, who gave a vaginal birth a week ago and said that breast milk was low, was talking about the benefit of herbal tea prepared by herself to increase the amount of breast milk. The number of views in these two videos was quite high. Although these two videos provide some useful information about breastfeeding, patients' sharing of their experiences may not only be misleading, but may cause life-threatening risks for both mothers and babies.

We used the VPI score for each video to evaluate the popularity of videos associated with breastfeeding education. Our results show that academic instutions have a higher VPI score. The durations of the videos uploaded by academic instutions were longer than the durations of the videos uploaded by private users according to our results. Similar to the findings of our study, Biggs et al. reported that useful videos were longer. ¹⁷ Unfortunately, the videos uploaded by the academic institutions constitute only 19.3% of the videos we have included in the study. In recent years, video sharing sites have become an important resource for patients to access information, which especially worries health institutions about the accuracy of the information on these platforms.^18,19 Missing or inaccurate information may cause patients to apply unscientific treatments or act in the opposite direction of the guidelines. ²⁰ To our knowledge, this is the first study to evaluate breastfeeding content on YouTube. Many important parameters about breastfeeding were not mentioned in YouTube videos. Our results showed that YouTube videos about breastfeeding were a limited source for the mothers.

There were some limitations in our study. First, when a different keyword was used, different results were listed. Second, since the YouTube is a dynamic platform, the search results vary in a different time and hour. Another limitation of our study was that the educational content quality scoring was performed by two observers. Although this situation caused bias in our selection of videos according to their quality, the fact that the kappa score among the observers was 0.94 would dispel our concerns.

Conclusion

Although health care providers do not have the opportunity to check the quality of the uploaded videos, our analysis results showed that the videos most liked and commented by the viewers were professionally prepared and almost all were uploaded by academic institutions.

In light of these findings, we will advocate increasing the number of videos uploaded by academic institutions on social media sites such as YouTube, which are popular sources for health information and we advocate more regulatory oversight of the content of the videos.