Quality of Online Mastitis Management Information for Dairy Farmers: An Evaluation of YouTube Videos

Technical Summary: Quality of Online Mastitis Management Information for Dairy Farmers

Problem Statement
Mastitis is a prevalent and economically devastating disease affecting dairy animals in India, causing significant losses in milk production, quality, and farmer livelihoods. While timely, evidence-based management is critical, the traditional public extension system faces severe bottlenecks, with a farmer-to-extension-worker ratio far exceeding recommended levels. Consequently, farmers increasingly rely on Information and Communication Technologies (ICTs), particularly YouTube, for informal extension. However, the platform operates in an unregulated environment where content is user-generated and prioritized by popularity algorithms rather than scientific accuracy. This creates a risk of disseminating unverified, low-quality, or misleading veterinary information, potentially leading to suboptimal farm-level decision-making and public health risks. There is a lack of systematic evaluation regarding the quality of veterinary health information on digital platforms in India.

Methodology
This study conducted a cross-sectional evaluation of Hindi-language YouTube videos regarding mastitis management in dairy cattle and buffaloes.

• Data Collection: On March 6, 2026, the researchers utilized the YouTube Data API v3 with a custom Python script to search for the top 50 videos for four specific Hindi queries (e.g., "Mastitis in milch animals," "Treatment of mastitis"). This approach minimized personalization bias.
• Sample Selection: From an initial pool of 200 videos, duplicates were removed, and videos shorter than two minutes, purely promotional content, or those not intended for farmers were excluded. The final analysis included 63 videos.
• Quality Assessment Instruments: Four validated tools were employed:
  1. Modified DISCERN (mDISCERN): Assessed reliability (objectives, source, balance, uncertainty).
  2. JAMA Benchmark Criteria: Assessed accountability and transparency (authorship, references, currency, disclosure).
  3. Global Quality Scale (GQS): Assessed overall educational value.
  4. Mastitis Content Score (MCS): A purpose-built instrument developed by the authors to evaluate comprehensiveness (symptoms, diagnosis, treatment, prevention, antibiotic usage, etc.).
• Metrics: The study also calculated the Video Power Index (VPI) for popularity and Engagement Ratio (ER) for audience interaction.
• Statistical Analysis: Two independent experts rated the videos. Inter-rater reliability was confirmed using Intraclass Correlation Coefficients (ICC). Non-parametric tests (Kruskal-Wallis, Mann-Whitney U) were used to compare groups, with Bonferroni correction applied for multiple comparisons.

Key Results

• Overall Quality: The majority of videos were of low quality. Mean scores were low across all instruments: mDISCERN (1.96/5), JAMA (1.21/4), GQS (1.98/5), and MCS (2.44/5). Only a small fraction of videos achieved high-quality ratings (e.g., 0% for mDISCERN ≥4, 3.2% for GQS ≥4).
• Producer Disparities:
  • Independent Creators: Produced the majority of videos (52.38%) and achieved the highest popularity (Video Power Index: 2.32). However, they scored the lowest on clinical comprehensiveness and quality metrics.
  • Veterinary Professionals: Produced fewer videos (17.46%) but achieved the highest quality scores (mDISCERN: 2.55; JAMA: 1.55) and the highest engagement ratios (4.26), suggesting a trust-driven audience.
  • Media Houses: Scored highest on GQS and MCS but had low popularity.
• Correlations:
  • A significant negative association was observed between popularity (VPI) and quality metrics, indicating that highly viewed videos are not necessarily reliable.
  • A significant positive correlation existed between video duration and the Mastitis Content Score (r = 0.265, p = 0.036), suggesting that comprehensive information requires longer formats.
  • Engagement ratios showed a positive, though not always statistically significant, trend with quality scores.

Key Contributions

1. First Application in Veterinary Domain: To the authors' knowledge, this is the first study to apply established human health evaluation tools (mDISCERN, JAMA, GQS) to veterinary health information on YouTube.
2. Baseline Evidence: The study provides the first systematic baseline data on the quality of Hindi-language digital extension content for mastitis in India.
3. Identification of the "Popularity-Quality Gap": The research empirically demonstrates a divergence where high-reach content (often from independent creators) lacks the clinical depth and reliability found in lower-reach content produced by veterinary professionals.
4. Development of MCS: The introduction of a specific Mastitis Content Score (MCS) tailored to the disease's management requirements.

Significance and Claims
The paper argues that while digital platforms have democratized access to information, they have not ensured the dissemination of high-quality, comprehensive content. The findings highlight a critical distribution problem where the most visible content to farmers is statistically likely to be the least technically comprehensive.

The authors claim that this "popularity-quality gap" poses a significant risk to animal health and public health due to potential misinformation regarding antibiotic use and disease management. They conclude that bridging this gap requires a strategic shift in digital extension:

• Capacity Building: There is an urgent need to train field-level veterinarians and paraveterinarians in digital communication skills and content creation.
• Strategic Empowerment: Extension organizations should sensitize active professional creators to quality standards (mDISCERN, JAMA, GQS) to help them achieve both credibility and mass appeal.
• Transformation of the Ecosystem: By empowering professionals to integrate technical rigor with audience-centric presentation, YouTube can be transformed from a source of unverified viral advice into a standardized, reliable clinical resource for veterinary extension.

The authors remain modest regarding their claims, acknowledging limitations such as the single-language scope (Hindi), the cross-sectional nature of the data (a snapshot in time), and the non-standardized nature of the MCS tool. They suggest that future research should include longitudinal studies and demand-side surveys to understand farmer behavior and the long-term impact of these digital dynamics.