Quantitative microbial risk assessment of human H5N1 infection from consumption of fluid cow's milk

This quantitative microbial risk assessment study concludes that while pasteurization effectively eliminates the risk of human H5N1 infection from cow's milk, raw milk consumption poses a measurable risk that can be mitigated through bulk tank testing, though further research is needed on dose-response dynamics and consumption patterns.

The Gist

Imagine the dairy industry as a massive, bustling highway system where milk travels from farms to your local grocery store. Recently, a dangerous "traffic jam" of a virus called H5N1 (a type of bird flu) has appeared on this highway, specifically infecting the cows.

Scientists wanted to know: If you drink this milk, how likely are you to catch the virus? To answer this, they built a giant, digital "crystal ball" called a Quantitative Microbial Risk Assessment (QMRA). Think of this model as a super-powered simulation game where they ran millions of scenarios to see what happens when people drink different types of milk.

Here is what their "crystal ball" revealed, broken down into simple terms:

1. The Pasteurization Shield (The "Firewall")

Imagine pasteurization (heating milk to kill germs) as a super-strong, impenetrable force field.

• The Result: When the scientists simulated drinking pasteurized milk, the chance of getting sick was so incredibly tiny it's almost impossible to imagine. They described the odds as 7.66E-19.
• The Analogy: That number is so small it's like winning the lottery every single day for a million years and still not getting sick. It confirms that the heat treatment used in standard milk processing is a perfect shield against this virus.

2. The Raw Milk Risk (The "Open Gate")

Now, imagine drinking raw milk (milk that hasn't been heated) as walking through an open gate where the virus might be hiding.

• The Result: The risk here is higher than with pasteurized milk, but it's still very low. However, it's not zero.
• The Analogy: If pasteurized milk is a fortress, raw milk is like a house with the front door unlocked. You probably won't get robbed, but there is a real, calculable chance that a thief (the virus) could sneak in. The study found that if you drink a glass of raw milk from a farm store or a grocery store, there is a small, but non-zero, chance of infection.

3. The Detective Work (Testing the Tanks)

The scientists also asked, "How can we make the raw milk safer?"

• The Solution: They tested the idea of using a high-tech "sniffer dog" (a test called qrRT-PCR) to check the big milk tanks at the farm before the milk leaves.
• The Result: This acts like a security checkpoint. If the sniffer dog barks (finds the virus), that batch of milk is stopped. This significantly reduces the number of "thieves" that get through the gate.

4. What We Still Don't Know (The Missing Puzzle Pieces)

Even with this powerful crystal ball, the scientists admitted they are missing a few puzzle pieces:

• The Dose: We don't know exactly how many virus particles it takes to make a human sick. It's like knowing a car is dangerous, but not knowing if hitting it at 5 mph or 50 mph is what causes the injury.
• The Habits: We don't have a perfect map of exactly how many people drink raw milk and how often. It's like trying to predict traffic accidents without knowing how many cars are actually on the road.

The Bottom Line

This study is a reassurance and a warning:

1. Reassurance: If you buy standard milk from the store (pasteurized), you are safe. The heat treatment is doing its job perfectly.
2. Warning: If you drink raw milk, you are taking a gamble. While the odds of getting sick are low, they are real.
3. Call to Action: We need better data on how many people drink raw milk and exactly how the virus affects humans so we can build even better safety nets for the future.

In short: Pasteurized milk is a safe, heated fortress. Raw milk is an open gate that needs better security checks.

Technical Summary

1. Problem Statement

The emergence of the H5N1 clade 2.3.4.4b strain in dairy cattle has created a novel public health concern regarding the safety of the fluid milk supply. While there have been no reported cases of foodborne H5N1 infection in humans to date, the virus has infected at least 70 people (primarily through direct animal contact), and milk from infected cows has been proven to be infectious to other species via ingestion. Given the potential for spillover, there is an urgent need to quantify the risk of human infection through the consumption of both raw and pasteurized milk within the United States supply chain to inform regulatory and public health responses.

2. Methodology

The study employs a farm-to-table Quantitative Microbial Risk Assessment (QMRA) framework utilizing stochastic modeling. The approach includes:

• Model Parameterization: The models were calibrated using emerging data specific to the current H5N1 outbreak in dairy cattle, including viral load data and transmission dynamics.
• Exposure Scenarios: The assessment evaluates the probability of infection per 240-mL serving across three distinct pathways:
  1. Pasteurized milk.
  2. Raw milk purchased directly from farms (farmstore).
  3. Raw milk purchased at retail outlets.
• Intervention Analysis: The model simulates the impact of various mitigation strategies, specifically focusing on the efficacy of pasteurization and the implementation of quantitative real-time reverse transcriptase-polymerase chain reaction (qrRT-PCR) testing on bulk tank milk.
• Statistical Output: Results are presented as median probabilities with 5th and 95th percentiles to account for uncertainty and variability in the data.

3. Key Contributions

• First QMRA for H5N1 in Milk: This study provides the first stochastic, farm-to-table risk assessment specifically quantifying the human infection risk of H5N1 via fluid milk consumption.
• Comparative Risk Analysis: It offers a direct quantitative comparison between pasteurized and raw milk consumption, distinguishing between different retail channels (farmstore vs. retail).
• Intervention Evaluation: The paper moves beyond risk identification to evaluate specific control measures, demonstrating the quantitative impact of qrRT-PCR testing on bulk tanks as a risk reduction tool for raw milk.
• Identification of Knowledge Gaps: The study explicitly highlights critical data deficiencies, specifically regarding human dose-response relationships for H5N1 via ingestion and detailed patterns of raw milk consumption, which are essential for refining future models.

4. Key Results

The study calculated the median probability of infection per 240-mL serving for the three scenarios:

• Pasteurized Milk: The risk is negligible, with a median probability of $7.66 \times 10^{-19}$ (95th percentile: $4.02 \times 10^{-17}$). This confirms that standard pasteurization processes are highly effective at eliminating H5N1 infectivity.
• Farmstore-Purchased Raw Milk: The risk is significantly higher, with a median probability of $1.56 \times 10^{-7}$ (95th percentile: $1.28 \times 10^{-5}$).
• Retail-Purchased Raw Milk: The risk is comparable to farmstore raw milk, with a median probability of $1.40 \times 10^{-7}$ (95th percentile: $1.13 \times 10^{-5}$).

Scenario Analysis Findings:

• Implementing qrRT-PCR testing of bulk tank milk was identified as an effective intervention to numerically reduce the risk associated with raw milk consumption by filtering out contaminated batches before distribution.

5. Significance and Implications

• Validation of Pasteurization: The findings provide robust scientific evidence supporting the continued reliance on pasteurization as a critical barrier against H5N1 transmission via milk, reassuring the safety of the standard commercial milk supply.
• Risk Communication: The quantification of risk for raw milk consumers offers a data-driven basis for public health messaging regarding the dangers of consuming raw milk during this specific outbreak.
• Policy Guidance: The study suggests that if raw milk consumption is to continue, mandatory qrRT-PCR testing of bulk tanks is a viable mitigation strategy.
• Future Research Directions: The paper underscores the necessity for further research into human dose-response kinetics for ingested H5N1 and more granular data on raw milk consumption behaviors to improve the precision of future risk assessments.
• Methodological Framework: It demonstrates the utility of mechanistic epidemiologic models in guiding public health responses during emerging zoonotic outbreaks with foodborne potential.