It seems that Groton, MA, is still abuzz over the decision by its Board of Health to reverse the town’s long-standing ban on raw milk. There have been some postings on the town list serve  and some emails directed my way, all concerning the confusion—expressed by a couple of board members at last Monday’s meeting—on how to interpret data about raw milk’s dangers. 

I’ve posted a number of times about the sad state of data about food-borne illness–the data seem to be in roughly the same shape the U.S. Centers for Disease Control is now finding its labs to be in: neglected and carelessly tended to.  

Other towns and states will no doubt be asking questions similar to what Groton’s health board members were asking, as new legislative proposals are brought up to liberalize access to raw milk. To help them work through the fog, I have put together a brief Q&A.

Amazingly, the CDC doesn’t present annual data on reported illnesses from any food, including raw milk, to allow trends to be graphed on a year-to-year basis. The only way to obtain data is to extract it yourself from something called the Foodborne Outbreak Online Database maintained by the CDC. 

When it comes to illnesses from raw milk, the CDC and FDA seem only to provide data in clumps, based on five, eight, or thirteen years.  As one example, this from a CDC FAQ: “From 1998 through 2011, 148 outbreaks due to consumption of raw milk or raw milk products were reported to CDC. These resulted in 2,384 illnesses, 284 hospitalizations, and 2 deaths.”

Does anyone have the data?

I have gone through the tedious process of extracting annual data from the CDC database. When you examine CDC data over the last decade, you find that there are generally between 25 and 200 reported illnesses from raw milk each year. In 2008 I came up with 132, in 2011 I found 60.  There is no long-term trend up or down (though from 2006-2010 the trend was steadily up, before the number of illnesses plummeted in 2011).

So what I did was take a typical recent year for both raw milk illnesses, and total foodborne illnesses as reported by the CDC. That way, we are comparing apples and apples.

What did you find?

What I found was as follows:

*In 2008, there were 23,000 total foodborne illnesses reported by CDC.
*132 came from raw milk and raw milk cheese.
*Slightly over one half of one per cent of the total reported illnesses was from raw milk.

That is it? No one else breaks it down this way?

There is a web site run by a food safety lawyer Bill Marler, known as realrawmilkfacts.com, which maintains data on illnesses from raw and pasteurized milk. It shows graphics on annual reported illnesses from pasteurized milk and raw milk; it also provides a list of the actual outbreaks (type of facility, location) and illnesses from raw milk extracted from the CDC database I alluded to above.

Where it gets more confusing is after 2010—for those years, it uses a combination of CDC data and other information –what it extracts from newspaper articles and online reports of illnesses; some of these may subsequently be found to not be associated with raw milk.

Is there anything else we don’t know about the data on raw milk illnesses?

Yes, we don’t know how many of the illnesses attributed to raw milk come from “pre-pasteurized” milk—that milk intended for processing, not intended to be served raw. As one example, in 2011 a mom took some milk from the bulk tank of a Wisconsin dairy that sells pre-pasteurized milk to a school birthday party. Some 18 people got sick, most of them children, with campylobachter. Food safety officials in Wisconsin essentially excused the farmer for the “accident,” since he didn’t intend to distribute unpasteurized milk.

However, those 18 illnesses were tallied in CDC’s databases as illnesses attributed to raw milk for 2011. So, of the 60 illnesses reported by the CDC in 2011, nearly one-third came from pre-pasteurized milk….that we know about. Put another way, had those 18 not been included in the tally, there would have been 42 illnesses attributed to raw milk.

Are there other such situations that distort the data?

There almost certainly are. As another example, some illnesses attributed to raw milk come from tainted queso fresco cheese—a soft cheese that is popular in the Hispanic community. It isn’t aged 60 days as required by the FDA for cheese, and it is often made from pre-pasteurized milk (milk intended for pasteurization)….in bathtubs, of all places (hence its nickname, “bathtub cheese.”)

So you are saying this has inflated the data on raw milk illnesses?

Yes, these sorts of tabulation glitches appear to have inflated illnesses from raw milk. But beyond that, the only deaths associated with raw milk over the last 20 years were from queso fresco cheese. The FDA and CDC will often say that two people have died from tainted raw milk over the last 20 years. It turns out those two died from bad queso fresco cheese. Raw milk advocates complain that using those deaths to denigrate raw milk is highly misleading.

Have the CDC and FDA shown any interest in adjusting or fixing the data?

Quite the contrary. The CDC and FDA are focused on highlighting as many illnesses as they can possibly associate with raw milk, and then sensationalizing the data. As one example, the CDC in 2012 conducted an analysis claiming to the media that raw milk is 150 times more dangerous than pasteurized milk. What the agency didn’t clarify sufficiently for media people not familiar with the complexity associated with data on foodborne illnesses is that their data, based as it was on outbreaks, was much different than it would have been had the agency been consistent with other reports, and focused on actual illnesses.

As another example, the CDC in 2012 and 2013 sponsored a first-of-its-kind study in Minnesota to try to prove there were many more illnesses from raw milk than were officially reported. And, predictably, the CDC’s study concluded that the officially reported 21 illnesses from raw milk over the first decade of this century in Minnesota was way too low; in its new estimation, there were more than 20,000 illnesses from raw milk over the decade. 

Why is the data on outbreaks a problem?

An outbreak is two or more illnesses. When there is an outbreak from raw milk, it generally involves just a few people, less than fifty. But when there is an outbreak from pasteurized milk, it can, and has, involved thousands of people, because pasteurized milk is so much more widely consumed. For example, a 1987 outbreak of salmonella tied to pasteurized milk sickened as many as 197,000 people, according to a report in the Journal of the American Medical Association.  And a 1994 salmonella outbreak sickened 224,000 people, and was eventually traced to tainted pasteurized-milk ice cream.

So while there is a wide disparity in the number of outbreaks between raw milk and pasteurized milk illnesses, there is less of a disparity in the number of illnesses. 

That does it for this episode of “Cutting Through the Fog on Raw Milk Illnesses”.