Insights and Analysis
CDC Stays Mum On Release Of Tomato/Salmonella Data
Jim Prevor’s Perishable Pundit, June 6, 2008
One of the big questions we don’t know the answer to is whether any warning on consumption of tomatoes is justified at all. We know that sometimes by the time the government acts, the crisis has long passed. When the FDA imposed its “import alert” on Agropecuaria Montelibano, a Honduran grower and packer, it was March 22, 2008, and the last person known to have gotten sick was on March 5, 2008. Since Salmonella has an incubation period of 12 to 72 hours after infection, we know that by the time the FDA issued its Import Alert, the outbreak was over.
We don’t have enough information to say definitively what the situation is with the Salmonella Saintpaul outbreak and tomatoes. The CDC info sheet says that “Among the 38 persons who have been interviewed, illnesses began between April 23 and May 27, 2008.”
Now information can come in irregularly on this type of thing, and until all the information is in we won’t know the final result. Still, what can be useful is knowing the trend. Of the 38 people who have been interviewed, if we started off with one person sick on April 23, two sick on April 24, three sick on April 25 and then hit a peak on, say, May 4, with a decline every day thereafter to a May 27, when one person fell ill, this would give us some notion that we are near the end of a bell curve.
On the other hand, if the numbers had been increasing since April 23, and May 27 was the peak day, this is a strong indication that we still have a ways to go on this outbreak.
We asked Mira Slott, Pundit Investigator and Special Projects Editor, to find out more:
Centers for Disease Control and Prevention (CDC),
Q: Could you clarify the time line from the date when the first person reported getting ill to the most recent date a person reported getting ill? In that time line, could you provide us with data on how many people reported being ill by dates and location?
A: We are in the middle of the investigation. It is too early to establish a time line. There is a range when people reported that they got sick. You’ll see in CDC’s release of June 4, words are chosen judicially and specifically. Among 38 persons interviewed, their illnesses began between April 23 and May 27, 2008. This was a case study.
The epidemiology was conducted by New Mexico and Texas and Indian Health Service. There are 29 additional persons with the outbreak strain of salmonella Saintpaul in other states. Investigations are going on in those states to determine if those illnesses are linked to tomatoes. States are still monitoring and working through Pulsenet. This is the latest data CDC has collected.
Q: How does the scientific and epidemiological process work?
A: There is self reporting by people of when they got sick. Part of the epidemiological work is case/controlled studies of those who became ill. The case study by New Mexico, Texas and Indian Health Service identified the link to tomatoes. That same kind of work is now being done in those other states. The state health department asks a series of questions to determine what made them sick, in addition to the lab work of the genetic fingerprint, when did they get sick, what products purchased, where, etc.
The phrasing in CDC’s release of “when illnesses began” is when the person reported they got ill. That’s a wide range as well. Look at Salmonella; it develops between 12 to 72 hours after the infection, usually lasts four to seven days, and most of these people may not even report it. These people who report actually got sick enough to go to the doctor, and the private physician or hospital does a stool sample and waits to get the lab results back. CDC is taking great care to report what they report.
In the case of the spinach E. coli outbreak, the system worked perfectly and information came together in two weeks. With this one, you don’t even have two weeks from May 27.
Q: I still don’t understand why CDC couldn’t provide the data or a chart showing what you know at this point. Why not provide the number of people reporting illnesses each day or each week by region or state, based on what you’ve determined as the onset April 23 to the most recent on May 27? You can always update the information as you learn more.
A: It is difficult to develop analysis without knowing the methodology used; you have to know the scientific methods and the epidemiological work involved. You can’t start manipulating our data. We’re still at an early stage.
The numbers reported may change at the end of the day after we talk to the states. They may come up. It takes additional time, depending on the Pulsenet systems in each state; data may not have reached the U.S. Health Department yet from the individual states. This is quick, real time information. There may be people sick during that time frame whose lab results haven’t presented themselves.
Q: The longer it takes to hone in on the problem, the greater the damage to consumers and the industry. It seems the more information you could provide, the faster people might be able to narrow down the problem and pinpoint the source.
A: This is an ongoing investigation, with epidemiology studies continuing in states and here. We are actively involved in lab work, and haven’t gotten to a documentation of how many people got sick in different time periods. With outbreaks, it’s hard to say a bell curve until it’s over. Tomatoes do have a certain shelf life, and also FDA is doing a traceback investigation.
We will leave aside the issue of whether the epidemiology “worked perfectly” during 2006 spinach crisis. Others have different opinions.
What concerns us here is that CDC is not acting in a spirit of openness and transparency most likely to produce optimal results. We are all adults; we know this dataset will change every day as new people are interviewed.
But there is something wrong with CDC trying to maintain a monopoly on information in a situation such as this. Why not tell what we know now, with the understanding that we are learning more every day? In this case, it would be very easy for CDC to post a graph with the onset date of the reported illnesses.
This is an area where the industry’s government relations staffs might want to look at legislation to compel the release of this kind of information in a timely fashion.
We have nothing but respect for the CDC and know it has many intelligent and hard-working people trying to do the right thing.
That is more likely to happen if its decisions and those of the FDA are subject to contemporaneous scrutiny by the best minds in the private sector. This means that the bias should be toward sharing information, not attempting to control it.
As it is right now, the FDA imposed its warning for Texans and New Mexicans not to eat certain types of tomatoes on June 3. The last person known to have become sick did so on May 27. With a 12 to 72 hour incubation time, this means the outbreak may well have been over by the time FDA got around to issuing its alert.
It is one thing to harm an industry to enhance the public health; it is another thing entirely to severely damage an industry so that FDA can look like it is doing something to enhance public health.
Taxpayers deserve to know which it is and the CDC should tell them voluntarily.
We thank Ms. Russell for the time she spent helping the industry become better informed.