According to the CDC, for every one person who is a stool-culture positive victim of salmonella in the United States, there a multiple of 38.5 who are also sick, but remain uncounted.  (See, AC Voetsch, “FoodNet estimate of the burden of illness caused by nontyphoidal Salmonella infections in the United States,”Clinical Infectious Diseases 2004;38 (Suppl 3):S127-34).  This means that in this salmonella Saintpaul tomato outbreak of 228 (this number is bound to rise), we are missing 8,550 people so far.  (By the way, I am still in London and visited Saint Paul’s Cathedral yesterday).

So, how does such a miscount actually happen?  Frankly, some people who became ill simply suffer in silence or were not sick enough to seek medical treatment.  Those that do seek medical attention are also not that likely to have a doctor order a stool culture for salmonella.  Without a stool culture they do not get counted in the “official” CDC numbers (to date 228).

On where in the world is tainted tomato?  The U.S. Food and Drug Administration added New Mexico, Indiana and one Mexican state to its list of places that are cleared of being the source of the salmonella outbreak.  The FDA has now cleared 37 states, Puerto Rico and parts of Florida as the source of the outbreak.

Thus far we have been contacted by over 50 people in the US who either are stool-culture positive salmonella Saintpaul or who believe that they became ill after eating tomatoes in April or May.  Several have identified the same restaurant chain. Because we have not see the health department documentation of the illnesses, we will not divulge that name of the chain – yet.  Interestingly, on June 13, the FDA said that nine people fell ill at two restaurants that were part of the same chain.  The agency also declined to name the chain or its locations. Hmm, wonder if I beat the FDA to the punch?

Anyway, I still hope someone answers my questions:

1. Why does it take CDC so long to figure out an outbreak is going on?*

2. Why does it take so long for FDA to figure out where the tainted tomatoes came from?*

3. What if this was a bio-terrorism act instead of a likely food handling error?

*  Note – I think this more has to do with resources than it does will or ability.

  • I got this comment via email from a very bright, but unknown source:
    1. CDC rarely figures out that outbreaks are happening—-states or local health departments do. So the real reason why it takes CDC so long is that they don’t have anybody to call. They don’t get case names, phone numbers, or any other identifiers, unless some state is so feckless as to turn their cases over to them, which is pretty rare. Is that a bad thing? I don’t think so—-would you rather have relatively inexperienced junior people or trainees (i.e., CDC) doing these investigations? In this case NM figured out that an outbreak was going on as soon as they got 2 or 3 Saintpauls. You want us to figure this out with the first case? Most of the first cases lived on the Navajo Reservation. For the most part they did not have telephones, and it took 1-3 hours of driving to find out if anyone was home. If no one was home, someone had to go back later. Controls in this “neighborhood” were similarly difficult to locate and interview.
    2. This is a more difficult question for epidemiologists to answer. Part of the answer is that most consumers are poor historians—-meaning that they don’t remember very clearly when, where, or what kind of tomatoes they ate 2 or 3 weeks before they were contacted. When they do remember, it is sometimes in error, and often ambiguous. So the poor slobs at the FDA get shitty information from us, and they have to make the best of it. Tomatoes are handled like commodities, which means that they are often mixed and intermingled at one or more of the middleman stops along the continuum from farm to store, restaurant, or taco stand. It is worse than almost any other kind of produce. And yes, the FDA does need a lot of additional resources. And it wouldn’t hurt if they would be more forthcoming with epidemiologists instead of doing everything in secret.
    3. What if it was? Why would a terrorist put a relatively low dose of a relatively innocuous pathogen into a tomato? And moreover, what makes you think it was likely a food handling error? It might have been, but do you call a bird flying over a field or a snake crawling through it a food handling error?
    4. Given the unfortunate prevalence of both diarrhea and tomato consumption, the prior probability that someone without lab confirmation already (e.g., someone who ate tomatoes and “later” got sick) is a part of this outbreak is very low—-probably

  • Melissa

    Most people have no idea how hard it is to get properly diagnosed. Last year I had a bout of what I thought was food poisoning, but the symptoms did not subside even though they got milder. I was already diagnosed with irritable bowel syndrome, so doctors just assumed this was also IBS. However, none of the medicine helped. Finally I ended up in the E.R. where they finally did a stool culture…which turned up Salmonella St. Paul. I never figured out where it came from despite the fact I was, at that time, keeping a detailed food diary to control my IBS. No one ever followed up about tracing it either. Most food poisoning comes and gos mysteriously without the culprit being held accountable.

  • Another Email:
    You have to understand how these multipliers work: They are just
    WAG’s (“Wild-Arsed Guesses”).
    You are trying to estimate the unknown and unreported cases, knowing
    only the reported ones. You have no information upon which to base an
    estimate, other than opinion. What if I told you that, for every
    eyewitness that reports a crime, there are 12.6 who don’t. How would
    you verify this estimate?
    The reported cases are typically the most serious ones. They involve
    the following steps: 1) Patient has serious or chronic symptoms; 2)
    Stool culture is positive; 3) Resolution is difficult, or
    hospitalization occurs; 4) Sequelae or death may occur; 5) The
    disease is reportable, and someone reports it.
    There is a very good chance that all salmonellosis cases that result
    in hospitalization are reported. A large fraction of those resulting
    in death are also reported. I’d guess 1:1 for hospitalization and 1:2
    for death, perhaps.
    For acute symptoms that resolve upon treatment over 1-4 weeks, such
    as bloody diarrhea, I’d guess that maybe 1:5 or 1:10 are actually reported.
    For mild symptoms that are inconclusive, such as watery diarrhea, the
    reporting rate would be low, perhaps 1:30.
    For mild symptoms with no stool culture, the reporting rate would be zero.
    The high estimates of FBI are based on massive multipliers for mild
    cases. But these have little cost to society.
    The important cases are those with acute symptoms and sequellae.
    These are much more likely to be reported.
    You can determine these multipliers objectively from epidemiological
    investigations. For each “confirmed” case, there are typically 2-5x
    the number as “probable” cases.
    For the instant outbreak, if there are 228 reported cases, there are
    probably 400-1000 “probable” cases out there. Other theoretical cases
    are mild enough to be ignorable from a public health point of view.

  • Bix

    38.5 unreported cases for each reported case. Wow.
    I read the study to see what their methods were.
    Not a bad job of estimating non-confirmed cases, I’d say. Especially since their estimate of culture-confirmed cases was in line with actual reported data.
    It’s also nice to see that there was a system (FoodNet) established (in part) to document the effectiveness of HACCP.

  • Steve

    Mr. Marler,
    I stumbled upon this blog by my desire to satisfy my curiosity. This morning it dawned on me that if I was a terrorist, and I wanted to test how effectively I could poison the food supply, this would be a good technique.
    I “Googled” tainted-tomatoes-terrorism, and here I am. Shockingly, this was to only dominant URL on terrorism and the current tomato issue.
    Personally, the lax sanitation protocols in 3rd world nations that we import our food from seems to be a more likely reason for this outbreak. Nonetheless, I am confident that other entities that mean us harm here in the U.S. are very interested in studying the mechanics of this problem.

  • anne

    I agree with Mr. Marler. I believe it could possibly be terrorists figuring out how to taint our food supply. Look at the Salmonella map and you will see the most cases of outbreak are in Texas. They either need to look at all foods being exported from Texas for the origin of Salmonella or significant changes need to be made in how our food is monitored due to possible terroism. If you think about it, using poison in our food would be an easy way for terroists to attack our country and I think that is what they are going to do next.