Three men, one uncommon snack, and no sample left to test

On 1 December 2017, PulseNet found **three Listeria monocytogenes
clinical isolates** with indistinguishable patterns, **within three allele
differences** by whole-genome multilocus sequence typing — close enough to
indicate one source.

The three cases were in Illinois, Iowa and Michigan, isolated between **15
and 29 October**. All three patients were men, aged 55 to 71 (median 69),
and all three were hospitalised. No deaths.

What connected them

All three had eaten prepackaged caramel apples in the month before falling
ill — bought at retail, all the same brand, sold as **three apples on
sticks in a plastic clamshell**.

The statistical test is the nice part. Their exposure was compared against
186 sporadic listeriosis cases from the same states reported since 2006:

**Caramel apple consumption was significantly higher among the outbreak
patients** — odds ratio 21.7 (95% CI 2.3–infinity).

An interval running to infinity means the comparison group barely ate them at
all. That is exactly what makes an uncommon, seasonal food such a strong
signal: three people who all ate the same unusual thing is not a coincidence
the way three people who all ate lettuce would be.

And then nothing could be found

The timing explains most of it. The shelf life is 15 days, the illnesses
were in October, and the cluster was identified on 1 December. By then the
product that made three men ill had been eaten, thrown away or spoiled weeks
earlier.

**Because there was no evidence of ongoing risk, no public warning was
issued** — the honest conclusion when the implicated food can no longer be on
a shelf.

What was concluded anyway

**Although the outbreak strain was not isolated from caramel apples or their
production environment, the epidemiologic evidence indicated that caramel
apples were the suspected vehicle.**

Three people, one indistinguishable strain, one uncommon brand-specific
product. That is a conclusion drawn from people rather than from samples, and
it is the normal state of affairs for a food with a two-week life.

It had happened before: **caramel apples caused a large multistate listeriosis
outbreak in 2014–2015, from contamination of the whole apples.** Which is why
the recommendation aims upstream — **ready-to-eat food processors should treat
the introduction and persistence of Listeria in production environments as a
hazard**, with environmental monitoring and preventive controls, and **further
research into controlling it in fresh produce, including fresh apples**, might
find ways to remove it before it reaches a stick.

Source: Centers for Disease Control and Prevention, MMWR.