Raw and Undercooked Sprouts

Food · §336

The little nest of alfalfa on your sandwich is the one item in the produce aisle that has put thousands of people in hospital and run the largest kidney-failure cluster ever recorded — in 2011 it was a single batch of fenugreek seed from Egypt, four years of dialysis for some of the survivors, fifty-four dead Buchholz 2011. The per-meal risk for a healthy adult is genuinely small. The catch is that when the lot is bad, it is catastrophic, and the people who can least afford it — pregnant women, anyone over sixty-five, anyone immune-suppressed — are the ones the bacteria hit hardest. Saying no sprouts please at the counter costs nothing and closes a tail you didn't know was yours.

Avoid · As-needed Evidence Strong Chapter Food

For most healthy adults, skipping raw sprouts is a small, almost invisible win — a low-probability illness you never have. For a pregnant reader, a chemotherapy patient, a parent in their seventies, it is a much sharper one. The cost is one sentence at the sandwich shop and a habit of scanning grain-bowl menus. Thirty years of outbreak data, the FDA and CDC fully aligned: this is about as settled as food-safety advice ever gets.

A sprout is a seed kept warm and wet for three to seven days until it puts out a shoot. Those conditions — room temperature, continuous moisture, no sunlight, no cold — are also the textbook growing conditions for Salmonella, the Shiga-toxin-producing strains of E. coli, and Listeria. A handful of bacterial cells on the seed at the start of the run become millions by the time the sprouts hit the clamshell Taormina 1999. The germination tank is, biologically, a small incubator.

The bacteria don't arrive at the sprouting facility. They arrive on the seed — picked up in the field from irrigation water, manure, wildlife, or rodent contact months earlier, often a continent away. A single contaminated seed lot can supply dozens of growers; one bad batch in 1999 produced 157 Salmonella cases across eight U.S. states from seed distributed to 33 sprout growers in ten Taormina 1999. That is the structural problem: the contamination is upstream of every control the grower has.

And the bacteria are not only sitting on the surface of the seed, where a wash could reach them. They work their way into the seed coat and the developing root. The standard industrial disinfection step — a soak in a strong chlorine solution — knocks the load down by a thousand-fold on average, but the variability across batches is large enough that no single treatment can be trusted to produce a clean lot every time Ding 2013. The U.S. Food and Drug Administration treats this as the unavoidable ceiling of the current toolkit, not a problem with a clean engineering fix FDA 2019.

Thirty years of the same story

Between 1996 and 2020, the FDA documented 52 separate foodborne-illness outbreaks linked to sprouts in the United States, producing more than 2,700 illnesses FDA 2022. The CDC's outbreak system shows roughly the same shape: 53 sprout outbreaks between 2000 and 2020, with 1,498 illnesses, 179 hospitalizations, and five deaths in U.S. data alone CDC 2022. Two events sit on top of the global record and tell you what bad looks like.

The U.S. clusters between those two events are smaller but constant. 235 Salmonella cases from alfalfa sprouts in 2009. A multi-state bean-sprout Salmonella outbreak in 2014. Alfalfa-sprout Salmonella in 2016. A national sandwich chain whose clover sprouts produced repeated E. coli outbreaks across the 2010s, until the FDA sent a warning letter in early 2020 and the chain pulled sprouts from 2,800 stores permanently. The pattern is not a series of one-off failures — it is the same biology surfacing the same way over three decades.

The risk is shaped wrong

If you are a healthy adult who eats sprouts on sandwiches twice a week, you will most likely never get sick. The odds of any one meal carrying a bad lot are low. What makes this entry worth reading anyway is the shape of the risk, not its size.

When the lot is bad, it is bad in ways most other food-poisoning episodes aren't. The Shiga-toxin E. coli strains that have ridden sprouts into the outbreak record don't just give you a few days of diarrhea — in roughly one infection in twenty, the toxin destroys red blood cells and shuts the kidneys down, an emergency called hemolytic uremic syndrome. About half the children who develop it need dialysis during the acute illness, and about five percent of survivors never get full kidney function back and end up on dialysis or a transplant list for life. Adults over sixty have the highest death rate of all Frank 2011.

For a pregnant reader, the relevant pathogen is different and the asymmetry is even sharper. Listeria exploits a quirk of pregnancy: the immune system is partially turned down so the body doesn't reject the fetus, which means a pregnant woman is about ten times more likely to develop a serious Listeria infection than another healthy adult of the same age. When the bacteria cross the placenta, roughly a quarter of those infections end in miscarriage, stillbirth, or a newborn fighting sepsis in intensive care FDA 2024. The mother often barely feels sick. The first sign that something has gone wrong is sometimes the ultrasound.

The honest framing isn't you will get sick — for most readers, you won't. The framing is: the same garnish that, on most plates, does nothing at all is, on a tiny fraction of plates, the worst week of someone's life. The people who pull the bad draw are disproportionately the people who can least afford it.

What to actually do

The advice is binary. There is no safe-sourcing trick that meaningfully lowers the residual risk — the contamination is on the seed, the same warm-humid germination phase amplifies it regardless of how clean the facility is, and even the standard chemical seed wash leaves enough variability across batches that the next bad lot is always possible. So the practical action collapses to two things: at restaurants, ask for sprouts to be left off; at home, if you buy them, cook them all the way through — into a stir-fry hot enough to wilt them completely, or dropped into a soup or curry, not the lukewarm "warmed through" finish you get on a grain bowl.

For four groups, the floor is higher. The CDC and FDA both say plainly: pregnant women, anyone over 65, anyone whose immune system is suppressed by chemotherapy, biologics, transplant medications, or HIV, and young children should not eat raw or lightly cooked sprouts of any kind — including alfalfa, clover, radish, mung bean, and onion sprouts CDC 2022 FDA 2022.

Three things people get wrong

Organic and local are not safer. The 2011 German outbreak, the biggest in the history of this food, was traced to certified-organic fenugreek seed; the Sakai outbreak was a single small Japanese farm. Contamination happens at the seed-production stage, often abroad, sometimes years before the sprout shows up in your kitchen — and an "organic" certificate covers the growing practices, not the bacterial load on the seed.

Growing them at home is not safer. The kitchen-bench sprouter has the same problem the commercial facility does — bacteria already inside the seed, amplified by the same warm-wet germination phase — and without the industrial disinfection step that at least lowers the starting load Ding 2013. A clean countertop and good hand hygiene help with the things you bring to the seed; they do nothing about what the seed brings to you.

Rinsing doesn't fix it. A vigorous wash under the tap is what most home cooks reach for instinctively, and it is genuinely useful on, say, a head of lettuce. It does very little to sprouts. By the time the shoot is out, the bacteria are not sitting on the surface where water can reach them — they are inside the seed coat and the developing root Ding 2013.

If you wanted them for the crunch — or for the nutrition

For the texture and the fresh-bright look on a sandwich, anything that gets thin and crisp works: shaved cabbage, julienned carrot or radish, shredded lettuce, ribbons of cucumber, sliced fennel, a fistful of herbs. None of them requires three days in a warm cupboard.

For the nutritional reach often claimed for raw sprouts — vitamin K, folate, fiber, a clean greens-y profile — any leafy green delivers comparable amounts per gram. The interesting outlier is broccoli sprouts and a compound called sulforaphane, which the young sprouts genuinely concentrate at levels mature broccoli doesn't. Even there, you can get meaningful sulforaphane by lightly steaming mature broccoli florets for two to three minutes, or by taking a glucoraphanin supplement — without the seed-coat-internalization window that drives the outbreak record.

What you get back

Most of the payoff is invisible — the foodborne illness you never have, the renal ward you never see. For the typical healthy adult that is the entire reward, and it is honest to say it doesn't feel like anything. You will not have more energy. You will not look different. Nobody around you will notice.

The payoff turns concrete in the windows where the risk does. The pregnant reader walking past the salad-bar sprouts at a work lunch and quietly choosing the cucumber instead is, statistically, removing one of the small set of foodborne routes that produce a fetal-loss event during the months her immune system is least equipped to contain them. The reader whose father is on chemotherapy, doing the family grocery run, leaving the clamshell of alfalfa on the shelf, is doing the same thing. None of those decisions will ever obviously pay off — the events they prevent are by construction the ones that didn't happen — but the math is clean and the choice is free.

A few related things readers often ask about. Sprouted-grain bread is a different food with a different safety profile — the wet phase is short, then the grain is milled and baked, and the outbreak record looks nothing like raw sprouts. Microgreens are harvested a week or two later than sprouts, after the warm-humid germination phase is over, and don't share the same risk. If you're pregnant or immunocompromised and reading this, the broader list of foods to skip — unpasteurized soft cheeses, deli meats served cold, raw fish, raw eggs — runs on the same logic and is worth a longer look.

Substance and claimed effects

The substance is raw or undercooked seed sprouts consumed as a salad or sandwich topping: alfalfa, clover, radish, mung bean, and the broader sprouted-bean / sprouted-legume category (fenugreek, lentil, soy). Distinct from sprouted grains used in breads (where the wet phase is followed by milling, baking, or extended cooking) and from microgreens (harvested later, after the cotyledons have hardened and the warm-humid germination phase is over). The claim under investigation is not that sprouts are nutritionally bad — the dossier accepts the raw-sprout nutrient profile as roughly comparable to other young greens, with the broccoli-sprout sulforaphane case as a real outlier. The claim under investigation is that the germination process that produces a sprout is also a near-ideal incubator for any bacteria already on the seed, and that this has produced a 30-year run of Salmonella, Shiga-toxin E. coli, and Listeria outbreaks disproportionate to how much of the food supply sprouts represent. Consequences scored: acute foodborne illness risk (health_short_term), rare-but-catastrophic mortality and end-stage renal disease from hemolytic uremic syndrome (longevity), with severity multipliers for pregnancy, age over 65, and immunosuppression.

Evidence by addressing question

mechanism

Sprouting requires 3–7 days at roughly 20–25 °C in continuously moist conditions. These are the textbook growth conditions for Salmonella enterica, Shiga-toxin-producing Escherichia coli, and Listeria monocytogenes — doubling times of 20–40 minutes. The defining mechanistic problem is that the contamination route is almost always the seed, not the sprouting facility: low-level field contamination from irrigation water, manure, wildlife, or rodent contact at the seed-production stage seeds the inoculum, and the germination tank then amplifies it by 4–8 log over the growing run (Taormina, Beuchat, Slutsker 1999; Ding et al. 2013). Pathogens are not only on the seed surface — they internalize into the seed coat, the cotyledon, and the developing root, where surface disinfectants cannot reach them. This is why the FDA-reference seed treatment of 20,000 ppm calcium hypochlorite reduces bacterial load by only ~3 log on average with a standard deviation of ~2 log, and cannot guarantee a clean product (Ding et al. 2013; FDA 2019). A single contaminated seed lot can supply dozens of growers; the 1999 Salmonella Muenchen outbreak traced 157 cases across 8 states to one seed lot distributed to 33 sprout growers in 10 states (Taormina et al. 1999).

evidence

The outbreak record is one of the most consistent in the produce literature. Between 1996 and 2020, the FDA documented 52 sprout-associated foodborne-illness outbreaks producing more than 2,700 illnesses in the United States alone (FDA 2022); CDC's National Outbreak Reporting System recorded 53 sprout outbreaks between 2000 and 2020, with 1,498 illnesses, 179 hospitalizations, and 5 deaths (CDC 2022). Two events dominate the global record. Sakai City, Japan, July 1996: white radish sprouts from a single farm caused E. coli O157:H7 illness in 6,309 schoolchildren and 92 school staff across 62 elementary schools, with 101 cases of hemolytic uremic syndrome and three child deaths (Michino et al. 1999). Germany, May–July 2011: fenugreek sprouts grown from a single Egyptian seed lot caused 3,816 cases of Shiga-toxin E. coli O104:H4 infection, including 845 cases of hemolytic uremic syndrome — the largest HUS cluster ever recorded — and 54 deaths, with most of the HUS occurring in healthy adult women rather than the usual pediatric profile (Buchholz et al. 2011; Frank et al. 2011). Smaller U.S. clusters are recurring rather than isolated: 235 Salmonella Saintpaul cases from alfalfa sprouts in 2009; the 2014 Salmonella Enteritidis bean-sprout outbreak across 12 states; the 2016 Salmonella Reading and Abony alfalfa-sprout outbreak; multiple clover-sprout-mediated E. coli outbreaks tied to one national sandwich chain across the 2010s, culminating in a 2020 FDA warning letter and the chain pulling sprouts from 2,800 stores (CDC 2022). The pattern reads as a structural property of the food, not as a series of one-off failures.

protocol

The intervention is binary and behavioural rather than dose-titrated. Cook thoroughly to ≥ 71 °C / 160 °F — the temperature at which Salmonella and E. coli are reliably killed in a few minutes of contact (FDA 2022; CDC 2022). In practice this means stir-fried mung bean sprouts in a hot wok, sprouts dropped into boiling pho broth long enough to wilt fully, sprouts cooked into a curry or soup — not the brief "wilt" that leaves them crunchy. Lightly cooked (the lukewarm "warmed through" finish characteristic of many salads and grain bowls) does not clear the bar; CDC explicitly groups lightly cooked sprouts with raw sprouts as a hazard (CDC 2022). Rinsing has no meaningful effect on internalized pathogens (Ding et al. 2013). For the casual reader the protocol simplifies to: at restaurants and sandwich shops, ask for sprouts to be left off; at home, if buying them, cook them into something hot.

contraindications

The CDC and FDA both explicitly name four groups as needing to avoid raw and lightly cooked sprouts of any kind: pregnant women, children, adults over 65, and people with weakened immune systems (HIV, organ transplant, active cancer treatment, biologics) (CDC 2022; FDA 2022). The risk multiplier is not symbolic. Pregnant women are roughly 10× more susceptible to Listeria infection than other healthy adults, because gestational immune tolerance suppresses cell-mediated immunity that normally clears the organism; transplacental infection causes miscarriage, stillbirth, preterm delivery, or neonatal sepsis, with fetal mortality near 25% once invasive listeriosis crosses the placenta (FDA 2024). For E. coli O157:H7 and O104:H4, the case-fatality rate of hemolytic uremic syndrome runs 1–4% overall, with adults over 60 carrying the highest mortality in the 2011 German outbreak (Frank et al. 2011); approximately half of children who develop HUS require dialysis during acute illness, and ~5% of survivors progress to end-stage renal disease requiring dialysis or transplant for life.

misconceptions

Three myths recur. First, that "organic" or "locally grown" sprouts are safer — the 2011 German outbreak was traced to certified-organic fenugreek seed, and the 1996 Sakai outbreak to a single small Japanese farm (Buchholz et al. 2011; Michino et al. 1999). Contamination is a seed-level upstream problem, agnostic to growing-facility quality. Second, that home-grown sprouts are safer than commercial. They are not: the warm-moist germination phase amplifies whatever inoculum the seed brings, and the kitchen-bench grower has the same internalized-pathogen problem the commercial facility does, without the FDA-mandated seed-disinfection step (Ding et al. 2013). Third, that a thorough rinse before eating clears the risk. Internalized bacteria in the seed coat and cotyledon are not removed by surface washing (Ding et al. 2013).

stakes

Anchored on the typical reader, the absolute per-meal risk of foodborne illness from a sprout-containing sandwich is low — outbreak-implicated lots are a small fraction of the total. The honest stake is asymmetric severity, not high frequency. The healthy 35-year-old who eats sprouts on sandwiches twice a week will likely never get sick; the 1-in-a-few-thousand bad lot that arrives one day produces, at worst, a few days of bloody diarrhea and recovery. The stake bites at the tails: a pregnant reader trading a Listeria exposure for a topping she didn't want anyway, and ending the pregnancy; an older parent on a sprouted-grain-bowl kick whose Shiga-toxin E. coli illness escalates to HUS and dialysis; the rare healthy adult who pulls the 2011-Germany draw of a single contaminated lot and ends up in the 845-person HUS cluster. The stake is the variance, not the mean.

payoff

For the typical reader, the payoff of skipping is almost entirely invisible — the absence of a low-probability event you never see. For vulnerable readers (pregnancy, immune compromise, age over 65) the payoff is concrete: one fewer plausible foodborne route during a window where the immune system cannot reliably contain it. The nutritional cost of skipping is trivially small — alfalfa and mung bean sprouts contribute little that other vegetables don't (the broccoli-sprout sulforaphane case is the exception, and even there sulforaphane is available from cooked broccoli and from glucoraphanin supplements).

alternatives

For the crunchy-fresh-on-a-sandwich texture, alternatives that don't require a 3–7 day warm-humid germination phase: thinly shaved cabbage, julienned carrot or radish, shredded lettuce, cucumber ribbons, sliced fennel, fresh herbs (cilantro, mint, basil). For the nutritional reach often claimed for sprouts: any leafy green delivers comparable vitamin K, folate, and fiber per gram. For sulforaphane specifically — the one bioactive raw sprouts genuinely concentrate — broccoli florets briefly steamed (2–3 minutes), or a glucoraphanin / sulforaphane supplement, give meaningful intake without the seed-coat-internalization risk window.

practicalities

Raw sprouts appear by default in three places most readers encounter them: sandwich shops (where they are a free topping the reader can decline), grain bowls and salads at fast-casual chains (where they are mixed in and may be harder to spot or remove), and grocery clamshells in the produce section (where the consumer choice is whether to buy). Restaurants and chains have responded to outbreak history at uneven rates — at least one major U.S. sandwich chain permanently pulled sprouts in 2020 after repeated outbreaks and an FDA warning letter, while others continue to serve them. The practical action is to ask, and to assume that a restaurant claim that sprouts are "tested" or "from a reputable supplier" doesn't actually reduce the seed-borne risk that drives the outbreak record.

history

The first large outbreak that put sprouts on the food-safety map was a 1995 multistate Salmonella Stanley cluster in the U.S. and Finland traced to alfalfa sprouts. Sakai 1996 made it international. By 1999 the CDC's Emerging Infectious Diseases review article was already calling sprouts "an international concern" and naming the warm-humid amplification mechanism (Taormina et al. 1999). The 2011 German O104:H4 outbreak crystallized two things: that a single contaminated seed lot can become a continent-scale public-health event, and that the usual pediatric profile of HUS does not apply when the pathogen reaches the adult population through a commonly eaten food. FDA's 1999 first guidance and 2017–2019 final rules on seed disinfection and sprouting-facility standards have measurably tightened the commercial pipeline, but outbreaks continue at a steady cadence and the underlying mechanism — pathogens internalized in seed before disinfection can reach them — is not fully solvable with the current toolkit (FDA 2019).

out-of-scope

Adjacent topics this entry deliberately does not cover but that share infrastructure or risk profile: raw oysters and Vibrio; unpasteurized milk and soft cheeses for the Listeria exposure during pregnancy; the broader pregnancy-food-avoidance list (deli meats, raw fish, raw eggs). Sprouted-grain bread is a distinct substance with a distinct safety profile (the wet phase is short and is followed by milling and baking) and warrants its own entry.

The credibility range

Optimist case

Per-meal absolute risk is low. The vast majority of sprout-containing meals do not produce illness. Commercial growers operating under FDA's 2017 produce rule are required to treat seed with an antimicrobial step, test spent irrigation water for pathogens, and monitor growing surfaces for Listeria — these controls demonstrably reduce contamination. Nutrient profile and bioactives (sulforaphane in broccoli sprouts most notably) are real. For a healthy non-pregnant adult outside the vulnerable groups, eating sprouts occasionally is a defensible food choice, and many cultures (East and Southeast Asian cuisines especially) eat lightly cooked or raw sprouts routinely without obvious population-level harm.

Skeptic case

The outbreak record over three decades does not reflect a series of correctable failures — it reflects a structural property of the food. The amplifying biology cannot be engineered away while the warm-humid germination phase exists; seed disinfection reduces but does not eliminate internalized pathogen, with variance large enough that any given lot can be the one. Population-level data: 30 years of guidance, multiple FDA rule revisions, and the outbreak cadence has not closed. The nutritional case for raw sprouts specifically (as opposed to other young greens or cooked sprouts) is thin enough that the risk-benefit math points the same way for most readers regardless of subgroup. Public-health agencies that have to be conservative are unusually direct here: the CDC and FDA explicitly recommend vulnerable groups avoid raw sprouts entirely, and recommend everyone cook them thoroughly (CDC 2022; FDA 2022).

Author's call

The skeptic case is stronger. The evidence is high (30 years, two continents, ~3,000 documented illnesses in the U.S. alone, the largest HUS cluster ever recorded, regulatory consensus that does not flex) and the controversy is essentially zero — public-health agencies agree, industry has accepted the rules, the science is not in dispute. The entry takes the FDA/CDC line: vulnerable groups should not eat raw or lightly cooked sprouts; everyone else should default to cooked-thoroughly when they eat them at all. The honest framing for the typical healthy adult is asymmetric severity, not "you will definitely get sick" — the risk is real but tail-shaped.

Stakeholder and incentive map

Commercial sprout producers — industry trade group (International Sprout Growers Association) advocates that sprouts are safe when grown under FDA-rule-compliant conditions; the rules they comply with did not exist until repeated outbreak history forced them.
Sandwich and salad chains — sprouts are a low-cost topping and a perceived "healthy" signal. The mixed response — at least one major chain pulled sprouts permanently in 2020 after an FDA warning letter; many chains continue to serve them — reflects brand-protection vs. menu-tradition tension.
Regulators (FDA, CDC, USDA, state health departments) — the consumer-facing recommendation has been stable and aligned for decades; no agency has softened on the vulnerable-groups advice.
Wellness / raw-food community — counter-pressure from a culturally important segment of the produce-eating audience, where raw sprouts carry symbolic weight as a "living food." This community sometimes downplays outbreak data as overblown.
Pregnancy and maternal-health clinical groups — uniformly aligned with the FDA/CDC line; the listeriosis-in-pregnancy data is unambiguous (FDA 2024).

Population variability

Pregnant women — ~10× increased Listeria susceptibility; transplacental infection mortality near 25%; FDA advises avoiding raw sprouts categorically (FDA 2024).
Adults over 65 — highest case-fatality in E. coli HUS clusters (most clearly demonstrated in the 2011 German O104:H4 outbreak); reduced immune containment of invasive Listeria and Salmonella (Frank et al. 2011).
Young children — highest HUS rate among E. coli O157:H7 cases; the Sakai outbreak gave the dominant evidence: child-only deaths, 101 pediatric HUS cases (Michino et al. 1999).
Immunocompromised adults — HIV with low CD4 counts, organ-transplant recipients on tacrolimus or cyclosporine, active chemotherapy, high-dose corticosteroids, anti-TNF biologics — all increase invasive-infection risk for Listeria and Salmonella.
Healthy non-pregnant adults 18–64 — lowest absolute risk; the bulk of sprout-containing meals are eaten by this group without incident, but the 2011 Germany cohort demonstrated that healthy adults can be the dominant HUS profile when an unusually virulent strain reaches a common food.
Geographic exposure variability — East and Southeast Asian cuisines feature mung bean sprouts heavily, often briefly stir-fried or boiled (which clears most pathogen if hot enough for long enough); North American exposure is dominated by raw alfalfa and clover on sandwiches and salads, which carries the higher residual risk.

Knowledge gaps

Several open questions. What share of contaminated seed lots evade FDA-rule treatment — the published efficacy data on calcium hypochlorite seed treatment shows mean ~3 log reduction with SD ~2 log; the variance distribution at the tail (what fraction of treated lots still carry a problematic inoculum) is not well characterized in public data (Ding et al. 2013). Whether irradiation of seed could close the residual risk — technically promising, regulatory and consumer-acceptance barriers have kept it from scaling. Population-level effectiveness of the FDA 2017 produce rule — outbreak counts have not obviously declined since 2017; whether this is because the rule isn't working, because pre-rule contamination rates were already plateauing, or because reporting sensitivity has increased, is not clearly resolved in published evaluations. Whether the broccoli-sprout sulforaphane benefit case shifts the risk-benefit math for narrowly that one sprout type and that one bioactive — sulforaphane availability from cooked broccoli or supplements suggests no, but the sprout community continues to debate it.

Scope vs brief. The brief named alfalfa, mung bean, clover, and radish sprouts on salads and sandwiches; the nutrient-profile contrast with sprouted grains; the link to Salmonella and E. coli outbreaks; heightened risk for pregnant / older / immunocompromised; and cooking and sourcing practices that lower risk. All of these are covered. The sprouted-grain distinction is handled in the dek's framing and in out-of-scope rather than as a standalone addressing section, since the entry's spine is the raw-sprouts safety case and grains are a different food. Listeria is included alongside the two named pathogens — the pregnancy case requires it.

Rating calls worth flagging.

health_short_term at 2 (not 3) reflects honest per-meal asymmetry: most readers will never get sick, even with frequent exposure. Pushing to 3 would overstate the typical-reader case. The vulnerable-group risk does the heavy lifting in stakes and the warning callout, not in the score.
longevity at 2 captures tail-shaped mortality avoidance — HUS / end-stage renal disease / fetal loss are rare but catastrophic. A higher score would imply population-level mortality move, which the data doesn't support.
evidence at 5 is unambiguous: two named landmark outbreaks (Sakai 1996, Germany 2011), FDA/CDC alignment over 30 years, well-characterized mechanism. Meets the "name 2+ rigorous trials" bar.
applicability at 4 uses the avoidance / decision-audience lift — most adults choose whether to accept sprouts on sandwiches, even if they don't currently buy them at the grocery store. Naive prevalence (current sprout buyers) would be lower; the wider decision audience is real.
pull at 1 — the act of saying "no sprouts" produces almost no felt reward in the moment. The relief lever is real but mostly retrospective.

Dream narrative written despite sub-40 score (≈ 28). The relief / not-being-conned lever is honest and load-bearing for the dek and tagline; writing the narrative produced the sharper "skip the topping, close the tail" hook the dek and tagline now use. Below the obligatory floor, written by choice per dream-narrative.md §1.

Contraindications. Listed pregnancy and breastfeeding because raw sprouts are categorically contraindicated for these life stages (FDA Listeria-in-pregnancy guidance is unambiguous). The closed vocabulary doesn't carry tokens for "over 65" or "immunocompromised" — these groups are handled in the protocol section's warning callout and the dedicated audience prose.

Sandwich chain unnamed in the article body. The 2020 FDA warning letter and store-wide sprout removal involved a specific national chain; the dossier names it, the reader-facing prose deliberately doesn't. The entry is about the food, not about litigation-adjacent branding.

Separate-entry candidates surfaced during the write.

Listeriosis in pregnancy — the full food-avoidance list. Deli meats, soft cheeses, raw fish, raw eggs, pâté. The pregnancy section in this entry handles raw sprouts only; the broader list deserves its own entry.
Broccoli sprouts for sulforaphane. Specifically called out in alternatives; the bioactive case (10–100× glucoraphanin density vs mature broccoli) is real and the risk-benefit math is different enough from the general sprout case that it warrants standalone treatment, especially if paired with sulforaphane supplementation.
Hemolytic uremic syndrome — recognize and act. The presentation (bloody diarrhea progressing to pallor and reduced urine output in a child, 5–10 days after Shiga-toxin exposure) is one of those rare-but-serious emergency-recognition entries that scores high on the applicability-lift rule.

Future links to wire in. Once they exist: listeriosis-in-pregnancy, broccoli-sprouts-sulforaphane, recognizing-HUS, and the broader pregnancy-food-avoidance entry.

Excluded with reason.

Industrial seed-disinfection chemistry detail (calcium hypochlorite concentrations, soak durations, peroxyacetic acid alternatives). Belongs in the dossier (and is there); reader-facing it would flatten the felt voice without changing the action.
Country-by-country regulatory comparison (FSMA vs EU produce rules). Too detailed for the entry's actionable spine.
The full list of every documented U.S. sprout outbreak. The pattern is the point; an enumeration would read as a literature review and isn't what the reader needs.