Salmonella in Peanut Butter: Supply Chain Traceability Case Study

Background

Over a four-month period, public-health laboratories in eleven states reported 63 confirmed cases of Salmonella Typhimurium with an indistinguishable pulsed-field pattern. Ages ranged from 2 to 81. Epidemiologists ran a case-control study and found one exposure with overwhelming statistical strength: a national brand of creamy peanut butter, eaten by 78% of patients versus 14% of controls.

The peanut butter was made in a single large plant in the southeastern United States, running around the clock, roasting and grinding roughly 500 tonnes of peanuts a week into jars shipped nationwide. The product was shelf-stable, low-moisture — exactly the kind of food where Salmonella, once introduced, survives for months because the low water activity (around 0.3) protects the cells rather than killing them. A roasting step at 150–170 °C should have delivered a validated kill, but something downstream had gone wrong.

The plant’s own records showed no finished-product positives in the prior year. Its supplier certificates of analysis for incoming raw peanuts all read negative. On paper, the system worked. In the jars, it did not.

Investigation

Federal and state investigators entered the plant with a traceback mandate: follow the contamination from the jar backwards. They sampled widely — 180 environmental swabs across the facility, plus retained samples of finished product from multiple lots and raw peanuts from current inventory.

The finished-product results were damning: 9 of 42 retained jars tested positive for Salmonella Typhimurium, and the isolates matched the outbreak strain. Raw peanut samples from the warehouse were negative, which pushed attention to the post-roast environment.

Environmental positives told the story. Salmonella was recovered from a dust collection duct above the cooling conveyor, from the inside of a peanut-butter transfer auger, from a floor drain in the grinding room, and — critically — from the roof area above the packaging line where rainwater had been leaking through a damaged section for months. Maintenance logs showed the roof repair had been deferred three times.

The leak mattered because of what it enabled. Water dripping onto a ledge above the line created a damp pocket in an otherwise dry plant. Salmonella thrives in such intermittent-wet niches. Dust from the raw-peanut receiving area — where Salmonella is an expected hazard — travelled through shared air handling into the post-roast zones. The plant’s hygiene zoning, which should have separated raw and post-lethality areas with physical barriers, dedicated equipment and airflow controls, existed only as coloured lines on a floor plan. In practice, forklifts, maintenance staff and air moved freely between zones.

Investigators also reviewed the roasting validation. The roast had been validated five years earlier on a different peanut variety with different moisture content. No re-validation had been performed after the plant switched suppliers and increased line speed by 15%, which reduced dwell time in the roaster. The kill step’s margin of safety had quietly eroded.

Root Cause

The root cause was post-process contamination of a low-moisture product, enabled by failed hygiene zoning and a neglected building envelope. Salmonella entered the post-roast environment via dust migration from the raw area and found harbourage in a water-damaged zone above the packaging line. The un-revalidated roast may also have allowed occasional survival of incoming contamination, though the environmental pattern pointed primarily to recontamination after the kill step.

Underneath sat a management failure: deferred maintenance on the roof, a zoning programme that existed on paper only, and supplier certificates accepted without verification. The plant had never audited its peanut suppliers’ Salmonella controls, relying entirely on certificates of analysis from the suppliers themselves.

Corrective Actions

The company issued a nationwide recall of all peanut butter produced over a seven-month window — more than 12,000 tonnes — and shut the plant for a full remediation. The corrective plan included:

• Building repairs: the roof section was replaced, the water-damaged zone above the packaging line was stripped back to structure and rebuilt, and the dust collection system was re-engineered with HEPA-filtered exhaust so raw-area dust could not reach post-roast zones.
• True hygiene zoning: physical walls now separate raw receiving from roasting and post-roast areas, with separate air handling, dedicated tools colour-coded by zone, captive footwear, and a personnel flow that prevents backtracking from raw to clean areas.
• Roast re-validation: the kill step was re-validated with inoculated-pack studies using Salmonella surrogates at the current line speed and with the current peanut varieties, establishing a minimum time-temperature combination with a documented safety margin. Roast parameters are now continuously monitored with automated alarms.
• Environmental monitoring: a formal programme of 100 weekly swabs with zone-based escalation, plus finished-product testing of every lot with hold-and-release pending results.
• Supplier verification: on-site audits of all peanut suppliers’ Salmonella controls, and a shift from accepting supplier certificates alone to periodic independent testing of incoming lots.

The plant reopened after eight weeks following regulatory review of the corrective actions and three rounds of negative environmental testing.

Lessons Learned

Low-moisture foods are not low-risk foods. Salmonella survives for months at water activities below 0.4, so a single post-roast contamination event can affect enormous volumes of product with a long shelf life — which is why peanut butter outbreaks grow so large before detection.

The broader lessons: hygiene zoning must be physical and enforced, not drawn on a map; building maintenance is food safety (a leaking roof is a hazard, not a facilities issue); kill steps must be re-validated whenever raw materials, equipment or line speeds change; and certificates of analysis are only as good as the verification behind them. In low-moisture processing, the dry environment is your ally only if you keep it dry — one persistent damp niche can undo everything.

This case study is an educational reconstruction based on common outbreak investigation patterns. Facility names and specific details are fictionalised; it does not describe any single real company.