6.3 Evaluating HACCP Plans & Administrative Documentation
Key Takeaways
- Blueprint tasks B.9 and B.14 require evaluating HACCP plans and reviewing administrative documentation during an assessment.
- Evaluating a plan means checking that the flow diagram matches reality, that the hazard analysis is complete, that critical limits are scientifically supported, and that records show the plan is actually executed.
- The most common plan defect is a critical limit with no validation evidence; the most common execution defect is a monitoring log filled in retrospectively.
- Records to review include monitoring logs, corrective action records, verification and calibration records, employee health agreements, date marking, and supplier documentation.
- Identical handwriting, uniform ink, perfectly regular values, and entries for a day the establishment was closed are the classic indicators of falsified records.
6.3 Evaluating HACCP Plans & Administrative Documentation
Exam Tip: Chapter 4 taught you to build a HACCP plan. Blueprint task B.9 asks you to evaluate someone else's. The mindset is different: you are testing whether the document describes the operation you are standing in.
Step 1: Does the Plan Describe This Establishment?
Start with the flow diagram, and walk it physically.
The single most common finding in HACCP evaluation is a plan that describes a process the establishment no longer runs — a step removed, a new piece of equipment added, a product line changed, a supplier switched. The Code and the FSMA rules both require that the flow diagram be verified on site. A plan whose diagram omits the blast chiller sitting in the corner has not been reanalyzed since the chiller was installed.
Check also:
- Product description — is the finished product, its packaging, its intended shelf life, and its intended use current?
- Intended consumer — has the establishment begun serving a highly susceptible population since the plan was written?
- Date of last reanalysis — FSMA requires reanalysis at least every three years and whenever a significant change occurs or a new hazard emerges.
Step 2: Is the Hazard Analysis Complete?
Evaluate whether every reasonably foreseeable hazard was considered at every step. Under FSMA's HARPC framework, that means biological, chemical (including radiological and food allergen), and physical hazards, plus hazards that may be intentionally introduced for economic gain.
Red flags:
- Allergens absent from the analysis entirely
- Only pathogens considered, with no chemical or physical hazards
- Every hazard rated "not reasonably likely to occur" with no justification
- No consideration of hazards introduced by the facility environment, such as Listeria harborage
Step 3: Are the CCPs Correct?
A critical control point is a step at which control can be applied and is essential to prevent, eliminate, or reduce a hazard to an acceptable level. Two symmetric failures:
- Too many CCPs. Designating every step a CCP dilutes attention and buries the genuinely critical steps in paperwork. Steps controlled by prerequisite programs — general sanitation, pest control, employee hygiene — are usually not CCPs.
- Too few CCPs. A cooling step with no CCP in a Process 3 operation is a serious gap.
Apply the decision tree logic: is there a control measure at this step? Is the step specifically designed to eliminate or reduce the hazard? Could contamination occur at or increase beyond acceptable levels here? Will a subsequent step eliminate it?
Step 4: Are the Critical Limits Defensible?
A critical limit must be measurable, specific, and scientifically supported.
| Weak critical limit | Why it fails | Defensible version |
|---|---|---|
| "Cook thoroughly" | Not measurable | "Internal temperature 165°F, instantaneous, verified in the thickest part" |
| "Cool quickly" | Not measurable | "135°F to 70°F within 2 hours; 135°F to 41°F within 6 hours total" |
| "Sanitizer at proper strength" | Not specific | "Quat at 200 ppm, verified by test strip at start of shift and every 4 hours" |
| "pH low enough" | Not specific | "Equilibrium pH ≤ 4.6, measured on a macerated composite with a two-point calibrated meter" |
Ask for the validation. Where did the number come from? Acceptable answers: the Food Code, a USDA or FDA performance standard, a published challenge study, a process authority letter, or a validated predictive model. Unacceptable answer: "that's what we've always used." Under FSMA, process preventive controls must be validated, and the validation must be performed within 90 calendar days after production of the applicable food first begins.
Step 5: Do the Records Show Execution?
| Record | What to check |
|---|---|
| CCP monitoring logs | Values within critical limits; frequency matches the plan; the monitor is identified; gaps explained |
| Corrective action records | Every out-of-limit monitoring value has a matching corrective action record showing product disposition, cause, and preventive measure |
| Verification records | Thermometer and pH meter calibration; supervisory record review with signature and date; internal audits; any product or environmental testing |
| Employee health | Signed reporting agreements on file for all food employees |
| Date marking | Ready-to-eat TCS held over 24 hours marked with a discard date ≤ 7 days at 41°F, day of prep = day 1 |
| Receiving | Temperature logs, invoices, shellstock tags retained 90 days |
| Supplier documentation | Approved supplier list, specifications, certificates of analysis, FSVP records for imports |
| Training | Rosters, competency verification, current certificates |
| Pest control | Service reports, device maps, trend logs, applicator license, pesticide labels |
| Cleaning | Master cleaning schedule with sign-off, sanitizer concentration checks |
Under USDA-FSIS HACCP (9 CFR 417.5(c)), records must be reviewed, dated, and signed by a trained individual prior to shipping the product. FDA's seafood and juice HACCP rules use a different clock — review within one week of the day the records are made. Do not merge the two.
Recognizing Falsified Records
Record falsification is common enough that recognizing it is a professional skill.
| Indicator | What it suggests |
|---|---|
| Identical handwriting across a whole month on a log that should have three shifts of monitors | One person filled it in at once |
| Uniform ink and pressure across dates | Same sitting |
| Implausibly regular values — 38, 38, 38, 38 all month | Values written, not measured |
| No out-of-spec readings ever | Real processes drift; a log with zero deviations across a year is a log nobody read |
| Entries for a day the establishment was closed | Fabrication, and easy to prove |
| Corrections made with correction fluid, erasure, or overwriting | Record integrity violation regardless of intent |
| Signature dates preceding the activity | Pre-signed forms |
| Log values that contradict the data logger or thermostat chart | Direct conflict with instrument data |
The legitimate correction method is to draw a single line through the incorrect entry so it remains legible, write the correct value, and initial and date the change. Never white out, never erase, never overwrite.
When falsification is suspected, document precisely what you observed — photograph the record, note the specific dates and values, and record the PIC's statements verbatim rather than characterizing them. Falsification findings can support enforcement action, and the documentation must be able to stand on its own later.
An auditor walks a facility's process and finds a blast chiller in daily use that does not appear anywhere on the HACCP flow diagram. What is the significance?
A facility's cooling log shows 41°F recorded at every check for eleven consecutive months with no deviations of any kind. What should this prompt?
Under USDA-FSIS HACCP regulations at 9 CFR 417.5(c), when must CCP monitoring and corrective action records be reviewed and signed?
A HACCP plan lists the critical limit for a cook step as "cook until done." What is the primary deficiency?