Discussion Topics. Some process engineering considerations

Size: px
Start display at page:

Download "Discussion Topics. Some process engineering considerations"

Transcription

1

2 Discussion Topics Definition of verification Introduction of ICMSF sampling terms Purposes of microbiological testing Maximizing the value of verification data Some process engineering considerations

3 Definition: Verification The application of methods, procedures, tests and other evaluations, in addition to monitoring to determine compliance with the HACCP plan. Codex Alimentarius Commission CAC/RCP -69, Rev. - - Annex

4 Potential Verification Procedures Validation of process effectiveness Calibration of equipment Review of records Targeted sampling and testing Visual inspection of equipment Environmental monitoring nd and rd party audits

5 Product Sampling and Testing Periodic verification may also include targeted sampling and laboratory testing of: Ingredients In-process materials Finished products

6 Microbial Testing Means Different things to different people: Volumes of data to study Detective game to identify unknown or causative agent Product is either good or bad Data as information to evaluate & verify process control & improvement OR

7 The Purpose of a Test Determines: The target Indicator or pathogen The method The sample The frequency The interpretation The action Time to results, accuracy, repeatability, etc. Environment, line residue, end product, location collected, size/ number of samples Daily, weekly, monthly, etc. or event triggered Investigational, routine, regulatory, etc. Rejection, process adjustment, recall, outbreak investigation, etc.

8 When & Where to Test for Food Safety Management ONLY: when there is good evidence that: There is a microbiological problem Food safety or quality Historical or current AND Testing will help to control the problem Micro problem if and Then: Test! Testing will help control

9 Key ICMSF Sampling Plan Terms n c m M Number of sample units to be analyzed Maximum number of sample units with marginal but acceptable results (i.e., between m and M) Concentration separating good quality or safety from marginally acceptable quality Concentration separating marginally acceptable quality from unacceptable quality or safety

10 Relative proportion of sample units in a lot Choosing m and M No concern m -- C -- Some concern M Decisive concern/ defective Sometimes m and M are specified by customers and/or regulations Mean log count

11 ICMSF Sampling Plans ICMSF sampling plans apply when: Information indicates a potential for contamination OR Production conditions and history are not known These are within-lot testing approaches

12 Probability of Acceptance..8.6 Sample Size Influence on Probability of Acceptance % % 86% 7% % (m = ) n = n = n = 6.. % Defective

13 Process Example Process Packaging Line A Ingredients Process Packaging Line B Process What action do you take when an unacceptable result is found on Line B?

14 Log (CFU/g) Result Format Influences Information Provided Presence/Absence Quantitative Positive M Negative Lot Number Lot Number

15 Log CFU/g Process Control Example 6 Loss of control 9% Upper CL Criterion (E.coli counts) 9% Lower CL 6 Time (weeks) From: ICMSF Microorganisms in Food 8, pg 8

16 Control Chart Example Example: Figure X. Log() E. coli counts illustrating an increase in CFU per ml (From AOAC methods, Appendix F. SPC, 6)

17 Control Chart Example Example: Hypothetical example of x-bar and R control charts using a process capability study that measured Total Aerobic Plate Count for data subgroups each with replicates. (Note: On these companion charts the middle solid line represents the target value the flanking lines are the +/- σ lines)

18 Log (CFU/g) Log (CFU/g) Log (CFU/g) Log (CFU/g) Trend Analysis Can Inform Process Control Lot Number Lot Number Lot Number Lot Number

19 Moving Range Log (CFU/g) Individual Value Log (CFU/g) Trend Analysis Can Inform Process Control SPC Charts # Individual Value SPC Charts of point above 7 9 Observation 7 9 UCL=.8 _ X=.97 LCL=.666 UCL=.6 Moving Range..8. MR= Observation 7 9 LCL= Lot Number SPC Charts of high variance 6 UCL=.6 _ X= LCL= Observation 7 9 UCL=.97 Lot Number 7 9 Observation 7 9 MR=.7 LCL=

20 Moving Range Log (CFU/g) Individual Value Moving Range Log (CFU/g) Individual Value Trend Analysis Can Inform Process Control SPC Charts # SPC Charts of Rising Trend.... UCL=.8 _ X=. LCL= Observation Observation 7 9 UCL=. MR=.667 LCL= Lot Number.6 SPC Charts of high log count trend UCL=.9..8 _ X= LCL=.9.6 Observation..8 UCL= MR=.7 LCL= Lot Number Observation

21 Sample Range Sample Mean STATISTICAL PROCESS CONTROL Charts Xbar-R Chart of Temp 6 Sample 6 Sample UCL=8.8 _ X=.6 LCL=.8 UCL=.88 _ R=.8 LCL= Verification can be supported by SPC charting of other (nonmicrobial) system parameters, e.g.: Temperature (cooking) Moisture (e.g., A w ) Monitoring these and examining trends can helpful in establishing process control and be investigative

22 Verification Summary Testing is recommended to generate meaningful data Validation of controls for quality or safety Verification of controls to direct corrective action Process verification data are most often more informative than end-product testing Statistical process control techniques can be used with quantitative data to continually monitor food systems and aids in investigations of overall processing systems level issues determining root causes Expertise is essential for reliable decisions

23 Questions?