Food Protection Trends - November/December 2024 - 552

can be applied to eliminate the most significant hazards or
reduce them to acceptable levels (10, 25, 44). Because of
this significance, the level of control of CCPs is expected
to be higher than that of other types of controls (OPRPs
and PRPs), and predetermined corrective actions should
be in place for " out of control " situations. These corrective
actions can be derived from an RCA. Earlier reviews of the
HACCP guidelines required only identification of the causes
of deviation from CCP critical limits to implement effective
corrective actions (10, 44). The most recent reviews of the
HACCP guidelines now mandate the completion of an RCA
when a deviation from the critical limit occurs (25).
CCPs and FMEA
Food researchers have illustrated the application of RCA
tools at CCPs. Seven of the 10 articles analyzed included a
fishbone diagram associated with each CCP. The preferred
tool was the 4M&1E approach: man, material, machine, and
method plus environment. These 10 studies demonstrated
the practical use of visualizing all possible causes of problem
variables at the CCP (2, 3, 17). An additional improvement
could be listing the measurement category of causes because
monitoring of critical limits typically requires measurements.
However, a fishbone diagram may not be the most appropriate
RCA method to meet the HACCP guidelines because it
does not include the corrective actions required to adjust a
process when it is out of control.
A more detailed analysis of the factors that affect the CCP
could be presented in a subsystem FMEA because it will
include actions to take in case of deviations. Several elements
of FMEA are ideal for managing out-of-control events at
CCPs (Table 4). The first element of an FMEA (function)
can be the best to describe the idea of selecting a CCP:
reducing food safety hazards. Failure modes can be defined
for variations above or below critical limits and the causes
of corrective actions. Regular FMEAs include controls and
recommended actions, but these may not be necessary for
this type of application.
CCPs and special causes of variation
Ann understanding of the concept of process control and the
special causes of variation becomes useful in this component
of the HACCP analysis. Control can be exercised via norms
and graphs; ideally, the norms should be linked to the causes
and have established actions to solve the problem and prevent
their recurrence (34). Early quality specialists established that
a process is controlled when all variation is predictable (51).
They also observed that there are always some unknown causes
of variation, referred to as common or chance causes, whereas
other causes that can be easily identified are called assignable
or special causes (51). identification and management of
special causes of variation are essential because these variations
can have more severe consequences than smaller variations due
to common causes (23).
552 Food Protection Trends November/December
Statistical process control aims to minimize variation by
eliminating causes that can be easily identified. A series of
tests and rules were developed to detect special causes of
variation in manual charts, known as Nelson or Western
Electric rules (45, 76). The association of these rules with
causes depends on the process being analyzed, but generic
causes were identified and associated with various forms
of variation in the early years (76). Variables also may not
occur separately, and there will be interactions that are not
understood, but when their combined effect is stable and can
be predicted statistically, the process can be considered under
control (23).
The HACCP guidelines suggest that identification and
reduction of the special causes of variation and bringing the
process under control are parts of the second, third, and
fourth principles of the HACCP method (25). Most special
causes of variation should have been identified during the
identification and validation of the CCP. Establishment of
these causes through use of an RCA method can facilitate
troubleshooting when deviations occur. As technology
advances, implementation of statistical process control
and management of special causes of variation will become
easier (52).
CAs and RCA
The requirements for identifying corrective actions differ
with the type of problem and control (CCP, OPRP, or
PRP). Corrective actions for the process are also required
for OPRPs and PRPs, but corrective actions for the product
must be evaluated on a case-by-case basis (22, 25). Thus,
completion of an RCA for corrective actions in OPRPs
and PRPs may not have been suggested yet. RCAs for one
significant or various minor corrective actions are considered
necessary (37).
Corrective actions can be predetermined for OPRPs and
PRPs, in a manner similar to that for CCPs. This can be
facilitated by using a structured management system and
RCA, which also makes problem solving easier. The user
must realize that predetermined corrective actions are also
problems that must be eliminated if their recurrence becomes
too frequent.
Typically, the corrective actions identified in an RCA
require a modification of the management system. The new
practices or root causes to control, identified via RCA, can
be more easily determined when the RCA and management
system follow the same structure. Use of a structured
approach such as that of Ishikawa to organize PRPs that
prevent root causes can facilitate the integration of solutions
for root causes.
CONCLUSIONS
At least six types of well-developed RCA methods
were found in the articles reviewed. These methods were
fundamentally different in their approach to uncovering root
Food Protection Trends - November/December 2024

Table of Contents for the Digital Edition of Food Protection Trends - November/December 2024
