ISO9001 System Document Package (45) | Work Instruction for Implementation and Verification of Corrective Actions
Document Description: This document is a third-level document (work instruction) that serves ISO 9001:2015 Clause 10.2 "Nonconformity and Corrective Action." It provides detailed operational guidelines for the second-level procedure document "Nonconformity and Corrective Action Procedure." The procedure document answers "who initiates and approves corrective actions under what circumstances," while this work instruction answers "how to handle each step of a corrective action from acceptance to closure, what level of detail is required to be considered qualified, and what data is needed to prove effectiveness." It is applicable to manufacturing, service, and engineering companies that have established a quality management system (QMS), especially for organizations where corrective actions are often "only corrections without measures," "measures are written but not verified," or "the same issues recur." Before using this document, please confirm that your company has issued the "Nonconformity and Corrective Action Procedure" and clearly defined the department responsible for corrective action form numbering.
1. Purpose
To standardize the entire process of corrective actions from problem acceptance, cause analysis, measure formulation, implementation tracking to effectiveness verification and closure, ensuring:
a) For nonconformities that have occurred (including complaints, internal audit nonconformities, process anomalies, supplier incoming quality issues, etc.), in addition to addressing the nonconformities themselves, the causes of their occurrence are eliminated to prevent recurrence;
b) The effectiveness of corrective actions is determined by objective data or physical evidence, not by subjective statements such as "training has been conducted," "emphasis has been placed," or "a meeting has been held";
c) The records of the entire corrective action process are complete and traceable, meeting the evidence requirements for internal audits, certification audits, and customer audits;
d) Effective practices confirmed in corrective actions are solidified into documents, control plans, and work standards, and horizontally deployed to similar products, processes, production lines, and suppliers, achieving an upgrade from "solving problems" to "eliminating the root causes of problems."
2. Scope of Application
2.1 Applicable to the management of corrective actions for all nonconformities and potential nonconformities within the scope of the company's QMS, including:
a) Nonconformities issued by internal audits, external audits (certification audits, customer audits, second-party audits);
b) Customer complaints, returns, claims, major customer complaint warnings;
c) Nonconforming batches or repetitive defects found during process inspections, finished product inspections, and first article inspections;
d) Process anomalies such as insufficient process capability, control chart anomalies, and SPC data exceeding limits;
e) Nonconforming incoming materials from suppliers and declining supplier quality performance;
f) Problems identified from management review outputs, data analysis, and declining customer satisfaction;
g) Risk items identified after changes in regulations, standards, major design, or process changes;
h) Quality issue leads from employee proposals and safety/environmental incidents.
2.2 Not applicable to: Nonconformities that can be addressed through on-site rework, repair, sorting, or concession acceptance, and are confirmed to be isolated and not systemic after review (such nonconformities should be handled according to the "Nonconforming Product Control Procedure," but the reasons and approval conclusions for "not initiating corrective actions" must be recorded in the "Nonconforming Product Disposition Form").
2.3 For issues that will not recur in the current instance but may recur in the future, handle them according to the preventive action clause in the "Nonconformity and Corrective Action Procedure." The methods in this work instruction (cause analysis, error-proofing measures, effectiveness verification) are equally applicable.
3. Responsibilities
3.1 Management Representative: Approve corrective actions for A-level (major) issues; adjudicate cross-departmental disputes, resource allocation, and re-initiation of measures if they are ineffective; ensure that the results of corrective actions are used as inputs for management review.
3.2 Quality Department:
a) Uniformly accept, number, register, maintain the ledger, and remind of deadlines for corrective action forms;
b) Organize or guide cause analysis and determine whether the root cause conclusion is acceptable;
c) Review the technical rationality and error-proofing level of measures, and the sufficiency of the effectiveness verification plan;
d) Track the progress of measures, initiate upgrades for overdue items, and organize horizontal deployment and document solidification.
3.3 Responsible Department (the department where the issue occurred): Responsible for completing cause analysis, formulating and implementing corrective actions, and providing implementation evidence within the specified time limit; the department head is responsible for the authenticity of the measures.
3.4 Verifier: Appointed by the Quality Department and has no direct responsibility for measure formulation or implementation (A-level issues are assigned by the Quality Department supervisor or management representative); responsible for collecting data during the specified observation period after measures are completed, independently judging effectiveness, and signing the verification conclusion on the corrective action form.
3.5 Internal Auditor: Track and verify corrective actions for internal audit nonconformities, following the verification methods in the "Internal Audit Implementation Work Instruction."
3.6 Document Administrator: Process changes and issue new versions of related documents, inspection standards, and control plans based on the conclusions of corrective actions to prevent the continued use of old versions on-site.
3.7 Department Heads: Implement horizontal deployment requirements within their departments and ensure that department personnel understand the measure requirements related to their positions.
4. Work Procedures
4.1 Clarifying the Boundaries of Three Concepts (Clarify Before Acting)
Distinguishing concepts is the first step in ensuring the quality of corrective actions. In practice, many "ineffective measures" stem from confusing corrections with corrective actions.
| Concept | Definition | Typical Expression | Need for Effectiveness Verification |
|---|---|---|---|
| Correction (Disposal) | Measures taken to eliminate or downgrade the identified nonconformity itself | Rework, repair, sorting, scrap, concession acceptance, re-inspection | No, the verification object is "the disposal result meets requirements" |
| Corrective Action | Measures taken to eliminate the cause of nonconformity and prevent recurrence | Adding error-proofing devices, revising work standards and training, changing parameters and verifying | Yes, must be judged by data or physical evidence |
| Preventive Action | Measures taken to eliminate the cause of potential nonconformity and prevent occurrence | Control method design after risk identification, preventive deployment for similar products | Yes, judged by risk assessment conclusions |
When filling out the corrective action form, the "Temporary Disposal" section (Section 4) and the "Corrective Action" section (Section 6) must be filled out separately and not combined. If both types of content are mixed on a single form, the Quality Department should return it for re-filling.
4.2 Problem Sources and Acceptance
4.2.1 Problem sources are identified according to items a) to h) in Section 2.1. Any problem, regardless of its source, must be reported by the discovering party within 4 hours (during the day shift) or before the end of the shift, by filling out the "Problem Description" section of the "Corrective Action Request Form" and submitting it to the Quality Department.
4.2.2 The Quality Department completes the following within 1 working day after acceptance:
a) Number and register the "Corrective Action Ledger";
b) Preliminary grading (see Section 4.3);
c) Assign the responsible department, responsible person, verifier, and deadlines for each node.
4.2.3 Numbering rule: CAR-YYYY-NNN (CAR is the code for the Corrective Action Request Form, YYYY is the year, NNN is a three-digit serial number, reset each year). Internal audit nonconformities are marked as CAR-YYYY-NNN (Internal Audit - Serial Number) for easy statistical tracking.
4.2.4 Multiple nonconformities caused by the same reason should be handled on a single corrective action form, but each affected item must be listed in the "Affected Scope" table. They should not be split into multiple forms to avoid deadline and statistical issues.
4.3 Problem Description and Grading
4.3.1 The problem description must answer the five elements (a simplified version of 5W2H), and any missing element will result in the form being returned:
a) What nonconformity (defect phenomenon, technical indicator deviation);
b) Where discovered (process/customer/supplier/audit occasion);
c) When (occurrence date, discovery date, batch or time period);
d) How much (quantity, batch, defect rate or proportion);
e) Impact scope (whether it involves work-in-progress, inventory, in-transit, or delivered products).
Using vague descriptions such as "some products have issues," "quality is unstable," or "improper operation" is prohibited.
4.3.2 Grading criteria:
| Level | Judgment Conditions (meet any one to be classified as such) | Acceptance → Root Cause | Root Cause → Measure Implementation Completion | Observation Verification Period | Approver |
|---|---|---|---|---|---|
| A-level (Major) | Customer bulk returns or claims; customer line stoppage; involves safety, regulations, environmental protection, personal injury; systemic failure (same cause in multiple departments); severe nonconformity in certification audit | Within 3 working days | Within 15 working days | Continuous 3 batches or 1 month, whichever is stricter | Management Representative |
| B-level (General) | Customer complaint but no bulk returns; process defect rate exceeds control limits; consecutive two batches of similar defects; general nonconformity in internal audit | Within 5 working days | Within 20 working days | Continuous 3 batches or 2 weeks, whichever is stricter | Quality Department Supervisor |
| C-level (Minor) | Isolated defects but with a trend of repetition; non-standard practices found on-site; single batch of nonconforming incoming materials from suppliers | Within 7 working days | Within 30 working days | Continuous 2 batches or 1 week, whichever is stricter | Quality Department Supervisor or Authorized Engineer |
Note: Grading is proposed by the Quality Department, and A-level issues must be confirmed by the Management Representative. In case of grading disputes, the higher level is applied; deadlines are counted in natural "working days." If external supplier cooperation is required, an extension can be requested, but only once and not exceeding 50% of the original deadline.
4.4 Emergency Containment Measures (Temporary Measures)
4.4.1 Measures can be initiated before the cause is identified, with the aim of preventing the spread and escalation of defects. The deadline is within 24 hours after acceptance (8 hours for A-level issues).
4.4.2 Containment must cover five directions, each confirmed with a tick, and any missing direction is considered incomplete:
a) Work-in-progress on-site: Isolation, labeling, and stopping flow;
b) Inventory: Full inventory count in the warehouse and line-side storage, batch freezing;
c) In-transit: Notify logistics to intercept or hold shipments;
d) Delivered to customers: Assess risk and decide whether to notify customers, conduct on-site sorting, or replace products;
e) Supplier end: If the issue is with incoming materials, notify the supplier to simultaneously pause shipments and conduct a self-check.
4.4.3 Containment measures must specify failure conditions, such as "This measure is effective until the new work instruction is issued and training is completed," to prevent temporary measures from being used long-term without transitioning to permanent measures.
4.4.4 The labor hours and material losses incurred during containment, such as sorting and rework, should be recorded in the "Impact and Loss" section of the corrective action form, serving as cost input for management review.
4.5 Root Cause Analysis
4.5.1 Conclusions must be provided for the three levels of cause analysis, and only one level is considered incomplete:
a) Occurrence Cause: Why did this nonconformity occur (technical, equipment, material, method level);
b) Flow Cause: Why was it not detected internally but flowed to the next process or customer (inspection method, sampling plan, judgment criteria, inspection capability level);
c) System Cause: Why did this cause persist for a long time without being detected by the management system (planning, responsibilities, monitoring, training, documentation level).
4.5.2 A root cause is valid only if it meets all four conditions, and any missing condition is unacceptable:
a) It can explain all phenomena, including differences in quantity, timing, and batch distribution;
b) It is verifiable: supported by on-site reproduction tests, data regression, or comparison tests;
c) After eliminating this cause, similar nonconformities should not reoccur (provable during the subsequent verification period);
d) It points to a level where measures can be taken: if the conclusion is "employee negligence," "lack of awareness," "weak sense of responsibility," or "insufficient training," it is not a root cause and further questioning is required to determine "why negligence is not prevented."
4.5.3 Common tools and recommendations for use:
| Tool | Applicable Scenario | Key Requirements |
|---|---|---|
| 5Why Layered Questioning | Clear single cause-and-effect chain technical issues | At least question to the management or system level, do not stop at "operator error" |
| Fishbone Diagram (6M: Man, Machine, Material, Method, Environment, Measurement) | Unclear cause direction, complex issues involving multiple factors | Each end cause must be labeled "verified/pending verification" and the verification method |
| Cause-Effect Matrix (Scoring Method) | Multiple candidate causes requiring ranking and selection | Scoring criteria agreed in advance, at least 3 scorers |
| Fault Tree Analysis (FTA) | Issues with severe consequences involving safety or functional failure | Decompose from the top event, quantitatively provide occurrence probability or likelihood level |
| Data Stratification and Comparison | Repeated issues with historical data | Stratify by machine, shift, batch, supplier, mold, etc. |
| Reproduction Test | Unclear technical mechanism | Consistency between test conditions and on-site conditions must be recorded |
4.5.4 Records of cause analysis must be accompanied by evidence: test data, measurement records, on-site photos, comparison samples, stratified data results, etc. Evidence is provided by the responsible department and verified for authenticity by the Quality Department.
4.5.5 Conclusions of root cause analysis for A-level issues must be reviewed by a cross-departmental team (Quality, Technology, Production, and suppliers if necessary), and both the review conclusions and differing opinions must be recorded.
4.6 Formulation of Corrective Actions
4.6.1 Measures must be proposed separately for the root causes confirmed in Section 4.5 (including occurrence, flow, and system levels), forming a complete countermeasure table. The countermeasure table should include: sequence number, corresponding root cause, measure content, measure type, responsible person, completion date, required resources, and expected results.
4.6.2 Priority order for error-proofing measures (from strongest to weakest, prefer the first option):
| Strength | Measure Type | Explanation and Examples |
|---|---|---|
| 1 | Elimination (Inherently Safe/Fail-Safe Design) | Cancel error-prone steps or components in design; parameters locked by the system and not modifiable on-site |
| 2 | Substitution | Use materials, processes, or equipment less prone to errors; use scanning instead of manual verification |
| 3 | Engineering Control (Error-Proofing Devices) | Positioning pins, foolproof fixtures, sensor interlocks, weight checks to prevent missing parts, color differentiation, unique sockets |
| 4 | Management Control | Dual verification, first article inspection, increased inspection frequency, release sign-off |
| 5 | Training and Communication | Training on work instructions, warning signs, pre-shift emphasis |
In principle, the fifth option should not be used alone as a corrective action. If training is the only feasible measure, it must be accompanied by "how to prove effectiveness with data after training," such as subsequent sampling pass rates, distribution of assessment scores, and inspection results of consecutive batches. Otherwise, the form will be returned for re-formulation.
4.6.3 Measure descriptions must be specific and verifiable, including the object, action, and standard. Example comparison:
| Nonconforming Description (Returned) | Conforming Description (Accepted) |
|---|---|
| Strengthen employee quality awareness | Install a weight sensor for the instruction manual at the packaging station, which alarms and stops the line if missing; installation and verification to be completed by the Equipment Department by March 20 |
| Strengthen incoming inspection | Add hardness testing for the end cap seal ring (frequency: 5 pieces per batch), inspection procedure to be revised and issued by March 15 |
| Improve work standards | Revise the "Packaging Work Instruction," adding a shift change inventory count clause, document number WI-PK-007 to be upgraded to Version B, training and examination to be completed by March 25 |
4.6.4 Measures involving product design, process parameters, inspection standards, or supplier changes must follow the "Change Management Procedure" to assess new risks before completion, preventing "solving one problem and creating another."
4.6.5 Measures should not rely solely on "increasing final inspection frequency" to replace source control; if objective constraints require dependence on inspection, the sampling plan, judgment criteria, and duration must be specified in the measures and re-evaluated in the next year's risk assessment to determine if they should be upgraded to engineering controls.
4.7 Implementation and Tracking
4.7.1 The responsible department implements measures according to the completion dates in the countermeasure table, recording the completion status on the corrective action form for each completed item, and attaching evidence (photos, drawings, revised documents, training records, equipment acceptance records).
4.7.2 The Quality Department updates the "Corrective Action Ledger" weekly, reminding responsible departments 3 working days before the deadline for items nearing completion, and for overdue items:
a) Overdue within 1 week: Issue a written reminder, the responsible department head signs a new completion date in the ledger;
b) Overdue for more than 1 week or expected to be unable to complete on time: Escalate to the Management Representative, who coordinates resources or adjusts the plan, and records the escalation handling opinion in the ledger;
c) If a department has 3 or more overdue items in the same period: Include it in the department's management review input as a basis for performance evaluation.
4.7.3 If errors in the root cause analysis conclusion are discovered during implementation, implementation should be halted and re-analysis conducted. Do not "fix and see" to avoid on-site chaos. The re-analysis conclusion and reasons should be recorded together, retaining the original analysis records for traceability.
4.8 Effectiveness Verification
4.8.1 Verification timing: After all measures are completed and self-inspected by the responsible department, enter the observation verification period (see Table in Section 4.3.2), and the verification is conducted by the verifier at the end of the observation period.
4.8.2 Verification methods (select based on issue type, use at least two methods, and at least one must be objective data):
| Issue Type | Primary Verification Method | Secondary Verification Method |
|---|---|---|
| Inspection/Assembly Omissions | No similar nonconformities during the observation period, and inspection records are batch-verifiable | Comparison of defect rates before and after measures; sampling to verify inspection accuracy |
| Dimension/Performance Nonconformities | Comparison of measurement data capability or mean/range after measures | Comparison of process capability indices; control chart anomalies |
| Process Parameter Issues | Parameter monitoring records show control, and key characteristics of finished products are qualified | Comparison test results; equipment inspection records |
| Supplier Incoming Material Issues | Supplier's three consecutive batches of incoming materials are qualified | Improvement reports and self-check data provided by the supplier; on-site audit conclusions |
| Customer Complaint Issues | No similar customer complaints during the observation period; quality data of similar products shipped is normal | Customer follow-up or satisfaction feedback; no interception records |
| Document/System Issues | Revised documents are issued, old versions are recovered, and relevant personnel are qualified | Document issuance and recovery records; on-site spot checks for execution consistency |
4.8.3 Verification sample size requirements: For batch-produced products, at least three consecutive batches (or at least 30 pieces if the batch size is less than 30) during the observation period; for single-piece or project-type products, at least two consecutive similar projects or one month of operation data. If the sample size is insufficient, the observation period is automatically extended until the requirement is met.
4.8.4 Independence of the verifier: The verifier must not be the measure formulator, implementer, or their direct superior; the verification conclusion for A-level issues must be reviewed and signed by a person at or above the Quality Department supervisor level.
4.8.5 Verification conclusions have only three options, and "basically effective" or "generally improved" are not acceptable:
a) Effective: All verification indicators meet the criteria, no similar nonconformities recur, sign off and close, proceed to Section 4.9 for standardization;
b) Partially Effective: Primary indicators meet the criteria but residual risks or individual batch nonconformities exist, additional measures and extended observation period, re-verify;
c) Ineffective: Similar nonconformities recur during the observation period, or data does not show improvement. Ineffective handling: do not close, return to Section 4.5 for re-analysis of root causes, re-initiate, and use this record as input for the new issue; also assess whether to upgrade the issue level.
4.8.6 It is strictly prohibited to close a corrective action form based on "measures have been completed." The only basis for closure is a verification conclusion of "effective."
4.9 Closure and Standardization
4.9.1 Within 5 working days after verification of effectiveness, complete the standardization actions, confirming each item:
a) Document Solidification: Incorporate effective measures into corresponding work instructions, inspection standards, control plans, FMEA, and equipment procedures; for document changes, follow the "Document Control Procedure," recover all old versions, and retain only the effective version on-site;
b) Horizontal Deployment: Identify whether similar products, equipment, processes, shifts, suppliers, or other sites have the same risk, assess and implement each item; if "no deployment needed," provide a reason;
c) Knowledge Accumulation: Organize the issue phenomenon, root cause, measures, and verification data into a case, enter it into the knowledge base or include it in annual training materials for similar positions to learn from;
d) Personnel Confirmation: For changes in work methods, complete training and assessment for relevant positions, and archive the assessment results;
e) Ledger Closure: The Quality Department records the closure date, verification conclusion, and standardization implementation status in the "Corrective Action Ledger."
4.9.2 After closure, include in management review input: Statistically summarize the number of cases, on-time completion rate, effective closure rate, and recurrence rate (the proportion of the same root cause re-initiated). The recurrence rate is the most sensitive indicator of system effectiveness. If a certain type of issue is re-initiated within the same year, it must be explained in a special topic during management review.
4.9.3 Archiving: The corrective action form and all attachments are archived by the Quality Department according to the "Record Control Procedure," with a retention period of no less than 3 years (or as required by contracts, regulations, or product life, whichever is stricter).
4.10 Process Flow Diagram (Text Version)
Problem Discovery (Internal Audit/Customer Complaint/Inspection/Process Anomaly/Supplier/Management Review)
│ Submit within 4 hours
▼
Quality Department Acceptance → Number and Register in Ledger → Preliminary Grading (A/B/C) → Assign Responsible Department and Verifier
│ Within 1 working day
▼
Emergency Containment (Within 24 hours, A-level within 8 hours): Check five directions for work-in-progress/inventory/in-transit/delivered/supplier
│
▼
Cause Analysis: Occurrence Cause + Flow Cause + System Cause (conclusions required for all three levels)
│ Tools: 5Why / Fishbone Diagram / Cause-Effect Matrix / Fault Tree / Reproduction Test
▼
Root Cause Determination (Four conditions: explains all phenomena / verifiable / eliminable / points to a measurable level)
│
▼
Measure Formulation (Countermeasure Table) → Error-Proofing Priority: Elimination > Substitution > Engineering Control > Management Control > Training
│ A-level requires cross-departmental review
▼
Measure Implementation and Tracking (Weekly ledger updates, overdue reminders → Escalate to Management Representative)
│
▼
Effectiveness Verification (Observation period: 3 batches / 1 month / 2 weeks, depending on level; independent verifier)
│
┌────┴────┬─────────────┐
▼ ▼ ▼
Effective Partially Effective Ineffective
│ │ │
Close and Standardize Add Measures and Extend Return for Re-Analysis
Document Solidification/Horizontal Deployment/Knowledge Accumulation Re-initiate and Assess Upgrade
│
▼
Include in Management Review Input (On-time Rate / Effective Closure Rate / Recurrence Rate)
4.11 Summary of Time Limits
| Stage | A-level | B-level | C-level |
|---|---|---|---|
| Containment Measure Initiation | Within 8 hours | Within 24 hours | Within 24 hours |
| Acceptance and Registration | Within 1 working day | Within 1 working day | Within 1 working day |
| Root Cause Analysis Completion | Within 3 working days | Within 5 working days | Within 7 working days |
| Measure Implementation Completion | Within 15 working days | Within 20 working days | Within 30 working days |
| Observation Verification Period | Continuous 3 batches or 1 month | Continuous 3 batches or 2 weeks | Continuous 2 batches or 1 week |
| Closure and Standardization | Within 5 working days after verification | Within 5 working days after verification | Within 5 working days after verification |
4.12 Forms and Records Referenced
The forms and records involved in this work instruction are listed in Chapter 5. The fields, filling instructions, and examples for each form are detailed in Chapter 5 and can be directly copied as templates for the company.
5. Related Records
5.1 Corrective Action Request Form (CAR-01)
One form per issue, follows the measures, and is the sole main record for audit evidence. Field list as follows:
| Column | Field | Filling Instructions | Example |
|---|---|---|---|
| 1 | Form Number | CAR-Year-Serial Number | CAR-2026-018 |
| 2 | Source | Customer Complaint/Return/Internal Audit/External Audit/Process Anomaly/Supplier/Management Review/Proposal | Customer Complaint |
| 3 | Discovery Date / Occurrence Date | Both must be filled out to determine the latency period | Discovered 2026-03-08 / Occurred 2026-03-01 |
| 4 | Problem Description | Five elements complete: phenomenon, location, time, quantity, impact scope | Missing instruction manual found in the packaging line of a certain model, 4 out of 6 batches shipped from March 1 to 7, 2026, had 12 units missing the manual, affecting 2 delivered customers |
| 5 | Temporary Disposal | Containment measures, five-direction confirmation, failure conditions | Freeze 320 units in inventory for full inspection; notify customers for on-site sorting; measures effective until the new work instruction is issued and training is completed |
| 6 | Problem Grading | A/B/C, with judgment basis | B-level: Customer complaint without bulk returns |
| 7 | Cause Analysis | Occurrence Cause / Flow Cause / System Cause, each with evidence number | Occurrence: Not counting according to the list during shift change; Flow: No random checks of instruction manuals in finished product inspection; System: No shift change clause in the work instruction and no error-proofing device. Evidence: Surveillance video, inspection record spot checks |
| 8 | Corrective Action | Sequence number, corresponding root cause, measure content, type, responsible person, completion date | ① Install a weight sensor for the instruction manual at the packaging station and interlock (Engineering Control), Equipment Department, 3-20; ② Revise the work instruction to add a shift change inventory count clause (Management Control), Process Department, 3-18 |
| 9 | Impact and Loss | Labor hours, materials, customer compensation, freight, etc. | Full inspection labor hours 48 person-hours, customer sorting freight 1200 yuan |
| 10 | Implementation Record | Completion status and evidence attachment number for each item | Measures ① and ② completed on 3-19 and 3-18, with attached photos and document issuance records |
| 11 | Effectiveness Verification | Method, sample size, observation period, data, conclusion (Effective/Partially Effective/Ineffective) | Observation period 3 batches (starting from 3-20), 0 occurrences of missing instruction manuals, inspection records verifiable by batch, conclusion: Effective. Verifier: Zhang San, Quality Department |
| 12 | Standardization | Confirmation of document solidification, horizontal deployment, and knowledge accumulation | Revised WI-PK-007 to Version B; 3 similar lines have been equipped; case entered as No. 27 in the knowledge base |
| 13 | Signature | Responsible department, Quality Department, verifier, approver (by level) | Responsible department 3-19 / Quality Department 3-22 / Verifier 3-25 / Approved 3-26 |
Filling requirements: Do not use pencils or erasable pens; modifications must be crossed out and signed with the date; attachments and the main form are archived together; electronic forms must retain modification logs.
5.2 Corrective Action Ledger (CAR-02)
Used for statistics and tracking, fields: form number, source, problem summary, level, responsible department, responsible person, case initiation date, root cause completion date, measure completion date, verification conclusion, closure date, on-time, and recurrence (mark the original form number if the same root cause is re-initiated). The ledger is exported by the Quality Department at the end of each month and used as input for management review and monthly quality meetings.
5.3 Cause Analysis Record Form (CAR-03)
Fields: form number, analysis tool, participants and roles, list of candidate causes, verification method, verification data and conclusion, determined root cause (divided into Occurrence/Flow/System), reasons for not adopting causes, and cross-departmental review opinions (for A-level issues). All candidate causes must have verification conclusions, not just the final adopted one.
5.4 Effectiveness Verification Record Form (CAR-04)
Fields: form number, verifier and independence declaration, verification method, observation period start and end, sample size (number of batches/pieces), baseline data before measures, actual measurement data after measures, judgment criteria, conclusion (Effective/Partially Effective/Ineffective), subsequent handling, attachment list, and signature with date.
6. Related Documents
a) "Nonconformity and Corrective Action Procedure" (second-level document, corresponding to 10.2);
b) "Nonconforming Product Control Procedure" (second-level document, corresponding to 8.7, for the disposal of nonconforming products themselves);
c) "Internal Audit Procedure" (second-level document, corresponding to 9.2) and "Internal Audit Implementation Work Instruction" (third-level document, this series' 42nd document);
d) "Customer Complaint Handling Work Instruction" (third-level document, this series' 41st document, source and interface for customer complaint leads);
e) "Data Analysis and Statistical Tool Application Work Instruction" (third-level document, this series' 44th document, data methods for root cause verification and effectiveness judgment);
f) "Record Filling and Archiving Work Instruction" (third-level document, this series' 39th document) and "Document Preparation and Approval Work Instruction" (third-level document, this series' 38th document);
g) "Change Management Procedure" (second-level document, corresponding to 6.3/8.5.6, synchronous requirements when measures involve changes);
h) "Management Review Procedure" (second-level document, corresponding to 9.3, statistical input for corrective actions);
i) Related record form templates (fourth-level document, this series' 48th document for management record form templates).
Usage Instructions
1. Key Points for Modification Based on Company Actualities
Adjust deadlines according to company rhythm, but do not cancel them. The deadline table in Section 4.11 is a general recommendation. For small single-shift companies, the root cause analysis for A-level issues can be extended to 5 working days, but the deadlines for acceptance, containment, verification, and closure must be clearly defined to prevent corrective actions from hanging indefinitely. Please note the modifications in the "Version/Revision" section of this document.
Re-grade criteria based on industry risk. For industries involving safety, regulations, food and pharmaceuticals, automotive functional safety, it is recommended to extend the observation period for "safety/regulation" issues to continuous 5 batches or 3 months; for pure service companies, the "quantity/batch" criteria can be changed to "repetition frequency or customer impact level."
Do not simplify form fields. The 13 fields in Section 5.1 are the minimum set corresponding to each clause of ISO 9001:2015 Clause 10.2. For small companies, the "Impact and Loss" field can be combined with "Problem Description," but the three levels of cause analysis, measures, and verification cannot be combined.
Adapt to organizational structure. For companies without an independent Technology Department, the responsibility for "engineering control measures" can be assigned to the Equipment/Process position; for service companies without an Equipment position, replace "engineering control" with "system process solidification" (such as mandatory system fields, automatic prompts, secondary review nodes), and the error-proofing priority order remains valid.
Integrate with existing systems. If the company has implemented a QMS/MES system, the CAR form can be made into an electronic process, but it must retain: unique numbering rules, modification logs, attachment retention, independent verifier signatures, and automatic reminders for overdue items. Otherwise, electronic processes may reduce traceability.
2. Audit Focus Points (Typically pursued by auditors)
Whether corrections and corrective actions are recorded separately—Check if the "Temporary Disposal" and "Corrective Action" sections are filled out separately.
Whether root causes are supported by evidence—Auditors often ask, "How do you know this is the root cause?" At this point, records of reproduction tests, data stratification, or comparison tests are needed. Only textual conclusions will be judged as inadequate analysis.
Whether measures are error-proofed—If the measure list is full of "training," "strengthening inspection," and "emphasizing," auditors will generally question why engineering means are not considered.
Whether effectiveness is judged by data—If the verification section only states "improved, no abnormalities" without data comparisons, sample sizes, or observation periods, it is a frequent nonconformity.
Whether closure and standardization are linked—After measures are effective, whether the corresponding work instructions, inspection standards, and control plans are actually revised, old versions are recovered, and on-site execution is consistent with the new versions.
Recurrence rate—Auditors will check the ledger to see if the same root cause is re-initiated. If so, it indicates that the previous measures were ineffective but closed, which is a system effectiveness issue.
Independence of the verifier—The verifier must not be the same person as the responsible person, and the approval level for A-level issues must be implemented.
3. The Most Common Twelve Errors
Treating rework and repair as corrective actions. The form only states "fully inspected and reworked" without cause analysis or measures, equivalent to not implementing 10.2. Countermeasure: Separate "Temporary Disposal" and "Corrective Action" sections, and verify each section upon acceptance by the Quality Department.
Stopping root cause analysis at "operator negligence/awareness deficiency." This is a symptom, not a cause. Countermeasure: Force a three-level analysis—why negligence is not prevented, why there is no error-proofing, why there is no verification? Push the conclusion to the system and control method level.
Analyzing only the occurrence cause, not the flow cause. The issue is fixed, but similar issues may still flow to customers. Countermeasure: All three levels of conclusions are required, and the flow cause must be addressed in the inspection method and judgment criteria.
Writing measures as slogans. "Strengthen management" and "improve responsibility" are not verifiable. Countermeasure: Follow the expression requirements in Section 4.6.3, including the object, action, and standard, otherwise return the form.
One measure covering multiple root causes. It appears efficient but makes it impossible to determine which measure is effective. Countermeasure: Countermeasure table items correspond to each root cause, and observe each item separately during verification.
Closing measures upon completion. Signing off without entering the observation period is equivalent to no verification. Countermeasure: The only basis for closure is a verification conclusion of "effective," and the observation period sample size must meet the criteria.
Lack of baseline data for verification. Without pre-measure data, improvement cannot be assessed. Countermeasure: Establish baseline data (from the last 3 batches or 1 month) at the case initiation stage and include it in the CAR form.
Insufficient sample size for verification. Judging effectiveness based on 2 qualified items. Countermeasure: Follow the sample size requirements in Section 4.8.3, with at least 30 items if the batch size is less than 30.
Verifier is the measure implementer. Self-verification. Countermeasure: The Quality Department appoints an independent verifier, and A-level issues are reviewed by a person at or above the supervisor level.
Documents not revised after measures are effective. On-site personnel still use old versions, and the issue reoccurs after six months. Countermeasure: Make "document solidification + recovery of old versions" a prerequisite for closure, confirmed by the document administrator.
No horizontal deployment. Only one line is changed, while other lines with the same risk remain. Countermeasure: Complete horizontal deployment assessment and retain records before closure, and provide reasons if "no deployment needed."
Scattered archiving of corrective action forms, making traceability difficult. Countermeasure: Archive by form number, with consistent attachment numbers and the main form; retain modification logs for electronic forms; and follow the stricter retention period based on contracts, regulations, or product life.
Measures must be error-proofed, effectiveness must be data-driven, and closure must be independent.
Knowledge code: 2.3.1
Version: v20260809
Author: QTank QTank is dedicated to providing systematic professional knowledge, methodologies, and practical tools to quality management practitioners, helping companies continuously improve their quality capabilities.