FDA QSIT Guideline

11

GUIDE TO INSPECTIONS OF

QUALITY SYSTEMS

GUIDE TO INSPECTIONS OF

QUALITY SYSTEMS

August 1999

FOOD AND DRUGADMINISTRATION

2

3

This document was developed by the Quality System Inspec-tions Reengineering Team Members

Office of Regulatory Affairs

Rob RuffGeorgia Layloff

Denise DionNorm Wong

Center for Devices and Radiological Health

Tim Wells – Team LeaderChris NelsonCory Tylka

Advisors

Chet ReynoldsKim Trautman

Allen Wynn

Designed and Produced by Malaka C. Desroches

Guide to Inspections of

Quality Systems

4

Foreword

This document provides guidance to the FDAfield staff on a new inspectional process thatmay be used to assess a medical devicemanufacturer’s compliance with the QualitySystem Regulation and related regulations. Thenew inspectional process is known as the“Quality System Inspection Technique” or“QSIT”. Field investigators may conduct an ef-ficient and effective comprehensive inspectionusing this guidance material which will helpthem focus on key elements of a firm’s qualitysystem.

Note: This manual is reference material for investi-gators and other FDA personnel. The document doesnot bind FDA and does not confer any rights, privi-leges, benefits or immunities for or on any person(s).

5

Performing Subsystem Inspections. . . . . . . . . . . . . . . 7

Pre-announced Inspections. . . . . . . . . . . . . . . . . . . ..13

Getting Started. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

Management Controls. . . . . . . . . . . . . . . . . . . . . . . .17Inspectional Objectives 18Decision Flow Chart 19Narrative 20

Design Controls. . . . . . . . . . . . . . . . . . . . . . . . . . . .31Inspectional Objectives 32Decision Flow Chart 33Narrative 34

Corrective and Preventive Actions (CAPA). . . . . . . .47Inspectional Objectives 48Decision Flow Chart 49Narrative 50Medical Device Reporting 61

Inspectional Objectives 62Decision Flow Chart 63Narrative 64

Corrections & Removals 67Inspectional Objectives 68Decision Flow Chart 69Narrative 70

Medical Device Tracking 73Inspectional Objectives 74Decision Flow Chart 75Narrative 76

Production and Process Controls (P&PC). . . . . . . . 79Inspectional Objectives 80Decision Flow Chart 81Narrative 82Sterilization Process Controls 91

Inspectional Objectives 92Decision Flow Chart 93Narrative 94

Sampling Plans. . . . . . . . . . . . . . . . . . . . . . . . . . . 103Instructions 104Tables (Backpage) 107

Table of ContentsTable of Contents

6

This reference is intended to be used in conjunction withthe:

r Compliance Program Guidance Manual for Inspection of MedicalDevice Manufacturers (CP 7382.845).

r Investigations Operations Manual (IOM).

r Code of Federal Regulations, Title 21 (21 CFR) Part 820 QualitySystem Regulation; Part 803 Medical Device Reporting; Part 806Medical Device Corrections and Removals; Part 821 Medical DeviceTracking.

r Compliance Policy Guides (CPG) for devices (Sub Chapter 300).

r Guideline on General Principles of Process Validation, FDA, May1987.

Other references include:

r The Federal Food, Drug, and Cosmetic Act; The Safe Medical DevicesAct (SMDA) of 1990 and the Medical Device Amendments of 1992.

r Medical Device Quality Systems Manual: A Small Entity ComplianceGuide.

r The FDA Worldwide Quality System Requirements Guidebook forMedical Devices.

r Other device specific guidance documents prepared by CDRH for themedical device industry.

r FDA Recognized Standards.

These additional guidances are posted to the CDRH InternetWorld Wide Web Home Page at http://www.fda.gov/cdrh.See IOM Chapter 10, References, for additional information.

http://www.fda.gov/cdrh

7

The Guide to Inspections of Quality Systems provides in-structions for conducting medical device quality system/GMP inspections. It is to be used in conjunction with thecompliance program entitled Inspections of Medical DeviceManufacturers (7382.845). The guide was prepared bythe Food and Drug Administration (FDA) Office of Regula-tory Affairs (ORA), and the Center for Devices and Radio-logical Health (CDRH). It provides guidance for inspectingmedical device manufacturers against the Quality SystemRegulation (21 CFR Part 820) and related regulations.

This process for performing subsystem inspections is basedon a “top-down” approach to inspecting. The subsystemapproach is designed to provide you with the key objec-tives that can help determine a firm’s state of compliance.The process was designed to account for the time con-straints placed on field investigators when performing de-vice quality system inspections. If you can focus your ef-fort on key elements of a firm’s quality system, you canefficiently and effectively evaluate that quality system.

Performing Subsystem

Inspections

8

When you begin an inspection by looking at one or moreinstances of quality problems, such as nonconforming de-vice reports, and work your way back through the firm'squality system, you are doing a “bottom-up” inspection.This method has been helpful in zeroing in on specific prob-lems, and evaluating the firm’s actions relating to thoseproblems. However, with the “top-down” approach, weare looking at the firm’s “systems” for addressing qualitybefore we actually look at specific quality problems. In the“top-down” approach, we “touch bottom” in each of thesubsystems by sampling records, rather than working ourway from records review backwards towards procedures.

The “top-down” approach begins each subsystem reviewwith an evaluation of whether the firm has addressed thebasic requirements in that subsystem by defining and docu-menting appropriate procedures. This is followed by ananalysis of whether the firm has implemented the require-ments of that subsystem.

The illustration provided inside the front cover of this bookshows the seven subsystems, along with related satelliteprograms. Based on discussions between the device in-dustry and the agency, we have chosen four major sub-systems that are the basic foundation of a firm’s qualitysystem. Those four major subsystems are ManagementControl; Corrective and Preventive Actions (CAPA) (withsatellites Medical Device Reporting, Corrections and Re-movals, and Medical Device Tracking); Design Controls;and Production and Process Controls (P&PC) (with satel-lite Sterilization Process Controls). We have provided a

9

suggested technique for inspecting each of these four sub-systems. In addition, following the chapter of the relatedsubsystem we have provided suggested techniques forinspecting the satellite programs.

The satellite programs were included in the QSIT Inspec-tion due to their correlation in the inspection process withthe related subsystem. For instance, the CAPA subsystemis the logical “jumping-off” point to begin inspecting forMedical Device Reporting, Corrections and Removals, andMedical Device Tracking programs which relate to a firm’spostmarket activities. In the case of the CAPA subsystem,if you are covering the satellite programs in your inspec-tion, approximately half a day should be added to yoursubsystem inspection timeframe.

Rather than check every aspect of the firm’s quality sys-tem, the subsystem approach focuses you on those ele-ments that are most important in meeting the requirementsof the quality system regulation and which are key qualityindicators. Between 6-15 inspectional objectives are pro-vided for the review of each subsystem. The review in-cludes both a (broad) review of whether the firm has pro-cedures in place, and appears to meet the requirements,and a closer (detailed) review of some records to verifythat the requirements have been implemented in actualproduction, design and daily quality assurance situations.

One similarity between “top-down” and “bottom-up” inspec-tional approaches is record review. Both approaches in-volve review of raw data, or individual records. In the“top-down” approach, however, we are asking you to use

10

a sampling approach to the record review. With the “top-down” approach, you will sample records in many of thesubsystems to verify whether or not the firm is in compli-ance. In other words, you are doing the raw data reviewas you did in the past, but in a more controlled manner.We have provided sampling tables to assist you in deter-mining how many records you need to review, and whatconfidence you can have in the potential prevalence ofthe observed conditions.

One new feature in the “top-down” inspection techniqueis the use of inspectional objectives and flow diagrams toguide you during the inspection. We have provided in-spectional objectives and flow diagrams that are useful ininspecting the four major subsystems. The flow diagramsprovide a quick overview of how the inspection of eachsubsystem should occur.

In addition to the inspectional objectives and flow dia-grams, we have provided a narrative description describ-ing how to perform the inspection of each subsystem. Thenarrative description includes a discussion on how toachieve each inspectional objective and reflects the ques-tions contained within the flow diagrams. You are notbound to follow each and every sentence in the narrative.Rather, you should inspect the subsystem with the narra-tive guidance in mind.

The Quality System Regulation (21 CFR 820.3(k)) de-fines “Establish” as “define, document (in writing or elec-tronically), and implement”. The Quality System Inspec-tion Technique uses the “establish” approach in conduct-

11

ing the inspection. For each subsystem, you will first deter-mine if the firm has defined and documented the require-ments (CAPA, Design, etc.) by looking at procedures andpolicies, and then you will bore down into records, using thesampling tables, where appropriate, looking at raw data todetermine if the firm is meeting their own procedures andpolicies, and if their program for executing the requirementis adequate.

The duration of inspection is related to the depth of the in-spection. Keep in mind that the subsystem approach pro-vides you with the key inspectional objectives that can helpdetermine a firm’s state of compliance. At the same time,the guidance was designed to accomplish a complete re-view of all four subsystems in approximately one week.While the length of your inspections will vary, using key in-spectional objectives will help assure that you look at themost important elements of the firm’s quality system duringthe inspection.

Most device firms are inspected more than once. By prob-ing different subsystems, different devices or different pro-cesses each time, FDA will eventually have covered mostof the firm’s quality system. You are not expected to covereverything in the firm and in the narrative each time. Youare expected to evaluate the firm’s quality system, but alsoto do it in an efficient and focused manner. Thus, you shouldlimit the depth of coverage when necessary to meet thetime frame suggested. As a general rule of thumb, one dayshould be sufficient to cover each subsystem when usingthe “top-down” approach described within this document.In practice, you may find that the inspection of a certain

12

subsystem may take half a day, while another may takeone and a half days. This situation would still reflect anoverall one day per subsystem time frame.

By directing your attention to the major areas in a firm’squality system, you should be better able to determine ifthe firm’s quality system is in control. Using the subsystemapproach, you may find less opportunity to cite minor de-viations from the quality system regulation than in the past.However, you will be citing more serious (systemic) devia-tions from the regulation.

13

PREANNOUNCED

INSPECTIONS

The ORA Medical Device Industry Initiatives program en-compasses preannounced medical device inspections, FDA483 Annotation and Postinspectional Notification.

The instructions for Preannouncement (including the crite-ria to be used in determining when preannouncement isappropriate), FDA 483 Annotation and Post-inspection No-tification were provided in an April 3, 1996, Federal Regis-ter Notice (Volume 61, Number 65). Refer to the Investiga-tions Operations Manual (IOM) for further information.

When contacting the firm for the preannounced QSIT In-spection, the investigator should ask for a copy of the firm’sQuality Policy and high level Quality System Procedures(including Management Review Procedures), QualityManual, Quality Plan or equivalent documents to previewprior to the inspection. The firm is not required to supplythese documents. The investigator should tell the firm thatthe preview of these procedural documents would facili-tate the inspection. The documents would be returned atthe time of the inspection. If you find deficiencies in thesedocuments, you should request copies of the original docu-ments after you initiate the inspection.

14

15

GETTING STARTED

It is essential that the firm establishes and maintains a qualitysystem that is appropriate for the specific medical devicebeing manufactured and meets the requirements of theQuality System Regulation. The Management Represen-tative has the responsibility to ensure that the requirementsof the Quality System Regulation have been effectively es-tablished and maintained. Prior to your review of any sub-system, interview the Management Representative (or des-ignee). The objective of this interview is to obtain an over-all view of the subsystem as well as a feel for management’sknowledge and understanding of the subsystem. An im-portant linkage for this activity is Management Controls(820.20 Management Responsibility).

16

17

ManagementControls

Subsystem

18

Management Controls

Inspectional Objectives

1. Verify that a quality policy, management review andquality audit procedures, quality plan, and quality sys-tem procedures and instructions have been defined anddocumented.

2. Verify that a quality policy and objectives have beenimplemented.

3. Review the firm's established organizational structureto confirm that it includes provisions for responsibili-ties, authorities and necessary resources.

4. Confirm that a management representative has beenappointed. Evaluate the purview of the managementrepresentative.

5. Verify that management reviews, including a review ofthe suitability and effectiveness of the quality system,are being conducted.

6. Verify that quality audits, including re-audits of deficientmatters, of the quality system are being conducted.

At the conclusion of the inspection....

7. Evaluate whether management with executive respon-sibility ensures that an adequate and effective qualitysystem has been established and maintained.

1919

Management Controls

Narrative

1. Verify that a quality policy, management review andquality audit procedures, quality plan, and qualitysystem procedures and instructions have beendefined and documented.

Prior to the start of the inspection, preferably at the timeyou make the preannouncement of the inspection (if pre-announced), you should ask the firm to send you their overall(or top level) quality system policies, objectives, and proce-dures. This should include their management review pro-cedures, quality policy, and quality plan. If not receivedprior to the start of the inspection, you will need to reviewthese documents at the start of your inspection.

Purpose/Importance

The purpose of the management control subsystem is to provide adequateresources for device design, manufacturing, quality assurance, distribution,installation, and servicing activities; assure the quality system is functioningproperly; monitor the quality system; and make necessary adjustments. Aquality system that has been implemented effectively and is monitored toidentify and address problems is more likely to produce devices that functionas intended.

A primary purpose of the inspection is to determine whether managementwith executive responsibility ensures that an adequate and effective qualitysystem has been established (defined, documented and implemented) at thefirm. Because of this, each inspection should begin and end with an evalua-tion of this subsystem.

21

Quality Policy and Objectives

The firm must have a written quality policy. The definition ofquality policy is provided in the Quality System Regulation.It means the overall intentions and directions of an organi-zation with respect to quality. The firm is responsible forestablishing a clear quality policy with achievable objec-tives then translating the objectives into actual methods andprocedures. Management with executive responsibility (i.e.has the authority to establish and make changes to the com-pany quality policy) must assure the policy and objectivesare understood and implemented at all levels of their orga-nization. The policy does not need to be extensive. Per-sonnel are not required to be able to recite the policy butthey should be familiar with it and know where to obtain it.

Management Review and Quality AuditProcedures

Management reviews and quality audits are a foundationof a good quality system. Assure that the manufacturer haswritten procedures for conducting management reviews andquality audits and there are defined intervals for when theyshould occur. The firm’s quality audits should examine thequality system activities to demonstrate that the proceduresare appropriate to achieve quality system objectives, andthe procedures have been implemented. A successfulimplementation of the firm’s procedures should result in thefirm achieving its quality policy and associated objectives.Whether the quality policy and objectives are “good” maybecome evident as the other subsystems are reviewedduring the inspection.

Quality Plans

The firm must have a written quality plan that defines thequality practices, resources and activities relevant to thedevices that are being designed and manufactured at thatfacility. The manufacturer needs to have written proceduresthat describe how they intend to meet their quality require-ments.

For firms that manufacture devices as well as other prod-ucts, there must be a quality plan that is specifically rel-evant to devices. Much of what is required to be part of theplan may be found in the firm's quality system documenta-tion, such as, the Quality Manual, Device Master Record(s),production procedures, etc. Therefore, the plan itself maybe a roadmap of the firm's quality system. The plan in thiscase would need to include reference to applicable qualitysystem documents and how those documents apply to thedevice(s) that is the subject of the plan.

Quality plans may be specific to one device or be genericto all devices manufactured at the firm. Quality plans canalso be specific to processes or overall systems.

Quality System Procedures and Instructions

All manufacturers of medical devices are required to es-tablish and implement a quality system tailored to the de-vice manufactured. Each manufacturer must prepare andimplement all activities, including, but not necessarily lim-ited to the applicable requirements of the Quality SystemRegulation, that are necessary to assure the finished de-vice, the design process, the manufacturing process, andall related activities conform to approved specifications.

The term "quality system" as specified in the Quality Sys-tem Regulation encompasses all activities previously re-ferred to as "quality assurance" which were necessary to

23

assure the finished device meets its predetermined designspecifications. This includes assuring manufacturing pro-cesses are controlled and adequate for their intended use,documentation is controlled and maintained, equipment iscalibrated, inspected, tested, etc. Some manufacturers mayuse the terms "quality control" or "GMP Control" or "qualityassurance" instead of quality system. It doesn't matter whatterm is used as long as the quality system concept is un-derstood and implemented.

Written quality system procedures and instructions are re-quired. Any FDA 483 observation regarding Quality Sys-tem procedures must be specific and point out the controlsthat are missing or believed inadequate.

2. Verify that a quality policy and objectives have beenimplemented.

One way to determine whether personnel are familiar withthe quality policy is to ask employees directly. This shouldnot be done when the employee is engaged in the actualperformance of his/her duties, but could be done when he/she is at break or when he/she has finished a task andbefore he/she begins his/her next task.

You can also look to see how management has made thepolicy available. For example: Is it in their Quality Manualor another part of their written procedures? Is it posted atpoints throughout the building? It doesn't matter how theymade the policy known, only that personnel know that thereis a policy and where they can read the policy for them-selves.

A review of employee training records to show they havebeen trained in the firm’s quality policy and objectives canalso be done. In particular, this should be done for thoseemployees involved in key operations.

24

3. Review the firm's established organizational struc-ture to confirm that it includes provisions for re-sponsibilities, authorities and necessary resources.

The firm's organizational structure must be adequate toensure devices are designed and manufactured in accor-dance with the Quality System Regulation. The organiza-tional structure should ensure the technical, administrative,and human factors functions affecting the quality of a de-vice are controlled. These functions may involve hardware,software, processed materials or services. All such controlshould be towards the reduction, elimination, or ideally, theprevention of quality nonconformities.

To determine what the firm's organizational structure is, startby asking the authority and responsibility questions that arethe start of every FDA inspection. Review the firm's orga-nizational charts.

The firm's procedures should describe the functional areasor people responsible for performing certain tasks governedby their quality system. They should also include provisionsfor resources and designating a management representa-tive.

Determine whether personnel involved in managing, per-forming or assessing work affecting quality have the nec-essary independence and authority to perform those tasks.Organizational freedom or independence does not neces-sarily require a stand-alone group. However, the responsi-bility, authority and independence should be sufficient toattain the firm's stated quality objectives.

Adequate resources must be available for the quality sys-tem to assure the firm's stated quality objectives can beachieved. Resources include money, supplies, personnel,etc. One approach to confirm that adequate resources areavailable is to ask the management representative how re-sources are obtained and allocated.

25

4. Confirm that a management representative has beenappointed. Evaluate the purview of the managementrepresentative.

The firm must appoint a management representative whois responsible for ensuring the quality system is effectivelyestablished and maintained, and who will report on its per-formance to management with executive responsibility forreview. The appointment must be documented.

To determine whether there is in fact a documented man-agement representative, review the firm's organizationalchart(s) or their Quality Manual.

Determine whether the appointed management represen-tative actually has the purported responsibility and author-ity granted to him/her by the firm's procedures or organiza-tional structure. Ways of reaching this determination include:Whether he/she has sign-off authority for changes to docu-ments, processes, or product designs; whether the peopleconducting quality audits report or provide him/her with theirresults; and noting how he/she interacts with corrective andpreventive actions, relative design control issues, com-plaints, MDRs, in-process or finished product failures, etc.In other words, his /her responsibility and authority shouldbe apparent through the review of the other subsystems.

Verify that the management representative is reporting backto the management with executive responsibility on theperformance of the quality system. These reports shouldeither be the subject of the management reviews or at leastprovide the framework for those reviews.

NOTE: The agency’s policy relative to the review of quality audit results isstated in CPG 7151.02 (CPG Manual subchapter 130.300). This policyprohibits FDA access to a firm’s audit results. Under the Quality SystemRegulation, this prohibition extends to reviews of supplier audit reports and

management reviews. However, the procedures and documents that show conformancewith 21 CFR 820.50, Purchasing Controls, and 21 CFR 820.20(3)(c), Management Reviews,and 21 CFR 920.22 Quality Audit, are subject to FDA inspection.

5. Verify that management reviews, including a reviewof the suitability and effectiveness of the qualitysystem, are being conducted.

Management reviews must measure the firm’s quality sys-tem against the Quality System Regulation and the firm’sown stated quality objectives as defined in their qualitypolicy. Management reviews must be documented. Theremust be written procedures for conducting managementreviews. These procedures can be inspected and the firmmust certify in writing, if requested, that the firm has com-plied with this Quality System Regulation requirement.

Review the firm’s management review schedule to confirmmanagement reviews are being conducted with sufficientfrequency. Management reviews should be frequent enoughto keep them informed of ongoing quality issues and prob-lems. During your review of the CAPA subsystem, if youfind that there are quality issues that do not seem to beknown to executive-level management, then the reviewsmay not be occurring with sufficient frequency.

The dates and results of management reviews must bedocumented to show dates conducted and whether man-agement with executive responsibility attended the reviews.It is not permissible as explained above for an FDA Investi-gator to review the firm's actual management review docu-mentation. However, the firm should be able to show youhow the reviews are to be documented. Management re-view procedures or instructions should include a require-ment that the results of the reviews be documented anddated.

27

6. Verify that quality audits, including re-audits of de-ficient matters, of the quality system are being con-ducted.

Review the firm’s quality audit schedules to assure qualityaudits are being conducted with sufficient frequency. It isrecommended that the time between quality audits not ex-ceed a 12-month period. More frequent audits may be rec-ommended if the firm has a serious Quality System Regu-lation problem.

Quality audits should consist of a formal, planned check ofall elements in the quality system. They are NOT productaudits. Quality audits must be conducted using adequatedetailed written procedures by appropriately trained indi-viduals. If conducted properly, a quality audit can detectsystem defects and, through isolation of unsatisfactorytrends and correction of factors that cause defective prod-ucts, prevent the production of unsafe or nonconformingdevices. Without an effective quality audit function the qualitysystem is incomplete and there is no assurance the manu-facturer is consistently in a state-of-control.

Evidence of inadequate auditing may exist without gainingaccess to the written quality audit reports. This evidencemay be obtained by relating the audit program to deficien-cies observed in other subsystems. If significant qualitysystem problems have existed both before and after thefirm's last self-audit, then you should critically review thewritten audit procedures. The audit procedures should covereach quality system, and should be specific enough to en-able the person conducting the audit to perform an adequateaudit. The auditors must be adequately trained. If it is nec-essary and possible to interview an auditor, ask how theaudits are performed; what documents are examined; howlong audits take; etc.

Audits should be conducted by individuals not having di-rect responsibility for matters being audited. One personand other very small firms must generally establish inde-pendence, even if it means hiring outside auditors, becausethe failure to have an independent auditor could result in anineffective audit. If there are significant FDA 483 observa-tions, and independent audits are being performed, but de-ficiencies are apparently not being identified by the auditor,then an FDA 483 should contain an observation indicatinga lack of adequate audits.

Determine whether corrective action by upper managementis being taken. Auditors may be asked if they observed anyof the ongoing Quality System Regulation deficiencies dur-ing their prior audits (ongoing Quality System Regulationdeficiencies may also be identified by reviewing prior FDA483's). If the answer is yes, check the written audit sched-ule, if available, to determine if a follow up audit is sched-uled for the deficient areas. Check the written audit proce-dure for instructions for review of audits by upper manage-ment . For example, do the procedures require quality au-dit results to be included in the management reviews? Verifythat the procedures contain provisions for the re-audit ofdeficient areas if necessary. A failure to implement follow-up corrective actions, including re-audits of deficient mat-ters may be listed as a Quality System Regulation deficiencyon the FDA 483.

NOTE: Re-audits of deficient matters are not always required, but where oneis indicated, it must be conducted. The reaudit report should verify the recom-mended corrective action(s) was implemented and effective.

29

7. Evaluate whether management with executive re-sponsibility ensures that an adequate and effectivequality system has been established and main-tained.

At this point in QSIT, you stop your review of the manage-ment system. You continue your inspection by evaluatingthe other subsystems. While you evaluate the other sub-systems, keep thinking about what you are finding andwhether it indicates that management is appropriately car-rying out responsibilities for providing adequate resourcesand overseeing the quality system to detect problems andaddress them.

From your review of the other subsystems, you have a bet-ter idea on whether the management representative hasthe appropriate authority and responsibility, whether theorganizational structure is adequate, whether the qualityaudits and management reviews are sufficient, whether thequality policy has really been implemented, and whetherthe training being provided is sufficient.

You need to take the time after reviewing the other sub-systems, to evaluate the inspectional findings of the man-agement and other subsystems. You need to determinewhether the management representative and managementwith executive responsibility are ensuring the adequacy andeffectiveness of the quality system and whether that sys-tem has been fully implemented at this firm.

If you found major nonconformances (as defined in the Com-pliance Program, Part V) in your review of the manage-ment or other subsystems that indicate management withexecutive responsibility is not ensuring the establishmentand maintenance of an adequate quality system, you maycite this deficiency on your FDA 483. This cite should not

30

be used routinely, but should be used in those situationswhere major portions of a quality system have not beenestablished and maintained or whenever there is a totallack of a quality system.

When you have made that determination and have com-pleted your FDA 483, or decided no FDA 483 is needed,you may proceed to your final discussion with Management,or the official closeout meeting with the firm.

31

DesignControls

Subsystem

DesignControls

Subsystem

32

Design Controls


1. Select a single design project.

Note: If the project selected involves a device that contains software, consider reviewingthe software's validation while proceeding through the assessment of the firm's designcontrol system.

2. For the design project selected, verify that design control proce-dures that address the requirements of Section 820.30 of the regu-lation have been defined and documented.

3. Review the design plan for the selected project to understand thelayout of the design and development activities including assignedresponsibilities and interfaces.

Note: Evaluate the firm's conduct of risk analysis while proceeding through the assess-ment of the firm's Design Control system.

4. Confirm that design inputs were established.5. Verify that the design outputs that are essential for the proper func-

tioning of the device were identified.6. Confirm that acceptance criteria were established prior to the per-

formance of verification and validation activities.7. Determine if design verification confirmed that design outputs met

the design input requirements.8. Confirm that design validation data show that the approved de-

sign met the predetermined user needs and intended uses.9. Confirm that the completed design validation did not leave any

unresolved discrepancies.10. If the device contains software, confirm that the software was vali-

dated.11. Confirm that risk analysis was performed.12. Determine if design validation was accomplished using initial pro-

duction devices or their equivalents.13. Confirm that changes were controlled including validation or where

appropriate verification.14. Determine if design reviews were conducted.15. Determine if the design was correctly transferred.

3333

34

Design Controls

Narrative

1. Select a single design project.

Note: If the project selected involves a device that con-tains software, consider reviewing the software's vali-dation while proceeding through the assessment of thefirm's design control system.

The design control requirements of Section 820.30 of theregulation apply to the design of Class II and III medicaldevices, and a select group of Class I devices. The regula-tion is very flexible in the area of design controls. The typeof design control system and the precise details of imple-mentation are left for each firm to decide based on the com-plexity and risks associated with their devices.

Purpose/Importance

The purpose of the design control subsystem is to control the designprocess to assure that devices meet user needs, intended uses, andspecified requirements. Attention to design and development planning,identifying design inputs, developing design outputs, verifying that de-sign outputs meet design inputs, validating the design, controlling de-sign changes, reviewing design results, transferring the design to pro-duction, and compiling a design history file help assure that resultingdesigns will meet user needs, intended uses and requirements.

35

If design control requirements are applicable to the opera-tions of the firm, select a design project. Unless the inspec-tion assignment directs the inspection of a particular de-sign project, select a project that provides the best chal-lenge to the firm's design control system. This project willbe used to evaluate the process, the methods, and the pro-cedures that the firm has established to implement the re-quirements for design controls.

Do not inspect a device under design control requirementsto determine whether the design was appropriate or safeand effective. This is precluded under Section 520(f)(1)(A)of the Act. However, if based on information obtained dur-ing an evaluation of the firm's design controls, it appearsthat the device is unsafe or ineffective, then report thosefindings in the EIR.

The requirement for software validation is included in Sec-tion 820.30(g) Design Validation. However, if the projectselected involves a device that contains software, considerreviewing the software's validation while proceeding throughthe assessment of the firm's design control system.

If the firm has not completed a design project, has no ongo-ing or planned design projects, and has not made a designchange, proceed to the narrative discussion under Objec-tive 2 and limit your review of design controls to those in-structions.

36

2. For the design project selected, verify that designcontrol procedures that address the requirementsof Section 820.30 of the regulation have been de-fined and documented.

Firms, including small firms and those who design simpledevices, who are subject to Section 820.30 of the regula-tion, are required to define, and document, either in writingor electronically, procedures which address the require-ments of the regulation. These procedures serve to set thestructure for the firm's design control system.

However, if the firm has not completed any design projects,has no ongoing or planned design projects, and has notmade a design change, it is only required to maintain a de-fined and documented design change procedure.

Review the firm's design control procedures and verify thatthey address the specific requirements of the regulation.As examples, determine if the design input procedures in-clude a mechanism for addressing incomplete, ambiguous,or conflicting requirements; the design output proceduresensure that those design outputs that are essential for theproper functioning of the device are identified; and the de-sign review procedure ensures that each design review in-cludes an individual(s) who does not have direct responsi-bility for the design stage being reviewed.

In order to determine if the firm's design control procedureshave been implemented, use the selected design project toexercise the firm's procedures and accomplish the fol-lowing objectives.

37

3. Review the design plan for the selected project tounderstand the layout of the design and develop-ment activities including assigned responsibilitiesand interfaces.

Note: Evaluate the firm's conduct of risk analysiswhile proceeding through the assessment of thefirm's Design Control system.

The firm's development of concepts and the conduct of fea-sibility studies are not subject to the design control require-ments of the regulation. However, once the firm decidesthat a design will be developed, a design plan must be es-tablished. A firm will determine when it will begin to applydesign controls. However, design controls must be appliedno later than the time the firm approves its first set of in-puts.

Utilize the firm's design plan as a road map for the selecteddesign project. Plans include major design tasks, projectmilestones, or key decision points. It is not necessary forplans to show starting or completion dates for activities cov-ered by the plan. Plans may vary depending on the com-plexity of the project and the degree of risk associated withthe device. Plans may take the form of a simple flow chartfor less complex projects or may be expressed as ProgramEvaluation and Review Technique (PERT) or Gantt chartsfor larger projects. However, plans must define responsi-bility for implementation of the design and development ac-tivities and identify and describe interfaces with differentgroups or activities.

While the requirement for the conduct of risk analysis ap-pears in Section 820.30(g) Design Validation, a firm shouldnot wait until they are performing design validation to beginrisk analysis. Risk analysis should be addressed in thedesign plan and risk should be considered throughout thedesign process. Risk analysis must be completed in de-sign validation.

38

When conducting risk analysis, firms are expected to iden-tify possible hazards associated with the design in bothnormal and fault conditions. The risks associated with thosehazards, including those resulting from user error, shouldthen be calculated in both normal and fault conditions. Ifany risk is deemed unacceptable, it should be reduced toacceptable levels by the appropriate means, for exampleby redesign or warnings. An important part of risk analysisis ensuring that changes made to eliminate or minimizehazards do not introduce new hazards.

Common tools used by firms to conduct risk analyses in-clude Fault Tree Analysis (FTA), and Failure Modes andEffects Analysis (FMEA).

4. Confirm that design inputs were established.

Inputs are the requirements of a device. They must be docu-mented. Review the sources used to develop inputs. De-termine that relevant aspects were covered. Examples ofrelevant aspects include: intended use, performance char-acteristics, risk, biocompatibility, compatibility with the en-vironment of intended use including electromagnetic com-patibility, human factors, voluntary standards, and sterility.

5. Verify that the design outputs that are essential forthe proper functioning of the device were identi-fied.

Design outputs are the work products or deliverables of adesign stage. Examples include, diagrams, drawings, speci-fications and procedures. The outputs from one stage maybecome inputs to the next stage. The total finished designoutput consists of the device, its packaging and labeling,and the device master record. Important linkages to con-sider are Sections 820.80 Receiving, in-process, and fin-ished device acceptance, 820.120 Device labeling, and820.130 Device packaging.

39

Design projects can produce a large volume of records.Not all of the records generated during the project are de-sign outputs and as such do not need to be retained in thedesign history file. Only approved outputs need to be re-tained.

Outputs must be comprehensive enough to characterizethe device design to allow for verification and validation.Also, design outputs which are essential for the proper func-tioning of the device must be identified. Typically a risk analy-sis tool such as FTA or FMEA is used to determine essen-tial outputs. For the selected project, verify that essentialoutputs have been identified. In addition, review the firm'sprocess for determining how the essential outputs wereidentified and determine if it was done in accordance withtheir design output procedures. Important linkages to con-sider are Sections 820.50 Purchasing controls, and 820.100Corrective and preventive action.

6. Confirm that acceptance criteria were establishedprior to the performance of verification and valida-tion activities.

Verification and validation activities should be predictiverather then empiric. Acceptance criteria must be stated upfront. Review the documentation associated with a sampleof verification activities and a sample of validation activi-ties as determined using the Sampling Tables. If possible,select activities that are associated with outputs identifiedas essential to the proper functioning of the device. Con-firm that acceptance criteria were established prior to per-formance of the verification or validation activity.

40

7. Determine if design verification confirmed that de-sign outputs met the design input requirements.

Design verification activities are performed to provide ob-jective evidence that design output meets the design inputrequirements. Verification activities include tests, inspec-tions, analyses, measurements, or demonstrations. Activi-ties should be explicit and thorough in their execution. It isthe firm's responsibility to select and apply appropriate veri-fication techniques. Complex designs can require more anddifferent types of verification activities than simple designs.Any approach selected by the firm, as long as it establishesconformance of the output to the input, is an acceptablemeans of verifying the design with respect to that require-ment.

Review the documentation of the verification activities as-sociated with a sample of inputs and outputs as determinedusing the Sampling Tables. If possible, select activities thatare associated with outputs identified as essential to theproper functioning of the device. Confirm that design out-puts met design input requirements.

8. Confirm that design validation data show that theapproved design met the predetermined user needsand intended uses.

Design validation is performed to provide objective evidencethat device specifications (outputs) conform with user needsand intended use(s). Design validation must be completedbefore commercial distribution of the device.

Design validation involves the performance of clinical evalu-ations and includes testing under actual or simulated useconditions. Clinical evaluations can include clinical investi-gations or clinical trials, but they may only involve otheractivities. These may include evaluations in clinical or non-

41

clinical settings, provision of historical evidence that similardesigns are clinically safe, or a review of scientific litera-ture. Validation activities must address the needs of all rel-evant parties (i.e. patient, health care worker, etc.) and beperformed for each intended use. Validation activities shouldaddress the design outputs of labeling and packaging.These outputs may have human factor implications, andmay adversely affect the device and its use.

If possible, review the evaluations (clinical or other activi-ties) performed to assist in validating the device design.

9. Confirm that the completed design validation didnot leave any unresolved discrepancies.

Design validation may detect discrepancies between thedevice specifications (outputs) and the needs of the user orintended use(s) of the device. All discrepancies must beaddressed and resolved by the firm. This can be accom-plished through a change in design output or a change inuser need or intended use.

10. If the device contains software, confirm that thesoftware was validated.

As previously noted, design validation includes the require-ment for software validation. If the selected device is soft-ware controlled its software must be validated.

11. Confirm that risk analysis was performed.

As previously noted, risk analysis must be completed indesign validation.

42

12. Determine if design validation was accomplishedusing initial production devices or their equivalents.

Initial production units, lots, or batches, or their equivalentsare to be used in design validation. Confirm that such pro-duction devices or their equivalents were used by review-ing the design validation documentation. If production de-vices were not used, the firm must demonstrate equiva-lency to production devices. When the so called "equiva-lent" devices are used in design validation the manufac-turer must document in detail how the device was manu-factured, and how the manufacturing is similar and possi-bly different from initial production. Where there are differ-ences, the manufacturer must justify why design validationresults are valid for production units, lots or batches. Theregulation is flexible and it does allow for the use of equiva-lent devices, but the burden is on the manufacturer to docu-ment that the units were indeed equivalent.

Process validation may be conducted concurrently withdesign validation. Production devices used in design vali-dation may have been manufactured in a production runduring process validation.

13. Confirm that changes were controlled includingvalidation or where appropriate verification.

Change control is not a new requirement. The 1978 GMPregulation Section 820.100(a)(2) required approval ofchanges made to specifications after final design transfer(post-production changes). The Quality System regulationclarified and relocated the requirement into Section820.30(i). It expanded the requirement to include changesmade during the design process (pre-production changes).

The documentation and control of design changes beginwhen the initial design inputs are approved and continues

43

for the life of the product. Examples of the application ofchange control include: changes made to approved inputsor outputs such as to correct design deficiencies identifiedin the verification and validation activities; labeling changes;changes which enhance the device's capabilities or thecapabilities of the process; and changes resulting from cus-tomer complaints.

Product development is inherently an evolutionary process.While change is a healthy and necessary part of productdevelopment, quality can be ensured only if change is con-trolled and documented in the development process, as wellas in the production process.

The degree of design change control is dependent on thesignificance of the change and the risk presented by thedevice. Manufacturers may use their routine post-produc-tion change control procedure for pre-production designchanges. However, most post-production change controlprocedures may be too restrictive and stifle the develop-ment process. Firms may use a separate and less strin-gent change control procedure for pre-production designchanges.

Post-production design changes require the firm to loopback into the design controls of Section 820.30 of the regu-lation. This does not mean that post-production changeshave to go back to the R&D Department for processing.This track is dependent on what the firm specifies in theirchange procedure. It is acceptable for the manufacturingdepartment to process the entire design change and toimplement the controls of Section 820.30.

The design change control section is linked to and is re-dundant with Section 820.70(b) Production and processchanges of the regulation.

44

All design changes must be verified. Design changes mustalso be validated unless the performance of only verifica-tion can be justified and documented by the firm. Where adesign change cannot be verified by subsequent inspec-tion and test, it must be validated. For example, a changein the intended use of the device will require validation.However, if a firm was making a design change in the ma-terial used in the device, then verification through analysismay only be required. The burden is on the firm to justifyand document why verification only is appropriate in lieu ofvalidation.

Review a pre-production and a post-production designchange.

14. Determine if design reviews were conducted.

Formal design reviews are planned and typically conductedat the end of each design stage or phase, or after comple-tion of project milestones. The number of reviews is depen-dent on the complexity of the design. A single review maybe appropriate at the conclusion of the design project for asimple design or a minor change to an existing product.Multiple reviews are typically conducted for projects involv-ing subsystems or complex designs.

Design reviews should provide feedback to designers onexisting or emerging problems, assess the progress of thedesign, and confirm the design is ready to move to the nextphase of development. Reviews should focus on the abilityto produce the design and whether the design meets theinput requirements.

The design review process should account for risk analysisand change control where relevant.

45

Full convened meetings with an agenda, minutes, etc. neednot take place for all design reviews. Meetings may not benecessary for reviews involving simple designs or minorchanges. In these cases desk reviews and sign-offs by thevarious organizational components including an individualnot having direct responsibility for the design stage beingreviewed may be appropriate. However, such reviews muststill be documented and covered by defined and docu-mented procedures.

Review the records of one design review and confirm thatthe review included an individual without direct responsibil-ity for the design stage being reviewed. Also, confirm thatoutstanding action items are being resolved or have beenresolved.

15. Determine if the design was correctly transferred.

The transfer process must be a part of the design plan. It isnot uncommon for the design to be transferred in phases.Production specifications typically consist of written docu-ments such as assembly drawings, inspection and testspecifications, and manufacturing instructions. However,they can also consist of electronic records, training materi-als such as video tapes or pictures, and manufacturing jigsand molds.

Review how the design was transferred into productionspecifications. Review the device master record. Samplethe significant elements of the device master record usingthe Sampling Tables and compare these with the approveddesign outputs. These elements may be chosen based onthe firm's previously identified essential requirements andrisk analysis.

46

47

Corrective andPreventive

Actions(CAPA)

48

Corrective and Preventive Actions(CAPA)


1. Verify that CAPA system procedure(s) that address the requirements ofthe quality system regulation have been defined and documented.

2. Determine if appropriate sources of product and quality problems havebeen identified. Confirm that data from these sources are analyzed toidentify existing product and quality problems that may require correctiveaction.

3. Determine if sources of product and quality information that may showunfavorable trends have been identified. Confirm that data from thesesources are analyzed to identify potential product and quality problemsthat may require preventive action.

4. Challenge the quality data information system. Verify that the data re-ceived by the CAPA system are complete, accurate and timely.

5. Verify that appropriate statistical methods are employed (where neces-sary) to detect recurring quality problems. Determine if results of analy-ses are compared across different data sources to identify and developthe extent of product and quality problems.

6. Determine if failure investigation procedures are followed. Determine ifthe degree to which a quality problem or nonconforming product is inves-tigated is commensurate with the significance and risk of the nonconfor-mity. Determine if failure investigations are conducted to determine rootcause (where possible). Verify that there is control for preventing distribu-tion of nonconforming product.

7. Determine if appropriate actions have been taken for significant productand quality problems identified from data sources.

8. Determine if corrective and preventive actions were effective and verifiedor validated prior to implementation. Confirm that corrective and preven-tive actions do not adversely affect the finished device.

9. Verify that corrective and preventive actions for product and quality prob-lems were implemented and documented.

10. Determine if information regarding nonconforming product and quality prob-lems and corrective and preventive actions has been properly dissemi-nated, including dissemination for management review.

49

50

Corrective and Preventive Actions(CAPA)

Narrative

1. Verify that CAPA system procedure(s) that addressthe requirements of the quality system regulationhave been defined and documented.

Review the firm's corrective and preventive action proce-dure. If necessary, have management provide definitionsand interpretation of words or terms such as “nonconform-ing product”, “quality audit”, “correction”, “prevention”,“timely”, and others. It is important to gain a working knowl-edge of the firm's corrective and preventive action proce-dure before beginning the evaluation of this subsystem.

Purpose/Importance

The purpose of the corrective and preventive action subsystem is tocollect information, analyze information, identify and investigate prod-uct and quality problems, and take appropriate and effective correctiveand/or preventive action to prevent their recurrence. Verifying or vali-dating corrective and preventive actions, communicating corrective andpreventive action activities to responsible people, providing relevant in-formation for management review, and documenting these activities areessential in dealing effectively with product and quality problems, pre-venting their recurrence, and preventing or minimizing device failures.

One of the most important quality system elements is the corrective andpreventive action subsystem.

NOTE: Corrective action taken to address an existing product or quality problemshould include action to:

- Correct the existing product nonconformity or quality problems and;- Prevent the recurrence of the problem.

51

The CAPA procedure should include procedures for howthe firm will meet the requirements for all elements of theCAPA subsystem. All procedures should have been imple-mented.

Once you have gained a knowledge of the firm's correctiveand preventive action procedure, begin with determining ifthe firm has a system for the identification and input of qualitydata into the CAPA subsystem. Such data includes infor-mation regarding product and quality problems (and poten-tial problems) that may require corrective and/or preven-tive action.

2. Determine if appropriate sources of product andquality problems have been identified. Confirm thatdata from these sources are analyzed to identifyexisting product and quality problems that may re-quire corrective action.

The firm should have methods and procedures to input prod-uct or quality problems into the CAPA subsystem. Productand quality problems should be analyzed to identify prod-uct and quality problems that may require correctiveaction.

The firm should routinely analyze quality data regardingproduct and quality problems. This analysis should includedata and information from all acceptance activities, com-plaints, service, and returned product records. Determine

NOTE: In accordance with Agency policy (CPG 7151.02), do not requestrecords regarding the results of internal quality audits, management reviews,third party audits (including ISO audits), or supplier audits. However, you willbe reviewing raw data that is used by the firm when conducting their quality

audits, management reviews, etc. Trending information and results of analyses are generallypart of evaluations under the corrective and preventive action requirements. This informationis utilized in internal audits and management reviews. Information or data utilized in internalaudits and management reviews are considered raw data and should be available for routinereview.

52

if the firm is capturing and analyzing data from acceptanceactivities relating to component, in-process and finisheddevice testing. Information obtained subsequent to distri-bution, which includes complaints, service activities andreturned products, as well as information relating to con-cessions (quality and nonconforming products), qualityrecords, and other sources of quality data should also becaptured and analyzed. Examples of other sources of qual-ity data include quality audits, installation reports, lawsuits,etc.

3. Determine if sources of product and quality infor-mation that may show unfavorable trends have beenidentified. Confirm that data from these sources areanalyzed to identify potential product and qualityproblems that may require preventive action.

Determine if the firm is identifying product and quality prob-lems that may require a preventive action. This can beaccomplished by reviewing historical records such as trend-ing data, corrective actions, acceptance activities (compo-nent history records, process control records, finished de-vice testing, etc.) and other quality system records for un-favorable trends. Review if preventive actions have beentaken regarding unfavorable trends recognized from theanalysis of product and quality information. Product andquality improvements and use of appropriate statistical pro-cess control techniques are evidence of compliance withthe preventive action requirement.

Determine if the firm is capturing and analyzing data re-garding in-conformance product. Examples include cap-turing and analyzing component test results to detect shiftsin test results that may indicate changes in vendor pro-cesses, component design or acceptance procedures. Iden-tification of these indicators may necessitate a vendor in-vestigation as a preventive action. Monitoring in-process

53

and finished device test results may reveal additional indi-cators of potential quality problems. For devices where sta-bility is an issue, test results of reserve samples are con-tinually monitored. These monitoring activities may triggerprocess changes, additional training activities and otherchanges required to maintain the process within its toler-ances and limits.

Determine if the firm is using statistical control techniquesfor process controls where statistical techniques are appli-cable. An example would be “Statistical Process Control”(SPC). SPC is utilized to monitor a process and initiateprocess correction when a process is drifting toward a speci-fication limit. Typically, SPC activities are encountered withlarge volume production processes such as plastic mold-ing and extrusion. Any continuing product improvements(in the absence of identified product problems such as non-conforming product) are also positive indicators of preven-tive actions. Important linkages for this activity include820.70 Production and Process Controls and 820.250 Sta-tistical Techniques.

4. Challenge the quality data information system.Verify that the data received by the CAPA systemare complete, accurate and timely.

Select one or two quality data sources. Using the samplingtables, review records from the chosen data sources todetermine if the data were entered into the CAPA system.In addition, determine whether the data are complete, ac-curate and entered into the CAPA system in a timely man-ner.

Important linkages for this activity include 820.80 Accep-tance Activities, 820.90 Nonconforming Product, 820.170Installation, 820.198 Complaint Files and 820.200 Servic-ing.

54

5. Verify that appropriate statistical methods are em-ployed (where necessary) to detect recurring qual-ity problems. Determine if results of analyses arecompared across different data sources to identifyand develop the extent of product and quality prob-lems.

The analysis of product and quality problems should in-clude appropriate statistical and non-statistical techniques.Statistical techniques include Pareto analysis, spread-sheets, and pie charts. Non-statistical techniques includequality review boards, quality review committees and othermethods.

The analysis of product and quality problems should alsoinclude the comparison of problems and trends across dif-ferent data sources to establish a global, and not an iso-lated view, of a problem. For example, problems noted inservice records should be compared with similar problemtrends noted in complaints and acceptance activity infor-mation.

The full extent of a problem must be captured before theprobability of occurrence, risk analysis and the propercourse of corrective or preventive action can be determined.

6. Determine if failure investigation procedures are fol-lowed. Determine if the degree to which a qualityproblem or nonconforming product is investigatedis commensurate with the significance and risk ofthe nonconformity. Determine if failure investiga-tions are conducted to determine root cause (wherepossible). Verify that there is control for preventingdistribution of nonconforming product.

Review the firm's CAPA procedures for conducting failureinvestigations. Determine if the procedures include provi-

55

sions for identifying the failure modes, determining the sig-nificance of the failure modes (using tools such as risk analy-sis), the rationale for determining if a failure analysis shouldbe conducted as part of the investigation, and the depth ofthe failure analysis.

Discuss with the firm their rationale for determining if a cor-rective or preventive action is necessary for an identifiedtrend regarding product or quality problems. The decisionprocess may be linked to the results of a risk analysis andessential device outputs.

Using the sampling tables, select failure investigationrecords regarding more than one failure mode (if possible)and determine if the firm is following their failure investiga-tion procedures.

Confirm that all of the failure modes from your selectedsample of failure investigations have been captured withindata summaries such as reports, pie charts, spreadsheets,Pareto charts, etc.

Determine whether the depth of the investigation (wherepossible) is sufficient (root cause) to determine the correc-tive action necessary to correct the problem. Select onesignificant failure investigation that resulted in a correctiveaction and determine if the root cause had been identifiedso that verification or validation of the corrective action couldbe accomplished.

Using the sampling tables, review a number of incompletefailure investigations for potential unresolved product non-conformances and potential distribution of nonconformingproduct. Unresolved problems that could be of significantrisk to the patient or user may require product recall if theproblem cannot be resolved.

56

Using the sampling tables, review records regarding non-conforming product where the firm concluded corrective orpreventive action was not necessary. As noted above, verifythat the firm is not continuing to distribute nonconformingproduct. This may be an important deficiency based onthe class of, and the risk associated with, the product. Im-portant linkages for these activities include 820.20 Man-agement Responsibility, 820.25 Training, 820.30 DesignControls, 820.90 Nonconforming Product and possibly820.250 Statistical Techniques.

Using the sampling tables, review nonconforming productand quality concessions. Review controls for preventingdistribution of nonconforming products. Product and qual-ity concessions should be reviewed to verify that the con-cessions have been made appropriate to product risk, withinthe requirements of the quality system and not solely tofulfill marketing needs. Important linkages regarding theseactivities include 820.20 Management Responsibility and820.90 Nonconforming Product.

7. Determine if appropriate actions have been taken forsignificant product and quality problems identifiedfrom data sources.

Where appropriate, this may include recall actions, changesin acceptance activities for components, in-process and fin-ished devices, etc.

Using the sampling tables, select and review significantcorrective actions and determine if the change or changescould have extended beyond the action taken. A signifi-cant action would be a product or process change to cor-rect a reliability problem or to bring the product into con-formance with product specifications. Discuss with the firmtheir rationale for not extending the action to include addi-tional actions such as changes in component supplier, train-

57

ing, changes to acceptance activities, field action or otherapplicable actions. Investigators should discuss and evalu-ate these issues but be careful not to say anything that couldbe construed as requesting a product recall.

8. Determine if corrective and preventive actions wereeffective and verified or validated prior to implemen-tation. Confirm that corrective and preventive ac-tions do not adversely affect the finished device.

Using the selected sample of significant corrective and pre-ventive actions, determine the effectiveness of these cor-rective or preventive actions. This can be accomplishedby reviewing product and quality problem trend results.Determine if there are any similar product or quality prob-lems after the implementation of the corrective or preven-tive actions. Determine if the firm has verified or validatedthe corrective or preventive actions to ensure that suchactions are effective and do not adversely affect the fin-ished device.

Corrective actions must be verified and (if applicable) vali-dated. Corrective actions must include the application ofdesign controls if appropriate.

Good engineering principles should include: establishing averification or validation protocol; verification of productoutput against documented product requirements and speci-fications; ensuring test instruments are maintained and cali-brated; and that test results are maintained, available andreadable. Important linkages regarding this CAPA elementinclude 820.30 Design Control and 820.70(b) Productionand Process Control.

58

9. Verify that corrective and preventive actions forproduct and quality problems were implemented anddocumented.

Using the sampling tables, select and review records of themost recent corrective or preventive actions (this samplemay consist of or include records from the previously se-lected sample of significant corrective actions). To deter-mine if corrective and preventive actions for product andquality problems and changes have been documented andimplemented it may be necessary to view actual processes,equipment, facilities or documentation.

10. Determine if information regarding nonconformingproduct and quality problems and corrective andpreventive actions has been properly disseminated,including dissemination for management review.

Determine that the relevant information regarding qualityproblems, as well as corrective and preventive actions, hasbeen submitted for management review. This can be ac-complished by determining which records in a recent CAPAevent were submitted for management review. Review theraw data submitted for management review and not theactual results of a management review.

Review the CAPA (and other procedures if necessary) andconfirm that there is a mechanism to disseminate relevantCAPA information to those individuals directly responsiblefor assuring product quality and the prevention of qualityproblems.

Review information related to product and quality problemsthat has been disseminated to those individuals directlyresponsible for assuring product quality and the preventionof quality problems. Using the sample of records from Ob-jective 9 above, confirm that information related to product

59

and quality problems is disseminated to individuals directlyresponsible for assuring product quality and the preventionof quality problems.

An important linkage to this CAPA element is 820.20 Man-agement Responsibility.

60

61

Corrective and PreventiveActions(CAPA)

Medical DeviceReporting

62

Medical DeviceReporting


1. Verify that the firm has MDR procedures that addressthe requirements in 21 CFR Part 803.17.

2. Verify that the firm has established and maintainsMDR event files that comply with 21 CFR Part 803.18.

3. Confirm that the appropriate MDR information is be-ing identified, reviewed, reported, documented andfiled.

4. Confirm that the firm follows their procedures and theyare effective in identifying MDR reportable deaths, se-rious injuries and malfunctions.

63

64

Medical Device Reporting

Narrative

1. Verify that the firm has MDR procedures that ad-dress the requirements in 21 CFR Part 803.17.

Review and confirm that the firm’s written MDR proceduresaddress:

A. Internal systems that provide for the timely and effec-tive identification, communication, and evaluation ofevents that may be subject to medical device report-ing.

Purpose/Importance

The Medical Device Reporting (MDR) Regulation requires medical device manu-facturers, device user facilities and importers to establish a system that ensuresthe prompt identification, timely investigation, reporting, documentation, and filingof device-related death, serious injury, and malfunction information.

The events described in Medical Device Reports (MDR’s) may require the FDA toinitiate corrective actions to protect the public health. Therefore, compliance withMedical Device Reporting must be verified to ensure that CDRH's Surveillance

Program receives both timely and accurate information.

65

B. A standard review process/procedure for determiningwhen an event meets the criteria for MDR reportingand ensuring the timely transmission of complete de-vice reports to FDA.

C. Documentation and recordkeeping regarding: infor-mation evaluated to determine if an event is reportable;all MDR reports and other information submitted to theFDA; and systems that ensure access to informationthat facilitates timely follow-up and inspection by FDA.

2. Verify that the firm has established and maintainsMDR event files that comply with 21 CFR Part 803.18

Using the sampling tables, select a number of MDR eventfiles. Review and verify that the MDR event files (hard copyor electronic) are prominently identified and easy to access.MDR files may be maintained as part of the 820.198 com-plaint file IF the two aforementioned criteria are met.

Confirm that the MDR event files contain: information fromany source that describes a device-related death, seriousinjury or malfunction; the firm’s evaluation of this informa-tion including decisions to submit or not to submit an MDRreport; and copies or references to supporting documenta-tion (e.g., failure analysis, lab reports, etc.).

Decisions not to submit an MDR report for a device-relateddeath, serious injury or malfunction must be documentedin the MDR file.

When applicable, the files will also contain copies of MDRdeath, serious injury, malfunction and five-day reports sub-mitted on FDA form 3500A, Supplemental Reports (3500A),Baseline Reports (3417) and MDR-related correspondence.

66

3. Confirm that the appropriate MDR information isbeing identified, reviewed, reported, documentedand filed.

Using the sampling tables, select a number of MDR reportsthat were submitted to the FDA.

Compare the firm’s written procedures to the way it identi-fied, processed, evaluated, reported and filed the reports.Note any discrepancies between the firm’s practice/writtenprocedures and any failure to follow or obtain informationrequired by the regulation and form 3500A (e.g., timely re-porting, complete investigation, consistency, etc.)

4. Confirm that the firm follows their procedures andthey are effective in identifying MDR reportabledeaths, serious injuries and malfunctions.

Using the sampling tables, select a number of unreportedcomplaints and records from one additional source of qual-ity data (service reports, repair reports, returned goods files,etc.).

Review these records and confirm that they do not containinformation relating to MDR reportable events (device-re-lated deaths, serious injuries or malfunctions).

If unreported events are identified, determine the firm’s ra-tionale for not submitting MDR reports. If the firm has failedto identify these events, or does not provide an adequaterationale for not submitting an MDR report (an adequaterationale may be that the firm’s investigation determinedthat it was in fact another manufacturer’s device involvedin the event), then this may be a significant MDR relatedobservation.

67


Reports of Corrections andRemovals

68

Reports ofCorrections and Removals


1. Determine if corrections or removals of a device wereinitiated by the manufacturer.

2. Confirm that the firm's management has implementedthe reporting requirements of 21 CFR Part 806.

3. Verify that the firm has established and continues tomaintain a file for all non-reportable corrections andremovals per 21 CFR Part 806.20. Also verify that thefirm is complying with the other file-related requirementsof 21 CFR Part 806.

69

70

Reports ofCorrections and Removals

Narrative1. Determine if corrections or removals of a device were

initiated by the manufacturer.

If the firm has not initiated any corrections or removals, noinspection under Reports of Corrections and Removals isnecessary, proceed to the inspection of Medical DeviceTracking. However, state in the EIR that Reports of Cor-rections and Removals were considered for inspection.

If the firm has initiated any corrections or removals, pro-ceed to Objective 2.

Purpose/Importance

The Corrections and Removals (CAR) Regulation requires med-ical device manufacturers and importers to promptly notify FDA ofany correction or removal initiated to reduce a risk to health. Thisearly notification improves FDA's ability to quickly evaluate risksand, when appropriate, initiate corrective actions to protect the pub-

lic health.

71

2. Confirm that the firm’s management has implementedthe reporting requirements of 21 CFR Part 806.

Using the sampling tables, select a number of files relatingto corrections or removals that have been reported to theFDA.

Review the files and verify that the firm: (1) is submittingwritten correction and removal reports to the appropriateFDA District Office within 10 days of initiating the actions;and (2) has provided all the information required in the writ-ten report per 806.10.

Using the sampling tables, select a number of correctiveaction files in general (e.g., CAPA files). Review the files. Ifyou identify any apparent Class I or Class II recalls thathave not been reported to the appropriate FDA District Of-fice, discuss the discrepancy with the firm. It may be nec-essary to list unresolved discrepancies on your FDA 483.All observations must be consistent with current FDA poli-cies and procedures.

3. Verify that the firm has established and continues tomaintain a file for all non-reportable corrections andremovals per 21 CFR Part 806.20. Also verify that thefirm is complying with the other file-related requirementsof 21 CFR Part 806.

Using the sampling tables, select a number of files relatingto non-reportable corrections or removals (806.20 files).

NOTE: Part 806 does not require firms to establish and maintain files forcorrections and removals reported to the FDA. However, documentation ofcorrective actions is required by the Quality System Regulation (21 CFR 820.100,Corrective and Preventive Action and 21 CFR 820.198, Complaint Files).

72

Review the 806.20 files and verify that the records containall the information required in 806.20. This review mustinclude confirmation that the files are retained for the ap-propriate period of time (2 years beyond the expected lifeof the device).

Confirm that these files also do not contain evidence of un-reported (apparent) Class I or Class II recalls. Determinewhether the files contain evidence of unreported (appar-ent) Class III voluntary recalls under 21 CFR Part 7. Also,verify that the firm is complying with the other file-relatedrequirements of 21 CFR Part 806.

Confirm any claims for exemption from 806 as a result of asubmission under either the MDR regulation or Radiologi-cal Health requirements. If you need assistance, contactthe District Recall Coordinator.

If the device has been sold to another firm, verify that the806.20 files have been transferred to the new manufac-turer or importer.

If compliance with the above requirements cannot be con-firmed, discuss the discrepancy with the firm. It may benecessary to list unresolved discrepancies on your FDA 483.All observations must be consistent with current FDA poli-cies and procedures.

NOTE: If the device has been sold to a firm that is not in your District, forwardan assignment request to the appropriate District Office requesting confirmationthat the 806.20 files have been transferred to the new manufacturer or importer.

73


Medical DeviceTracking

74



1. Determine if the firm manufactures or imports a trackeddevice.

2. Verify that the firm has established a written standardoperating procedure (SOP) for tracking that complieswith the requirements in 21 CFR Part 821.25(c).

3. Verify that the firm’s quality assurance program includesaudits of its tracking system within the appropriate time-frames specified in 21 CFR Part 821.25(c)(3).

75

76


Narrative

1. Determine if the firm manufactures or imports atracked device.

Ask the Management Representative (or designee) whetherthe firm manufactures or imports any device subject to theMedical Device Tracking Regulation (21 CFR Part 821). Ifthe firm does not manufacture or import a device subjectto the tracking regulation, you can terminate your trackinginspection.

If the firm does manufacture or import a device subject tothe tracking regulation, verify via discussions with the Man-agement Representative (or designee) or the review of es-

Purpose/Importance

The purpose of the Medical Device Tracking Regulation is to en-sure that manufacturers and importers of certain medical devicescan expeditiously locate and remove these devices from the mar-

ket and/or notify patients of significant device problems.

77

tablished procedures, that the firm is aware of its trackingobligations.

Verify that the firm is aware of its obligation to: (1) notifyFDA if it goes out of business and provide copies of its track-ing records to its FDA District Office; (2) transfer trackingrecords to a firm purchasing its tracked device(s); and (3)continue tracking a device the firm stops manufacturing orimporting if the firm remains in business.

If the firm's tracked device was purchased from anotherfirm, confirm (where applicable) that the firm has obtainedand maintains the prior manufacturer's tracking records orequivalent information.

2. Verify that the firm has established a written stan-dard operating procedure (SOP) for tracking thatcomplies with the requirements in 21 CFR Part821.25(c).

Review the firm's written tracking SOP(s) and confirm (ifpossible) that they address the firm's capability to: (1) iden-tify the location and other required data, for tracked devicesundistributed to a patient within three working days after arequest by FDA, and (2) identify the location and otherrequired data for tracked devices distributed to a patient,within 10 working days after receipt of a request from FDA.

If applicable, select one or two files containing tracking in-formation requested by the FDA and confirm that the ap-propriate information required by 821.25(a)(1) – 821.25(a)(3)was provided within the appropriate time-frame(s).

Confirm that the written tracking SOP(s) address the re-maining 821.25(a), 821.25(b), and 821.25(c) requirementsfor the collection, maintenance and auditing of tracking data.

78

NOTE: The agency’s policy relative to the review of quality audit results is stated inCPG 7151.02 (CPG Manual Sub Chapter 130.300). This policy prohibits FDA ac-cess to a firm’s quality audit results. However, the audit procedures and documentsthat demonstrate that the audits have been conducted at the appropriate time inter-vals are subject to FDA inspection.

3. Verify that the firm’s quality assurance programincludes audits of its tracking system within theappropriate time-frames specified in 21 CFR Part821.25(c)(3).

Confirm that the audit procedure addresses both the func-tioning of the tracking system and the accuracy and com-pleteness of the data within the system.

Confirm that the firm has conducted audits of its trackingsystem at the appropriate time intervals (no less than ev-ery six months for the first three years of tracking and an-nually thereafter).

79

Productionand

Process ControlsSubsystem

80

Production and Process Controls


1. Select a process for review based on:

a. CAPA indicators of process problems;b. Use of the process for manufacturing higher risk devices;c. Degree of risk of the process to cause device failures;d. The firm’s lack of familiarity and experience with the process;e. Use of the process in manufacturing multiple devices;f. Variety in process technologies and Profile classes;g. Processes not covered during previous inspections;h. Any other appropriate criterion as dictated by the assignment

Note: If the process chosen is sterilization, evaluate the process accord-ing to the “Sterilization Process Controls” chapter of this handbook.

2. Review the specific procedure(s) for the manufacturing process selectedand the methods for controlling and monitoring the process. Verify thatthe process is controlled and monitored.

Note: Control and monitoring procedures may include in-process andor finished device acceptance activities as well as environmental andcontamination control measures.

3. If review of the Device History Records (including process control andmonitoring records, etc.) reveals that the process is outside the firm’stolerance for operating parameters and/or rejects or that product noncon-formances exist:

a. Determine whether any nonconformances were handled appropri-ately;

b. Review the equipment adjustment, calibration and maintenance;and

c. Evaluate the validation study in full to determine whether the pro-cess has been adequately validated.

4. If the results of the process reviewed cannot be fully verified, confirmthat the process was validated by reviewing the validation study.

5. If the process is software controlled, confirm that the software was vali-dated.

6. Verify that personnel have been appropriately qualified to implementvalidated processes or appropriately trained to implement processeswhich yield results that can be fully verified.

81

82

1. Select a process for review based on:

a. CAPA indicators of process problems;b. Use of the process for manufacturing higher risk

devices;c. Degree of risk of the process to cause device

failures;d. The firm’s lack of familiarity and experience with

the process;e. Use of the process in manufacturing multiple de-

vices;f. Variety in process technologies and profile classes;g. Processes not covered during previous inspec-

tions;h. Any other appropriate criterion as dictated by the

assignment

Production and Process Controls

Narrative

Purpose/Importance

The purpose of the production and process control subsystemis to manufacture products that meet specifications. Develop-ing processes that are adequate to produce devices that meetspecifications, validating (or fully verifying the results of) thoseprocesses, and monitoring and controlling the processes are allsteps that help assure the result will be devices that meet speci-fications.

83

Note: If the process chosen is Sterilization, evaluatethe process according to the “Sterilization ProcessControls” chapter of this handbook.

In order to meet the Production and Process Control re-quirements of the Quality System Regulation, the firm mustunderstand when deviations from device specificationscould occur as a result of the manufacturing process orenvironment.

Discuss with the Management Representative (or desig-nee) the firm's system for determining whether deviationsfrom device specifications could occur as a result of themanufacturing process or environment. The firm may ac-complish this requirement via Product and Process RiskAnalyses. Important linkages for these activities include820.20 Management Responsibility and 820.30 DesignControls.

Select for evaluation a manufacturing process where de-viations from device specifications could occur as a resultof the process or its environment. The selection of themanufacturing process for evaluation should be based uponone or more of the criteria listed above. Important linkagesto consider at this point include 820.30 (g) Design Valida-tion (risk analysis) and 820.100 Corrective and PreventiveAction.

2. Review the specific procedure(s) for the manufac-turing process selected and the methods for con-trolling and monitoring the process. Verify that theprocess is controlled and monitored.

Note: Control and monitoring procedures may includein-process and or finished device acceptance activitiesas well as environmental and contamination controlmeasures.

84

All processes that may cause a deviation to a device's speci-fication and all validated processes must be monitored andcontrolled in accordance with established procedures. Justbecause a process is validated, does not mean verificationactivities utilized to monitor and control the process areunnecessary. Examples of some verification activities as-sociated with validated processes include review of pro-cess parameters, dimensional inspections, package per-formance tests, sterility and EO residual testing.

For the process chosen, confirm that the established Pro-cess (and where applicable Environmental and Contami-nation) Control, Monitoring and Product Acceptance Pro-cedures maintained by the shop floor are the most currentapproved revision contained within the Device MasterRecord (DMR). Most firms maintain a "Master List" of themost currently approved documents. This list can be veri-fied against the DMR and brought to the shop floor to com-pare with the currently available documents.

Verify that the building is of suitable design and containssufficient space to perform necessary operations.

Verify that the control and monitoring activities demonstratethat the process is currently operating in accordance withthe DMR. This should be done on the shop floor by review-ing work instructions, product acceptance criteria and re-sults, control charts, etc.

While on the shop floor, make note of one significant pieceof process equipment and one significant piece of inspec-tion, measuring or test equipment (preferably from a fin-ished device acceptance activity). Prior to concluding the

NOTE: Control and monitoring procedures may include in-process and/orfinished device acceptance activities as well as environmental and con-tamination control measures.

85

inspection, confirm that applicable maintenance activities(preventive maintenance, cleaning, adjustment etc.) areperformed as scheduled for the chosen piece of process-ing equipment. Also confirm that the piece of inspection,measuring or test equipment was controlled and calibrated.

Once you’ve reviewed the process control and monitoringactivities on the shop floor, use the sampling tables andselect for review a number of Device History Records(DHR's including monitoring and control records, etc.) fromrecent production runs. If the process is run over morethan one shift, your review should include DHR's from allshifts. Verify that the product was manufactured in accor-dance with the Device Master Record.

This verification must include a review of the purchasingcontrols and receiving acceptance activities regarding atleast one component or raw material (preferably determinedessential for the proper functioning of the device).

In addition, this verification must include a review of in-pro-cess and final finished device acceptance activities andresults as well as environmental and contamination controlrecords (if applicable). Verify that sampling plans for pro-cess and environmental control and monitoring activitiesare based upon a valid statistical rationale.

If your review of the device history records reveals noanomalies proceed to Objective 4.

If evidence that the process or environment are not con-trolled and monitored (no control and monitoring activities,not operating within most currently approved parametersor reject limits, etc.) is observed, this may be a major pro-duction and process control deficiency. Important linkagesto consider at this point include Documents, Records &Change Controls, (820.40 Document Controls, 820.180

86

Records, 820.181 Device Master Record, 820.184 DeviceHistory Record,), Facilities and Equipment Controls (820.72Inspection, Measuring, and Test Equipment), Material Con-trols (820.50 Purchasing Controls, 820.60 Identification,820.65 Traceability, 820.80 Receiving, In-process, and Fin-ished device acceptance, 820.86 Acceptance Status,820.130 Packaging, 820.140 Handling, 820.150 Storage,820.160 Distribution) and 820.250 Statistical Techniques.

3. If review of the Device History Records (includingprocess control and monitoring records, etc.) re-veals that the process is outside the firm’s toler-ance for operating parameters and/or rejects or thatproduct nonconformances exist:

a. Determine whether any nonconformances werehandled appropriately;

b. Review the equipment adjustment, calibrationand maintenance; and

c. Evaluate the validation study in full to determinewhether the process has been adequately vali-dated.

If process or product nonconformance(s) are identifiedbased upon these activities, determine whether thenonconformance(s) were recognized by the firm, handledappropriately and fed into its CAPA system. Review (if ap-propriate) the firm’s nonconforming product control, reviewand disposition activities and any CAPA's indicated. If thefirm's Quality System failed to recognize the process or

NOTE:1.If the firm engages in a number of manufacturing processes, Investiga-

tors should avoid repeatedly selecting the same process every time thefirm is inspected.

2.If Device Labeling is the process chosen, include in your inspectioncoverage of the requirements of “820.120 Device Labeling”.

87

product nonconformance(s) or take appropriate CAPA, thismay be a major CAPA deficiency.

Review the firm's equipment adjustment, maintenance andcalibration records for the process and (if appropriate) com-prehensively evaluate the Validation Study as described inthe "Note" contained within the narrative discussion of Ob-jective 4. These activities may provide further insight intothe cause of the nonconformance. If the firm has recog-nized and implemented appropriate CAPA's regarding theobserved nonconformance(s), then the quality system waseffective. Proceed to Objective 5. Important linkages toconsider at this point include Corrective and PreventiveAction, Material Controls (820.90 Nonconforming product),and Facilities and Equipment Controls (820.72 Control ofinspection, measuring and test equipment).

4. If the results of the process reviewed cannot be fullyverified, confirm that the process was validated byreviewing the validation study.

If the results of the process can be fully verified, proceed toObjective 5.

If the chosen process requires process validation, reviewthe established Process Validation Procedure(s). The regu-lation does not require a general Process Validation Proce-dure. Therefore, separate procedures may be establishedfor each individual Process Validation Study. Remember,the definition of "Product" contained within the regulationincludes components, in-process devices and finished de-vices. Verify via a review of the Process Validation StudySummary (if available) and Approval, that objective evi-dence has demonstrated that the process will consistentlygenerate a product or result meeting its predeterminedspecifications. With respect to process validation, an ex-

88

ample of a “result” is a Sterility Assurance Level (SAL). If aValidation Study Summary and Approval is not available, areview of objective evidence within the validation study willbe necessary.

NOTE: If there are indications (via review of DHR’s, the Process Valida-tion Study Summary and Approval, the assignment, CAPA system, etc.)of unresolved, potential problems with a validated process, in additionto a review of process monitoring and control activities, a comprehen-

sive validation study review should be conducted. This review should include determin-ing whether: 1. The instruments used to generate the objective evidencewere properly calibrated and maintained prior to the validation study; 2. Predeterminedproduct specifications were established; 3. Test sample sampling plans were basedupon a statistically valid rationale; 4. Objective evidence demonstrates predeterminedproduct specifications were met consistently; 5. Process tolerance limits were chal-lenged; 6. Process equipment was properly installed, adjusted and maintained; 7.Process monitoring instruments are properly calibrated and maintained; 8. Changes tothe validated process were appropriately challenged; and, 9. Process operators areappropriately qualified.

If the objective evidence demonstrates that the process is not capable of consistentlyproducing a product or result meeting its predetermined specifications, this is a majorprocess validation deficiency. Important linkages to consider at this point includeManagement Responsibility (including 820.25 Personnel), Design Controls (820.30(h)Design Transfer), Corrective and Preventive Action, and Facilities and EquipmentControls (820.72 Inspection, Measuring and Test Equipment) and 820.250 StatisticalTechniques.

89

5. If the process is software controlled, confirm that thesoftware was validated.

If the process chosen is NOT controlled with software, pro-ceed to Objective 6.

If the process chosen is automated with software, reviewthe software requirements document, software validationprotocol, software validation activities, software changecontrols and software validation results to confirm that thesoftware will meet user needs and its intended use. If mul-tiple software driven systems are used in the process, chal-lenge one based upon significance. An important linkageto consider at this point is Material Controls (820.50 Pur-chasing Controls). For example, for software developedelsewhere, confirm that appropriate software and qualityrequirements were established and provided to the vendorand that purchasing data (and validation results) supportthat the requirements were met.

6. Verify that personnel have been appropriately quali-fied to implement validated processes or appropri-ately trained to implement processes which yieldresults that can be fully verified.

Using the sampling tables, select a number of training andqualification records for process operators and employeesconducting Q.C. activities related to the chosen process.Where a process is operated over more than one shift, train-ing records from all shifts should be included within yourreview. Confirm that the employees are aware of the de-vice defects that may occur as a result of improper perfor-mance of their assigned responsibilities. Confirm that em-ployees conducting Q.C. inspections and tests are awareof the defects and errors that may be encountered whileperforming their assigned responsibilities. An important link-age to consider at this point is Management Responsibility(820.25 Personnel).

90

91

Production and ProcessControls

Subsystem

Sterilization ProcessControls

92

Sterilization Process Controls


1. Confirm that the sterilization process was validatedby reviewing the validation study.

2. Review the specific procedure(s) for the sterilizationprocess selected and the methods for controlling andmonitoring the process. Verify that the process iscontrolled and monitored.

3. If review of the Device History Records (includingprocess control and monitoring records, acceptanceactivity records, etc.) reveals that the sterilizationprocess is outside the firm’s tolerance for operatingor performance parameters:

a. Determine whether the nonconformances werehandled appropriately; and

b. Review the equipment adjustment, calibrationand maintenance

4. If the sterilization process is software controlled,confirm that the software was validated.

5. Verify that personnel have been appropriatelyqualified and trained to implement the sterilizationprocess.

93

94

Sterilization Process Controls

Narrative

If you are inspecting a contract sterilizer, Inspectional Ob-jectives 2 through 5, described below, are applicable andmust be performed. Inspectional Objective 1 regardingvalidation is applicable only in so far as the contract steril-izer has assumed any responsibility for validation of theprocess, as indicated in the written agreement between thedevice manufacturer and the contract sterilizer.

1. Confirm that the sterilization process was validatedby reviewing the validation study.

Validation studies (according to established procedures) arerequired for sterilization processes.

Purpose/Importance

The purpose of the production and process control subsystem (includingsterilization process controls) is to manufacture products that meet speci-fications. Developing processes that are adequate to produce devicesthat meet specifications, validating (or fully verifying the results of) thoseprocesses, and monitoring and controlling the processes are all stepsthat help assure the result will be devices that meet specifications. Forsterilization processes, the primary device specification is the desiredSterility Assurance Level (SAL). Other specifications may include ster-ilant residues and endotoxin levels.

95

The review of the sterilization process validation study maybe limited to a review of the Validation Study Summary (ifavailable) and Approval if the complete validation study wasassessed during the previous inspection and there havebeen no significant changes in the process, product or pack-age that may impact sterilization effectiveness.

When conducting a complete sterilization process valida-tion study assessment, the items included in the narrativenote under Objective 4 of the Production and Process Con-trols chapter of this Handbook apply. A complete steriliza-tion process validation study assessment must include areview of the established validation procedures and verifi-cation (via a review of objective evidence) that: 1. Basedupon the bioburden of the product, the defined sterilizationprocess parameters will consistently be effective in obtain-ing a predetermined Sterility Assurance Level (SAL); and 2. The defined process parameters will not adversely aff-ect product and package performance.

Objective evidence that the sterilization process parameterswill consistently be effective in obtaining a predeterminedSterility Assurance Level (SAL) includes records document-ing: 1. The determination of product bioburden; 2. The es-tablishment of process parameters and tolerances; 3. Thedefinition of acceptance criteria for a successful validationstudy; 4. The process challenge studies (e.g. half cycle runsfor Ethylene Oxide, verification dose experiments for radia-

NOTE: Many firms sterilize their products according to the guidance provided withinconsensus standards (e.g. AAMI/ANSI/ISO standards). These standards are specificto various types of sterilization processes. FDA recognizes many of these standards.This means FDA finds them acceptable. A list of recognized sterilization standards

appears at FDA’s Center for Devices and Radiological Health (CDRH’s) web site located at:

www.fda.gov/cdrh/modact/recstand.html

Firms may elect to comply with these standards. However, compliance to the standards is voluntary.When a firm claims to comply with one of the recognized standards, the requirements of the standardmust be met. If a firm does not claim to comply with a recognized standard, it must provide a scientificrationale supporting the method used for validating and processing its sterilization loads.

http://www.fda.gov/cdrh/modact/recstand.html

96

tion, or media fills for aseptic processing); and 5. Theresults of process control and monitoring and acceptanceactivities (control charts, Biological Indicators, Dosimeters,etc.) used to demonstrate that predetermined acceptancecriteria had been met.

Objective evidence that process parameters will not ad-versely affect product and package performance includerecords documenting performance testing of the productand packaging following the sterilization process or mul-tiple sterilization processes (if applicable).

Determine whether periodic assessments (e.g.revalidations, sterility dose audits, etc.) of the adequacy ofthe sterilization process are conducted. Review the recordsof one periodic assessment of the adequacy of the steril-ization process.

NOTE: Many device manufacturers use contract steril-izers for sterilization of their devices. These manufac-turers retain the responsibility for the sterility of the fin-ished devices even though sterilization processing is not

performed at their own facilities. Therefore, your inspection of a manu-facturer that uses the services of a contract sterilizer must verify thatthe manufacturer has assumed that responsibility. Inspectional Ob-jectives 1 through 3 are applicable in this situation because the manu-facturer must be able to provide to you the documentation regardingsterilization validation and processing of its devices regardless of thelocation of these activities. Although the manufacturer may not havedetailed records regarding Objectives 4 and 5 for the contractor’ssoftware and personnel, he must have assured the adequacy of theseactivities by the contractor, through activities such as an audit of thecontractor, visits to the contractor, or review of documentation fromthe contractor. Objective 5 regarding qualifications of themanufacturer’s own Q.C. personnel should be covered during yourinspection of the manufacturer.

97

2. Review the specific procedure(s) for the steriliza-tion process selected and the methods for control-ling and monitoring the process. Verify that theprocess is controlled and monitored.

The sterilization process must be validated. However, thisdoes not mean that verification activities utilized to monitorand control the process are unnecessary.

If performed at this location, confirm that the sterilizationprocess, associated environmental and contamination con-trols, and monitoring and acceptance procedures main-tained by the shop floor are the most current approved re-vision contained within the Device Master Record (DMR).Most firms maintain a “Master List” of the currently approveddocuments. This list can be verified against the DMR andbrought to the shop floor to compare with the currently avail-able documents.

Verify that the building is of suitable design and containssufficient space to perform necessary operations.

Verify that the control and monitoring activities demonstratethat the process is currently operating in accordance withthe DMR. Sterilization parameters which may need to bemonitored and controlled include: time, temperature, pres-sure, load configuration, and humidity. Several of theseparameters may require monitoring and control prior to,during and after sterilization processing (e.g. precondition-ing, conditioning and aeration in Ethylene Oxide process-ing). Verification activities used to monitor and control thesterilization process may include: bioburden testing, Bio-logical Indicator (BI) testing, Chemical Indicator (CI) test-ing, process control record review, sterilant residue test-ing, and endotoxin testing.

98

Additionally, packaging integrity verification activities mustbe reviewed for every inspection during which sterilizationis covered. This review of the control and monitoring ac-tivities should be done on the shop floor by reviewing workinstructions, product acceptance procedures, control charts,etc.

While on the shop floor, make note of one piece of signifi-cant sterilization process equipment and one significantpiece of inspection, measuring or test equipment (prefer-ably from a finished device acceptance activity). Prior toconcluding the inspection, confirm that the applicable main-tenance activities (preventive maintenance, cleaning andadjustment, etc.) are performed as scheduled for the cho-sen piece of sterilization process equipment. Also, confirmthat the piece of inspection, measuring, and test equipmentwas controlled and calibrated.

After you have reviewed the process control and monitor-ing activities on the shop floor, use the sampling tables andselect for review a number of Device History Records(DHRs, including monitoring and control records, accep-tance testing records, etc.) from recent production runs. Ifthe process is run over more than one shift, your reviewshould include DHRs from all shifts. Verify that the productwas sterilized in accordance with the DMR. Your review ofthe selected records should include all applicable verifica-tion activities (see above) including records of process pa-rameter monitoring, and in-process and final device accep-tance activities and results.

Your evaluation must also include a review of the firm’spurchasing controls and receiving acceptance activitiesregarding at least one component, material or service.Examples include: the sterilant, sterilization indicators, andservices provided by contract sterilizers or contract labora-tories. In addition, review environmental and contamina-

99

tion control records (e.g. bioburden sampling, testing andresults). Verify that the sampling plans for process and en-vironmental control and monitoring activities are based upona valid statistical rationale.

If your review of the Device History Records reveals noanomalies, proceed to Objective 4.

If evidence that the process or environment are not con-trolled and monitored (no control and monitoring activities,not operating within most currently approved parameters,etc.) is observed, this may be a major production and pro-cess control deficiency. Important linkages to consider atthis point include: Documents, Records and Change Con-trols (820.180 Records, 820.181 Device Master Record,820.184 Device History Record, 820.40 Document Con-trols); Facilities and Equipment Controls (820.72 Inspec-tion, Measuring, and test Equipment); Material Controls(820.50 Purchasing Controls, 820.80 Receiving, In-process,and finished device acceptance, 820.140 Handling, 820.150Storage, and 820.160 Distribution, and 820.250 StatisticalTechniques, 820.60 Identification, 820.65 Traceability,820.86 Acceptance Status); 820.130 Packaging; and820.250 Statistical Techniques.

3. If review of the Device History Records (includingprocess control and monitoring records, accep-tance activity records, etc.) reveals that the steril-ization process is outside the firm’s tolerance foroperating or performance parameters:

a. Determine whether the nonconformances werehandled appropriately; and

b. Review the equipment adjustment, calibrationand maintenance

If process or product nonconformance(s) are identified

100

based upon these activities, determine whether thenonconformance(s) were recognized by the firm, handledappropriately and fed into its CAPA system.

Review (if appropriate) the firm’s nonconforming productcontrol, review and disposition activities and any CAPA'sindicated. If the CAPA included a retest, review the firm’srationale for invalidating the original test results. If the CAPAincluded resterilization, confirm that the effects of theresterilization process on the product and package are un-derstood. For example, did a validation study provide ob-jective evidence that resterilization was acceptable?

If the firm's Quality System failed to recognize the processor product nonconformance(s) or take appropriate CAPA,this may be a major CAPA deficiency. Review the firm'sequipment adjustment, maintenance and calibration recordsfor the process. These activities may provide further insightinto the cause of the nonconformances.

Examples of nonconformances and sterilization processfailures the investigator may encounter include: Test Fail-ures (e.g. Positive Biological Indicators, high EO residues,high bioburdens, out of specification endotoxin results);Parametric Failures (process failures such as unspecifieddwell times, low pressure, low EO gas weights, loss of hu-midity, etc.); and Packaging Failures. Packaging Failuresmay be an indication of a sterilization process parameterproblem (vacuum) or a packaging process problem (vali-dation, sealer set up, etc.).

Important linkages to consider at this point include Correc-tive and Preventive Actions, Material Controls (820.90 Non-conforming product), and Facilities and Equipment Controls(820.72 Control of inspection, measuring, and test equip-ment).

101

4. If the sterilization process is software controlled,confirm that the software was validated.

If the sterilization process chosen is NOT controlled withsoftware, proceed to Objective 5.

If the sterilization process is automated with software, re-view the software requirements document, software vali-dation protocol, software validation activities, softwarechange controls and software validation results to confirmthat the software will meet user needs and its intended use.If multiple software driven systems are used in the steril-ization process, challenge one based upon significance.An important linkage to consider at this point is MaterialControls (820.50 Purchasing Controls). For example, forsoftware developed elsewhere, confirm that appropriatesoftware and quality requirements were established andprovided to the vendor and that purchasing data (and vali-dation results) support that the requirements were met.

5. Verify that personnel have been appropriately quali-fied and trained to implement the sterilization pro-cess.

Using the sampling tables, select a number of training andqualification records for process operators and employeesconducting Q.C. activities related to the sterilization pro-cess. Where a process is operated over more than oneshift, training records from all shifts should be included withinyour review. Confirm that all employees are aware of thedevice defects that may occur as a result of improper per-formance of their assigned responsibilities. Confirm thatemployees conducting Q.C. inspections and tests are awareof the defects and errors that may be encountered while

102

NOTE: Information that must be reported with the Establishment In-spection Report (EIR) includes: 1. The identification of all sterilizationprocesses used by the firm (e.g. Ethylene Oxide, Gamma irradiation,etc.); 2. The identification of the sterilization process covered; 3. The

identification of any standard that the firm claims to follow for the process covered(if applicable); 4. The location of the sterilization sites; 5. The division of responsi-bilities for sterilization services (e.g. contract testing labs, sterilizer, finished devicemanufacturer, packaging, labeling etc.); 6. The SAL ; and, 7. whether or not para-metric release is utilized.

performing their assigned responsibilities. An important link-age to consider at this point is Management Responsibility(820.25 Personnel).

103

SamplingPlans:

Instructions&

Tables

SamplingPlans:

Instructions&

Tables

104

Sampling Plan Instructions

1. Select the table based upon how sure you wantto be about what is observed. For example, ifyou are reviewing Device History Records of alife supporting device, you may choose to useTable 2 (99% Confidence). You may choose touse Table 1 (95% Confidence) for the review ofDevice History Records regarding a device withlower risk.

2. Select a sample size. If the population ofrecords to be sampled is small (approximatelythirty or less), you may choose to review all ofthe records.

3. Review the sample of records selected. Youcan terminate your review of the entire sampleif you observe objectionable conditions beyondthe number stated in the column header1. How-ever, if you do not review all of the records inthe sample, you may not report additional infor-mation that could be useful in further under-standing the potential prevalence of the objec-tionable condition observed, or you may notrecognize whether other objectionable condi-tions exist.

Note: Factors to consider when selecting a samplingtable and sampling size may include the risk of thedevice being inspected or the records being sampled,and the amount of time you have allocated to this por-tion of the inspection.

105

1 If you choose to terminate your review prior to completing the review of the entiresample, in addition to the information contained in instruction 4, report in theEstablishment Inspection Report how many records were reviewed prior to yourtermination of the review.

2 The information requested in instruction 4 must be reported whenever an OfficialAction Indicated (OAI) endorsement is considered. Reporting this informationmay not be necessary when Voluntary Action or No Action is indicated. However,caution is advised when using this reporting discretion because Voluntary ActionIndicated endorsements are sometimes elevated to Official Action Indicated.

NOTE:A. There are no “acceptable” violations of the Quality System Regulation. All

Quality System Regulation violations encountered must be handled appro-priately according to current FDA policies and procedures. When using the“1 out of:” and “2 out of:” columns, it does not mean no more than thatnumber of Quality System Regulation violations per the appropriate samplesize is acceptable. It will only give you an initial understanding of how preva-lent the problem may be.

B. When at all possible, all samples should be chosen at random.

4. When objectionable conditions are observedbased upon samples chosen using these tables,report in the Establishment Inspection Report:(a) the total number of records included in thepopulation from which the sample was chosen;(b) the table used to select your sample; (c) therow used to select your sample; and, (d) thesample size selected2.

106

108

Management Controls Subsystem

Design Controls Subsystem

Corrective and Preventive Actions (CAPA)Subsystem

Medical Device Reporting(CAPA Satellite)

Corrections & Removals(CAPA Satellite)

Medical Device Tracking(CAPA Satellite)

Production and Process Controls (P&PC)Subsystem

Sterilization Process Controls(P&PC Satellite)

Sampling Plan Instructions & Tables

Important Linkages