Вы находитесь на странице: 1из 11

Device Validation Forum.

John E. Lincoln

Issues with Medical Device


Part 11 Electronic Records;
Electronic Signatures
John E. Lincoln

Device Validation Forum discusses regulatory


requirements, scientific principles, strategies, and
approaches associated with medical device validation that are useful to practitioners. We intend this
column to be a valuable resource for daily work
applications. The key objective for this column is
useful information.
Reader comments, questions, and suggestions
are needed to help us fulfill our objective for this
column. Suggestions for future discussion topics or questions to be addressed are requested.
Case studies illustrating principles associated with
medical devices submitted by readers are also most
welcome. We need your help to make Device
Validation Forum a useful resource. Please send
your comments and suggestions to column coordinator John E. Lincoln at jel@jelincoln.com or to
journal managing editor Cale Rubenstein at crubenstein@advanstar.com.

INTRODUCTION
As the medical device industry moves toward electronic records (ER) and signatures by in-house systems and/or cloud/web-based systems, and away from
paper documentation, 21 Code of Federal Regulations
(CFR) Part 11, Electronic Records; Electronic Signatures
(ES) verification and validation (V&V) activities and
documentation become mandatory. These issues are
not only a regulatory/Part 11 concern but also a user/
customer concern.
These requirements should not be viewed as unnecessary bureaucratic red tape. All industries, not just

For more Author

gxpandjvt.com/bios

gxpandjv t.com

magenta
cyan
yellow
black

What are these Issues?


There are several issues pertinent issues that need to
be described. They discuss how a company can verify
or validate compliance to those portions of Part 11
that are applicable to their operations.
Large software applications having a current good
manufacturing practice (cGMP) impact include enterprise resource planning (ERP) systems. ERP validations typically involve both cGMP activities/records
and purely business/non-cGMP activities/records.
These typically impact all areas of a company. They
often present the most complex challenge for a Part 11
V&V project. Since that is the case, a Part 11 validation
will typically only use test cases/scripts that address
specifically the cGMP functions that the software
performs, when they can be separated. Each test case
is developed from the software requirements specification (SRS) or its equivalent, which should only list
those requirements that are cGMP-specific (for the
purpose of the cGMP/Part 11 software V&V).
The companys 21 CFR Part 11 ER/ES requirements
would be included in the SRS. However, the author
recommends that the purely Part 11 requirements be
addressed by test cases in the operational qualification

information,
go to

US Food and Drug Administration-regulated ones,


are increasingly faced with these issues. The requirements of Part 11 are very similar to those that banking, finance, legal, and other business entities face.
All must strive to ensure the integrity of electronic
records/signatures as these increasingly replace paperbased records and documentation systems.

ABOUT THE AUTHOR


John E. Lincoln is a principal consultant for J.E. Lincoln and Associates LLC, which assists companies in the
design and implementation of complete 21 CFR 111, 210, 211, 820 and ISO 13485 quality management systems
that are fully cGMP-compliant and will have pass FDA audits. He may be reached by e-mail at jel@jelincoln.com.

Journal

of

Validation t echnology [Autumn 2012]

ES156867_IVTJVT1112_015.pgs 11.21.2012 00:44

15

ADV

Device Validation Forum.

(OQ), including those addressed by non-software/


offline systems and references to relevant standard
operating procedures (SOPs), manual logs, or similar
documentation. The rationale is that the V&V of Part
11 requirements generally focuses on the existence/
initialization (installation qualification [IQ] or operational qualification [OQ]) of each applicable element
of Part 11 rather than its repeatability performance
qualification (PQ). Where proof of repeatability is
a concern, test cases could be added to the PQ runs
as well.
As with any validation, a line in the sand must be
drawn prior to start. This means that once the decision
is made to validate, the software must be frozen
in time, with any future changes performed under
revision/release number/change control. Any changes
must include consideration of the degree of effect the
change may have on any previous verification/validation activities. Where such change control is relatively
easy with hardware, it is increasingly difficult with
software, especially cloud or web-based software
(e.g., applications or data warehousing/storage that
can almost automatically be upgraded, patched,
or have a service pack added by the vendor over the
Internet without notification or input from the using
company).
Whenever this author undertakes such a validation
with a client, a meeting is arranged with the companys information technology (IT) department and
quality assurance (QA) team to initiate systems and
capture and hold all such incoming changes for joint
IT/QA review against existing V&Vs. The appropriate
decision and method of implementation, regression
testing required, and/or similar actions can be decided, documented with supporting rationale, signed,
dated, and implemented under change control. Without such a system in place, any validations are merely
a waste of time and valuable resources.

ELECTRONIC RECORDS/SIGNATURES
AREAS REQUIRING V&V
The following are the type of electronic records and/or
e-signatures that require validation under 21 CFR Part
11. These may be exclusive cGMP records or records
used for cGMP decision-making (regardless of the
company written policy):
Any cGMP document that an SOP states is documented by a controlled hard/paper copy with
manually entered signatures (this includes personnel actually not using these hard copies but
referring to their computers in order to make
quality control [QC]/cGMP decisions [i.e., it is
16

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

not what a company says, but what it is actually


being done])
Management reviews of quality policies, systems,
organization/staffing, audits, etc.
Internal quality audits
Training: conduct, subject matter, and records
Proof of design control activities (an electronic
design history file [e-DHF])
Any cGMP document approval using e-documents
and/or e-signatures
Change control
Documentation of suppliers, evaluation/audits
rankings, and purchasing/quality data
Inventory identification, traceability, and status
Electronic SOPS (e-SOPs)
Monitoring/control of production processes electronically with e-reports
Environmental controls (heating, ventilation,
and air conditioning [HVAC], vector/pest, et al)
Post monitoring (PM) and/or calibration
scheduling
Record of equipment inspections
Control of manufacturing materials (e.g., lubricating oils, cleaners)
Test equipment control, including the above
Validation records
Incoming, in-process, and finished goods inspections: data, acceptance status, quarantine
Non-conformance reports, controls, reviews, dispositions, and approvals
Corrective and preventative action (CAPA) system documentation, including complaint and
MDR files, failure investigations, and root cause
analysis
Labeling design, control/storage, and issuance/
counts
Packaging documentation
Distribution records
All cGMP e-records (if primary records, as defined
by usage)
Device master record (DMR)
Device history (batch/lot) record (DHR)
Any electronic/computer statistical analysis tools
related to making cGMP decisions (e.g., product
release, which may require additional software
V&V)
As per above, all software systems, independent
of cGMP records/signatures used in manufacturing or part of medical devices, require their
own V&V per other guidance documents (820.30
design controls [product validation] and 820.70
[i] automated processes)
iv tnetwork.com

ES156871_IVTJVT1112_016.pgs 11.21.2012 00:44

ADV

John E. Lincoln.

Table I: Subpart 11.10: Verify Records Input and Retention.


Action Initiated

Expected Outcome

Meet Outcome

Can invalid or altered records be


determined?

Invalid/altered records can be


determined.

Yes/No
Attachment #

Is system capable of producing


accurate/complete hard/paper
copies of electronic records?

System produces accurate/


complete hard copies of ER.

Yes/No
Attachment #

Are records readily retrievable


throughout their retention period
(user to define records/data bases
involved and retention periodone
year from shipment, minimum)?

ERs are readily retrievable


throughout their retention period.

Yes/No
Attachment #

Is system access limited to


authorized personnel (by
password, SOP, and user-provided
and physical security)?

System access is limited to


authorized personnel (state
method).

Yes/No

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table II: Verify Audit Trail.


Action Initiated

Expected Outcome

Meet Outcome

Does system create/maintain a


secure, time-stamped audit trail?

System creates/maintains a
secure, time stamped audit trail.

Yes/No
Attachment #

Does it record date/time, entries/


actions for any activity that
creates, modifies, or deletes
electronic records (documents
to be controlled by user, and 21
CFR Part 820 Quality System [QS]
Regulation/medical device cGMPs).

System records date/time, entries/


actions for any ER creation,
modification, or deletion of cGMP
records.

Yes/No
Attachment #

Are changed or deleted records


archived and retrievable (records to
be defined by user)?

Changed or deleted records are


archived and retrievable.

Yes/No
Attachment #

Is the audit trail retrievable throughout


that records retention period?

Audit trails are retrievable


throughout the ERs retention
period.

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

ERP software used to control movement and storage of inventory, as per above
Any other cGMP/QA/QC approval action and/
or status record.
gxpandjv t.com

magenta
cyan
yellow
black

RISK-BASED BLACK BOX V&V


Since most projects usually involve commercial offthe-shelf (COTS) software, the scripts/test cases are run
black box and also involve hardware functionality.
Journal

of

Validation t echnology [Autumn 2012]

ES156875_IVTJVT1112_017.pgs 11.21.2012 00:44

17

ADV

Device Validation Forum.

Table III: System features/checks.


Verify data installation is completed, correct, and readily retrievable.
Action Initiated

Expected Outcome

Meet Outcome

Can it be reviewed/copied by FDA?

ERs can be reviewed/copied by


FDA.

Yes/No
Attachment #

Does the system enforce sequence


of steps/events if required (e.g., no
release to inverse of non-approved
components/specific steps)?

The system enforces the sequence


of ERP events per referenced flow
charts.

Yes/No
Attachment #

Are only authorized individuals


allowed access to the system,
permitted to sign records, access
the operation/input/output device,
alter records, and perform other
operations (e.g., defined by password
and level of authority/access)?

Only authorized individuals are


allowed access, sign records,
installation/operation access,
records altering, and similar
operations that affect ER accuracy/
retention/retrieve ability.

Yes/No
Attachment #

Does the system check the validity


of the data source if multiple
sources for such data exist?

The system checks one source


therefore there are no checks
between sources.

Yes/No

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table IV: Training.


Verify data installation completed per SOP.
Action Initiated

Expected Outcome

Meet Outcome

Is training of all involved personnel


conducted and documented
(user issue/vendor assist)?

Training is conducted periodically


and documented.

Yes/No
Attachment #

Do written policies address the


accountability and responsibility of
individuals actions initiated under
their electronic signature (user
issue)?

Written policies/SOPs address ES


accountabilities/responsibilities.

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Of course, all software V&V is product risk-based. It


is recommended that an International Organization
for Standardization (ISO) 14971 product risk management file/report (and use the same format for non-device
industries) be developed prior to developing the V&V
documentation. The degree of risk tied to the user of the
companys products can be used to determine the amount
of test case/script elements necessary to prove compli18

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

ance. Tie test cases to specific risk document references


by a traceability matrix or commonality of numbering
between hazard/risk entry and test case/script to justify
the degree of verification elements addressed in each test
case. It is crucial to draw that line in the sand on the
software with the support of the company IT department
to prevent non-approved updates, patches, etc. to the
relevant software that could impact the V&V downstream.
iv tnetwork.com

ES156880_IVTJVT1112_018.pgs 11.21.2012 00:44

ADV

John E. Lincoln.

Table V: Systems documentation control.


Verify data installation completed per SOP.
Action Initiated

Expected Outcome

Meet Outcome

Is systems operation/maintenance
documentation controlled
(user and password limits)?

Systems operation/maintenance
documentation is controlled
(reference method[s]).

Yes/No
Attachment #

Is system documentation under


formal change control with a
time-sequenced audit trail for
changes (Also see other audit
trail questions/comments)?

System documentation is under


formal change control with an audit
trail.

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table VI: Subpart 11.30.


Verify data installation is completed, correct, and readily retrievable.
Action Initiated

Expected Outcome

Meet Outcome

Is open system data

Open system data is encrypted.

N/A

Are open system signatures


digitized?

Open system signatures are


digitized (or reference any alternate
method(s).

N/A

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

The following test case elements are extracted


directly from 21 CFR Part 11.

ELECTRONIC RECORDS AND ELECTRONIC SIGNATURES


Subpart A--General Provisions 11.1 - Scope. 11.2 Implementation. 11.3 - Definitions.
Subpart B--Electronic Records 11.10 - Controls for closed
systems. 11.30 - Controls for open systems. 11.50 - Signature manifestations. 11.70 - Signature/record linking.
Subpart C--Electronic Signatures 11.100 - General requirements. 11.200 - Electronic signature components and controls. 11.300 - Controls for identification codes/passwords.

DEVELOPING THE TEST CASES/TEST


SCRIPT
Each element of the subparts of Part 11 are reframed
into questions or statements for which an answer in
the companys Part 11 software or offline systems
will have to be found. Subparts 11.1, 11.2, and 11.3
gxpandjv t.com

magenta
cyan
yellow
black

provide background information and requirements


of the CFR, and consideration is for reference only.
Beginning with subpart 11.10, the suggested
approach described below can be implemented. In
some instances, assumptions have been made regarding the element to verify. When using this example,
the user will have to adjust actual test cases/scripts to
match the systems/applications elements that apply
to their application. Higher risk applications would
require expansion of the number of test case/test script
entries to resolve or verify function of each element.

SOFTWARE VERIFICATION/VALIDATION
PROTOCOL FORMAT EXAMPLES
The following should be considered as very basic templates. Applicable test cases or test case elements should
be expanded depending upon the applications being
verified/validated. These present one possible method
among many that could be acceptable in validating electronic records and electronic signatures to 21 CFR Part 11.
Journal

of

Validation t echnology [Autumn 2012]

ES156872_IVTJVT1112_019.pgs 11.21.2012 00:44

19

ADV

Device Validation Forum.

VERIFICATION SCRIPT: ELECTRONIC


SIGNATURES

Table VII: Subpart 11.50: Electronic Signatures Features.


Verify data installation is completed, correct, and retrievable.
Action Initiated

Expected Outcome

Meet Outcome

Do electronic signature
manifestations include the printed
name, date/time of signing, and
meaning of signing (approval,
review, responsibility, and feature
is available generally by level of
password-protected /defined level
of access)?

ESs include stated requirements.

Yes/No
Attachment #

Is the signature supporting


information mentioned above
displayed and printed on hard
copies of the electronic record?

ESs are displayed, printed, or


obviously linked on hard copies
printed of the ER.

Yes/No
Attachment #

Are signatures linked to the


respective electronic record
to prevent cut/copy/transfer/
falsification (are signatures
imbedded in the actual record/
document or stored in another file
and flagged)?

ESs are either linked to, or


embedded in, the respective ER.

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table VIII: Subpart 11.100: Unique Electronic Signature.


Action Initiated

Expected Outcome

Meet Outcome

Are electronic signatures unique to


an individual (through specific login
and password)?

Unique ESs exist.

Yes/No
Attachment #

ID verified before issue?

Verification performed by system to


prevent duplicate IDs.

Yes/No
Attachment #

Are electronic signatures reused or


reassigned to others (controlled user
SOP number/user setup)?

ESs are not reused

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

20

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

iv tnetwork.com

ES156869_IVTJVT1112_020.pgs 11.21.2012 00:44

ADV

John E. Lincoln.

Table IX: Subpart 11.200: Secure Electronic Signature.


Action Initiated

Expected Outcome

Meet Outcome

Is the signature made up of at least


two components (e.g., code, card,
password combinations, ID and
password, and use physical ID
components [e.g., cards])?

ES has minimum of two


components (describe).

Yes/No
Attachment #

Must the password be executed at


each signing and during a multiple
signing (continuous) session?

Password must be entered for


each signing in multiple signings.

Yes/No
Attachment #

Does the capability exist to be


defined by user?

Capability is defined by user


(describe).

Yes/No
Attachment #

Recommend reentry of password


wherever a new physical signature
would be required rather than a
multiple/continuous-signing feature.

Password reentry is required for


any new signature.

Yes/No
Attachment #

If not continuous, must both


components of the signature be
executed (to be user-defined)?

Describe number of components


required for a signature to be
entered (user ID/user password).

Yes/No

Are non-biometric signatures only


used by their genuine owners (user
SOP defined/a user security issue)?

Describe method used for control


of non-biometric ES: issue of single
user ID/user defined password.

Yes/No

Has it been shown that biometric


signatures can only be used by
their genuine owner (are biometric
signatures [retina or fingerprint
scans, etc.] utilized)?

Biometric signatures are not


currently used in this ERP.

Yes/No

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table X: Planned security breach.


Action Initiated

Expected Outcome

Meet Outcome

Would an attempt to falsify an


electronic signature require the
collaboration of at least two
individuals (only in the sense that
one of the two just have been
careless in allowing another to
steal and use his/her password)?

Purposeful falsification of an
ES requires two or more willing
individuals

Yes/No

Is the software configured


to require a minimum of two
passwords to accomplish a defined
action (e.g., document changes)?

A minimum of two ESs are required


for the approval of a cGMP ER.

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________
gxpandjv t.com

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

ES156868_IVTJVT1112_021.pgs 11.21.2012 00:44

21

ADV

Device Validation Forum.

Table XI: Subpart 11.300: User ID/Passwords.


Action Initiated

Expected Outcome

Meet Outcome

Are controls in place to assure


the uniqueness of each code/
password combination?

User IDs/passwords are controlled;


system prevents user ID to be reissued/re-used

Yes/No
Attachment #

Do procedures require the periodic


checking of the validity of ID codes
(user SOP issue: does software do
this automatically)?

This is controlled in Windows OS;


passwords expire per defined
intervals.

Yes/No
Attachment #

Do passwords periodically expire


and require revision (see above)?

Passwords are controlled by


Windows OS with periodic
expiration/revision.

Yes/No
Attachment #

Is there a procedure to recall ID


codes/passwords when someone
leaves/is transferred?

User IDs/passwords are recalled/


retired when the owner leaves/is
transferred per SOP XXX.

Yes/No
Attachment #

Is there a procedure to
electronically disable any ID code/
password that has been potentially
compromised/lost?

User IDs/passwords can be


disabled if they are suspected of
having been compromised/lost per
SOP XXX.

Yes/No

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

Table XII: Outside systems breech/hacking.


Action Initiated

Expected Outcome

Meet Outcome

Is there a procedure to detect


attempts at hacking and inform
security (primarily by the audit trail
feature after the fact)?

Hacking can be detected and is


acted upon by reviewing IT or
advanced security logs daily.

Yes/No
Attachment #

Is there a procedure for reporting


repeated or serious attempts at
unauthorized use to management
(could be by means of audit trail
review or user SOP)?

Attempts at unauthorized use


(see above) are documented and
reported to management (describe
method).

Yes/No
Attachment #

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

22

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

iv tnetwork.com

ES156870_IVTJVT1112_022.pgs 11.21.2012 00:44

ADV

John E. Lincoln.

Table XIII: Loss management.


Action Initiated

Expected Outcome

Meet Outcome

Is loss management defined/


practiced for lost or stolen devices
(only by user SOP)?

Loss management of any


applications-accessible devices is
practiced (describe).

Yes/No
Attachment #

Is there a procedure to
electronically disable a device if
its lost/stolen/compromised (by
password access/user alternative)?

Describe any method to disable a


compromised device.

Yes/No
Attachment #

Are there controls for issuance


of temporary and permanent
replacements?

Describe any controls in the


issuance of temporary or
permanent replacement devices.

Yes/No
Attachment #

Is there initial and periodic testing


of tokens/cards?

Describe or N/A

N/A

Does this check for unauthorized


alterations?

Describe or N/A

N/A

Verified By
Initial & Date

Comments: __________________________________________________________________________________
QA Reviewed by: _____________________________________ Date: _________________________________

AN IMPORTANT CAVEAT

CONCLUSION

A company may believe and proclaim that it is not


using electronic records and/or electronic signatures.
It may base this on the fact that its SOPs define that
controlled records are paper documents with manual
signatures, and those hard copies are routed for
approval and used for cGMP actions and retains/files.
However, the real test is how records are actually being
used to make cGMP decisions in the company. FDA
consumer safety officers (CSOs)/auditors have been
seen observing company personnel using their computer screen to pull up records and SOPs and then
make cGMP decisions from that image. If this is the
companys practice, even if they are controlling hard
copies and state that in their SOPs, the auditor will
rightly conclude that an e-record is being used and
expect to see 21 CFR Part 11 validation performed.

The use of electronic records and electronic signatures is increasingnot just in regulated industries.
These types of issues will be seen in all industries that
require legally binding documentation. Most professionals already deal with encrypted transactions on
the Internet and hope that companies have similar
systems in place to ensure integrity versus the growing danger of identity theft. The type of information
and verification/validation required in 21 CFR Part
11 will be replicated and expanded upon worldwide,
not only in medical products, but in finance, legal,
and all business entities desiring a viable global business model. JVT

gxpandjv t.com

magenta
cyan
yellow
black

Journal

of

Validation t echnology [Autumn 2012]

ES156882_IVTJVT1112_023.pgs 11.21.2012 00:44

23

ADV

Device Validation Forum.

GLOSSARY
Black box

CDRH
cGMPs
CFR
COTS
CSO
ERP
FDA

24

magenta
cyan
yellow
black

Journal

of

Review/verification of software algorithm/coding by observing the softwares operation of the hardware, without access to the actual software code,
as opposed to white box or glass
box testing (see white box below)
Center for Devices and Radiological
Health
Current good manufacturing practices
(for devices it is 21 CFR Part 820)
Code of Federal Regulation
Commercial off-the-shelf software
Consumer safety officer (i.e., the FDA
compliance auditor)
Enterprise resource planning
The United States Food and Drug
Administration

Validation t echnology [Autumn 2012]

ISO
IT
IQ
OS
OQ
PQ

International Standards Organization


Information technology
Installation qualification
Operating system
Operation qualification
Performance qualification (generally
three or more as needed by inherent
system inputs, et al, variability)
QA
Quality assurance
RA
Regulatory affairs
R&D
Research and development
SOP
Standard operating procedure
SRS
Software requirements specification
V&V/V[T]&V Verification [Testing] and Validation
White box
Code review for logic and adherence to
conventions with no observable problems (same as glass box review).

iv tnetwork.com

ES156878_IVTJVT1112_024.pgs 11.21.2012 00:44

ADV

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Вам также может понравиться