Per 21CFR 820.198 (b) - Justification for no investigation

#1
Per 21CFR 820.198 (b) "When no investigation is made, the manufacturer shall maintain a record that includes the reason no investigation was made and the name of the individual responsible for the decision not to investigate." I realize we can reference a previous investigation for the issue/product, but if there isn't a previous investigation, what are some reasons why an investigation would not have to be done? The company I work for manufacturers low risk medical devices and cosmetics.
 
Elsmar Forum Sponsor

Tidge

Trusted Information Resource
#2
I can think of a few options, YMMV. Examples are off-the-cuff and not meant to be fully realized defenses of the options.

1) The complaint is not be attributable to your product. This is absolutely clearest when the complaint is literally for something you are not responsible for. A word of caution: it is natural to want to 'pass the blame' for complaints; if your device is implicated in any way it would be appropriate to investigate (or reference a previous investigation for similar circumstances).

2) If your complaint handling process is robust (rather than ad hoc), it could be that the established process has pre-determined thresholds for certain types of complaints which only trigger an investigation if the threshold is crossed. For example: if your batch production methods for lip gloss were filling tubes with 4.30 +- 0.02 g of lip gloss but your labeling indicated 4.25 g on the tube, you could expect some number of complaints that a customer didn't get the full 4.25 g which they were promised by labeling. You could establish a threshold minimum number of such complaints below which you could have essentially already attributed the complaint to a known production variance. This is a bad example because failing to meet labeling should trigger an investigation under 820.198(c), so try to imagine something mathematical that doesn't directly involve labeling or a device specification.

Caveat: my personal opinion is that it is difficult to establish a sufficient level of robustness (in procedures, with data backing up thresholds) a priori. it is a more likely scenario that some sort of corrective action has been done (on a specific type of complaint) such that as part of the implementation and effectiveness check such thresholds would be established and monitored.

3) For some types of complaints, you may require product to be returned for a complete investigation. In the specific case of bulk cosmetics perhaps lots samples are retained? But I digress. It is possible (but IMO highly unlikely) that you could rationalize no investigation for some circumstances. I caution against this, as some minimal investigation (e.g. asking questions of the user) is probably warranted for those cases where you wanted a device to be returned.

TL; DR: See (1) above. There may be other options (with a lot of rationalization).

Ultimately, investigations can be as deep or as superficial as established by the manufacturer. The manufacturer has to be prepared to pass a 'red-face' test when asked about the investigation records. The most important elements of investigation of complaints are (IMO) to (a) not violate the expectation for reportable complaints and (b) to appropriately feed into a corrective action system (820.100) to support analyses or trigger corrective actions. If the investigations are deep enough to do those two things, you are probably on a good path.
 
Last edited:
#6
I can think of a few options, YMMV. Examples are off-the-cuff and not meant to be fully realized defenses of the options.

1) The complaint is not be attributable to your product. This is absolutely clearest when the complaint is literally for something you are not responsible for. A word of caution: it is natural to want to 'pass the blame' for complaints; if your device is implicated in any way it would be appropriate to investigate (or reference a previous investigation for similar circumstances).

2) If your complaint handling process is robust (rather than ad hoc), it could be that the established process has pre-determined thresholds for certain types of complaints which only trigger an investigation if the threshold is crossed. For example: if your batch production methods for lip gloss were filling tubes with 4.30 +- 0.02 g of lip gloss but your labeling indicated 4.25 g on the tube, you could expect some number of complaints that a customer didn't get the full 4.35 g which they were promised by labeling. You could establish a threshold minimum number of such complaints below which you could have essentially already attributed the complaint to a known production variance. This is a bad example because failing to meet labeling should trigger an investigation under 820.198(c), so try to imagine something mathematical that doesn't directly involve labeling or a device specification.

Caveat: my personal opinion is that it is difficult to establish a sufficient level of robustness (in procedures, with data backing up thresholds) a priori. it is a more likely scenario that some sort of corrective action has been done (on a specific type of complaint) such that as part of the implementation and effectiveness check such thresholds would be established and monitored.

3) For some types of complaints, you may require product to be returned for a complete investigation. In the specific case of bulk cosmetics perhaps lots samples are retained? But I digress. It is possible (but IMO highly unlikely) that you could rationalize no investigation for some circumstances. I caution against this, as some minimal investigation (e.g. asking questions of the user) is probably warranted for those cases where you wanted a device to be returned.

TL; DR: See (1) above. There may be other options (with a lot of rationalization).

Ultimately, investigations can be as deep or as superficial as established by the manufacturer. The manufacturer has to be prepared to pass a 'red-face' test when asked about the investigation records. The most important elements of investigation of complaints are (IMO) are to (a) not violate the expectation for reportable complaints and (b) to appropriately feed into a corrective action system (820.100) to support analyses or trigger corrective actions. If the investigations are deep enough to do those two things, you are probably on a good path.
Thanks for the feedback! Seems like it's better/easier to do an investigation rather than trying to justify why we don't need to do one :) Isolated complaints that happen only a few times a year that aren't a risk to safety seems like they wouldn't warrant an investigation too, but is it enough to just say it's "isolated?"
 
Last edited by a moderator:

Ronen E

Problem Solver
Staff member
Moderator
#7
Isolated complaints that happen only a few times a year that aren't a risk to safety seems like they wouldn't warrant an investigation
How would you know that they "aren't a risk to safety" if you haven't investigated?
 

Watchcat

Trusted Information Resource
#8
Rant alert...

I don't care what any regulation, standard, or guidance says. Every complaint warrants investigation. The idea you can justify no investigation just leads to trouble. (Trust me, I know.)

If you don't investigate, you can't know it's not your device. If you don't investigate, how do you know there was a previous investigation. To my mind, the industry has this preconceived/inflexible/limited/myopic notion of what it means to "investigate." It may come from a regulation, standard, or guidance that has a preconceived/inflexible/limited/myopic notion of what it means to "investigate."

As for not investigating because of a previous investigation, sounds like Tidge is on the trail of a situation in which this might make some sense (except I think the math is wrong?), but I've seen big trouble here, too. A previous investigation of what? The "same' complaint? How do you know it is really the same if you don't investigate?

Rant over.
 

Tidge

Trusted Information Resource
#9
If you don't investigate, you can't know it's not your device.
Part of complaint intake always includes a description of the device itself. If the device is literally not something that you are not responsible for, there is no need to investigate. Believe me, it is possible to receive complaints for a competitors device.
As for not investigating because of a previous investigation, sounds like Tidge is on the trail of a situation in which this might make some sense (except I think the math is wrong?), but I've seen big trouble here, too. A previous investigation of what? The "same' complaint? How do you know it is really the same if you don't investigate?
There are a spectrum of possibilities why a previous investigation might not make a new investigation unnecessary. For example: there may be an active recall addressing the complaint.
 

Mark Meer

Trusted Information Resource
#10
I don't care what any regulation, standard, or guidance says. Every complaint warrants investigation. The idea you can justify no investigation just leads to trouble. (Trust me, I know.)
...To my mind, the industry has this preconceived/inflexible/limited/myopic notion of what it means to "investigate."...
Agree that there is some greatly needed clarification as to what is meant by "investigate", because depending on your defined scope, I may agree or disagree with your initial statement that "every complaint warrants investigation".

If it's just a matter of assessing risk, previous/repeat incidents, MDR, and whether some pre-defined further investigation/corrective action threshold is met, then I agree every complaint should be "investigated".

However, if you mean that a root-cause investigation should be performed, I disagree, as the resources necessary would make this totally impractical and unnecessary to perform for every single complaint.
 
Thread starter Similar threads Forum Replies Date
ebrahim QMS as per ISO 13485, Clause 4.2 Requirements for regulatory purposes for Medical Devices Authorized Representatives. ISO 13485:2016 - Medical Device Quality Management Systems 3
P Should eIFU link per ISO 15223-1:2016 be added to labels out of scope of Reg 207/2012? EU Medical Device Regulations 1
S Bend allowance formula to calculate per degree per radii Manufacturing and Related Processes 2
R Risk control measures as per ISO 14971 ISO 14971 - Medical Device Risk Management 6
D Is PMCF really a continuous activity per Annex XIV,Part B? EU Medical Device Regulations 5
K New supplier audit as per V3.1 by French Automotive OEM General Auditing Discussions 2
S Fogging test as per DIN 75201(50ug/g ) VDA Standards - Germany's Automotive Standards 3
Sravan Manchikanti Software Risk Management & probability of occurrence as per IEC 62304 IEC 62304 - Medical Device Software Life Cycle Processes 8
K Applicability of eIFU as per EU MDR 2017/745 Annex 1 23.1 CE Marking (Conformité Européene) / CB Scheme 0
Dean Bell Implementation of Controls as per SOA for Stage 2 Audit IEC 27001 - Information Security Management Systems (ISMS) 0
M Case study help as per ISO 9001: 2015 ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 18
M Case study solution help required as per ISO 9001 : 2015 ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 1
R Declaration of conformity as per EU MDR (2017/745) CE Marking (Conformité Européene) / CB Scheme 5
K %GRR was between 10-30% so we have to have a "backup plan" per auditor IATF 16949 - Automotive Quality Systems Standard 15
K What is mean by Oxygen Rich Environment as per the IEC 60601-1 clause no 11.2.2 IEC 60601 - Medical Electrical Equipment Safety Standards Series 5
K Dielectric strength test as per IEC 60601-1 -Infant incubator IEC 60601 - Medical Electrical Equipment Safety Standards Series 2
B Accessible parts per 8.4.2 b) - Touch currents in combination with MOOP or MOPP? IEC 60601 - Medical Electrical Equipment Safety Standards Series 3
K Diagnostic X-ray devices - Applicability of Biocompatibility Testing per ISO 10993-1 Manufacturing and Related Processes 7
CPhelan Best metrics for monitoring low volume (<1000) parts per vendors Supply Chain Security Management Systems 1
D Supplied Product on Control Plan - Lot ranging in 40,000 pcs per product FMEA and Control Plans 3
K EIFU indicator symbol - Class I labels as per EU MDR EU Medical Device Regulations 3
B Determining SAT Offsets vs TUS Offsets per SAE AMS 2750E AS9100, IAQG, NADCAP and Aerospace related Standards and Requirements 0
E Equipment Qualification - IQ/OQ per ISO 13485:2016 section 7.5.6 Process validation ISO 13485:2016 - Medical Device Quality Management Systems 7
J RAMS and LCC assessment per ISO/TS 22163 Other ISO and International Standards and European Regulations 0
D Performance Qualification per GHTF Guidance Other Medical Device Related Standards 12
Nicole Desouza Sampling plan for a customer who wants AQL 1.0 (per ANSI Z1.4) AQL - Acceptable Quality Level 5
Sam Lazzara Medical Device File (MDF per 13485:2016 4.2.3) versus FDA Device Master Record (DMR) ISO 13485:2016 - Medical Device Quality Management Systems 3
A Glass and brittle plastic Policy - Required per FSSC 220002-4? Food Safety - ISO 22000, HACCP (21 CFR 120) 0
B Is it necessary to conduct product audit of entire product which is manufactured in the company as per IATF IATF 16949 - Automotive Quality Systems Standard 4
B More than one Risk Report per Medical Device ISO 14971 - Medical Device Risk Management 3
M Aluminum - What is the difference between 6061-T6 and 6061-T651 (both per ASTM B211)? Manufacturing and Related Processes 4
Sidney Vianna 40 miles per gallon...on an airplane?!?! 65,000 ft altitude....with a propeller? Meet the Celera 500L Coffee Break and Water Cooler Discussions 1
V Thinning percentage in sheet metal forming per SES (Suzuki Engineering Standard) Manufacturing and Related Processes 1
R Where to get information - M6 per WA900 Manufacturing and Related Processes 1
R The difference b/w FMEA & Risk analysis as per iso 14971 ISO 14971 - Medical Device Risk Management 8
B Stability testing of medical devices as per ASTM F1980-16 CE Marking (Conformité Européene) / CB Scheme 0
N Per the FDA regulations can a medical device be class I despite that it is sterile? Other Medical Device Related Standards 4
S Are Defibrillator Protection and Pacemaker pulse detection features mandatory as per IEC 60601-2-25 or 2-27...? IEC 60601 - Medical Electrical Equipment Safety Standards Series 11
PastorBee13 IATF 16949 Audit Nonconformance Responses per 5.1.1.2 - Format IATF 16949 - Automotive Quality Systems Standard 11
Roberticus Negative corner radius (per ISO 1101)? Inspection, Prints (Drawings), Testing, Sampling and Related Topics 1
T Risk Management Report as per MDR Requirements EU Medical Device Regulations 4
supadrai Just found out the hard way that only one registration per manufacturer is allowed in Yemen, regardless of OEM or private label arrangement. Other Medical Device Regulations World-Wide 0
S The procedure for replicate check in calibration lab as per ISO 17025 General Measurement Device and Calibration Topics 2
T Design responsibility for customized product per customer requirements Design and Development of Products and Processes 4
S Testing my ECG device as per IEC 60601-2-25 IEC 60601 - Medical Electrical Equipment Safety Standards Series 2
R CNC Software Validation requirements as per ISO 13485:2016 Other ISO and International Standards and European Regulations 8
C AS9100 rev D 8.5.1 c 2 - Defining the Machine in-process frequency per ANSI/ASQ Inspection, Prints (Drawings), Testing, Sampling and Related Topics 8
R Remote Sites per IATF 16949 IATF 16949 - Automotive Quality Systems Standard 1
Roberticus Reinstatement of Revoked Certification for NDT per NAS 410 8.4 AS9100, IAQG, NADCAP and Aerospace related Standards and Requirements 2
R Brexit - European Commission - The impact of the Brexit per March 29, 2019 EU Medical Device Regulations 3

Similar threads

Top Bottom