1. Top 10 Medical Device Citations
Regina A. Barrell
Food and Drug Administration
Office of Regulatory Affairs
Office of Enforcement and Import Operations
Division of Enforcement
Good Afternoon!
Source of information: Turbo software system
Format for talk:
- Describe inspectional approach
- Brief description of the top 10 citations for
- Show how top 10 have changed over the past years
- Examples of language for the top 5 cites
- Overview of all enforcement statistics for 2011

2. QSIT Approach to Inspections
Level 1 – Abbreviated –
2 subsystems;
Corrective and Preventive Actions (CAPA) plus Production and Process Controls (P & PC) or Design Controls.
Level 1 inspections are used for routine surveillance and initial inspections of all firms, other than Class III manufacturers. Initial inspections of Class II manufacturers are typically conducted used Level 2 inspectional guidance. Primarily for firms that have a previously compliant history: The last inspection did not result in any type of administrative or regulatory action.

3. QSIT Approach to Inspections
Level 2 – Comprehensive –
4 major subsystems;
Management Controls, Design Controls, CAPA and P & PC.
Level 2 Inspections are more complex or in-depth than Level 1:
For all initial inspections of Class III device manufacturers and, where possible, Class II device manufacturers
By assignment
For foreign inspections
For training
For Accredited Persons audits (PAC 82845P)
When an inspection, which started out as Level 1, reveals post market information and/or objectionable conditions which cannot be adequately assessed as a Level 1 inspection.  (Before converting to this more comprehensive level, district management should be informed.)
Where district work plan resources permit (Level 2 should be considered for any inspections of Class II and Class III device manufacturers.  The decision to use Level 2 inspections should be based on risk.)
The Level 2 QSIT approach was validated using the following inspectional sequence:  Management Controls, Design Controls, CAPA and P&PC.  This inspectional sequence allows the investigator to review design control issues and how the device specifications were established before reviewing the CAPA subsystem. 

4. QSIT Approach to Inspections
Level 3 – Compliance Follow-Up –
Special – For Cause –
Special – Risk Based Work Plan -
As directed by inspectional guidance and elements of QSIT.
Level 3 inspections are Compliance Follow-up Inspections.
Level 3 inspections are necessary after a firm is found to have Situation I conditions during a previous QS inspection which was classified Official Action Indicated (OAI).    
Level 3 inspections will also be performed when directed by assignment.
The QSIT Guide should be used for guidance, but the inspectional guidance provided by the assignment, the district compliance branch, and/or CDRH will guide the flow of the inspection.
During domestic Level 3 inspections:
(A) Verify that adequate correction(s) and corrective action(s) have been implemented to the quality system problems previously identified; and
(B) If the correction(s) and corrective action(s) were not implemented or were not implemented effectively, verify that the deficiencies continue to exist and provide adequate evidence to support a possible regulatory action. 
(C) Document any additional quality system problems observed during the inspection, and provide adequate evidence to support a possible regulatory action. 

5. QSIT Approach to Inspections
Compliance Follow-Up, For Cause and Risk Based Work Plan are dictated by the previous FDA-483 findings and other regulatory information and may differ from the typical QSIT approach.
Some program areas are considered satellites to the four major quality management system subsystems (Management Controls, Design Controls, CAPA, and P&PC):
CAPA Satellites:
MDR
Corrections & Removals
Tracking
Thus, MDR, Corrections and Removals, and tracking requirements (where applicable) should be covered when covering the CAPA subsystem.
So … why did I bring up the types of inspections to start my talk?
One thing that should be clear from the QSIT approach is the importance we’ve placed on CAPA and how it is a major factor in developing the inspectional approach. This emphasis can be seen in the 483 citations revealed in the following slides.

6. 2012 820.100(a) – Corrective and Preventive Action (Procedures)
(a) – Complaint procedures
– Written MDR procedures
(b) – Corrective and Preventive Action (Documentation)
– Purchasing Controls
These citations were taken from our Turbo database, the database that all inspectional results/reports are entered. Here you see the top five most frequently cited 483 items. Inspectional guidance with regards to how inspections are conducted require our investigators under the Compliance Program, as seen in the last 4 slides, to cover CAPA plus 1 major subsystem (P & PC or Design Controls) for Level 1 (Abbreviated inspections) or all 4 of the major subsystems for a Comprehensive or Level 2 inspection. MDR, Corrections and Removals, and Tracking (where applicable) should be covered when covering the CAPA subsystem. So, the association of why we typically see CAPA and MDR procedures appear in the top 5 citations may in fact be related to the emphasis of our inspectional approach.
These are very similar to what we have seen historically in the past – which we’ll review in just a minute
The second citation – (a) deals with the establishment and maintenance of complaint procedures. Although found under the Records Subpart, complaint procedures go hand-in-hand with the CAPA system.
The third citation – deals with developing, maintaining and implementing written MDR procedures.
The fourth citation – (b) again deals with CAPA, however this time the issue is the lack of documentation of actions taken and not issues with the procedures per se.
Finally the fifth citation deals with control of purchased products or services had not normally shown up in the top five 483 citations when we look at historical data trends. In the past, it had been down a few slots. We are seeing more and more that this is an area that some firms have not really tightened up with regards to their suppliers. We’ll discuss a little more the importance that purchasing controls play in quality systems, especially as they relate to recalls.

7. 2012 820.75(a) – Process validation
820.90(a) – Control of non-conforming product
(i) – Design change procedures
(c) – Complaint investigation
– Device master record not maintained
Citations 6 through 10 reveal a little more variability than the first five items.
#6 – (a) deals with Process Validation
#7 – (a) deals with non-conforming product
#8 – (i) deals with Design Change Procedures
#9 – (c) relates to the failure to investigate complaints .
#10 – is failure to maintain the device master record.

8. Total Turbo Citations vs. 2012 Citations
803.17
(a) 3
(b) 4 5
820.75(a)
820.50
As we normally have seen over the last few years, the top 4 -5 citations have been fairly consistent.
Here I’ve compared the 2012 citations against the cumulative history of citations – roughly from 1998 – What’s striking here is that the top 4 citations are fairly consistent. You can see some change in position for MDR procedures (803.17) vs. Complaint procedures ( (a)), but overall the rest are pretty stable. Again, (a) – CAPA procedures is #1, whereas documentation of CAPA activities is #4, and process validation activities and documentation, (a) – has historically been #5. One new difference for 2012 is the appearance of Purchasing Controls – for the first time in the top five. We’ll discuss this requirement further latter on, especially in the context of device recalls.

9. Total Turbo Citations vs. 2012 Citations6
820.22
820.75(a) 7
820.90(a) 8
820.30(i) 9
820.50
(c) 10
820.30(a)
This graph of the bottom five observations really shows some changes and perhaps a change in the emphasis regarding how inspections are conducted and what our investigators are looking at over this time period.
#6 – Quality Audit (performance are inadequate or not performed); (a)Process Validation activities and documentation
#7 – Quality Audit (procedures have not been established or are inadequate); (a) procedures not (adequately) established (non-conforming product)
# (i) Design change (procedures); both the same
#9 – Purchasing controls; (c) complaint failure investigation
#10 –820.30(a) Design control (procedures); Device Master Record

10. TOP FIVE FDA 483 CITATIONS FOR MEDICAL DEVICE FIRMS
Number One
21 CFR (a)
Procedures for corrective and preventive action have not been [adequately] established.
The next few slides represent the types of citations you may see as written on the 483. I’ve included the first five since these are fairly consistent over all the time period from
The first slide here is the stock language that is generated by Turbo
“Procedures for corrective and preventive action have not been [adequately] established.”
The investigator has the option of using the word “adequately” if the firm has procedures but they are lacking in certain respects. If there are no procedures at all, then the investigator would simply leave out the word “adequately”.

11. 21 CFR (a)
For example:
Failure to analyze processes, service records, quality audit reports, complaints, returned product, and other sources of quality data to identify existing and potential causes of nonconforming product;
The investigator then has to explain what about the firm’s procedures is lacking.

12. 21 CFR (a) contd.
Cause of nonconformities relating to product, processes and the quality system are not investigated;
Actions needed to correct and prevent recurrence of nonconforming product or other quality problems are not identified.
Some additional examples you may see written for this …

13. TOP FIVE FDA 483 CITATIONS FOR MEDICAL DEVICE FIRMS
Number Two
21 CFR (a)
Procedures for receiving, reviewing, and evaluating complaints by a formally designated unit have not been [adequately] established.
This citation is used where an investigator may have observed that complaints were not addressed readily by a firm – they take an inordinate amount of time to do an investigation into the complaint and it remains open for a long period of time or there is no documentation that the complaint was ever investigated.
Finally, this also ties back to , the firm has no documentation to show that they evaluated a complaint for MDR reportablity and this would be where an investigator would cite the observation.

14. 21 CFR (a)
For example:
Complaints are not processed in a uniform or timely manner;
are not documented; or
have not been evaluated for MDR applicability.
Some of the examples you may see under (a) …

15. TOP FIVE FDA 483 CITATIONS FOR MEDICAL DEVICE FIRMS
Number Three
21 CFR
Written MDR procedures have not been [developed] [maintained] [implemented].
This one is fairly self-explanatory …

16. 21 CFR
For example:
Your firm’s Complaint Handling procedure states that complaints shall be reviewed for MDR reporting per 21 CFR 803, however there are no additional MDR procedures.
An example of the type of citation under …

17. 21 CFR contd.
For example:
Your procedure does not contain a standardized process for determining when an event meets the criteria for MDR. There are no instructions for how your firm will evaluate information about an event to make MDR reportability determinations in a timely manner.
MDR procedures need to give clear instructions which are used in evaluating events. Sometimes we see fairly generic or general descriptions which really doesn’t meet the intent of the regulation. You need to insure that the procedure is explicit enough for your people to make good decisions.

18. TOP FIVE FDA 483 CITATIONS FOR MEDICAL DEVICE FIRMS
Number Four
21 CFR (b)
Corrective and preventive action activities and/or results have not been [adequately] documented.
This citation is fairly straight forward and typically this may be what you would see on a 483, followed by examples of some instances specifically where a firm may have told the investigator they took some remedial action but there was no documented evidence of it.
We like to say, “If it’s not documented, it wasn’t done.”

19. TOP FIVE FDA 483 CITATIONS FOR MEDICAL DEVICE FIRMS
Number Five
21 CFR
Procedures to ensure that all purchased or otherwise received product and services conform to specified requirements have not been [adequately] established.
contains several subparagraphs that include the requirement to evaluate suppliers, contractors and consultants (820.50(a)) and purchasing data (820.50(b)) to assure that the incoming materials/components or services provided meet required specifications or quality. And the evaluation must be documented, as always.

20. 21 CFR
For example:
Of 11 supplier records reviewed, 4 were not evaluated or included on your firm’s approved supplier list.
Here are a few examples you can find from some 483’s….

21. 21 CFR contd.
For example:
Your supplier, which was used to fill ampoules, was disqualified due to quality issues but was left on your approved suppliers list because the supplier was still approved for other services.

22. FDA Enforcement Statistics Summary Fiscal Year 2012
Seizures 8
Injunctions 17
Warning Letters
4,882
Recall Events
4,075
Recalled Products
9,469
Debarments 20
The next section I thought I would give you a quick overview of the enforcement statistics for 2012 – this slide is for all Centers, not just CDRH.
That figure for warning letters - 4,882 – seems rather large but in reality, the majority of those warning letters were issued by our newest Center – Center for Tobacco Products. These letters were issued to retailers for sales of tobacco products to minors

23. CDRH Enforcement Statistics Summary Fiscal Year 2012
Seizures 1
Injunctions 2
Warning Letters
210
Recall Events
1,190
Recalled Products
2,475
This slide however, is just for CDRH in Shows the total number of seizures, injunctions, warning letters and recalls taken in 2012 strictly for devices.

24. CDRH Seizures By Fiscal Year
This slide shows the number of device seizures the Agency took from 2007 – Note: we are missing the data for 2009, I specifically didn’t include it on the slide as I didn’t want to give the impression that the value was zero. For information, the total number of seizures for each of these fiscal years for all Centers: FY ’07 = 6, FY ’08 = 8, FY ’09 = 8, FY ’10 = 10, FY ’11 = 15 and FY ’12 = 8

25. CDRH Injunctions By Fiscal Year
This slide shows the number of device injunctions taken for the same time period – 2007 – (Again, data for 2009 was missing, so that year was intentionally left off). Again to contrast CDRH vs. All Centers – FY ’07 = 12, FY ’08 = 5, FY ’09 = 11, FY ’10 = 17, FY ’11 = 16 and FY ’12 = 17.

26. CDRH Warning Letters By Fiscal Year
Note that the total number of warning letters has stayed fairly consistent over this time period. The number of inspections however has increased, especially with regards to foreign inspections. So this is somewhat interesting given that fact. To compare CDRH to the rest of the Centers, for FY ’12, CDER = 95, CFSAN = 335, CBER = 20, CVM = 76, CTP = 4,146.

27. Total Recall Events - CDRH Fiscal Year 2007 - 2012
This slide shows total recalled events by year. There has been upward trend in the total number of recall events from 2007 – 2011 where the number of events almost doubled over this time. The number is 2012 dipped slightly from the 2011 number but still shows a trend upward.

28. Total Recall Products - CDRH Fiscal Year 2007 - 2012
This slide shows total recalled products by year. Also note that there is a correspondingly upward trend in the total number of products affected by the recalls from 2007 – The number of products affected almost tripled over this time. However in 2012, we see that the number declined back towards what it was in 2008.

29. CDRH Recalls – All Classes Fiscal Year 2007 - 2012
This slide shows both recall events and products by year. The Events are in bright blue whereas the products appear in the grey column.
This is a combination of the last two slides.

30. Recalls: Definition of Class I
Class I recall: a situation in which there is a reasonable probability that the use of or exposure to a violative product will cause serious adverse health consequences or death.
Next we’ll look at the number of recall events and products by classification. This is the definition of a Class I recall.
Class 1 – the most serious classification

31. CDRH Class I Recalls – Fiscal Year 2007 - 2012
This slide shows the number of Class 1 recalls by events and products. Note that the number of recall events has doubled over this time period, the number of products affected has also increased but fallen back considerably in 2012 from 2011.
It is uncertain why there has been an increase in Class 1 medical device recalls over the past few years. The standard of “reasonable probability of serious health consequences or death” has remained the same during this time period. The Center is going back through the old data to understand whether the standard has been upheld consistently or perhaps 10 years ago, we were less likely to classify something as a Class 1 – but the standard has not changed at all.
75% of all recalls fall into Cardiovascular, Chemistry, General Surgery, General Hospital, Orthopedics, and Radiology medical specialty areas. Recently, the most frequently seen recalls involve radiology and cardiovascular devices.

32. Recalls: Definition of Class II
Class II recall: a situation in which use of or exposure to a violative product may cause temporary or medically reversible adverse health consequences or where the probability of serious adverse health consequences is remote.
Class II recalls – the classification whereby the majority recalls fall.

33. CDRH Class II Recalls – Fiscal Year 2007 - 2012
Again this slide shows you Class II events and products by year. Again, you see that the trend is moving upward over time. In 2012, there were roughly twice the number of recall events and recalled products as there were in 2007, although the number both dipped slightly in 2012 from 2011.

34. Recalls: Definition of Class III
Class III recall: a situation in which use of or exposure to a violative product is not likely to cause adverse health consequences.
Finally Class III – recall least likely to result in adverse health consequences.

35. CDRH Class III Recalls – Fiscal Year 2007 - 2012
Finally, Class III recall events and products by year….pretty much the number have stayed fairly steady over this time period

36. Root-Cause Categories FY 2010 – FY 2012
This chart is for Class 1 recalls - Individual root causes were aggregated for review purposes into broader categories related to:
Design (includes device, software, packaging, labeling, process or component design)
Change Control (includes component, labeling, vendor, process, packaging, software, or finished device change control)
Process control (includes process, packaging process or reprocessing controls)
Material/Component (includes nonconforming material, contamination, mix-up, removal or release of material prior to testing) and
Packaging/Labeling (includes labeling mix-up, expired dating, labeling false and misleading, or error in labeling)
49% of recalls are related to design control – firms need to go back to and see what they can do better in the quality system design controls to better prevent these design control-related recalls. To prevent recalls, need better design validation and better testing up front. Additionally, there has been recent uptick with software-related recalls. Firms struggle with releasing high-quality software. Again, perhaps there needs to be better testing or better validation, and maybe more involvement with clinical people upfront in the design of the software so you get the correct algorithms. The increase in component-related recalls could be due to supplier control issues not being as rigorous as they need to be. Again, going back to our top five 483 citations, we saw that this year “Purchasing Controls” had moved up quite a bit in frequency over what had been seen in years past. This observation may be related to the increased number of component-related recalls.

37. QUESTIONS??