ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
Automated Visual Inspection of
Pre-filled Syringes
Wolfgang Weikmann
Director Pharmaceutical Production / Engineering
Vetter Pharma-Fertigung GmbH & Co. KG
24-25 February 2010, Tampa
Content
1. Manual Visual Inspection (MVI)
Basic requirements
2 Automated Visual Inspection (AVI) Machine
2.
overview
3. Qualification Approach
1. PQ: Pre-tests / Considerations
2. PQ 1: Knapp test (Test kit)
3 PQ 2: Knapp test (3 batches)
3.
4. Operational Approach
5. Conclusion
2
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
1. MVI Basic requirements
Enclosed dedicated room
Incandescent Lamp
(light intensity: 2000-3750 LUX)
Black & white background
Reject bin / Eject bin for measurements
Defect evaluation list (SOP)
Temperature monitoring of room
One product per room at a given time!
Inspection speed
Obvious separation and placement of
uncontrolled units, rejects and good pieces!
Scheduled Eye Breaks
2. AVI - Machine Overview
Glass Barrel Remover
(De-Trayer)
Inspection Machine
Glass Barrel Inserter
(Re-Trayer)
Good Pieces
Product for inspection
Ejection of syringes
4
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
2. AVI Machine Overview
Unit 2 (TV 2a/3a & 2b/3b)
Unit 3 (TV 4/5)
Unit 4 (TV 6/7)
Glass Barrel
Closure Part
Stopper
Missing closure part
Bent Needle Shield
Fit of closure part/Total
length
Foreign particles
Solution in rib
Dirt/abrasion
Damage/deviation
Stopper position
Upside-down (180)
Crooked stopper
Double stopper
Under- and Overfilling
Empty syringe
Dirt/Abrasion
Damage/Deviation
Scratches/Cracks zone 1
Airline over total length
Closure Part
Unit 5 (TV 8/9)
& 7 (TV 12/13)
Solution
Dirt/Abrasion
Damage/ deviation
Broken bridge
Pierced Needle Shield
Foreign particles
Unit 6 (TV 10/11)
0
OUTFEED
INFEED
Solution
Glass splinters
1
2
Unit 11 (HV)
3
4
Solution
Unit 1 (TV 1)
Leaking system
Glass Barrel
Damaged finger rest
Broken
Unit 8 (TV 14/15)
Unit 9 (CA) and 10 (OD)
Glass Barrel
Solution
Stopper
Scratches/Cracks zone 2
Discoloration, Turbidity
Missing stopper
2. AVI Machine overview
Eject 0
Eject 1
Eject 2
Eject 3
Glass Barrel
Damaged finger rest
B k
Broken
Solution
Discoloration, Turbidity
Solution
Foreign particle
Solution
Glass splinter
Under and overfilling
UnderEmpty syringe
Leaked system
Stopper
Foreign particle
Solution in rib
Dirt/ abrasion
Damage/ deviation
Stopper position
Crooked stopper
Double Stopper
Stopper
Missing stopper
OUTFEED
INFEED
1
2
3
4
Eject 4
After a machine failure or
power break down, all
remaining syringes in the
machine are ejected here.
Stopper
Upside-down (180)
Glass Barrel
Dirt/Abrasion
Damage/Deviation
Scratches/Cracks area 1
Scratches/Cracks area 2
Airline over total length
Closure Part
Dirt/Abrasion
Damage/Deviation
Broken bridge
Pierced Needle Shield
Missing closure part
Bent Needle Shield
Fit of the closure part/Total
length
Other defects
Wrong primary packaging
component
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3. Qualification / Validation Approach
Design Qualification
Installation Qualification
Operational Qualification
Performance Qualification
(DQ)
(IQ)
(OQ)
(PQ)
PQ 1: Knapp test with Test kit
PQ 2: Knapp test with 3 commercial lots
in this presentation, focus on PQ
7
3.1 PQ: Pre-tests / Considerations
Definition of Spinning speed
Definition of Inspection area / windows
Air bubbles
Evaluation of different turbidity levels
Evaluation of different colors of the
product solution
p
Detection level of foreign matter
depending on different filling levels
8
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3.1 PQ: Pre-tests / Considerations
Light intensity
Regular checks
Definition of the max. holding time
Measures to be taken in case of
decrease of light intensity special
maintenance / definition of acceptance
criteria
Handling of exchange of lighting
Burn-in time of LED spare parts to reach
acceptable stability
9
3.1 PQ: Pre-tests / Considerations
Heat-up of syringes during machine
stop Definition of max. break time
without removal of syringes
Definition of Defect categories
(Assignment of minor, Major and Critical
Creation of Batch Report)
Assignment of defects to Eject bins
Strategy for Manual Classification
(Re-inspection) of Ejects
10
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3.1 PQ: Pre-tests / Considerations
Handling / Risk assessment in case of
glass breakage at different locations
within the machine
Implementation of effective Line Cleaning
and Line Clearance
11
3.2 PQ 1: Knapp test (Test kit)
Creation of a product specific Test
kit (Rejects + Good pieces)
500 pieces in total comprising approx. 160
Rejects
Rejects selected from routine batches as
far as possible
Optional: Usage of standardized particles
for better reproducibility in case of
replacement due to destruction
12
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3.2 PQ 1: Knapp test (Test kit)
Performance of the Knapp test
MVI: 20 Visual Inspectors inspect each the
test kit (500 pieces) / routine inspection
conditions
AVI: Test kit (approx. 160 Rejects) are run
21 times in the production mode of the
AVI machine Aim: simulation of real
production conditions as listed in the
following:
13
3.2 PQ 1: Knapp test (Test kit)
AVI Efficiency Test
1. Reject Rate (RR) Rejects
1. Aim: Test of each Inspection unit per defect
(10 runs per defect)
2. AC: for each of the single defect categories:
RR AVI RR MVI
2. False Reject Rate (FRR) Good pieces
1 Aim: Test of each Inspection unit with 500
1.
good pieces
2. AC: tbd. for each inspection unit
14
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3.2 PQ 1: Knapp test (Test kit)
AVI Laboratory Test Run
1. Reject Rate Rejects
1. Testing each single Inspection unit
1. Aim: Test of each Inspection unit with
all related defects (10 runs per defect)
2. AC: RR AVI RR MVI
2 Testing all Inspection units
2.
1. Aim: Test of all Inspection units with all
defects (1 run with all defects)
2. AC: RR AVI RR MVI
15
3.2 PQ 1: Knapp test (Test kit)
3. False Reject Rate Good pieces
1. Aim: Test of all Inspection units with 2000
good pieces (1 run)
2. AC: tbd. for the whole machine
AVI Performance Run (Good pieces; at
least Placebo)
Aim: Determination of the repeatability
of detection within 3 technical runs of the
same lot ( 10.000 syringes recommended)
AC: tbd. dependent on Process Capability
16
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
3.3 PQ 2: Knapp test (3 batches)
Evaluation
M2.1 + M3.1 = RR AVI
M3.1 + M4.1 = RR MVI
Comparison
RR MVI vs. RR AVI
17
3.3 PQ 2: Knapp test (3 batches)
Acceptance criteria:
Different approaches
pp
are p
possible,, e.g.:
g
#1: RR AVI RR MVI for all defect categories,
not dependent on criticality of the defect
#2: Split into minor, Major and Critical:
Number of defects with (RR AVI RR MVI) (RR
AVI < RR MVI) per Criticality Class
18
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
4. Operational Approach
19
4. Operational Approach
2. Manual Visual
Classification
Eject 0
3. AQL
(Good Pieces of Eject 0-4)
Rejects
AQL V l
AQL-Values
according
di to
t MVI (ISO 2859)
Good Pieces (FRR)
Critical, Major, minor
Rejects
Eject 1
Eject 2
Good Pieces (FRR)
Lot size = Sum of syringes of Ejects 0-4
Rejects
Sampling randomized of the Good Pieces
(FRR) from Ejects 0-4:
Good Pieces (FRR)
n = 125 (Lot size less than 3.200)
Eject 3
Eject 4
Rejects
j
n = 200 (L
(Lott size
i b
between
t
3
3.201
201 and
d 10
10.000)
000)
Good Pieces (FRR)
n = 315 (Lot size between 10.001 and 35.000)
Rejects
Good Pieces (FRR)
Good Pieces (FRR)
20
10
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
4. Operational Approach
2. Manual Visual
Classification
Eject
j
0
Eject 1
Eject 2
Eject 3
Eject 4
3. AQL
4. Batch Evaluation
(Good Pieces of
Ejects 0-4)
Rejects
Good Pieces (FRR)
Evaluation according to AVI
Evaluation matrix
Rejects
Good Pieces (FRR)
Rejects
minor:
X limit Batch Release
X > limit Deviation/Measures
Major:
X limit Batch Release
X > limit Deviation/Measures
Critical:
X limit Batch Release
X > limit Deviation/Measures
Good Pieces (FRR)
Rejects
Good Pieces (FRR)
Rejects
Rejects (total)
Good Pieces (FRR)
21
5. Conclusion
Main topics to be considered upfront:
Product characteristics
Evaluation strategy / Assignment of
defects to Eject bins
Creation of the Test kit
Acceptable False Reject Rate
Re-Classification of Ejects
Performance of lighting
Preventive maintenance
Re-Qualification concept
22
11
ISPE Tampa Conference
22-25 February 2010
Tampa, Florida USA
Thank You!
Wolfgang Weikmann
Director Pharmaceutical Production / Engineering
g
g
Vetter Pharma-Fertigung GmbH & Co. KG
Schuetzenstrasse 87
D - 88212 Ravensburg, Germany
+49-(0)751-3700-2401
[email protected]
12
