Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 174 (2017) 686 – 692

13th Global Congress on Manufacturing and Management, GCMM 2016

Process Validation and Revalidation in Medical Device Production

Yincheng Zhaoa, Kuangjie Shengb, Zheng Wanga, Xilin Zhanga, HengyiYang a, Rui
Miaoa,*
a
School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
b
School of Mechanical Science and Engineering, Huazhong University of Science and Technology,Wuhan 430074, China

Abstract

In this paper, the author according to ISO13485:2003, YY / T 0287-2003 quality management system for medical device
regulatory requirements, and process validation guidance document GHTF-SG3-N99-10-2004, combined with the actual
implementation process in the enterprise, detailed the process and applications of process validation. Then share some the
statistical methods of data collection and analysis, describe how to do Design of Experiment experiments by Minitab software.
© 2016The
© 2017 TheAuthors.
Authors. Published
Published by by Elsevier
Elsevier Ltd. Ltd.
This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the organizing committee of the 13th Global Congress on Manufacturing and Management.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 13th Global Congress on Manufacturing and Management
Keywords: Process Validation, Minitab, Design of Experiment

1. Process Validation Introduction

The process validation is establishing by objective evidence that a process consistently produces a result or
product meeting its predetermined requirements. Since in some cases, such as some defect of product can’t be direct
verified by process inspection, but it will become manifest only after the product has been used. These processes
should be identified in advance or validation in order to ensure process capability and control all key parameters of
process.
Thus the process validation is to establish a documentary evidence that specific procedure can produce product of
meeting predefined specification and quality characteristic continuous with high confidence.

* Corresponding author. Tel.: +86-13671789106; fax: +86-21-64160192.

E-mail address: miaorui@sjtu.edu.cn

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the 13th Global Congress on Manufacturing and Management
doi:10.1016/j.proeng.2017.01.207
 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692 687

2. Implement of process validation

2.1. Process validation implementation step

Concrete steps of validation is as follows. First, the establishment of the validation team, clearly functions as a
member. Determine process validation purposes. Clear the need of process. Master the principles of the process for
determining the specific process parameters and desired output. Determine the evaluation and acceptance criteria,
test methods, test tools, statistical analysis tools.
Then write process validation protocols, implement IQ, OQ, PQ. Finally, determine ongoing process controls,
and process monitoring. When appropriate, get on revalidation. A flow chart shown in Fig. 1.

2.2. Process Validation Protocols (PVP)

Process validation protocols include a detailed scheme about how to implement Installation Qualification (IQ),
Operational Qualification (OQ), Performance Qualification (PQ), and revalidation.

Identification and
description of the process to
be confirmed

Start the process

validation

Making process
validation protocols

Installation
Qualification

Operational
Qualification

Performance
Qualification

Validation Summary
Report

Monitoring process

Revalidation decision

End

Fig.1 Process validation implementation step

688 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692

A detailed plan is the basic requirement to ensure the process is sufficient. Process validation program typically
includes the following components:
1) Objective of process validation
2) The scope of process validation including the operator, equipment, etc.
3) Responsibility and authority of personnel
4) Background and principles, including the description of the device, system, and process
5) Sample preparation programs
6) Process validation test items, test methods and acceptance criteria. Full accounting criteria of measurable and
subjective, defining which cases are substandard
7) OQ, PQ test program
8) The requirements about document output
9) Statistical methods of data collection and analysis
10) Consider maintenance equipment
11) Standard of revalidation

2.3. Installation Qualification (IQ)

Installation Qualification is to check whether the device is properly installed, the various system checking after
the equipment installed and documented work about technical data.
The contents of Installation Qualification include:
1) Installation conditions, such as: wiring, water supply, power supply, air supply, auxiliary equipment etc.
2) The installation environment, such as: clean room requirements, temperature, humidity and so on.
3) Checking supplier documents, prints, drawings, manuals and so on. Software-controlled device needs to
consider software backup and inspect installation environment;
4) Checking equipment / component operation and safety devices
5) Make sure the main technical parameters of the equipment, including software functional;
6) Analyzing the device parameters affect the degree of process. Parameter stability should be evaluated and
analyzed by status of real process. Find out the fluctuation range of key parameters, in order to confirm the
extreme conditions of process parameters in OQ.

2.4. Operational Qualification (OQ)

OQ is running various tests and documentation works to prove the device or system reaches the set requirements.
OQ stage need to determine the process parameters and its control range, such as: time, temperature, pressure, speed,
etc. Before the start of the test should be clear requirements of raw materials, personnel training requirements,
process procedures. OQ the actual implementation process can be divided into the following steps.

2.4.1. Parameters primaries, Identify key parameters, Preliminary experimental design

Identify critical process parameters can be used Design of Experiment (DOE) experimental methods.
Experimental design is a kind of statistical methods about designing and implementing experiment, which make
experiment can efficiently and economically determine the effect of a set of independent variables on the response
variables.
The following in a welding process, for example, experimental steps are as follows:
1) First, output controllable factor level table (Table 1)

Table 1. Controllable factor level table.

Factor Level Heating temperature Heating time Left push time delay
-1 240 100 20
0 250 150 45
1 260 200 70
 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692 689

In the parameter range of options use Full Factorial DOE. The center point was repeated twice, come to trial
schedule, and then follow the table pilot scheme to test, fill the results, test result data can be obtained (Table 2);

Table 2. Experimental design data table.

RunOrder Heating temperature Heating time Left push time delay Intensity
1 -1 -1 -1 36.89
2 -1 1 1 23.19
3 -1 -1 1 34.09
4 1 1 -1 73.87
5 0 0 0 38.22
6 -1 1 -1 51.53
7 1 -1 -1 22.26
8 1 -1 1 65.97
9 0 0 0 80.72
10 1 1 1 49.88

For 5 or more factors, due to the larger sample size of Full Factorial DOE, fractional factorial experiments can be
used to identify significant factors, and then to do Full Factorial DOE about significant factors.
2) Fit model
We can use Minitab software for data fitting model to obtain the following results (Table 3 and Figure 2):

Table 3. Analysis of variance of the regression model of welding strength

Source DF Seq SS Adj SS Adj MS F P
Main Effects 3 3074.67 3074.67 1024.89 13.48 0.015
2-Way Interactions 1 366.26 366.26 366.26 4.82 0.093
Curvature 1 23.15 23.15 23.15 0.30 0.611
Residual Error 4 304.21 304.21 76.05
Lack of Fit 3 302.84 302.84 100.95 74.16 0.085
Pure Error 1 1.36 1.36 1.36
Total 9 3768.28
R-Sq = 91.93% R-Sq(pred) = 41.80% R-Sq(adj) = 81.84%

Fig.2 Pareto chart of factor effect

690 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692

Fig.3 Main effects plot for pull strength

Fig.4 Interaction plot for pull strength

From the above analysis, model main effect is remarkable, P value is 0.015> 0.05, which means model is
effective as a whole. Curvature value is 0.611, no curvature. Lack of Fit value is 0.085, no lack of fit. R-Sq is close
to R-Sq (adjustment). That the model is appropriate, a significant factor is heating temperature.
If the model has a curved, it should be further response surface design (RSM) to fit.
3) Analysis of selected models
Analysis of the main effects of the model, interaction (Figure 3 and Figure 4):
As can be seen from the main effect of the figure to make the maximum breaking strength, three factors should
be taken of all high-level, and then we can get the optimal combination of parameters combined with Response
Optimizer. (FIG. 5)

Fig.5 The optimal response optimizer

 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692 691

Fig.6 The relationship between the limit of failure, process limit and tolerance limit

2.4.2. Determines the level of limit parameter

In the OQ stage, should challenge the limit parameters of process (the Worst Case), in order to ensure that even
under extreme conditions, the process output is still able to meet predetermined requirements.
Method of challenging limit parameter is as follows:
First challenge target specification limit (LSL) of response variable Y, given each significant factor makes
response Y get a new and less level value A1B1C1, implement A1B1C1, then obtain Y1 values. If Y1 more than
LSL, it indicates that this new level has not yet reached requirement. Should repeat the above steps to adjust the Y to
less level based on A1B1C1, made A2B2C2, until get AxBxCx to make the unqualified limit of Y is breakthrough
by Yx. Selected A (x-1) B (x-1) C (x-1) level as a group of worst case.
Then in the same manner, challenge the upper limit (USL) Y, gain the A (y-1) B (y-1) C (y-1). Whereby the
upper and lower limits of the parameter.

2.4.3. Optimization and Validation of key parameters

The optimal parameters of DOE should be verified by at least 30 samples and analyze its process capability.
Steps of process capability analysis is as follows:
1) Do normality test to the collecting samples test data, judged whether the sample is normality;
2) Draw control charts, judged whether the control charts are abnormal.
3) After confirming the process is in statistical control by control charts to calculate Cpk, Cpk at least greater
than 1.
Process Capability criterion (Table 4):

Table 4. Process capability criterion

Process capability Explain
Cpk>1.33 process capability is enough
1.0<Cpk<1.33 process capability is acceptable
Cpk<1.0 process capability is lack

2.4.4. Verify limit parameter

In the same manner, Do the sample validation and analysis capabilities to the resulting limit parameters. In order
to ensure that even under extreme conditions, the process output is still able to meet predetermined requirements.

2.5. Performance Qualification (PQ)

In the PQ stage, the key objective is to demonstrate the process under normal operating conditions can continue
to produce compliant products. Need to inspect the actual situation in production about the process parameters of
establishing by OQ stage, in order to verify the acceptability of the process. Continuous producing at least three
batches of samples by operators under normal production conditions. Sample making should consider the time span,
as well as the impact of equipment and personnel fluctuations as far as possible.

2.6. Validation Summary Report (VSR)

Validation Summary Report is the final conclusion includes IQ, OQ, PQ and validation activity. VSR should
outline the program and the results of process validation, and confirm the conclusion of the process. At the same
time, we need to determine the control method after the validation, such as SPC control about key parameter.
692 Yincheng Zhao et al. / Procedia Engineering 174 (2017) 686 – 692

3. Revalidation

Revalidation is to confirmed that verified status is no drift occurs. Due to the process is complicated and change
or the relevant regulations, some processes require periodic revalidation according to the characteristics of the
process. Such as the sterilization process.
Any process or product changes, including procedures, equipment, personnel, etc., should assess the impact of
the process and the need for re-confirmation, and determine the scope of the re-confirmation. For example, the
following occurs:
1) Change in the actual process that may affect quality or its validation status.3
2) Negative trend in quality indicators.
3) Change in the product design which affects the process.
4) Transfer of processes from one facility to another.
5) Change of the application of the process.
Even when the obvious change did not happen, it should be revalidation at intervals, in order to prevent small
changes can’t be aware to cumulative affect the results of the process.

4. Summary

With the improvement of living standards, the development of an aging population, the demand for medical
products will increase. How to ensure the safety and stability of medical products, is an issue for all medical device
manufacturers. Medical device product quality directly affects the safety of life of patients. Therefore, process
validation is more important to the medical device manufacture. It is not only for regulations, but also is a way to
ensure that the manufacturing process is continuing effective.

Acknowledgements

The authors gratefully acknowledge the financial support of the innovation practice program of Shanghai
(IPP12048), and the national natural science foundation, China (No. 71432006)

References

[1] Iso,13485:2003 Medical devices,Quality management Systems. 2003

[2] Ghtf, SG3/N99-10 Quality Management Systems-Process Validation Guidance, Jan. 2004
[3] D. C. Montgomery,” Design and Analysis of Experiments”, Topics in Biostatistics. 30(2000) 241-242
[4] T Pyzdek, “The Complete Guide to Six Sigma”, Tucson: Quality, 1999.
[5] C. Jin, “Statistical Process Control in Manufacturing Process”, Electronics Quality. 9(2007) 51-52
[6] D. Mast, Jeroen, Lokkerbol, Joran, “An analysis of the Six Sigma DMAIC Method from the Perspective of Problem Solving”, International
Journal of Production Economics. 139(2012) 604-614.