Human Factors in the GMP Inspection Process
Gregory W. O'Connell
Office of Compliance
Center for Devices and Radiological Health
Food and Drug Administration
Regulatory Authority for Design Controls
Congress, through the Safe Medical Devices Act (SMDA) of 1990, gave the Food and Drug Administration (FDA) the authority to impose good manufacturing practice (GMP) controls necessary to ensure proper device design. However, Congress made it a point to state that GMP controls did not extend to an evaluation of device safety and effectiveness. Congress made it clear that safety and effectiveness would be determined by the Office of Device Evaluation through its device review process.
Why did Congress feel it was necessary to amend the Medical Device Amendments of 1976? Between 1985 and 1989, FDA compiled data through its recall database and worked with the Office of Inspector General to generate a report that demonstrated that 45% to 50% of all device recalls stemmed from poor product design. (That included recalls because of problems with software.) Congress was alarmed at this information and asked FDA what was needed as an enforcement mechanism to address the issue. Our response was that design controls or enhanced GMP controls were needed. So Congress passed SMDA-90, which explicitly authorized and required FDA to impose design controls as part of the GMP. To address Congress's mandate, FDA launched its proposed revisions to the current GMP regulation in 1993. The revisions included a section that specifically dealt with design controls section 820.30 (of Title 21, Code of Federal Regulations).
FDA received a large volume of comments regarding the revisions, specifically the design controls. As a result, during the past 2 years, we have tried to take those concerns into consideration in revising the original proposal. In July 1995, the latest revision, the Working Draft of the Current Good Manufacturing Practice (CGMP) Final Rule, also entitled the Quality Systems Regulation, was launched. It is expected that the final rule will be released in 1996.
Who Will be Required to Comply With 21 CFR 820.30?
As currently written, manufacturers of all Class II or III medical devices, manufacturers of Class I automated devices with computer software, and manufacturers of the following Class I devices will be required to adhere to the design control section of the GMP.
|868.6810||Catheter, Tracheobronchial Suction|
|892.5650||System, Applicator, Radionuclide, Manual|
|892.5750||Sources, Radionuclide Teletherapy|
Source: Working Draft of the Current Good Manufacturing Practice (CGMP) Final Rule, page 99, July 1995.
What Will 21 CFR 820.30 Require Manufacturers To Do?
Generally speaking, it will require them to establish and maintain procedures to control and verify device design to ensure that design requirements are met. More specifically, it will require manufacturers to establish and maintain plans that describe or reference the design and development activities and indicate responsibility for their implementation. It will further require manufacturers to establish and maintain procedures to ensure that design requirements relating to a particular device are appropriate and address the intended use of the device, including the needs of users and patients. It will also require manufacturers to establish and maintain procedures to ensure that design output meets the design input requirements.
Further, Section 21 CFR 820.30 will require manufacturers to establish and maintain procedures to ensure that a formal, documented review process is conducted at appropriate stages during the design development process. Manufacturers will be required to establish and maintain procedures for verifying and validating device design. Procedures to ensure that the design basis for the device and its components are correctly translated into approved product specifications must also be established and maintained. Section 21 CFR 820.30 will require manufacturers to establish and maintain procedures for the identification, documentation, verification or validation where appropriate, review, and approval of design changes. Lastly, Section 21 CFR 820.30 will require that the manufacturer establish and maintain a design history file which either contains the required information or references where the information can be located
From FDA's perspective, where do human factors fit into the product design process? Human factors evaluation is most important at the design input stage. If you don't pay enough attention to human factors at this early stage, you will have more design changes down the line, which will increase your costs and time to market, and also increase the opportunity for error. So when you define your performance characteristics, safety and reliability requirements, environmental requirements, physical characteristics, labeling requirements, and so forth, you also need to conduct user interviews or usability testing and other tests and analyses to identify and formulate what your human factors objectives will be for that device.
One approach that has been employed successfully for a number of years in the automotive, aerospace, and nuclear industries, incorporates methods for human engineering evaluation of system and component design similar to those described in the Human Factors Design Handbook by Wesley E. Woodson. These methods include:
- Worker and consumer interviews. This is analogous to usability testing;
- Human-machine operational observation. The design team observes the user in the user setting/environment;
- Personal operation experience. When practical, this gives the designer a hands-on feeling for the issues;
- Time and motion studies. These are also important, although they are more typical of the automotive and aerospace industry;
- Preliminary design review;
- Mockup evaluations, whether static or dynamic. They should be full-scale and three-dimensional;
- Other methods. Simulations, prototype testing, field evaluation or trials, and human factors experiments can be employed.
In addition, there are guidelines specifically tailored to the medical device industry. AAMI has developed ANSI/AAMI HE48-1993, Human factors engineering guidelines and preferred practices for the design of medical instrumentation, which is being considered for adaption as an International Organization for Standardization (ISO) standard, with addendums based on comments from the European community.
What Can Industry Expect During GMP Quality Systems Inspections?
Essentially, investigators will look at the overall device design development process to determine whether human factors were considered. If they were not, is there adequate justification for excluding human factors from the process? If human factors were considered part of the process, investigators will look at your design review process and procedures to ensure that the device output equals the device input with respect to human factors objectives.
In addition, investigators will examine the design review process to ensure that when design modifications were made, first and foremost, they were documented; then an assessment will be made about their impact on human factors objectives. Lastly, we will look at the design input and design verification and/or validation procedures to identify the methods that you as a manufacturer have employed for human factors evaluation in the device design development process.
What tools or guidances might investigators draw upon to aid them in their investigation? The following are several sources:
- Guidances that the Office of Device Evaluation develops as it increases its consideration of human factors' review of device submissions;
- Industry and academia publications that identify specific human factors issues in product lines or types;
- Recognized sources, such as AAMI and ISO guidelines, to determine if the firm's approach is consistent with standard practices.
Undoubtedly, there will be situations in which manufacturers will choose to employ an alternative method to demonstrate that they have addressed human factors in their design development process. In theory that will be fine, but what we will be looking for during an investigation is whether or not the rationale for employing that alternative approach is documented, and whether or not the alternative approach will produce results that are as meaningful and reliable as those one would expect from the recognized methodologies.
When will a human factors deficiency observation appear on Form FDA 483, Inspectional Observation? Several scenarios are described below; as we become involved in the evaluation process, more scenarios will arise. One possible case is when no consideration of human factors has been included in the design procedures and no justification has been provided. I consider this the "ignorance" scenario, for lack of a better term.
Another plausible situation, which will probably occur more often, is when consideration of human factors evaluation is provided in the procedures, but our investigator concludes that the procedures are inadequate. The procedures may be deficient because a generally recognized and accepted method of evaluation is not included, nor is there a rationale provided for why certain human factors issues are not addressed. This case might occur when an alternative group of methods are employed to address human factors.
The final scenario that we envisioned is when a firm establishes procedures for evaluating human factors but does not follow them. This case involves a deviation from standard operating procedures. A twist on that would be when, in addition to the deviation from standard operating procedures, there is no or is inadequate documentation to demonstrate that human factors design requirements were met.
Investigators will not assess the appropriateness of the human factors specifications themselves as part of their review. We believe that the ODE review process addresses that concern. However, if an investigator discovers a situation that in his judgment needs to be addressed by ODE (e.g. changes made to the device or manufacturing process that impact the original human factors objectives for the device), that information will be documented and forwarded to ODE. ODE will determine if there is a problem. I want to emphasize that investigators are not charged with evaluating device safety and effectiveness with regard to design controls, including human factors evaluation. Congress was clear that this is ODE's responsibility.
Undoubtedly, there will be a learning curve involved with the implementation of human factors in the design controls requirement. We will take what steps we can to provide our investigators with all available guidelines to make the inspectional process consistent, and to standardize, if you will, the degree of subjectivity that investigators will exercise. We envision that the development of those guidelines will involve not only FDA resources, but outside human factors specialists and industry input as well.
We believe that device safety and effectiveness will be improved by including design controls, specifically human factors consideration, in the GMPs. Initially, both industry and FDA will experience frustrations with the implementation of the design control requirements. There will be an increased work load. For some, particularly small manufacturers, there may be increased costs. In the end, though, FDA believes that it will be able to demonstrate to industry the benefits to device safety and effectiveness, to their bottom line, and to their ability to get devices to market quickly. The benefits should outweigh the initial difficulties.