MD+DI Online is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


Articles from 1997 In August

Medical Packaging Validation: Complying with the Quality System Regulation and ISO 11607

An MD&DI August 1997 Column

Cover Story

Packaging validation is a total process involving the identification and control of materials and processing variables that affect the ability of a packaged device to meet its acceptance requirements. The results of validation produce several benefits. Through identifying the optimum windows for each key variable, process control is achieved, as well as the confidence in meeting the device package requirements. Equally important are the financial benefits that can be realized through reduced inspection, increased output, fewer complaints, and minimized scrap and rework. In May 1997, validation was raised as a GMP requirement with the issuance of Guidelines on General Principles of Process Validation. The new quality system regulation now specifically lists process validation requirements, and ISO 11607 provides key validation steps specific to sterile packaging. This article provides an integrated approach for complying with these standards.

Packaging validation must address three basic elements: requirements, assumptions, and capability assessments (of materials, equipment, and processes); it examines variations within a package, from package to package, and from lot to lot. Validation also examines the interactions within the handling and use system, which encompasses the manufacturing system (including sterilization), human interaction with the package, and the distribution and storage system and environment. To help ensure that the anticipated results are achieved, validation must be performed by someone with the necessary education, background, training, experience, and qualifications for each particular function. At both onset and completion, the validation program must be documented and approved.

In current terminology, there are three possible approaches to validation: prospective, concurrent, and retrospective. Prospective validation is performed before the packaged device is commercially distributed. Concurrent validation is also performed before the device is commercially distributed or packaged but assumes that the devices produced during validation will be distributed. Obviously, these two types of validation significantly overlap, because packaged devices produced during prospective validation are also typically sold at commercial release. Concurrent validation could better be defined as a validation process applied to products of limited commercial applicability, produced only once or a few times a year. The validation will continue with each production run until the requirements have been satisfied. Both prospective and concurrent validation are used for new products and existing products that undergo significant changes; these methods are also used when a manufacturing process or piece of equipment experiences a change that can affect product characteristics or quality.

Retrospective validation is performed after the packaged device has been commercially distributed; it is based on the review of data collected and maintained during production. Retrospective validation is difficult to justify because it typically requires appropriate and accurate product data, generated by qualified test methods, with the corresponding manufacturing records, procedures, and continuous monitoring of key parameters (controllable and uncontrollable). For these data to be valid, the process must be functional, as evidenced by few rejections and complaints. Retrospective validation is generally only useful for confirming continued validation of an already validated process.

Not all processes require validation. Verification can be used for processes that allow product requirements to be fully evaluated by inspection and testing. For example, 100% automated inspection of a packaging process would qualify for verification; however, if 100% inspection is not used and variation within the process prohibits full confirmation of requirements, then validation must be used.

There is much confusion over the terms verification, qualification, and validation. For the purposes of this article, we shall assume that the combined test results for a requirement provide a verification that the product or process meets those requirements within a snapshot in time. A capability assessment requires a broader analysis than a requirement. To determine the capability of a material, equipment, process, or final product to consistently meet the package requirements, a combination of verifications to the requirements over time is necessary. This combination of verifications provides a qualification. Specific examples of packaging qualifications are: materials, initial design, equipment, process (performance), and product (performance) or final design (all of these qualification processes will be discussed later). The combination of the appropriate qualifications results in validation.


The validation process begins with a validation plan, consisting of individual plans for each qualification to be performed. To appropriately address the qualification requirements, the plan must be based on a thorough understanding of the package requirements as they relate to the material; the design, design output, and other functional requirements; and the manufacturing equipment and process. In general, the plan should identify all pertinent factors.

For example, the validation plan should delineate what is and is not covered by the study. This would include a list of products or product families. For a family of products, a worst-case product should be selected to be representative of the most difficult product to manufacture; a rationale for why that product was chosen should also be given.

The validation plan should also spell out clear and concise objectives with an understanding of what constitutes a successful validation. All assumptions should be identified. A key outcome should be process control; therefore, the process capability index (Cpk) should not fall below 1.33. Plan developers should specify the references to be used.

The validation plan should describe the package design configuration to be qualified. This description should include the final product information, such as mass and fragility levels, and the product unit of sale configurations to be evaluated. Process variables should also be addressed, such as the inherent variability of the primary package materials, additives, and manufacturing materials. The document should indicate the equipment and process parameters to be monitored and controlled--including the methods of monitoring--as well as the package requirements or characteristics to be monitored. Environmental conditions should also be defined, and rationales stating why certain conditions do not require control should be given.

The validation plan should address the validation process with its elements of qualifications and verifications. A determination of the test methods to be used should be supported by the rationale for each test along with the intended means for collecting accurate and complete data. Careful consideration should be given to the appropriateness, accuracy, reliability, precision, and bias of the test methods and procedures and to the ease with which the output can be measured. All preparations, samples, tests, and test sequences to be performed should be included, along with the acceptance criteria with measurable pass/fail end points for each evaluation. Determining an appropriate sample size is critical in achieving reliable data, and the evaluation must be based on sampling plans employing a sound statistical approach. Testing should be conducted under conditions that simulate actual product use. All tests should be analyzed both individually and within the context of the full process.

The plan should also cover manufacturing and distribution methods, systems, and environments, and storage environments. The document should define the full data analysis required for each phase of validation and its integration for the full validation assessment. Finally, the validation plan should define how results will be approved and documented. Before instituting this protocol, it should undergo a design review with the appropriate approvals.


The next step in packaging validation entails the qualification of materials and the initial package design qualification. A materials qualification plan should be developed to analyze the material requirements with respect to safety, product performance, sterilization compatibility, shelf-life stability, and suitability for the intended manufacturing, handling, distribution, and storage methods. In forming material requirements, lot-to-lot variations must be considered in order to establish the minimum performance requirements. To ensure reproducibility, the variability range, sampling plans, and test methods must be established and agreed upon with the supplier. Limiting values are to be determined not only for adverse physical interactions but also chemical interactions, such as potential migration and transfer between the package and device. A fingerprint or other identification should be documented for each material. To ensure that properties are maintained, all materials should be kept under proven storage conditions or those specified by the supplier.

The product design should be qualified to the product requirements before proceeding with the development project. The design should be reviewed to input requirements, and initial testing should address end-use requirements and device protection as well as the manufacturing and distribution requirements. Design performance testing should be conducted under actual use or conditions that simulate actual use. (Applicable evaluations are described within the product qualification section.)

Initial evaluation of both the material and design prior to process qualification can save a tremendous amount of effort and time. If either the material or design does not meet the requirements, process qualification is useless.


After process equipment is designed or selected, the installation must be qualified to establish confidence that the process equipment and ancillary systems meet the established requirements and that they can provide consistent operation within limits and tolerances. Software systems must also be validated.

A separate plan should be drawn up for installation qualification. The plan should include: a formal set of requirements for the equipment or process; documentation of equipment conformance to design, specifications, blueprints, and drawings; determination of the utilities required for proper operation; verification that the equipment has been installed to specification and codes; manufacturer's guidelines and other requirements that must be met to achieve specifications and other performance criteria; identification of critical equipment characteristics and systems; determination and verification of the required safety features; requirements for calibration, maintenance, spare parts inventory, and adjustments; a short-term reliability or capability study, typically performed at nominal or optimal settings; and an analysis of the contamination potential from wear debris, manufacturing materials, and external factors.

Upon completion of the installation qualification, the equipment can be released for operational qualification. Operational qualification is the dynamic test of a piece of equipment; it verifies that the equipment will operate as intended. Operational qualification normally includes a full functional test, verification of machine operating ranges, and experiments to begin to define process ranges. Operational qualification is the first step in developing standard operating procedures (SOPs) for monitoring and control; therefore, the equipment must be fully calibrated and able to monitor key parameters. Corresponding written procedures, specifications, and schedules should be in place along with certification of all relevant monitoring, sensing, and measuring equipment. Note that calibration and measurement requirements should be assessed during all phases of validation and should include a verification after validation.

A documented procedure should be established for the routine inspection of the forming, sealing, and other closure systems; tooling; and machine settings. Procedures and schedules for preventive maintenance, adjustment, and cleaning should be established. Documents should also define and describe initial setup, startup, and operating process procedures, and include documented operator training. Also, any inherent machine variability should be identified (e.g., the temperature along the sealing die, bar, or platen).

Operational qualification begins to identify the equipment elements that affect the package; the process also serves to establish environmental control and procedures and determine the range of operation.


Process qualification is a critical step toward achieving process control. Through an understanding of the key process parameters and their resultant monitoring and control, the product requirements can consistently and reliably be achieved. To begin, a qualification plan should be developed. Contained within this plan should be a description of the process along with a flowchart. This qualification plan should include an initial identification and assessment of key process parameters and their potential interaction for each step along the flowchart. Process qualification requires rigorous testing; thus, a quality assurance plan should be included along with the rationale for the methods, testing, and sampling. The qualification plan should identify the initial or draft setup, start-up, and operating procedures and specifications, with preliminary acceptance from operational qualification. This plan should also identify the requirements for operator training, defining and describing the process operating procedures and in-process and finished goods evaluations to be performed as well as actions to be taken. The language of these procedures and training must be specific and clear to ensure that the requirements are fully understood. Before starting process qualification, ergonomics and safety should be evaluated, because these issues can result in changes to the procedures.

All key process parameters should be monitored and documented, including settings and tolerances. Process parameters include those that are controlled during production as well as those that are not controlled through equipment or procedures (e.g., the environment). Important interactions should be identified to help center the process within the optimal processing window. To aid in the initial and subsequent identification of parameters that have the most effect on the process output and their potential interaction, the following tools are recommended: design of experiments, multivariate analysis, fault tree analysis, failure mode and effects analysis (FMEA), cause-and-effect diagrams, process capability studies, and (if available) historical information.

Process qualification challenges the process limits. Upper and lower control limits must be established for all key parameters, and worst-case or challenge conditions should be identified to establish process limits sufficiently removed from failure or marginal conditions. The qualification plan should include an explanation of how the worst case was determined and, if necessary, a rationale of why certain other items were deemed unimportant. Packaging processes typically involve more than one significant parameter for each step; therefore, there can be several combinations of extreme settings. For example, several combinations of temperature and dwell time provide a wide variety of extreme sealing conditions.

A minimum of three consecutive production runs, including setups, is recommended. Individual upper and lower control limits for each individual process parameter need not be run separately. A combination of the worst-case upper and lower control limits can be used to verify process reliability. The preferred operating conditions should also be included because the relationship between upper and lower control limits is not always linear. Each setup should be a distinct production or experimental run and not a continuation of previous setups. All acceptance criteria must be met during the test or challenge. The output must be in discrete terms--not a simple pass/fail rating--and a statistical comparison should be made between each trial. Variation due to all controllable factors should be identified and eliminated or reduced. The combined effect of the separate outputs on achieving the combined input requirements for the final product should be analyzed. In all cases, the output data must equal the input requirements.

Any failures or deviations from the acceptance criteria should be recorded in an issue log. An evaluation of each deviation should be conducted to determine the root cause of the failure and identify corrective or preventive actions. All information should be documented. Corrective or preventive actions should be verified with additional test runs and, in some cases, validated. Process qualification is a key transition into manufacturing. Consequently, it should be a test of the full manufacturing process, including operating procedures. This qualification should begin as a team effort and end with a full transition into manufacturing. The results of process qualification will be an established range of acceptable values for each key process parameter and the corresponding control procedures. With the goal of process control, key output from process qualification will be control charting of the significant process output values and measurement of Cpk. The result should be a minimum of 1.33 Cpk. To achieve this degree of process control for each specific attribute, process qualification typically is not a one-pass study. As information develops, further studies become increasingly focused on understanding and controlling key parameters.


Product qualification establishes confidence through appropriate testing that the finished packaged product manufactured through a specified process meets all release requirements. A product performance evaluation and stability plan should be developed. The test packages should be produced on fully validated manufacturing lines; however, in situations in which this is not possible, they must be produced on equipment that is fully representative of the final process. If neither the prototype or the final process is used, the manufacturer assumes the burden of proof of equivalence. Package performance testing should be conducted under actual use or conditions that simulate actual use. Both shock and vibration testing of the final packaged product should be considered.

Package seals must demonstrate continuity and impermeability. Seal strength must be determined at the upper and lower control limits of the process as well as at the preferred setting. All seals must demonstrate their suitability to the package materials, intended package requirements, and means of access. Physical test methods can be employed. Peelable seals must meet criteria concerning particulate generation, splitting, or tearing for aseptic presentation.

Final package testing must be performed using the maximum sterilization exposure and tolerance level identified for the product. For example, if resterilization is part of the requirements, the package must be evaluated under this double-exposure condition. Furthermore, all evaluations on irradiated products should be performed at the maximum tolerance level permitted by the process; cobalt 60 gamma radiation cycles, for example, can allow a 15-kGy variation. To achieve a minimum dose of 25 kGy, the exposure can be as high as 40 kGy. The package must permit attainment of sterilization, aeration, if applicable, and maintenance of sterility over the intended product shelf life.

Package integrity--a function of material properties, design, seals, and device mass and geometry--must be demonstrated under the full manufacturing, distribution, handling, and storage environment for the intended shelf life of the product. Limits for these conditions must be defined by the manufacturer.

In establishing storage conditions, temperature, pressure, humidity, and exposure to light (including UV), and their maximum rate of change, should be considered. Package stability should be demostrated by real-time aging to the worst-case storage conditions for a period of time equivalent to the intended shelf life. Accelerated aging can be used in parallel to real-time aging, but a rationale should be established for the accelerated-aging conditions selected. Product introduction can be based on accelerated aging studies as long as there is a correlation to real-time aging. In addition to the overall package, the package materials must also remain within the validated limits of the performance specification.

Test results should show that the process yields acceptable output in a consistent manner; documented evidence should be available demonstrating that the test results and conclusions are correct. Likewise, final product testing must show achievement of the product input requirements, durability in the manufacturing and distribution environments, and stability during storage. Handling and distribution testing is recommended at the end of stability testing. In addition to package integrity, verification of device retention and protection within the package must be demonstrated through appropriate testing. All of this information must be included in the final validation report.


A final step of the validation procedure is the certification of the equipment, process, and product through a documented review and approval process. All certification-supporting documentation must be included within the validation report.

The validation plan serves as the final documented review and approval of the validation process. Analysis of the data will establish the variability of the process and the adequacy of the equipment and process controls. The validation report should undergo a thorough final review before acceptance. Any process changes to equipment, product, components, materials, or process that can compromise the original validation and affect the package's ability to maintain sterility, safety, or efficacy should be revalidated. Additionally, there are a wide variety of other changes that may require revalidation. These include: process deviations; unexpected deviations (e.g., more rejects, stability failure); changes to specifications and those identified in process monitoring; complaints traceable to the process and an increase in returns, scrap, and rework; changes in supplier; equipment moves; and changes in equipment, environment, the order of operations, and process-control software. Note that if the root cause of problems can be isolated or if verification can show that there is no effect on the process, revalidation may not be required. In several cases, the entire process may not require revalidation for a specific change; however, in all cases the impact of the change should be assessed relative to the full process and the product. The need for revalidation should be considered on a periodic basis. This review can also be a part of the change-control procedure.


Documentation from validation activities should be maintained in the design history file. General validation protocols can be maintained in the quality system record. Validation documentation should include: the equipment and process validated, with dates; individuals performing the validation; the dated signatures of the individuals approving the validation; monitoring and control methods and data; and review and evaluation for possible revalidation. All process and product documents must be managed under a change-control procedure requiring analysis, verification or revalidation, and change approval.


In the validation of people-dependent packaging processes, a key element is the elimination of controllable sources of variation. All equipment, materials, and components should be prequalified for the packaging operation. Where possible, the use of fixtures, holders, and special equipment should be implemented to reduce variation. Operating procedures should be developed, and the packaging operation should be outlined using written descriptions, illustrations, photographs, and samples. An important part of these procedures is operator training and qualification. Typically, training and qualification can be accomplished by a combination of observation of conformance to procedure, formal and informal testing, and inspection and testing of the packaged product. To evaluate the system, the manufacturing procedure should be challenged by having people unfamiliar with procedure perform it; if they have difficulty, the procedure should be refined and retested. Acceptance criteria (e.g., the maximum number of defects) must be identified.

Once satisfied that the packaging procedure is appropriate, a minimum of three consecutive lots, batches, or runs without direct observation is recommended to confirm that the process consistently produces a product that meets specifications. In general, people-dependent processes typically require more than three runs. The package should be thoroughly inspected or tested to determine conformance to specifications and the number of defects. The process performance must also be continuously monitored to detect drift. Operators will periodically require retraining and should definitely be retrained when drift is detected. Any process changes must be assessed as to the impact on the operator's ability to perform.


Package validation for validation's sake is worthless. The result of package validation should be full process control and the corresponding confidence in consistently achieving the package requirements. An additional benefit will be in the form of increased efficiency, cost reduction, and reduced risk. The validation plan should be reviewed both at the start and the end to determine the benefits derived, and the approach should be refined for future validations.


Federal Register, 61 FR:52601­52662 (the quality system regulation).

"Guideline on General Principles of Process Validation," Rockville, MD, FDA, May 1987.

Packaging for Terminally Sterilized Medical Devices, ISO/DIS 11607, Geneva, International Organization for Standardization, 1995.

Denis G. Dyke is vice president, quality and regulatory affairs, Rexam Medical Packaging (Mundelein, IL).

Photo by Roni Ramos

Copyright ©1997 Medical Device & Diagnostic Industry

The Partial Art of Salary Prediction

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column

The salary estimator should not be used to determine real salaries or salary ranges. Some will no doubt ignore this warning--but they do so at their own risk.

Of all the articles we publish each year, few are more popular than our annual salary survey review. I learned this early in my career at MD&DI when I decided to publish the survey results every other year instead of annually. By the next year, the number of phone calls I'd received asking when the salary survey would appear had shown me my error. We've published it every year since then.

Nothing so popular can be without controversy, however. This seems to be especially true of the salary estimator that we include with each article (as well as on the Web in an interactive format at

The idea behind the salary estimator is simple, even if the terminology isn't. By means of multiple-regression analysis, our survey company, Readex, is able to identify a number of key variables that help determine an individual's salary. The resulting worksheet allows readers to answer a few questions and calculate a predicted salary based on those variables.

The worksheet is a lot of fun to play with--and that's the main point. We don't intend it to be used by human resources departments or anyone else to determine real salaries or salary ranges. Some will no doubt ignore this warning--but they do so at their own risk.

The caveats supplied in the article are worth repeating here. The salary prediction model is described as "moderately powerful," and the emphasis should be on the word moderately. The model explains about half of the variation in salaries for the people we surveyed. This is so because the survey cannot accurately measure all of the significant factors that affect salaries, most notably individual job performance. As a result, about a third of the time, the model's prediction of an individual's salary will be off by 27% or more. So if you find that your salary is less than predicted, please don't leap to the conclusion that you are in fact undercompensated.

The worksheet isn't just a toy, however. As noted in the article, the model "provides a sense of the relative importance of each factor in predicting salary." This year, for instance, it tells you that the size of the company you work for is likely to affect your salary, and that if you work in quality assurance, your contributions may well be less valued than those of other job functions.

The fact that we're sharing this information should not be construed to mean we are promoting them or that we necessarily approve of them. It's just what the numbers tell us.

Occasionally our readers don't realize this fact, and react with dismay or outrage at what the model says. In last year's worksheet, for instance, gender was listed as a factor determining salary. Noting this fact, one reader wrote to complain that we were promoting this inequity by including it in the model. Instead, the reader said, we should have simply omitted it from the calculation.

If we were presenting the model as a tool for determining salary, the reader would be absolutely right. But that is not at all the case. The purpose of the regression analysis is to identify the factors that contribute to salary, whether justified or not. To hide any factor could cover up a serious industrywide problem. If race had been listed in the worksheet, the reader said, there would have been a public outcry. As it happens, our survey does ask about race. And rest assured that if race were a factor, we would not cover up that fact.

This year, by the way, our analysis did not identify gender as a significant factor in determining salaries. I think its absence from the model this time would be less notable had we hidden it last year.

In any event, I encourage you to try out the worksheet--and, of course, to use it wisely.

John Bethune

[email protected]

Copyright ©1997 Medical Device & Diagnostic Industry

Specialty Distributors Promise Sales and Efficient Distribution

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column

To the Editor:

I applaud your focus on distribution in the May 1997 issue with the feature, "The Supply Chain: Operational Efficiency Is No Longer Just an Option." The importance of this often-overlooked function--getting products from manufacturers to end-users--can't be stated enough.

The article spoke primarily to manufacturers of commodity items, whose primary concern is moving large quantities of product to hospital customers. This begs the question: Who is actually selling end-users on the benefits of these products? Who is actually standing next to clinicians and showing them how to use these devices?

While I couldn't agree more that the entire supply chain must become more efficient, I also believe that small manufacturers and those with innovative, noncommodity devices are finding it more difficult than ever to get their products in front of end-users in the first place. That's what specialty sales and marketing organizations, such as the members of the Independent Medical Distributors Association, help manufacturers do.

IMDA members also offer excellent distribution capabilities such as 98% fill rates and next-day delivery to end-users. And for manufacturers, we perform many of the same functions as the third-party logistics companies mentioned in the article: transportation, warehousing, storage, processing, inventory management, and product fulfillment.

IMDA members are also outsourcing specialists. First and foremost, we are sales and marketing organizations. Our representatives spend every day with clinical and economic customers. Second, we are distribution experts. We accept the burden of stocking products, filling orders, providing customer service, and managing accounts receivable.

Efficiency is a must; our industry has to move forward on this score. But so is exciting, innovative medical technology. The only people bringing it to end-users are specialty sales and marketing organizations.

Steven Picheny
Stepic Medical
Long Island City, NY, and
Chairman of the Board
Independent Medical Distributors Association

Copyright ©1997 Medical Device & Diagnostic Industry

Canada's Medical Device Industry Faces Cost Pressures, Regulatory Reform

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


An interview with Kevin Murray, director of regulatory
affairs and communications at Medical Devices
Canada (MEDEC).

Estimated at $2.8 billion, Canada's medical device market is a small but significant part of the $120-billion worldwide market. And in Canada, major changes in medical device regulation are on the way. These changes, which will increase the effort and expense for bringing products to market, come at a time when manufacturers are already feeling the pressure of a tightening market. As the Canadian government looks to save money in many areas of its health-care system, medical device manufacturers are being called upon to produce less-expensive products. The new medical device regulation, which will take effect in February 1998, will make gaining product acceptance much more difficult. The time and effort required by the new regulation, not to mention user fees, will add to a manufacturer's difficulties.

Kevin Murray, the director of regulatory affairs and communications at the MEDEC trade association, spoke with MD&DI about the state of the industry in Canada and about the major changes that are reshaping its future.

What is the state of the health-care system in Canada?

Well, Canadians have always prided themselves on having the best health-care system in the world, because our system guarantees all Canadians access to health care, regardless of economic status. But the system has been publicly funded and the funds have been running out, so we're now going through a rather painful restructuring. Health-care providers are having to find ways to work with less money, and that is causing some difficulties.

In terms of market opportunities for industry, it's obviously having an impact because the hospitals are having to cut back, and one of the ways that they are cutting back is to purchase less or to force manufacturers to provide less-expensive products. So it is also having an impact on the industry.

What is the status of the recently proposed new medical device regulation, and what changes will it bring about?

The proposed regulation was published in the Canada Gazette 1 in February. This is the formal process by which regulations are brought along in Canada. After the proposal is published, there is a formal response period. The deadline for that response period was June 1. The Therapeutic Products Directorate may make changes based on these comments. Once all the changes have been made, the proposal has to go through the Justice Department and before the parliamentary committee. Finally, the minister of health will sign it, and it will be published in the Canada Gazette 2 in February 1998, at which point it becomes law.

The new regulation proposes a risk-based classification system consisting of four classes. Manufacturers and distributors are going to have to decide to what risk class their products belong. Also, all devices will be subject to some premarket requirements. Currently in Canada, manufacturers of most devices simply have to notify regulators within 10 days of putting the devices on the market, and this attests that the devices satisfy requirements. Under the new regulation, manufacturers will not be able to put a product on the market until they have demonstrated that it satisfies the relevant requirements.

Another major change is that all devices with the exception of Class 1 devices will have to be manufactured under a quality system based on ISO 9000. Also, all devices are going to have to be registered, and all establishments operating in Canada will have to be registered.

How will the regulatory changes affect Canadian manufacturers?

The regulatory burden is going to go up. And added to that is cost recovery or user fees. In addition to the increased regulatory requirements, which are going to drive costs up, manufacturers are also going to face user fees. The fees haven't been determined yet, but we assume that they're going to be fairly substantial. So the cost of bringing a product to the Canadian market is going to go up significantly.

Are user fees a source of contention among members of industry, and what is MEDEC's position on them?

User fees are a source of contention for every manufacturer, but we recognize that they are government policy and we're not able to change the government's position on them. The government of Canada is moving more and more toward cost recovery in many different departments, not just Health Canada, the department that regulates the sale of medical devices. We have tried to get the government to drop cost recovery on medical devices and have not succeeded, so what we've done now is to recognize that cost recovery is an inevitability and to work with regulators to develop an affordable cost-recovery fee schedule.

Will the new regulation have a major effect on U.S. manufacturers who market products in Canada?

Yes. Any U.S. manufacturer who is selling to Canada now will find that the regulatory requirements are going to go up significantly. To sell in Canada, manufacturers will have to be using a quality system that is based on ISO 9000. FDA's quality system regulation is based very much on ISO 9000, and if manufacturers are already following it they may have no problem. But if they aren't, they will have some very tough economic decisions to face in terms of the cost of exporting a product to Canada. Also, U.S. manufacturers must choose Canadian distributors who can meet the new requirements.

On the plus side, Canada is working to harmonize regulations with the European Union and with the United States. A number of projects are under way, such as a mutual recognition agreement that is being considered with the EU. Canada is also engaged in joint review with FDA for some high-risk devices. Somewhere down the road, if FDA reviews are recognized in Canada and vice versa, that's going to make things easier for manufacturers.

What is MEDEC planning to do to help manufacturers deal with the upcoming changes in device regulation?

We are lobbying the federal government to make sure the interests of industry are considered. In terms of cost recovery, for example, we're working to try to lessen the impact of the fees. We want to ensure that a fee schedule will be found that will be reasonable for industry. This kind of lobbying is what we've been doing for the last two to three years, and we will continue to do that.

In terms of educational programs for manufacturers, we have had some sessions for our members and have been providing them with newsletters to keep them current on what's happening. We also plan on doing a general regulatory informational session, probably sometime in the fall, which we would open up to all members and nonmembers.

How can Canadian manufacturers remain successful?

Canadian manufacturers have a lot of opportunities in the global medical device market. Some of the emerging markets in Asia and Latin America, for example, probably represent a great opportunity for Canadian manufacturers because Canadian manufacturers do provide unique and very effective device technologies. I certainly feel that if you're a Canadian manufacturer, you've got to be exporting, because the home market is relatively small. Most Canadian manufacturers see the United States as the first foreign market, because it is right next door, but the U.S. regulatory system is going through some changes now that may present some future barriers. I would say that Canadian device makers should be looking at emerging markets.

It's getting harder to sell in this country, because health-care reform in Canada is definitely going to have an impact on companies' ability to bring products to market. It's getting harder to get products through the system. We are seeing more and more buying groups; we're seeing hospitals' purchasing departments demanding more and more of suppliers. Hospital budgets are very limited, so it's getting more difficult to sell a product. That's not to say that there aren't opportunities, but the combination of health-care reform, new regulations, and cost recovery is going to see a restructuring of the companies that are operating here and the products they are going to bring into the market.

What is your overall prognosis for Canada's medical device industry?

I don't want to paint a gloomy picture. It's going to get tough, and you know there are times when we've said on the record some very strong things about what the government is doing. But we recognize that federal regulators have a job to do. We try to represent the interests of the industry, and we act as a buffer between the interests of the industry and those of the federal government. Companies should be prepared for the changes that are coming, but it's not going to be all doom and gloom. Companies will survive; companies will do what they have to do. Regulatory reform is happening all over the world, not just in Canada. Health-care reform is happening in the United States and it's happening overseas, and the medical device industry is going to change to adapt to it.

Copyright ©1997 Medical Device & Diagnostic Industry

Time-Lapse Video Offers More Information in Less Time

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


Using time-lapse video compresses lengthy documentation, providing designers with quantitative data more quickly.

FDA's evaluation processes now include a consideration of human factors. Part of this focus is on product design procedures. Although FDA does not mandate specific procedures, it clearly wants evidence that the product design process incorporates due consideration of human factors. Addressing this key component for medical products requires study of relevant real-world tasks for which the product will be used.

Figure 1. Time-lapse video controller system.

Video documentation is essential for conducting observational research. For example, to develop a new surgical instrument, a designer should fully understand the relevant procedure by watching it, and it is helpful to see it more than once. This can be difficult if a surgical procedure lasts several hours. Video documentation enables the designer to review the procedure repeatedly within a convenient time schedule. Further, time-lapse video provides a solution for analyzing procedures that run several hours.

Time-lapse video entails taking samples rather than taping continuously. For example, a 1-second sample every 10 seconds drastically compresses the amount of video to review. When the tape is played back, the procedure unfolds like the old Disney films of blooming flowers.

One advantage of time-lapse video is that it usually requires only one videotape. Although this may seem like a small consideration, it allows the observer to set up a video camera without the need to change tapes frequently. Another advantage is that time-lapse video documentation enables designers to easily perform quantitative analyses. For example, each sample can be treated as a data point which the designer can use to determine the percentage of time in which the surgeon is using both hands or the percentage of time the surgeon's wrist is bent beyond a given number of degrees. Time-lapse video provides a manageable document that makes such data more accessible more quickly.

A time-lapse video system should contain a device that allows the designer to set both the length of the video sample and the frequency of the samples. For example, the system shown in the figure contains four pulse generators. Generator 1 creates a regular short pulse, which can be varied to alter the sample rate. This pulse stimulates generator 2, and a longer pulse, which trips a relay to turn the camera on. In the meantime, the second pulse stimulates generator 3, which can be varied to determine how long the camera will stay on. The end of the third pulse stimulates generator 4, which creates a pulse to turn the camera off. The camera remains off until generator 1 creates the next pulse to restart the cycle.

It is usually necessary to use trial and error to find the correct sample duration and sample rate for a given procedure. It is imperative to determine settings that compress the procedure to a manageable length without losing crucial pieces of information. One approach is to determine the length of the final summary video and to work back from that figure. Fifteen minutes seems to be a standard tolerance level for an average viewer. Typical settings might be a 1-second duration with four samples per minute. Such settings result in a final tape length of 4 minutes per hour of the original procedure.

A multiple-camera, split-screen system increases the amount of information, providing the designer with an additional view of a procedure. This can be particularly useful when evaluating the user's needs for a complex device or to allow the viewer to see the tasks when one camera is blocked. This advanced system also allows simultaneous close-ups and overview shots.

Stephen B. Wilcox is a partner in the Chicago-based firm Insight Product Development, and he runs Design Science, Insight's human factors and design research consulting group in Philadelphia.

Copyright ©1997 Medical Device & Diagnostic Industry

FDA Confronts Mutual Recognition Reality

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


Disgruntled FDA and EU officials couldn't reconcile differences before the scheduled May 28 signing of the mutual recognition agreement, and they face significant problems in the three-year transition ahead.

James G. Dickinson

Hailed with enthusiasm by the medical device industry at the end of May, the mutual recognition agreement (MRA) between the United States and the European Union (EU) still poses significant transitional problems, especially for an FDA trying hard not to look like a dinosaur.

One of those problems with the MRA is that FDA's officers in the compliance and enforcement group simply don't like it. Being loyal soldiers, however, these FDAers can't openly disagree with their commander in chief, at whose behest the MRA was negotiated. They can, however, quit. Unless their concerns are met by the agency during the MRA's three-year phase-in period, many probably will.

Their disagreement with the MRA hinges mainly on what they see as excessively loose wording that they fear will give too much latitude for interpretation. As originally drafted with FDA field organization input, the MRA language was much tighter. In that form, it met strong EU resistance, including an eleventh-hour formal protest from EU Commission president Jacques Santer to President Clinton claiming obstruction by unnamed U.S. regulatory agencies (assumed by everyone to mean FDA).

Under the pressure of a twice-deferred signature deadline of May 28, when Clinton would be in Europe for a much ballyhooed signing ceremony, the negotiators revised the wording to provide the wiggle room that now worries FDA enforcement staff. But even that change wasn't enough to produce an overall conclusion in time for the intended May 28 ceremony, and Clinton went on with his tour without signing the MRA.

Other disagreements were snagging the talks, mostly to do with drugs. Six subjects had separate draft MRAs, all to be embraced in an umbrella MRA--medical device approvals and inspections, pharmaceutical inspections, telecommunications equipment, recreational craft, electrical safety, and electrical compatibility. The medical device MRA was largely settled to the agreement of all parties, at least officially.

In contrast to the private views of many FDA staffers, an FDA paper issued on May 28 put the agency's public position plainly:

Under the MRA, both the U.S. and the EU may be able to save resources by relying on inspections of manufacturers conducted by the other country, thereby saving overseas travel time and expense. However, the MRA would only become operational after a three-year transition period comprised of rigorous joint activities designed to ensure that the inspections conducted by either party would be equivalent. Moreover, each party retains the right to conduct its own inspections, if needed.

For medical devices, the MRA also provides for premarket review of designated low and medium risk medical devices, beginning with the devices covered under FDA's third-party pilot program. Under the MRA, European notified bodies would participate in FDA's third-party pilot program by reviewing 510(k) applications against U.S. requirements, and submitting the review to FDA for final action.

During the three-year transition period, FDA would determine which European notified bodies have demonstrated proficiency to conduct such product reviews in a manner consistent with FDA standards. FDA would prepare written guidance on product testing requirements to help assure consistency in the review of similar applications by different notified bodies. This effort parallels a domestic program already under way to pilot 510(k) reviews conducted by designated third parties in the U.S.

For both inspections and premarket review, the paramount issue is that each party retains full control over the regulatory criteria that apply to products marketed in their country. The MRA recognizes this, and seeks to conserve resources by relying in large part on equivalent inspections or product reviews conducted in the other jurisdiction. This streamlines the process while maintaining the current high standards for health and safety of these products.

The failure of the president to sign the final MRA didn't prevent the Health Industry Manufacturers Association (HIMA) from issuing a May 28 news release applauding "a major trade agreement concluded today...that will boost U.S. economic growth and jobs by promoting greater two-way trade, increasing U.S. competitiveness in global markets, and eliminating regulatory redundancy and cost."

An official news release from the Office of the U.S. Trade Representative (USTR) was more reserved, declaring that outstanding issues had been "largely resolved" and expressing hope that "we will conclude an agreement in the next few days."

On one aspect of the less-than-conclusive MRA, both HIMA and the USTR were in complete agreement: The package would "serve to increase U.S. exports by saving manufacturers up to 10% of the cost of delivering U.S. exports to Europe," said the USTR statement, while HIMA observed that the MRA would save U.S. taxpayers millions of dollars a year by having EU third-party bodies conduct GMP inspections in place of FDA.

A third May 28 news release was less sanguine about the failure of the president to sign the MRA, while still emphasizing the economic importance of the negotiations. The European-American Business Council said it was "disappointed that negotiators failed to meet another deadline" and that it would seek "to ensure that the new deadline of 'soon' is just that."

Stripped of rhetoric, the struggle to reach general MRA consensus reflects a classic conflict between marketplace forces and governmental protectionism. Business representatives see only gains through lowering of governmental barriers to trade, especially where such barriers are seen as duplicative and excessive. Government officials such as those in FDA's enforcement division, by contrast, see only risk and endangerment of the public if the regulatory barricades being lowered are their own.

There are other, less-obvious challenges to the success of the MRA as well. Among these are the cultural and political differences between countries that the MRA must first bridge, then blend. For example, for 30 years the United States has had a tradition of government openness, primarily expressed in the Freedom of Information (FOI) Act. The act is derided by many as the "legalized industrial espionage act" (indeed, almost 80% of all FOI requests received at FDA are from or for companies researching their competitors). The FOI Act has spawned a myriad of entities and special interests that use information derived from formerly secret government documents to advance their respective causes in various forms of public dialogue.

Europe has no equivalent to the FOI Act, a deficiency that almost killed the MRA negotiations. In Europe, factory inspections by regulators and third-party auditors are treated as confidential. In the United States, however, all details other than statutorily-defined trade secrets and commercial or confidential data are publicly releasable. Under the MRA, if FDA delegates to an EU counterpart an inspection it would in the past have conducted itself, could it release the resulting report pursuant to the FOI Act? No, said the EU. The stalemate became a potential deal-breaker, but at the eleventh hour the EU relented.

That agreement, however, opened a new can of worms. Which documents are releasable, and how will U.S. requesters know to request them? In the United States, an inspection becomes public knowledge in two ways: first, FDA posts in its FOI office all warning letters (each suitably purged of trade secrets and commercial or confidential content), and second, FDA posts in the same office a daily FOI log that includes, among other things, requests for information on inspections that the requester has either heard about through the grapevine or that occurred at the requester's own facility.

Thus, FDA has a formal mechanism for advising the world about its own inspections. No such mechanism exists in EU countries. Under the MRA, which doesn't address this issue, FDA presumably would need to create a mechanism for advising the world about EU inspections done on its behalf.

FDA's enforcement community does not like the MRA for another reason. At a time when the FDA field organization's budget is reeling from unprecedented cuts, an official estimate predicts that the MRA would require as many as 50 full-time-equivalent personnel from the field organization to assess and evaluate, during a three-year transition period, the equivalency of EU and FDA inspections. That's a meaningful diversion of resources for an organization with 1100 investigators.

FDAers may wonder whether, after the three-year transition period is over, the turn-of-the-century Congress will let the agency keep the 50 investigators who will no longer be needed to do inspections in Europe. "Only one thing can save us," a long-serving senior enforcement official said, not too hopefully: "A major public health disaster in the meantime."

From the device industry's point of view, FDA's field organization may only be getting its just desserts after too many years of unnecessary officiousness, arrogance, and disruption of company operations. In any event, the concerns of the MRA naysayers do not appear to have received much consideration in the offices of FDA's highest policymakers.

Deputy commissioner for external affairs Sharon Smith Holston made no bones about her commitment to the MRA concept. She could not conceive of a future FDA, she said, that did not have an MRA with Europe. Budget realities, combined with new responsibilities assigned by Congress, were making it increasingly unlikely that FDA could do its job without MRAs.

Some FDA staffers may well wish to keep their agency in the past. But if Holston's views are any indication, FDA is not likely to turn into a regulatory dinosaur.

FDA-wide reform on the back of medical device legislation seems to be the aim of a bill filed in the House on May 22 by Joe Barton (R­TX) and Anna Eshoo (D­CA). Their bill (HR 1710) would give the agency a new mission statement; require General Accounting Office reports on its work "approving drug, device, and food products"; and authorize an FDA dispute-resolution procedure. The Center for Devices and Radiological Health (CDRH) gets attention in sections on a revised classification system, third-party review for most submissions, and limits on device tracking and postmarket surveillance. On the sponsor list are 10 of the Commerce Committee's 23 Democrats and all 28 of its Republicans.

In the Senate, Labor and Human Resources Committee chairman Jim Jeffords (R­VT) unveiled an omnibus FDA reform bill at a staff press briefing on May 28; it was filed on June 2, but despite avowals of a speedy markup, two days later it ran into opposition from Democrats led by Senator Edward Kennedy (D­MA). The consensus-driven bill covers devices, drugs, and reauthorization of the Prescription Drug User Fee Act and removes most of the provisions FDA objected to last year.

James G. Dickinson is a veteran reporter on regulatory affairs in the medical device industry.

Copyright ©1997 Medical Device & Diagnostic Industry

Micromachines Making Headway in Medical Applications

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


Fighting to switch the industry focus away from automotive applications, medical-use micromachines are introducing some amazing new technologies that are cheaper than their standard-size counterparts.

The trick to making successful--that is, practical--micromachines may well be to resist the urge to make them too small.

For the better part of two decades, engineers made micromachines ever smaller, a quest that doomed them to being laboratory curiosities, explains Henry Guckel, a micromachining pioneer. "Six years ago at a conference, I told my students that we should have copyrighted the pictures of our inventions because that was the only value of the devices they were making," says Guckel, IBM-Bascom professor of engineering and electrical and computer engineering at the University of Wisconsin (UW) in Madison. "We produced devices that looked beautiful but had absolutely no application. Today the emphasis is on making devices that have potential uses."

The majority of micromachines currently available in medical applications are designed to obtain precise pressure readings. This pressure sensor is used with catheters. Its size can be compared to the surrounding salt crystals. Photo courtesy of Lucas NovaSensor (Fremont, CA).

Micromachines are to the medical device industry what fleas are to a flea circus--imperceptible to the eye, yet capable of accomplishing extraordinary feats. Most micromachines currently used in medical practice record pressure in different parts of the body, delivering precise readings unobtrusively and inexpensively. But the potential of micromachine technology goes well beyond that. Accelerometers, which improve the operation of pacemakers, are being integrated into the medical mainstream. On the horizon are devices that will play primary roles in health-care delivery, such as pumps that deliver microquantities of drugs or that assist damaged hearts. Perhaps most intriguing are micromachines that will quickly and effectively identify infectious disease.

"Medical applications are going to grow very dramatically, particularly with the growth of home diagnostics," predicts Roger Grace, president of Roger Grace Associates (San Francisco), a high-technology marketing consulting firm. "There are a lot of people working on applications for urinalysis and blood analysis in which MEMS [microelectromechanical systems] will be a basic element."

MEMS are a subset of micromachines sculpted from silicon, as are semiconductors, and are typically fabricated in batches from 2500 to 15,000. The current flag bearers of MEMS technology are already engineering marvels. An entire device may be only a few millimeters in diameter. The ultrathin membranes etched into the silicon of pressure sensors can respond to changes as small as 0.1 psi. Electronic resistors planted in membranes can record changes in blood pressure or respiration, or in intrauterine pressure during birth. Other sensors have been placed in infusion pumps and dialysis equipment, where exact pressure measurements are needed to ensure proper flow.

Micromachine use in pacemakers is still evolving. These sensors detect patient motion and signal pacemakers to increase heart activity. Accelerometers also use electronic resistors to measure changes. Rather than being embedded in a membrane, the resistors, supported by tiny beams, recognize changes in the position of a mass.

Companies with deep pockets have bought into these evolving technologies. Sales are being driven by rock-bottom prices. Disposable blood pressure sensors are typically less than $2, compared to reusable devices that may cost $250 or more.

Some of the major companies taking advantage of this technology include Abbott, Baxter, Bird Products, Cardiac Pacemakers, Inc., CAS Medical, IVAC, Medex, Utah Medical, and Zimmer Patient Care. Their products and services include disposable blood pressure sensors, respiration and ventilation equipment integrated with MEMS sensors, pacemakers guided by microaccelerometers, infusion pumps, dialysis, and platelet phoresis.

This disposable blood pressure sensor from Lucas NovaSensor (Fremont, CA) incorporates a micromachined silicon sensor.

But today's offerings are just the beginning. Medical applications still only account for a small portion of the micromachine market. Leading the technology's commercialization is the automotive industry, which uses accelerometers to trigger the inflation of air bags and pressure sensors to relay information about oil and fuel.

The micromachine industry could grow to worldwide annual revenues of $12 billion by the year 2000, according to industry estimates.

A number of companies, such as EG&G IC Sensors (Milpitas, CA) and Lucas NovaSensor (Fremont, CA), are aggressively pursuing medical opportunities. Lucas NovaSensor is the world's largest producer of disposable blood pressure sensors, with volumes in excess of 8 million pieces a year. EG&G is not far behind.

A major challenge is to make medical products in a large enough volume to be cost-effective, says Harold Joseph, director of sales and marketing at EG&G. "The real drive has been to get the cost low, particularly for hospitals, where the benefit of micromachines is to provide a disposable part versus one that is cleaned and reused," says Joseph. To achieve the necessary volumes, EG&G is developing multiuse sensors--accelerometers, for example--that can be used for both automotive and medical applications with minor modifications.

But for EG&G and Lucas NovaSensor, pressure sensors and accelerometers are just paying the rent until the next generation of technology is born. A particularly hot area involves microfluidics, the ability to move, store, and otherwise manage small quantities of fluids. Developed at the UW in Madison and named after a German acronym for lithography, electroplating, and molding, LIGA is used to build the tiny pumps needed to move these fluids. One of the companies licensing the LIGA process is manufacturing a disposable micropump, which is being incorporated into intravenous lines for drug delivery. "It improves the metering accuracy of a drug tremendously," explains Guckel. "I especially like the product because it is a throwaway."

Micropumps are stepping stones to a number of applications. One of the most exciting is being pursued by MEMStek (Vancouver, WA), a start-up company and licensee of the UW technology. Nearing commercialization at MEMStek is a pumping system that might be used to speed up in vitro diagnostics.

John Skardon, director of marketing at MEMStek, notes that there are several common functions in all laboratory and analytical tests from electrophoresis to immunoassay. "What the functions all have in common is the ability to move milliliters and microliters of sample from point A to point B," he says. "We have attacked this particular part of the testing by creating what we think is the smallest soon-to-be commercially-available pump."

Measuring just 5 mm in diameter and 9 mm in length, including the motor, this device can be commanded to pump precise quantities, from tens of microliters to more than a milliliter, in one minute. "If you had a budget that would allow you to build three or four existing pumps into a device, you might be able to afford to use 10 of our pumps and reduce the number of valves and interconnects and in the process make the finished device operate quite a bit faster," Skardon says.

One of the most advanced near-term opportunities for microfluidics is DNA analysis. The technology's potential is evident in research at the MIT Lincoln Laboratory (Lexington, MA), where engineers are modifying capillary gel electrophoresis to take advantage of micromachines. The speed at which conventional gels can separate samples is determined by the voltage applied to the gel, according to lab investigator Albert Young, PhD. High voltages create heat that can skew the results. But microcapillaries 30 to 100 µm in diameter--about the diameter of a human hair--can disperse more heat than conventional gels.

"The heat gets extracted to the side walls very rapidly, and this allows you to drive the process much harder and get very fast separations," says Young. "So if you need to do a test quickly--for example, identify a bacterium in minutes rather than hours or days--there are a lot of advantages to making the test equipment small."

To get the DNA for this testing requires a polymerase chain reaction (PCR), wherein a DNA sample is amplified into larger quantities. Typically PCR uses carriers comprising pockets that hold reagent. These carriers cycle between two temperatures, a process called thermal cycling.

Engineers at Lucas NovaSensor have etched as many as 48 microwells into a silicon block 11 * 11 mm, with each well wired to allow precise thermal cycling. This use of micromachining requires less reagent and produces results much more quickly. "It also means potentially making these systems portable, bringing them right on-site," says Brian Wirth, marketing director at Lucas NovaSensor.

Wirth and his colleagues pride themselves on being at the leading edge of technology. Lucas NovaSensor was founded in 1985 on early research conducted at Stanford University, and the transducer lab continues to explore micromachine research. Among its projects are technologies that promise to forge the ultimate man-machine link. "The potential is to implant a mesh of microelectrodes into a nerve, let the nerve regenerate through the mesh, and then have this mesh serve as part of a closed-loop feedback mechanism," Wirth says. "It's a step toward becoming a bionic man."

Cutaway of a silicon micromachined accelerometer (EG&G, Milpitas, CA) used in pacemakers, activity/sleep monitors, and motion studies.

Such meshlike devices have the potential to precisely control prosthetic limbs. In turn, these sensors might be connected to micromachines that move the smallest surfaces on a prosthesis, making the actions of an artificial arm indistinguishable from those of a biological one. Alternatively, these devices might produce microelectrical signals to provide hearing to the deaf and sight to the blind. Experiments on frogs and rats in the Stanford Integrated Circuits Laboratory have established neural interfaces with cut nerves that have grown together, forming stable electrical links between on-chip microelectrodes and axons in peripheral and auditory nerves.

There are enormous obstacles to mak-ing these and other forward-reaching microtechnologies practical. Foremost among them are fabrication challenges--making mass quantities of devices that optimally fit the task for which they are being designed. This challenge has not been easy to meet, even for micromachines used in current and near-term applications.

Developers of these products have used a variety of etching processes, including chemicals and plasma. The etching is conducted under computer control, but silicon has an annoying habit of allowing electrochemical etching to occur only at 55° angles along the crystallographic planes of single-crystal silicon. "This limits not only the shape of the product but also the size," explains Wirth, "because if you start etching at 55° angles outward, the surface area has to get pretty big by the time you etch a thin diaphragm at the top."

Progress has been made by using advanced fabrication methods, such as deep reactive ion etching (DRIE). This technique provides greater control over the etching process than conventional methods, making deeper grooves that reduce the horizontal chip size and allow more flexible designs in silicon and custom sculpting. The end result is dramatic improvement in the capability to build microdevices. "With DRIE, you can literally sculpt structures," Wirth says.

Packaging, assembling, and testing are other challenges high on the list for improving micromachines. The massive growth of the semiconductor industry led to the development of cost-effective methods for manufacturing silicon devices. But packaging, assembling, and testing can be up to 95% of the total cost of a product whose strength in the market is its disposability, which requires low price. "The cost the customer incurs is the cost of buying, testing, and fitting the product into a package," Joseph explains.

To keep these costs low, EG&G is making products that fit into housings that have already been developed by their customers. For example, last October EG&G introduced the Model 1620 pressure sensor, which can be dropped directly into a vendor's disposable blood pressure housing.

Such drop-ins address an important part of the financial equation but not all of it. In the end, the components must be made efficiently and effectively. EG&G is adapting world-class manufacturing techniques, including C-cell manufacturing, to accomplish this goal. Adoption of this continuous-flow technique has dramatically improved yields, cut costs, and enhanced product quality, shrinking production cycle times from two weeks to eight hours on one product line and from six weeks to three days on another.

C-cells are characterized by physically grouping all the manufacturing equipment and resources necessary to produce a specific product rather than centralizing the work process to support a wide variety of different products. With this process in place, says Tom Spies, director of operations at EG&G, the company can "implement daily improvement and rapid responses to change in every product line."

This rapid response to opportunities in the manufacture of micromachines and their applications will be critical in making micromachines a significant part of the medical device industry. The products that now characterize medical applications are only the start. The future of micromachines, says Wirth, "is really up to the imagination of the design engineer."

Greg Freiherr is a contributing editor to MD&DI.

Copyright ©1997 Medical Device & Diagnostic Industry

Ten Techniques for Trimming Time to Market

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


Savvy companies can keep up with rapidly changing regulatory requirements and still streamline processes in product development.

At the same time that FDA's new quality system regulation has heightened the emphasis on regulatory requirements in the medical device community, pressure to speed up the manufacturing process is also escalating. It may seem impossible to reduce the time to market with the numerous other demands on the engineering process. However, compliance with regulatory requirements does not preclude efficient development practices.

Relying on individuals to rescue lagging projects with hours of overtime is not a workable solution. A system-level strategy is needed to streamline the engineering process and achieve reduced development schedules. By effectively addressing the following 10 aspects of product development, medical device manufacturers can improve both regulatory compliance and time to market:

  • Design control reviews.
  • Documentation detail.
  • Integrated team environment.
  • Management.
  • Personnel.
  • Process evaluation.
  • Process procedures.
  • Prototypes.
  • Test strategy.
  • Tools.

The extent to which the following strategies for dealing with these aspects apply to a given manufacturer will vary based on the specific device and personnel experience. Manufacturers may also find additional steps are necessary in certain processes.


The cost benefits and quality improvements resulting from effective review practices can be substantial, such as a tenfold reduction in errors for some development efforts.1 Although reviews are mandated by the new quality system regulation, they also benefit engineering by reducing development times. Moreover, specification reviews help identify errors early, when they are relatively inexpensive and easier to correct. Dunn has suggested that cost savings from error identification and correction can be as much as a factor of 25.2

Formal reviews also help ensure consistency and understanding of key project objectives and provide a forum for effective communication between engineering and other departments as well as among engineering subdisciplines concerned with hardware and software.


The importance of documenting compliance with design control requirements is self-evident. But some companies have instituted documentation practices that go far beyond any requirements. These manufacturers falsely equate the weight of documentation produced with product quality. Their goal instead should be to shift the emphasis from documentation quantity to documentation that contributes to product quality.

One way to help achieve this goal is to determine the documentation requirements for each project, including only procedures and specifications that apply to that particular effort. In addition, the level of documentation detail should be based on the significance of the device requirements. For example, the design and development plan might specify that programmers should formally document and review unit and integration testing for safety-related requirements but document verification less formally for functions that do not affect safety.

To reduce the workload and simplify the training of new personnel on the appropriate documentation format required for each specification, try using automated templates. Encourage the use of tables and diagrams to reduce text and to make documents easier to understand.


Historically, the emphasis in product development was on the productivity of individuals and individual engineering disciplines. Engineering departments tended to develop their own subsystems, resulting in significant problems later in the development process. It is now well established that team focus and discipline are essential to prevent such bottlenecks. To cope with the increased complexity of today's products, an integrated team is an imperative.

To ensure that customer requirements are correctly captured; that the target device can be efficiently manufactured, installed, and supported; and that process compliance and submission requirements are met, the teams must include a range of disciplines. This will minimize the risk of important requirements being forgotten in the haste to compress the development schedule. It will also satisfy FDA's expectations that such disciplines as marketing, servicing, and quality assurance are included in product development.


Numerous studies have attempted to quantify the essence of successful project management, but no silver bullet has yet been identified.3 However, a risk-driven management style that continually tracks potential problems and tries to identify solutions can be highly effective.4 Early identification of ineffective approaches to a project allows adequate time to switch to more appropriate solutions.

Paying close attention to the schedule enables management to achieve ambitious deadlines. It's more common, however, to focus instead on the technical aspects of the development process and strict adherence to defined standard operating procedures (SOPs). The result of insufficient attention to tracking and monitoring progress is often the sudden realization that the project has fallen behind schedule, sometimes irreparably.

Many design and development teams believe that all requirements contribute equally to customer satisfaction, and marketing groups tend to believe that all requirements have the same cost and schedule implications. These assumptions are not necessarily true. Changes are often readily accepted without consideration of their impact on a project. An analysis should be done to assess any changes to see what effects they have on scheduling and costs. Decisions can then be made whether to implement specific changes and extend the project schedule or to maintain the schedule and postpone the change for a subsequent release.


Most regulatory requirements focus
on the process as the primary determinant of success and do not pay enough attention to personnel. Hiring qualified personnel is one of the most significant strategies to ensure productivity. The range of productivity among design engineers is perhaps most pronounced in software development. Studies have shown that one programmer may be up to 10 times more productive than another.5 Hiring practices, therefore, should address not only the educational background of potential employees but also include in-depth interviews and testing.6

Employee development must also be addressed. Training programs should include not only regulatory compliance, but also effective design practices that lead to reduced errors and higher productivity. Identifying the practices of the most efficient engineers in the company and sharing them with others is often effective in increasing overall productivity.

Finally, motivation is an essential element of efficient productivity. Individuals are more likely to perform at their peak when the work environment and compensation systems motivate them to meet schedules and quality objectives. When overtime is mandated as a solution to address tight schedules, the desired productivity gains are often overshadowed by the losses experienced as a result of reduced motivation.


Improving procedures is frequently suggested as a panacea for shortening development schedules. Published models such as the Software Engineering Institute (SEI) capability maturity model have been used by some companies hoping to develop more-efficient practices.7 However, these models often result in greater delays and costs before any real schedule savings are realized. Making improvements to the existing process rather than substituting a new one may sometimes yield more immediate improvement.

One way to proceed is to measure the effort associated with current procedures to determine their efficiency and the quality level they produce. Perform a Pareto analysis of the phases that require the greatest amount of time and find ways to reduce the schedule through concurrent development or increased automation. The quality level produced in each stage can be determined by examining errors in verification activities and final validation testing. This information can pinpoint the phase in which the most errors are introduced and help identify techniques to prevent future mistakes.

An experienced auditor can be invaluable for assessing current practices and offering strategies for improvement based on lessons learned from other departments or companies. Audits that focus on optimizing development schedules are perhaps the quickest way to identify process improvements that can reduce time to market.

Postmortem evaluations of recently completed projects can also be helpful. Identify which activities were successful and which impeded productivity. Use this analysis to define possible strategies for improvement, such as staff training, using subcontractors for problematic components, emphasizing up-front reviews, and procuring new automated tools.


Procedures for documenting the development process receive the most attention from auditors because they are the focus of regulatory standards (FDA and ISO). Nonetheless, SOPs should satisfy engineering needs first and regulatory requirements second.

Ensure that SOPs are flexible and can be individualized for any project. Most projects are a mix of new functions and existing code and hardware designs, plus integration of off-the-shelf components and subcontractor products. Flexible strategies should ensure that these subsystems are integrated and validated in a manner that achieves the desired quality levels. This objective cannot be satisfied through strict adherence to structured SOPs that place too great an emphasis on document-driven sequential models.

SOPs should be continually updated with methods taken from the most effective development teams. Process improvement teams should also be encouraged to go beyond SOP compliance and given resources to facilitate improved productivity. Many companies do this through use of an intranet system that provides libraries of sample SOPs and guidelines, automated tools, reusable software components, sample specifications, and recommendations on vendor products. Internal procedures should encourage the sharing of any resources that can contribute to process productivity improvements.


Although development techniques based only on prototypes have historically been criticized as a source of low-quality devices, manufacturers can use prototypes effectively to help compress development time. Early prototypes can ensure that the user interface and functionality envisioned by engineering is consistent with the user needs and intended use requirements. Soliciting user input with prototypes is much more effective than providing users with a technical specification that may be difficult to read and understand.

Prototypes can also help personnel address technical questions. Early tests can discover whether additional processing capacity is required or whether certain algorithms need to be optimized. Identifying these problems early with high-level prototypes can avoid costly schedule delays during implementation.

Early prototypes can be used in clinical studies to gain data on device effectiveness. However, prototypes must be validated to ensure safety and performance requirements before any clinical use. Feedback may be required in order to determine which parameters are most significant in predicting desired outcomes. Prototype models can often clarify these unknowns.


Testing has been described as the most poorly scheduled part of the development process and perhaps also the most poorly managed.8 Testing should be performed periodically throughout development, not just during the last phase. Early design activities should incorporate testing at the unit or component level followed by structured integration testing prior to the formal validation test. Managing these early phases can reduce the number of errors and lessen the formal testing required at the final test phase, when schedule pressures are the greatest.

Many projects progress well during development only to encounter significant delays throughout the final test phase. Progress is frequently measured by the volume of test cases completed with no other criteria evaluated to determine testing adequacy or completeness. Measurable test completion criteria should include, at a minimum, tests to ensure coverage for all specified requirements. Manufacturers have significantly reduced the number of test cases by tracing test procedures to specified requirements. Large volumes of test procedures that exercise the same requirements in different combinations have little likelihood of identifying new errors.


There has been phenomenal growth of computer hardware capacity during the past 30 years. The expanding hardware capacity has been closely followed by increasingly powerful software tools, including those that support engineering development. Failure to harness the power and flexibility of these tools will hamper a company's ability to achieve optimum reduction in development schedules.

Automated tools can be useful in the support of design, from analysis of material dimensions for hardware to management of software changes and version control. These tools can perform the mundane but essential tasks of the design and development process in a far more efficient and reliable manner than humans can.

Automated tools are excellent for catching errors early in design stages. Tools used in support of hardware design can analyze the strength of materials and the consistent definition of interface signal levels. Software tools can catch common errors in typing, parameter passing, and initialization. Even the most structured design process can be significantly enhanced by tools that can catch errors that are difficult for the designer to identify.


FDA regulations, ISO standards, and associated guidelines continue to increase the requirements for product approval. Customers expect to receive more product capability and quality at reduced cost. In conjunction with these external pressures, device complexity continues to expand at an alarming rate. Designing and developing new products to meet these demands is challenging. By using the 10 steps described above, it is possible to minimize costs and reduce scheduling while producing a quality product that meets FDA regulations and ISO standards.


1. Freedman DP, and Weinberg GM, Handbook of Walkthroughs, Inspections, and Technical Reviews, 3rd ed, New York, Dorset House,
p 12, 1990.

2. Dunn RH, Software Quality Concepts and Plans, Englewood Cliffs, NJ, Prentice-Hall, p 79, 1990.

3. Brooks FP Jr., "No Silver Bullet, Essence and Accidents of Software Engineering," IEEE Computer, August, pp 10­19, 1987.

4. Boehm BW, "A Spiral Model of Software Development and Enhancement," IEEE Computer, May, pp 61­72, 1988.

5. Keyes J, Software Engineering Productivity Handbook, New York, Windcrest/McGraw-Hill, p 513, 1993.

6. "Microsoft Software Development Procedures Test/Verification Procedures," Microsoft, Redmond, WA, Personnel Qualifications section, p 2, 1996.

7. Capability Maturity Model for Software, Version 1.1, CMU/SEI-93-TR-24, Pittsburgh, Software Engineering Institute, February 1993.

8. Brooks FP Jr., The Mythical Man-Month, Reading, MA, Addison-Wesley, p 20, 1975.

Daniel P. Olivier is the president of Computer Applications Specialists, Inc. (San Diego), a software validation and development services company.

Copyright ©1997 Medical Device & Diagnostic Industry

Effective Training for New Manufacturing Employees

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column


Clarence Semple, president of Sigma Four (Redondo Beach, CA), a usability engineering firm, explains how to best design, develop, and evaluate a cost-effective manufacturing training program.

How can one define training requirements for manufacturing departments, determine who is a qualified trainer, and fully document that an employee is trained and qualified to perform a manufacturing task?

Effective manufacturing training supports shortened learning curves, increased production rates, increased production quality standards, and enhanced profitability. The objective of training is to instill, effectively and efficiently, job-required skills, knowledge, and attitudes to support the manufacturing process.

Training design, development, fielding, and evaluation is a systems engineering process. Accordingly, well-designed training includes objectives, content, methods and media, measurable outcomes, standards of performance, and an evaluation of the process. The principles and methods of instructional system development (ISD) are widely accepted as a cost-beneficial way to develop highly efficient and effective training.

A job-centered task analysis is the training process foundation. Each job is divided into major behavioral tasks--structures describing the manufacturing job's main components--that workers must perform. Tasks are further analyzed into their component elements or steps. Tasks and elements new to the worker are singled out for further training analysis. Procedures to perform each selected task or element are listed, including cognitive (mental) procedures, such as recalling details from memory, and visual perception associated with worker inspections to assess meeting work standards.

Each task and its elements are analyzed for learning content including the prerequisite skills and knowledge workers must bring with them to the learning situation; skills, knowledge, and attitudes needed to effectively perform each task and element; standards of acceptable performance; special tools and equipment; personal safety issues; and manufacturing mistakes that may occur and their likely consequences for subsequent manufacturing operations, such as rework requirements. Design and manufacturing engineers, manufacturing supervisors, skilled production workers, and, in some cases, component vendors are valuable sources of task-analysis information.

The training should be designed around the behavioral and cognitive components to be performed and their standards of acceptable performance. The technical content also stresses how to recognize and correct common mistakes, use special equipment, and deal with safety issues in performing each job activity. Worker attitude is important to both productivity and quality control and is discussed in the context of job safety needs and the effects of undetected or unreported manufacturing mistakes on subsequent manufacturing operations and, more importantly, on end-users of the product.

Look for similar, but not identical, behaviors the worker may be familiar with from other manufacturing operations. The goal is to identify existing worker skills and knowledge that may interfere with the learning of new tasks. Such opportunities for negative transfer of training should be emphasized and discussed during instruction and closely monitored during training and initial on-the-job performance.

Training methods should be tailored to the particular skill, knowledge, and attitude to be learned. Lectures with discussion are common. Videos or demonstrations of skilled personnel performing the desired behaviors provide realistic models for learners to emulate. Hands-on training may involve mock-ups or actual equipment that can be either reused or economically discarded.

Trainers must have knowledge of and the ability to demonstrate all of the skills, knowledge, and attitudes to be taught. Their levels of attainment can be measured using the same test and evaluation methods used for students. Effectiveness in answering questions is another way to assess trainer competence, as are learner ratings. Trainers must possess a general knowledge of how people learn, how to teach, and how to motivate. They must understand that learner errors and initial lack of speed are normal steps in the learning process and are tools to be used constructively to guide learners to desired performance standards.

Consistent performance deficiencies suggest the need to modify training. When consistent deficiencies are eliminated and job performance is acceptable, the training has been content validated because the person demonstrates the needed job skills, knowledge, and attitudes.

A central ISD concept is that training results, including intermediate points in the learning process, must be measured objectively. Written and oral tests provide skill and knowledge measurement opportunities, as do standardized observations and measurements of training work samples. Observation and dialogue with workers provide assessments of whether safety and productivity attitudes are being learned. Having learners demonstrate that they meet or exceed required performance standards documents their qualifications and is used for quality control.

Most learning occurs somewhat erratically and along a curve. Thus, even learners who demonstrate required performance during and at the end of training should be monitored until supervisors are fully satisfied with the quality and speed of their work. Supervisor monitoring and criticism must not be intimidating, since this will only reduce worker willingness to ask questions. Manufacturing problems should be documented and analyzed so constructive feedback can eliminate the problems, often through refinements to the basic training program as well as highly focused remedial training.

"Help Desk" solicits questions about the design, manufacture, regulation, and sale of medical products and refers them to appropriate experts in the field. A list of topics previously covered can be found in our Help Desk Archives. Send questions to Help Desk, MD&DI, 11444 W. Olympic Blvd., Ste. 900, Los Angeles, CA 90064, fax 310/445-4299, e-mail [email protected]. You can also use our on-line query form.

Although every effort is made to ensure the accuracy of this column, neither the experts nor the editors can guarantee the accuracy of the solutions offered. They also cannot ensure that the proposed answers will work in every situation.

Readers are also encouraged to send comments on the published questions and answers.

Copyright ©1997 Medical Device & Diagnostic Industry


Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI August 1997 Column

The results of MD&DI's ninth annual salary survey reveal that the earnings of device professionals responding to the survey continue to grow at a healthy rate.

Daphne Allen

Not only are the wallets of medical device professionals bulging more than they did last year, but their salaries are growing at a higher rate than those of U.S. employees in general. According to the results of MD&DI's ninth annual salary survey, employees of medical device and in vitro diagnostic manufacturing companies received an average annual raise of 6.6%, with at least half the survey respondents receiving an increase of 5% or more. According to the 1997 survey, the average raise was higher than last year's, which was reported at an average 6.2% increase. According to figures from the U.S. Labor Department, the average hourly earnings of U.S. workers in all industries are only 3.8% higher than a year ago.

This year, to help readers locate statistics relevant to their positions, we have dedicated a full page to each of seven job functions: general and corporate management, marketing, product design engineering, production and manufacturing (including packaging and sterility assurance), quality assurance and quality control, regulatory and legal affairs, and research and development.

Average salaries for medical device professionals responding to our annual survey rose from last year's nearly $64,000 to almost $70,000 this year. This increase is even more dramatic when you consider what survey respondents earned in 1994--salaries averaged just under $60,000. Respondents have seen their salaries increase by almost $10,000 in just four years.

More than half of industry personnel appear to be pleased with their current positions, perhaps as a result of the steady salary increases. Sixty-two percent of all respondents report they are not considering new jobs. Those happiest with their positions are in general and corporate management, which is no surprise when their salaries and compensation are considered.

Survey results:

Similar to past surveys is the scarcity of women and minorities in the sampling. Nearly 90% of the respondents are white and more than 75% are male. But because the 600 respondents represent only a fraction of the entire industry, it would be wrong to assume that minority and women professionals are having difficulty obtaining positions in the medical device industry. Still, it is interesting to note that only 14% of the female respondents and only 5% of the minority respondents hold general and corporate management positions. The sparse number of women and minorities in these upper-level positions may be attributed to reasons other than discrimination, however. For instance, only 11% of the minority respondents have received postgraduate degrees.

This year's salary approximation worksheet differs only in parts from those of previous years. While experience, job function, and job responsibility continue to influence salary using this model, the importance of gender and region has varied from year to year. For instance, if an employee's primary job function involved production and manufacturing, using the worksheets from 1995 or 1996, he or she would need to subtract $6926 or $5828, respectively, from the base salary. This year production and manufacturing personnel still need to deduct from the base salary, but only by $2570.

Gender is not included in the worksheet this year, suggesting that it does not weigh as heavily as other factors. There is still a disparity between the average salaries of men and women--$72,700 for men, $55,400 for women--but factors such as education, experience, and primary job function may determine salary more than gender does.

Similar to survey results of 1994 and 1995, region plays a key role in predicting salary. This year, users of the survey worksheet must deduct a hefty $5930 if they work in the southern United States; their average salary is $7200 less than the industry average.


The data for this year's survey were obtained from a mail survey designed jointly by MD&DI and Readex Research, Inc. (St. Paul, MN), and conducted by Readex in March and April. Surveys were mailed to 1200 medical device professionals, 656 of whom provided usable responses.

The survey results are based on the responses of 600 individuals who identified themselves as full-time professionals working for companies that manufacture medical devices or in vitro diagnostics. These individuals were segmented according to the following seven job functions: general and corporate management, marketing, product design engineering, production and manufacturing (including packaging and sterility assurance), quality assurance and quality control, regulatory and legal affairs, and research and development.

Another key segmentation that recurs throughout the accompanying article categorizes individuals according to their level of responsibility: CEOs and presidents, vice presidents and directors, department heads and supervisors, and engineers and scientists.

The margin of error for percentages based on the 600 responses is ±4.0% at the 95% confidence level.


The Ninth Annual MD&DI Salary Survey is available as a bound reprint. Each 200-page volume includes a copy of this article and tabular breakdowns for the device industry as a whole, plus previously unpublished tabular breakdowns for each job function covered by the survey.

Copyright ©1997 Medical Device & Diagnostic Industry