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 1996 In November

Exploring the Long-Term Effects of Implants

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996


Research Scientist, Environmental Trace Substances Laboratory, Center for Environmental Science and Technology, University of Missouri, Rolla

More than 200,000 joint replacements are implanted every year in the United States. Most of these, especially artificial hips and knees, are placed in the elderly whose sedentary lifestyles minimize wear. For many of these patients, long-term effects of the implants are of little concern. But as shoulder, elbow, knee, and even hip implants are increasingly placed into younger patients, the effects that the implants may have 20 to 30 years down the road become important. For these patients, the wear and tear that may cause metal to shed from implants into the bloodstream, and the corrosion of materials that are in contact with body fluids are a concern. And while some of the elements used in artificial joints are known carcinogens at high levels, the long-term effects of low doses are unknown. Such concerns and long-term effects need to be addressed, however, and Lijian Yu, a research scientist at the Environmental Trace Substances Laboratory (ETSL) at the University of Missouri in Rolla, is working with other researchers to do just that.

As part of a five-year $750,000 grant from the National Institutes of Health, Yu is cooperating with researchers at Rush Presbyterian–St. Luke's Medical Center in Chicago who are documenting the wear seen in implants and then sending biological samples to ETSL. In this research, Yu will investigate what happens to the primary metals—aluminum, titanium, vanadium, cobalt, chromium, and nickel—used in artificial joints following implantation. He and his colleagues will also try to determine how these metals spread to other sites in the body. In this interview with MD&DI, Yu describes the study, the technology behind it, and how its results could affect the development of implants.

What is the relationship between Rush Presbyterian–St. Luke's Medical Center and your laboratory?

The research project begins at Rush Presbyterian–St. Luke's, where they are looking at systemic implications of total joint replacements. One of the researchers there, Jorge Galante, designed a lot of these total joint implants. Another is the principal investigator, Joshua Jacobs, who directs and coordinates the whole project. According to their experiences, these metal implants in the human body are very successful because the devices are usually used in people of advanced age. These devices help a lot of people who otherwise would not be able to walk or function normally.

With the success of these devices in older people, those in the industry and the medical field are wondering whether such devices can be supplied to younger patients. The objective is to develop better designs for total joint implants for use in older people as well as in younger people. We'd also like to determine whether corrosion of these implants, which cannot be avoided, will have a long-term effect on a human being. Perhaps by controlling some design components of the implant, it may be possible to reduce the amounts of trace metals deposited in the body.

So design is an important consideration in your work, because one design might cause more or less wear of the materials than another?

Yes. We'll be analyzing urine serum, as well as tissues, from these patients. These will indicate whether corrosion takes place. If there's corrosion, you would expect the corroded product to bind to the tissue, or to dissolve into body fluids carried around with blood or urine. My part in this research is to point out the extent of corrosion that has occurred. This may shed light on the role a specific design has played in corrosion.

What's the current status of the project?

We've just begun. We have not yet analyzed any samples from patients, but we have developed a method to study the very low amounts of aluminum, titanium, and other metals in serum using trace metal techniques. The problem is that these techniques—such as graphite furnace, atomic absorption, spectrometry, or neutron activation—need a pretreatment of a sample such as separation or preconcentration to eliminate interferences. And sometimes the sensitivity of the technique is not enough to detect the extremely low amounts of these elements inside the human body.

We're using the latest technology, which is called electrothermal vaporization inductively coupled plasma mass spectrometry (ICPMS). Basically, it vaporizes the sample with these elements into a high-temperature stream, and then a stream of gas carries these elements into a plasma at a temperature of about 6000°C. This high-temperature source basically ionizes everything, and these ions are then sent to a mass spectrometer where they are detected.

Where was this technology developed?

The basic technology was developed in the 1980s by groups in England, Canada, and the United States. From this, several sample introduction technologies have been developed. Electrothermal vaporization sample introduction technology is one of these.

What is meant by sample introduction technology?

ICPMS basically requires a dilute aqueous sample—a homogeneous type of sample. But the body fluid from human beings does not aspirate very well. It is thick and does not become an aerosol very readily. So in order to get the sample into the detector, it somehow needs to be transported into this high-temperature plasma source, and that's where the electrothermal vaporization comes into play.

In addition to using a lot of off-the-shelf technology, will you also have to develop modifications?

The next step will probably need this kind of work. Currently, we're just using the stock techniques that are available but we have to develop a method ourselves.

Where are you in terms of developing this method?

We have developed a methodology for the serum analysis. This took a long time because there are a lot of technical difficulties in analyzing the metal elements in serum, because the serum contains calcium, chlorine, and phosphorus, as well as carbon. These interfere with the analysis of aluminum, titanium, and vanadium. We have to devise ways to alleviate these interferences before any measurements can be done. The method has to be very stable because these samples are collected over a long period of time—say, five to seven years—and the analytical data collected during this period should be comparable to those collected seven years later in order to establish a clear trend as to whether or not the device works.

How are the laboratory data correlated to the clinical data?

Rush Medical Center is designing a statistical model for how these samples are collected during a specific period from certain numbers of patients. They will then send these samples to us, and we will analyze them here. These data will be collated in terms of numbers of patients as well as the period over which these data were collected, and then we'll come to a conclusion as to whether there is an effect.

So you will be looking for patterns that develop over the next seven years?

Rush Medical Center has a large library of samples collected from the patients over the years, so we may not need to wait seven years to get the desired samples. But in any case, we will be looking at increasing or decreasing concentrations of these particular metals over time.

You said you're looking at aluminum, titanium, vanadium, cobalt, chromium, and nickel. Is that a comprehensive list, or might you be looking at other metals?

Those are the basic metals because there are really only two types of alloys used in this industry. One is the titanium alloy, which is made up of titanium, aluminum, and vanadium; the other is a stainless- steel-based alloy, which is chromium, nickel, and cobalt.

When do you expect to start seeing results from this study?

Hopefully, we'll have some results in a year. We are currently at the stage of developing the methodology—and we're trying to develop all the methods at the same time so that these methods can be used by other laboratories and can provide the basis for similar research elsewhere.

Copyright© 1996 Medical Device & Diagnostic Industry

FDA Reform Legislation Does a Slow Fade

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996

James G. Dickinson

FDA reform legislation began dying as Congress adjourned for its August recess, but that doesn't mean that the old FDA will come back on the rampage, bigger and meaner than ever, as some reform advocates have warned. The truth is, that old dragon had already been fading, partly of its own accord and partly out of budget-borne necessity.

The demise of reform legislation in Congress comes as a bitter blow to the many earnest reform proponents in industry. Among them was Jeffrey Kimbell, executive director of the Medical Device Manufacturers Association. His disappointment showed in an obituary-like August newsletter report: "I regret to inform the membership that our vote on FDA reform, originally scheduled for July 30 in the U.S. Senate, has been delayed because of a very crowded legislative calendar. . . . This unfortunate postponement, coincid[ing] with the House not taking any action on the bill in subcommittee or full committee before the August recess, reduces the chance of enacting any meaningful legislation this year."

To speculate before the November election on the chances of successful legislation next year might seem reckless. But in the interregnum, more than a few knowledgeable observers have spoken loosely of the "pendulum"—their implication being that the political bias of national affairs is due to swing back from the right toward the center, something that House Speaker Newt Gingrich had apparently decided was indicated for his own public statements. That makes it less likely that even the current Congress, if returned intact, would have FDA reform as high on its agenda next year as this year. And if it is restructured by the voters, this hypothesis suggests it will be a more-leftward Congress and even less likely to resurrect FDA reform legislation.

While Congress stumbles over agency reform, FDA continues strenuously working—and networking—to make legislation unnecessary. Central to its efforts and those of its industry advisers has been the goal of achieving a major attitude adjustment, especially among field investigators. "I am alarmed to still hear regulators describe their job as to 'kick butt,'" wrote FDA Office of Regulatory Affairs human resource development director Gary German in August, in his debut as the editor of the Association of Food and Drug Officials (AFDO) newsletter.

AFDO is the association of both federal and state regulators. This year it formally adopted medical devices as a focus of collaboration, staging two regional workshops in San Francisco and San Antonio to help FDA's Division of Small Manufacturers Assistance orient device manufacturers to all aspects of medical device regulations, including the new medical device reporting (MDR) regulations that became final on July 31.

German's editorial reflects what is happening to FDA's attitude. Importantly, his office of resource management is located within the headquarters enforcement office that directs field operations nationally.

His editorial extolled the virtues of the reinvention of government "revolution," which he saw as coming from the American people themselves, through industry and government. "I have heard in both my professional and personal life many indications that the revolution is real," German wrote. "A news reporter talked about one of the biggest ills of society being a government that is not responsive to the people. The same reporter indicated that a large percentage of the American people are afraid of their government. A CEO reported how he longs for a dialogue with the agency that regulates his firm. He indicated his desire to be in compliance, but they [the regulatory agency] treat everyone like crooks."

German deplored this state of affairs. "The revolution appears to me to be the major issue facing our members and the regulatory agencies that are our members. We as regulatory officials must be day-to-day players in the reinvention efforts. Our industry members need to recognize these efforts as sincere and step forward to offer their assistance in the change."

At AFDO's San Antonio meeting, local resident post supervisor Jack Davis produced some interesting statistics about the developing fusion of FDA and state inspections of medical device companies. Driven by budget cutbacks—fostered by medical device industry interests seeking less money for FDA enforcement and more for device approvals—FDA has contracted out more than 45% of its Class II and Class III medical device facility inspections in Texas to the state's health department.

This arrangement is becoming more common across the country. Although the inspections are conducted under FDA auspices, performed by FDA-trained state employees, paid for by FDA, written up on FDA-483 report forms, and analyzed by FDA for any follow-up actions, they raise many intriguing questions. Are the inspections as thorough or as competent? Are the inspectors more or less reasonable or diligent? Is FDA's presence in regulatory situations diluted? Can any amount of training by an outside entity that is not your employer convey the depth of core culture that may help both sides when unusual conditions are encountered? Can inspected firms reach appropriate understandings as readily with an agent as they might with the principal? For many firms, contact with the local FDA inspector may be the only hands-on experience they have with FDA. Does the loss of this matter?

At AFDO's San Antonio workshop, presented by FDA professionals from Washington but attended by both FDA Southwest Region and Texas state officials, it was interesting to see how everyone mingled—or didn't—during the coffee breaks. The FDA people talked easily with industry personnel and seemed eager to answer questions privately. The Texas inspectors, on the other hand, did not mix, seeming to prefer to remain somewhat aloof, even coplike.

This impression was reinforced by the announcement at the workshop that the Texas department was not participating in FDA's pilot establishment inspection program. Under this relaxed program, advance notice is given of inspections, FDA-483s are annotated, and close-out letters are sent after inspection. But perhaps the decision was FDA's, to avoid potential complications to what was officially just an experiment.

As FDA's budget crunch bites deeper next year, assuming a continued de-emphasis on enforcement, we can expect more states to be contracted to do FDA inspections. Such arrangements also seem to be at the heart of FDA's plans to regulate cigarette vending machines and advertising aimed at teenagers. The agency has no resources to mount such an effort itself directly; sources indicate the agency will be looking to the states for help.

Philosophically, this trend seems to comport with the pendulum theory—an historic shift of federal power back to the states, in which FDA's role is only part of a whole.

Copyright© 1996 Medical Device & Diagnostic Industry

Canada, EU Experiment with Harmonization

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996

James G. Dickinson

They're next-door neighbors in North America, but Canada and the United States are courting the European Union (EU) from different directions on the thorny issue of reciprocal reviews of medical devices. Both countries' regulatory agencies (the Health Protection Branch [HPB] in Canada and FDA here) agree on at least one thing, however: it's early yet, and the European device approval system is still evolving. Actually, so is Canada's, that country having adopted an overhauled device approval system last year that doesn't become fully operational until next year.

None of this, however, deters either HPB or FDA from plunging into novel schemes aimed at harmonizing their approval procedures with whatever system Europe evolves.

In Canada, HPB held yet another meeting with its EU counterparts in October to set up an 18-month series of "confidence-building" experiments. In the words of the chief of HPB's office of legislative and regulatory processes, Don Boyer, the two sides would "exchange information on regulatory requirements, methods of conducting evaluations of products, postmarket reviews, and doing double-blind experiments on evaluating a manufacturer's submission." Boyer acknowledged that industry had not yet been consulted on these ideas, and that in the final analysis, it all would depend on industry's willingness to participate.

"When the time is appropriate," HPB will issue a letter to industry asking for volunteers to submit a bifurcated application for marketing approval in both Canada and the EU. HPB would review such submissions both for Canada and for Europe, and the EU Commission would assign to a notified body the review of the same submissions for the same two markets. Then each side would compare its test work with the other's.

A best-case outcome ultimately would allow applications to be submitted in either Canada or Europe for approval in both regions. The objective would be to save money for both sponsors and HPB.

In Washington, CDRH deputy director of device evaluation Kimber Richter advises that it's not likely that FDA will seek to join in the Canada-EU experiments, but that CDRH is interested in starting a reciprocal-review pilot with the Canadians. This would have Canadian and FDA reviewers sharing the workload on the same application. She believes FDA prefers the harmonization approach, in which countries work cooperatively on building one common set of approval requirements. "That way, whoever does the review, you've got an understanding that you're using the same kinds of criteria," Richter says.

Richter predicts that ultimately the different approaches would come to the same result, "because the notified bodies, if they start taking on different countries' regulations and doing multiple reviews, will pretty soon be seeing where the similarities are, and try to bring them together. It's just not efficient to do four reviews when they can do one comprehensive one."

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

Copyright© 1996 Medical Device & Diagnostic Industry

Intranets: Using Web Technologies in a Regulated Environment

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996

Morteza Minaee

Managing the documentation necessary to comply with FDA's good manufacturing practices (GMP) regulation and the ISO 9000 series of quality systems standards can place a heavy burden on device manufacturers.1–6 It's no wonder that companies are constantly on the lookout for a system that will simplify this task.

Now, the emergence of web-based intranets is promising to resolve many of the issues that have concerned manufacturers. In terms of the features most often needed by device manufacturers, intranet technologies compare favorably to both traditional, paper-based systems and other electronic alternatives. In addition, intranet systems hold the potential for doing much more than merely managing documentation.


FDA views extensive documentation—as required in the revised GMP regulation and elsewhere—as an essential tool for enabling the agency to fulfill its mandate for protecting the public health and safety. Recently, the documentation requirements of the GMP regulation have been reinforced by a number of trends that affect manufacturers and suppliers in many industries, but have particular application to such regulated environments as the medical device industry. Among these trends are the following:

  • Government and commercial customers have shifted away from emphasis on inspection to emphasis on surveillance. Such surveillance schemes require that manufacturers provide extensive objective proof that all requirements have been met.
  • Consumer advocates are urging more documentation as a means of providing information to consumers.
  • The risk of product liability suits is forcing manufacturers to establish defensive schemes of documentation.
  • The rapid globalization of many industries—including the device industry—has increased the impact made by worldwide acceptance of the ISO standards, making ISO 9000 certification a sine qua non for many suppliers and manufacturers.

Collectively, these trends are creating information needs that add up to a requirement for a large, complex network or system consisting of numerous interrelated subsystems. In consequence, many manufacturers are looking to expand their information systems with respect to documentation and configuration management, and to develop a comprehensive design of documentation that can meet all needs.


Although the GMP regulation and ISO 9000 standards require manufacturers to document many aspects of their product development processes, neither is prescriptive about how such documentation should be carried out. In deciding what type of documentation system is right for it, therefore, the first task of the device manufacturer is to define the basic features that its system should possess. These features may vary according to the complexity of the company's design process, manufacturing systems, finished product, or other variables. Once the basic set of desirable features has been established, the manufacturer can then meaningfully compare and contrast various alternatives for creating and maintaining a document control system.

Some characteristics of document control systems are almost universally desirable and can be used to provide a starting point for defining the basic features of a company's system. In any organization where personnel have diverse backgrounds and education levels, for instance, the acceptance and subsequent success of a documentation control system hinges on how easy it is to use. Thus, first and foremost, such a system must permit natural and easy retrieval of information. At the same time, it should provide a quick and simple update mechanism for those who create and revise these documents. Other basic features that manufacturers might expect to find in a document control system include the following:

  • Ability to integrate text and graphic images (e.g., drawings, blueprints, forms, and figures).
  • Minimal redundancy of stored objects, such that updates and revisions are restricted to a few copies (or, ideally, to a single master document).
  • Mechanisms that enable users to navigate easily within and between documents, and to retrieve a document simply even while operating on another document.
  • Key word searchability, even for unplanned searches.
  • Security mechanisms that permit only those with specific authorization to update the documents. All other users are limited to read-only access.


In paper-based document control systems, procedures and instructions are created and controlled using traditional methods of preparing and distributing hard copy. One advantage of such systems is that, for most people, the natural form of retrieving information is to read printed manuals and procedures. Searching is performed in the traditional manner of reading a table of contents or index. Another advantage is that paper documents allow easy integration of text and graphic images.

Among the disadvantages inherent in paper-based systems are the problems of updating and obsolescence. As information contained in the documents changes over time, organizations can incur significant expense to input corrections, output and print new documents, and distribute them. To avoid reprinting entire documents, minor changes are often printed on separate pages and attached to the original document. But even if the new pages clearly note the part of the original that was revised and superseded, there is no way of ensuring that users will take note of the changed sections. Moreover, as changes accumulate over time, it becomes increasingly difficult for users of the document to keep track of them. Therefore, the eventual reprinting and reissuance of the entire document is inevitable.

To minimize the risk of providing inaccurate information, a paper-based document control system must include policies to ensure that outdated documents are recalled and replaced with their updated versions. In some organizations, it is customary to designate a custodian for each document. Responsibility for generating, revising, and distributing a document is vested in its custodian, who typically maintains a checklist detailing the points of use for each document. Whenever a document is revised or reprinted, the custodian uses this checklist to replace all copies of the old version with the revised one. Further, to avoid unauthorized copying and to validate the authenticity of a document, a stamp or seal is placed on each legitimate copy.

Although such a system can work well for small companies or those whose product development processes are relatively simple, for medium-sized to large companies a paper-based update and reissuance process can become extremely time-consuming, costly, and difficult to manage. This is especially true for companies that must compile and maintain a large amount of documentation in order to comply with the GMP regulation or ISO 9000 standards. Such difficulties can become the biggest obstacles to implementing a paper-based document control system.


Over the past decade, there has been a growing acceptance of computer technology in the workplace. This can be attributed to the increased sophistication of computing hardware, the reduced price of both hardware and software, and the increased computer literacy of today's workforce. In turn, these developments have made it possible for many companies to consider computerized storage and retrieval a realistic alternative to paper-based document control systems.

An electronic document control system is one in which procedures and work instructions are created, stored, distributed, and retrieved using computerized media. Thus, mere use of a word processor to create and print manuals is not considered an electronic alternative. The primary infrastructure for an electronic document control system is the existence of an electronic network comprising dumb terminals connected to a central computer, personal computers connected via a local area network (LAN), or a combination of the two. A growing number of organizations have already established some form of computer network to enhance communication among employees and facilitate the sharing of data and expensive computer peripherals.7–9 An electronic document control system extends the uses of such networks.

Electronic systems offer advantages such as reduced complexity and potential cost savings that may overcome many of the difficulties inherent in paper-based systems. In a computerized environment, for instance, it is sufficient to store a single copy of each document, which is then electronically accessed by multiple users. This not only simplifies distribution, but also facilitates the updating process. Since changes are limited to a single copy, they can be performed quickly and with ease.

However, adoption of a computerized system is not without risks and disadvantages. Although the costs of electronic equipment have dropped significantly in the past few years, for instance, equipping a large company with appropriate hardware and software can still be expensive. Similarly, training employees to use the company's system can be time-consuming and costly, and can be made even more difficult if the system itself must undergo significant upgrades. Before deciding to implement an electronic document control system, therefore, companies should take careful stock of their requirements and explore all of the alternatives that might meet those requirements.

The sections that follow compare four electronic alternatives: word processing software; desktop publishing software; database management systems; and hypertext, the language of the World Wide Web.

Word Processing and Desktop Publishing. Intuitively, the alternative that appears to be most apt for use in a document control system is word processing software. Though document creation and updating is simplest with word processing software, in general such software does not possess the ability to merge text and graphic images effectively. Consequently, organizations have considered either the exclusive use of desktop publishing software, or the use of desktop publishing software together with a word processor. Familiarity with both these alternatives (especially word processing) is relatively high, and hence training time and costs are usually low.

However, these two types of software have similar weaknesses. First, neither handles unplanned key word searches efficiently. Although such software typically provides text search capabilities, the searching process is slow and limited to the document being viewed. This limitation can be overcome by storing all procedures, plans, and policies as a single document, but the document would then become unmanageably large, further reducing search speed. In practice, procedures are often separated into distinct electronic files. When a combination of word processing and desktop publishing software is used, searching for related information in separate files is, at best, cumbersome.

A second drawback of word processing and desktop publishing software relates to the integrity and security of documents. In general, such software packages do not, on their own, permit document retrieval in a purely read-only mode. Some word processors (e.g., WordPerfect) have a look option, but special document formatting is generally lost when the document is viewed via this option. More importantly, the menu from which the look option is selected can be used for retrieving and editing the document. Most word processing and desktop publishing software allows passwords to be used to prevent unauthorized access to documents, but once a document is retrieved it can be both viewed and edited. As a result, it is difficult to prevent a user from changing a document, whether intentionally or unintentionally. It should be noted that both FDA and accreditation agencies for ISO 9000 certification explicitly prohibit alterations to certain specific controlled documents without notifying them.

Database Management Systems. The key advantage of a database management system is its ability to search efficiently for key words or other user-specified search strings. In light of the documentation requirements of the GMP regulation and ISO 9000 standards, efficient search capabilities are considered critical. Thus, an attractive document control system alternative would be a user-friendly, page-oriented database management system that would allow integration of text and graphics.

Unfortunately, such features are difficult to implement using existing database management systems. Unlike most other kinds of organizational data, manuals and procedures tend to incorporate graphic images alongside text. Furthermore, unlike inventory, sales, or employee data, the information under consideration tends to involve multiple pages of information. Traditional database management systems are record oriented and prove to be ineffective for variable-length documents.

Hypertext Systems. In its most primitive form, a hypertext system can be viewed as a textual database with a simple user interface. The tool used to create such a system is hypertext markup language (html), the language of the World Wide Web, which is designed to make it easy to combine text, graphics, and other media onto screen pages.

A web-based document control system is one based upon Internet technology, but placed on private computer servers within a company and designed to permit access only to company employees. Formally, it can be defined as a computerized network of database objects connected by links. An object may be a section of text or a graphic image that is presented to the user in the form of a screen window.

The term intranet has recently come into use to describe the application of Internet technologies to internal corporate networks.10 This practice enables manufacturers to seamlessly integrate desktops, LANs, client/server applications, legacy systems, and the public Internet to create highly effective corporate information systems. Web technology is ideal for implementing an electronic document control system because it offers several useful capabilities, including the following:

  • A very simple user interface.
  • Ability to merge text, graphics, and multimedia in general.
  • Minimal redundancy.
  • Multinode navigational characteristics.
  • Controlled access levels for different types of users.

The following subsections provide further information about the advantages of these features.

User Interface. A key feature of any hypertext system is the simplicity of its user interface. As a result, the technology is particularly attractive for situations where the system is to be used by a variety of users with different levels of computer skills.

Merging Media In a hypertext web document, each procedure or work instruction can be stored as a distinct text object. Similarly, forms, figures, and engineering drawings can be stored as distinct graphic objects. To satisfy the training requirements of a GMP- or ISO 9000–compliant quality management system, the manufacturer can even add sound and video to the system. Because such sites typically run on internal networks, bandwidth doesn't affect performance as much as it does on the Internet, so sound and video can be added without resulting in painfully slow file loading.

Minimal Redundancy. Each object belonging to the document control system—whether a procedure, figure, form, or other type of recorded information—needs to be stored just once. This nonredundancy of stored objects greatly simplifies the updating process and ensures information consistency. Unlike paper-based systems, in which a change to a form referenced in multiple procedures would require it to be updated and copied for inclusion in every copy of those procedures, a web-based system requires only an update to the changed item. Other objects and links remain unchanged.

Navigational Characteristics. The idea behind hypertext is that instead of reading text in a rigid, linear fashion (such as one reads a book), one can easily skip from one point to another, get more information, go back, jump to other topics, and navigate through the text based on one's interests at the time.

Links among objects are usually represented as on-screen icons, buttons, or underlined text. Since each link has an explicitly defined source and destination object, it is possible for the user to navigate easily from one object to another and then back to the original. Unlike conventional text that is structured sequentially, the links in hypertext documents encourage nonsequential navigation. Users can move within a hypertext document in one of three ways:

  • Follow predetermined links and open windows successively to examine their contents.
  • Search the network of nodes for a particular key word or character string.
  • Select a specific node or object from a graphic display of the network of all nodes.

Access Control. Access control can be accomplished on web servers in at least two ways. To control write access to the intranet site, for instance, the manufacturer can restrict the number of people with authority to create pages for the server. This kind of control is usually handled in the same manner as assigning read/write permissions to computer network users.

Intranets also offer ways to control read access, so that unauthorized employees do not gain access to privileged information. One way is to use the password feature built into the web browser software. Another is to use Internet protocol (IP) or host-name filtering, which enables the system to be set up so that only specific IP addresses are allowed access to certain pages on the web server.


Assuming that a company can support the expense required to create a network of PCs or a centralized computer system, electronic media offer significant advantages over paper-based document control systems. Besides the cost savings that can accrue from not having to print, warehouse, and distribute paper documents at intervals, electronic media have the advantages of simple updating, significant reduction or elimination of the problems associated with obsolete documents, and easy dissemination of information. The following sections describe the use of an intranet-based document control system for key areas of concern to device manufacturers.

Storing Documents. One way to handle the documentation of a quality management system, as mandated by the GMP regulation and the ISO 9000 standards, is to treat the documentation as a two- or three-level hierarchical structure, with the amount of detail increasing with each successive level. At the highest level is the main directory, which can be used to provide an outline of the documentation system as a whole and a brief description of each document it contains. At the next level, the system would include a detailed description of each procedure or activity that is critical to quality management. Typically, entries at this level will answer the who, what, when, and where for each activity. In the third level, entries would include detailed work instructions. A web-based system facilitates structures such as this by making it possible to store each document, procedure, or work instruction as a distinct hypertext object. Similarly, forms, figures, and engineering drawings can be stored as distinct graphic objects.

Document Approvals. When using an intranet system, approvals for first-time issuance of documents—as well as for subsequent changes—are handled much the same as they are in paper-based or other electronic network–based systems. All such systems require manufacturers to have well-structured policies in place, and can make use of task segmentation to delineate responsibilities for documentation. A firm can define responsibilities such that the personnel who prepare documents, approve them, and actually bring them on-line are three distinct groups.

An intranet system can facilitate such distinct responsibilities by making the creation, transfer, and archiving of documents a seamless process. Once a document has been created in hypertext, together with appropriate graphics and links, the writing group can transfer the electronic file to the approval group for sign-off. Upon approval of the document, the approval group can then pass both electronic and paper versions to the uploading group. In turn, members of that group can place the electronic version on-line and keep the paper version with appropriate signatures on file.

Controlling Web Documents. To comply with the GMP regulation and ISO 9000 standards, manufacturers must be certain that their documentation is correct and up to date. Hypertext documents can be controlled in the same manner as nonhypertext documents. A needed change can be submitted as an engineering change request, and an approved change circulated as an engineering change notice. Once changes have been entered in the appropriate document, however, an intranet-based system can help control them by permitting them to be stored as read-only files, thus preventing accidental or unapproved changes.

Cross-Referencing. Even though the amount of cross-referencing among procedures, forms, and figures is dependent on the manner in which the procedures are constructed, some cross-referencing is inevitable. An intranet-based documentation system makes it easy by representing references as predefined links among documents. Therefore, if one procedure refers to another, or to illustrative material found elsewhere, predefined links enable users to retrieve the references by merely clicking on the link.

Printing Documents. One of the basic objectives of a document control system is to make up-to-date documents available. Creating printed copies of the on-line version of a document for subsequent use would defeat this objective. Instead, manufacturers can ensure that employees make use of the intranet's computer technology by disabling the system's printing feature.

However, there are situations in which printed copies are legitimately required—for instance, by regulatory agencies or certification bodies, by service personnel on field visits, or for a meeting that takes place in a conference room without computer display terminals. In such cases where a hypertext document must be printed, it is recommended that a footnote or stamp be used to validate each page of the printed document. Such a footnote might also state that the validity of the hard copy expires after a certain time.

Security. As more companies transmit sensitive materials over the Internet, the implementation of a corporate intranet can raise some worrisome issues. Most of the problems related to security have to do with use of the Internet and the interface between that external system and a company's own internal system. Among the issues raised by manufacturers are the possibility of transmitting software viruses, concern about the privacy of communications, and the need to safely encrypt data. One way to handle these concerns is to use an Internet service provider for the company's Internet connection. In this way, anyone accessing the company's web pages is actually connecting to a computer being operated by the Internet service provider and not one connected to the company's intranet system.

Laser-Manufactured Features in Medical Catheters and Angioplasty Devices

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996

Ronald D. Schaeffer

Medical devices incorporating laser-based processing as a step in their production are becoming increasingly common, especially as industry demands smaller feature sizes that may be impossible or uneconomical to realize using traditional manufacturing techniques. It is fair to say that a large percentage of the processing applications involve punching small holes into some type of plastic or organic material. The material, hole sizes, hole patterns, and other parameters change with the individual application. Typical applications include orifices for drug delivery, liquid or material removal, inflation devices, and analytical devices. Each has its own set of unique requirements and constraints, and, as in all industrial situations, technical objectives must be balanced with fiscal considerations.

This article briefly discusses the advantages of using lasers for manufacturing as well as the most common industrial lasers available.1 Three examples of applying the technology—profiling metal stents, drilling orifices in medical catheters, and drilling orifices in angioplasty balloons—are presented and discussed.


Lasers of some type have been industrial workhorses for more than 25 years, and their use in the manufacturing environment is growing significantly. This is particularly true in the medical product industry, where the use of lasers to manufacture devices is relatively recent and where the possibilities for future use in various areas seem unlimited. Three principal laser types are used in most manufacturing environments. These are the carbon dioxide (CO2) laser, the solid-state (primarily Nd:YAG) laser, and the excimer laser. Each has its own distinct advantages and disadvantages when being considered for manufacturing jobs.

The principal, fundamental wavelengths of emission for these lasers are 10 µm (CO2), 1 µm (Nd:YAG), and 0.2 µm (excimer). It is generally true that thermal considerations in the processing become more important with increasing laser wavelength, and penetration depths are generally greater, allowing for faster bulk material removal at longer wavelengths. Because achievable resolutions become smaller at shorter wavelengths, it is fair to say that longer-wavelength lasers should be used where bulk material removal considerations outweigh either precision or quality requirements, but shorter-wavelength lasers must be used when the highest precision or process quality is required. Table I illustrates these points.

A comparison of hole quality when drilled with the three lasers discussed is shown in Figure 1. The material is polyimide and, in truth, would tend to show the differences between the lasers more dramatically than ceramics or other plastics. However, the same general conclusions can be reached for the processing of most materials: achievable feature sizes are smaller and feature quality is better as the wavelength of the laser decreases.

Table I. Comparison of common industrial lasers.
CO2 Nd:YAG Excimer
Processing plane Focal Focal Image
Relative operating cost per photon Low Low High
Wavelength (µm) 9.6–10.6 1.06 0.19–0.35
Average power 0.3 to 20 kW 0.1 to 2 kW
Penetration depth (µm) >10 1 to 10 0.1 to 1
Ultimate feature resolution (µm) 10 >1 0.2
Practical feature resolution (µm) ~ 50 ~ 25 ~ 1
Material interaction Thermal Thermal Photochemical/thermal
Types CW, pulsed CW, q-switched, pulsed Pulsed

The two issues of primary importance when determining the applicability of any lasers to specific tasks are the material's ability to absorb the wavelength of light used and to carry away excess thermal energy efficiently without causing secondary problems. In general, if the material does not show a fairly strong absorption of the desired wavelength of light, the best course of action is to find a suitable wavelength that does absorb strongly—assuming one can be found. Teflon, for instance, does not really laser machine well at any of the wavelengths under discussion, but it machines well at 157 nm (F2 laser) and can be processed if fillers are added that retain the initial desirable characteristics of Teflon while promoting absorption of the photons.

Many plastics absorb all wavelengths of interest, but their low thermal conductivity prevents the use of longer-wavelength lasers for processing because the heat-affected zone exhibits symptoms of charring, melting, or cracking. This is also true of materials like glass. On the other hand, some ceramics absorb fairly well at all the wavelengths under discussion, and the choice of laser in these cases is dictated by speed, feature resolution, and quality.

It should also be noted that there are optical frequency–altering techniques commercially available for use with solid-state lasers that will double, triple, or quadruple the fundamental laser frequency. This frequency enhancement comes at the cost of losing some of the original output energy and stability, but provides an alternative way to make "green" and UV photons. In many cases, these new frequency-converted lasers—particularly at the UV wavelengths achieved when frequency tripling or quadrupling—may replace other UV sources in some applications, especially those requiring single-hole drilling of small, round orifices in soft material. As the technology for generating shorter-wavelength light from solid-state sources matures, the relative ease of use and low cost compared to other short-wavelength sources will be quite attractive to potential users.

It is the goal of system integrators to build lasers into micromachining workstations that are cost-effective, highly reliable, and easily operated as well as maintained. A complete laser system will include the laser; beam delivery to get the photons on the target; workpiece tooling or other motion systems for part delivery, manipulation, and exit; integrated and interlocked frame and support; computer control; laser support equipment; and sometimes vision systems or other options to make the system more flexible or increase throughput. Note that flexibility and high throughput do not necessarily correlate in equipment design.

Investment in any laser system is costly and requires attention to facility preparation and education of associated personnel. In cases where either financial or control considerations dictate the placement of a laser system in-house, typical system costs start at $100,000 for a small laser-based system with few options or functions to more than $500,000 for a large, sophisticated system that may require part handling and automated machine vision. Typically, the cost of a small, sealed CO2 laser or small solid-state laser is about $50,000. An integrated system would cost from about $150,000 to $200,000 and would include the laser, camera vision for viewing, motion control—usually at least an x/y table and possibly including rotary motion—beam delivery, computer control, and Class I laser operation (operator-level interface and safety equipment). Excimer laser systems are usually more expensive, with the laser costing about $130,000 and the full system about $300,000. The higher cost for the processing end of the system is due to the fact that, in order to take advantage of the higher resolution capabilities of the excimer laser, better-quality vision, motion control, and other accessories must be used than in a comparable CO2 or solid-state system.

In addition to the above laser system components, most industrial applications require some level of tooling to efficiently and accurately move the parts on and off the laser as well as to position them during processing. The exact level and type of tooling is very dependent on the specific application, so these components are almost always custom made for each user.

If there is a need for laser processing but the initial capital investment for a system is not immediately justifiable, a good option is to work with a reputable contract manufacturer. This is especially true for small-sized or start-up companies. In many cases, device manufacturers can maximize profitability by concentrating on their areas of expertise and outsourcing laser processing, which requires a greater amount of logistics. Areas of concern that need to be addressed to ensure lot-to-stock or other acceptable delivery requirements include process documentation, document and tracking control, and quality control and assurance.


Angioplasty operations are performed on patients who have suffered a major blockage to vessels in the circulatory system. More than 500,000 angioplasty procedures are currently performed each year in the United States alone. The procedure typically involves inflating a balloon in the area of the blockage, which breaks up the accumulated plaque and opens the vessel. While this technique works well in the short term, 30 to 50% of all angioplasty operations performed will need follow-up treatment within six months. This is due to incomplete plaque removal and the formation of scar tissue as a result of irritation of the vessel, known as restenosis. There has been a tremendous push in the health-care industry over the last few years to combat restenosis because repeat operations are expensive, inconvenient, and potentially life threatening. Lasers play an important role in several of the new technologies being developed by both large, well-known medical device companies and small, specialized high-tech firms.

One product that has received tremendous attention in the industry is the metal stent. Most surgeons performing angioplasty operations today insert these stents as a matter of course, since initial results seem promising. These stents are usually machined in intricate patterns to make them more flexible while still maintaining the mechanical rigidity necessary to keep the vessel walls from closing (see Figure 2). The devices are also used to minimize the problem of arterial blockage caused by plaque falling into the vessel after inflation.

Such metal stents are usually made from small-diameter, stainless-steel tubing using a Nd:YAG laser. Other methods used to manufacture the metal stents with high precision include electrodischarge machining (EDM) and etch techniques, but these methods have significant drawbacks and the laser process is the method of choice.

These devices have been used for only a few years and, although initial results look promising, long-term effects have not been fully investigated. It's important to note that, if restenosis should recur, there is no good method to simply remove the blockage, and follow-up treatments may require laser surgery or localized drug delivery using other angioplasty devices.

Because of these potential problems with metal stents, some researchers are looking at polymeric or even biodegradable ones. These materials usually require processing with an excimer laser to avoid thermal problems and heated-affected zones. In some cases stents have even been inserted after being coated in a restenosis inhibitor such as heparin.

An alternative to metal stents is the use of catheters with orifices located near the insertion end. These orifices are generally drilled using UV photons (in most cases from an excimer laser, although frequency-converted solid-state lasers appear to be a good alternative).2,3 The idea in this instance is to inject small amounts of a restenosis- inhibiting drug locally.

Heparin, usually used as an anticoagulant or blood thinner, has been found to be effective in controlling restenosis when used in high concentrations in localized areas. Since this drug is already well known in industry and among regulatory agencies, it has been used as a test case for localized delivery and proof of concept. Other drugs being investigated are more costly and are not as well documented, and systemic delivery can in some cases be toxic at the levels necessary for successful retardation.

Many organic materials are being used, including polyester, polyimide, polyurethane, nylon, and Mylar. Most UV laser processing is done with 248-nm photons, although 193 nm is a good wavelength when materials have a low absorption at 248 nm. Hole sizes range from 10 to about 100 µm in diameter. Most devices require multiple holes, which can be placed along the shaft in many different orientations. The most common arrangement is a set of linear orifices along the catheter shaft at 180°, 90°, or 45° from each other. Another arrangement is a set of orifices spiraling along the length of the shaft. The holes may be drilled in the central lumen of the catheters (many have only one lumen), or into smaller lumens running along the length of the main, central lumen.

While the typical configuration is a series of individual holes, it is also possible to image a mask containing a number of holes and to drill sets of holes at preferred locations along the lumen shafts. Frequently—especially when the orifices are drilled into the circumferential lumens—blocker material must be used to avoid the laser beam hitting the opposite lumen wall. This blocker is usually a small metal sliver inserted into the lumen during laser processing and then removed. The hole shapes can also vary from the more common circular to square or elliptical shapes.

Another approach is to drill the drug-delivery orifices directly into the inflatable balloon. In this case, the orifices are usually very small in diameter to allow the pressure buildup needed for inflation before actual delivery of the drug. Orifices can cover most of the balloon surface or only portions of it. Some of these devices are drilled by placing the balloon flat and then drilling through both layers simultaneously, while others are drilled by controlling the number of pulses to avoid drilling through the opposing side. In order to accurately control the number of pulses, a consistent and uniform material thickness is required.


A large number of high-tech applications for laser-manufactured devices involve angioplasty in some way, but there are also other applications that use lasers in the manufacturing process. One example is the use of CO2 lasers to drill large holes in simple catheters for liquid and material removal. Because quality in these applications is not as great a concern as functionality and low manufacturing cost, these devices can be made in high volumes at low cost.

Another use of lasers is in the manufacture of analytical catheters or devices that are inserted into the body and whose primary purpose is to give analytical feedback.4 These devices are being manufactured for precise monitoring of vital organs and body functions at unprecedented levels of accuracy—all made possible by the application of laser technology.


It is clear that the number of applications involving the use of lasers in the manufacturing process is increasing rapidly as advances are made in laser technology as well as in other technologies associated with medical device manufacturing. Two of the largest ongoing efforts by today's scientists and engineers are to "engineer life" (medical and biotechnology) and to make engineered devices more lifelike (microelectronics, robotics, artificial intelligence).5 These two efforts seem to converge on a microscopic scale, if anywhere, and both require novel micromachining capabilities that, in many cases, lasers can provide.

Two of the largest industrial users of precision micromachining lasers are the microelectronics and medical industries. This article has covered only angioplasty and catheter applications for laser technology; there are many more medical applications where devices are already in production or in initial engineering phases.

Of all the applications for laser micromachining, at present it seems as though applications in the medical industry will grow significantly over the next several years. Because of the laser industry's relative infancy and the long lead times associated with incorporating new technology into medical manufacturing, laser use in the device industry is still relatively uncommon. However, this is likely to change over the next few years as the industry experiences a period of rapid growth in which the future looks bright for both device manufacturers and the laser community.


1. Schaeffer R, "Laser Micromachining of Medical Devices," Med Plast Biomat, 3(3): 32–38, 1996.

2. Pokora L, "Excimer Lasers and Their Applications in Industrial Technology and in Medicine," Opto-Electronics Rev, 1: 13–20, 1993.

3. Gower MC, Rumsby PT, and Thomas DT, Novel Applications of Excimer Lasers for Fabricating Biomedical and Sensor Products, Bellingham, WA, International Society for Optical Engineering, vol 1835, pp 133–142, 1993.

4. Highlights, Lambda Physik newsletter, no 8, November 1994.

5. Kelly K, Out of Control, Reading, MA, Addison-Wesley, 1994.

Ronald D. Schaeffer is director of corporate development for Resonetics, Inc. (Nashua, NH). The author acknowledges the contribution of Jordan Bajor of LocalMed (Palo Alto, CA).

Copyright© 1996 Medical Device & Diagnostic Industry

Computer-Based Systems Streamline Manufacturing

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996


Michael J. Major

Increased competition, regulatory and cost pressures, and the ever-increasing speed of innovation are the business drivers for today's device manufacturers, says Phil Couling, product manager for Consilium, Inc. (Mountain View, CA), a producer of manufacturing software systems. "Adding to these global challenges are some of the biggest changes in the health-care products industry since the beginning of FDA regulation." According to Couling, key changes affecting the device industry include the following:

  • Over the past 20 years, manufacturing costs have doubled.
  • Buying decisions are driven primarily by the ability of the manufacturer to deliver on time, even when custom specifications and custom packaging are involved.
  • Lawsuits and heightened regulatory scrutiny have increased the need to reduce human error and the risk of liability resulting from insufficient controls, as well as the need to expand the margin of safety.

How a medical device company responds to these challenges will determine its long-term success or failure. To achieve success, these manufacturers are increasingly relying on computer-based systems to help them streamline their manufacturing processes. Leading device companies are employing new advances in computerization to achieve best-practice manufacturing benchmarks (see box, page 68).


The use of computerization to aid in medical device manufacturing is, of course, hardly new. "I don't know how a company can run a manufacturing facility without computers," says Kathy Reading, MIS administrator for Utah Medical Products, Inc. (Midvale, UT). Roy Peters, an engineering consultant with the California Manufacturing Technology Center (Anaheim, CA), agrees: "There are hardly any manufacturers today that don't rely on at least manufacturing resource planning [MRP]. It's a way of life. Companies have to track materials and control their inventories in order to control costs and become and remain competitive."

But it's these competitive pressures and, especially, FDA regulations that are driving computerization forward—both for manufacturers of medical devices and for the vendors that supply computer manufacturing systems. As an example of the latter, John Bridges, marketing communications manager for Industrial & Financial Systems (formerly Avalon Software; Tucson, AZ), says, "It's good business practice to track the manufacturing of a product and to know how the customers use it and whether they are happy with it. But in the device industry there's no other choice—it's the law. Previously, computerization has come in bits and pieces. Because the medical device market has not been big enough to support it, there has never been a complete, integrated, cradle-to-grave system for device manufacturing. But now, it's clear that product liability concerns will sooner or later force all other industries to have such a complete system. So it's worth our effort to have this kind of strength. The medical device market is pushing the development of our software systems."

To see how these systems are developing, it may be helpful to take a quick look at the current batch of acronyms. As with most computer acronyms thought up to convey the image of something new, the definitions are a bit elastic and can vary somewhat with the speaker. Reading, for instance, says, "ERP [enterprise resource planning] is just a new term to describe MRP." All the same, there are some basic differences between the key buzz words now in use to describe computerized manufacturing systems.

Manufacturing resource planning is the most established, of course. MRP has to do with the automation of information directly related to manufacturing, including inventories, bills of materials, and orders from purchasing. MRP II takes this process a step further into capacity requirement planning (CRP), in which the general information of MRP is put to more specific uses, such as scheduling how many devices can be built within a certain period.

ERP, as its name implies, purports to integrate manufacturing information with all data throughout the company—including those related to sales, marketing, accounting, shipping, and so on—to enhance a company's ability to plan, provide better control of operations, and enable products to get to market more quickly. Directly opposite ERP is process control or computerized numerical control (CNC), which has to do with the loading of data into the machines that actually make parts on the shop floor.

In between ERP and CNC is the manufacturing execution system (MES). Where ERP is basically a transactional, purely informational system, MES operates in real time, pulling together all of the information directly related to the manufacture of a product—including raw materials, equipment, personnel, work instructions, and stipulation of the correct processes. After downloading this information from the ERP system for the actual making of a product, the MES uploads new information, such as the amount of material used and the number of units made, back to the ERP.

These definitions are somewhat schematic, and any particular manufacturing software package may actually perform more or fewer of the various activities described. But before exploring some of the more sophisticated or all-encompassing claims of some software vendors, it's helpful to realize that for many device manufacturers a simple MRP system may be all that is needed.


"We've come to realize that there are a lot of small medical device manufacturers doing less than $20 million in sales each year," says William Urschel, president of Alliance Manufacturing Software (Santa Barbara, CA). "Many are involved in making one of the many niche products in the marketplace, such as breast implants, pain management implants, or various neurological devices. When some of these manufacturers hear the term enterprise resource planning, they may think they need it—even though they're not yet even using MRP II. If a company is doing $50 million in sales, then it might want to start adding some bells and whistles to its manufacturing systems. But if it's still on its first computerized accounting system, it makes no sense to jump way ahead." Urschel, whose company specializes in computerized systems for small manufacturers, points out that only 9% of all the manufacturers in the United States earn more than $100 million in annual sales. "This means that 91% of all manufacturers are in our market," he adds. "It's incredible, but a survey of our customers suggests that the vast majority need just basic computerization to fulfill their manufacturing needs."

Although it may be wise to start small, California Manufacturing Technology Center's Peters suggests that "it's very important for the small manufacturer to get an MRP system that is expandable. Unfortunately, a lot of small manufacturers don't investigate this, so they end up stuck with a rigid, obsolete system that can't adapt."

Utah Medical Products is one company that was stuck with an older system, which, as Reading explains, "was too much of a proprietary-based system. It was not easy to customize and did not allow us to interface with other databases." The company turned to a new ERP system from DataWorks, Inc. (San Diego).

"It was a big project to change systems," Reading says. "We changed to our new program about one-and-a-half years ago. We spent a year investigating different vendors and about six months actually making the transition. It represented a huge investment in both time and money." About 45 network users access Utah Medical's new software off the company's mainframe computer, while another 100-plus users are able to access other software packages such as CAD systems and various spreadsheets. "It will take us a long time to bring up all the bells and whistles on the new system, but we are quite pleased with it."

Reading says that Utah Medical, which focuses on the manufacture of gynecology devices, was looking for a vendor "that we felt had positioned itself to look toward the future and to adapt to changing technology. The worst thing one can do is to put a lot of time and money into a system that cannot change, and then be stuck with it."


While moving into more sophisticated systems may be beneficial, or even necessary, it's not necessarily easy. "Our products have had the most success with those customers who fully understand the scope of their endeavor up front," says Consilium's Couling. "There are real benefits to be gained from implementing a manufacturing process within an electronic system, but one should never underestimate the complexity of the implementation."

Not only can the implementation of such systems be complex, but medical device manufacturers must also undergo a complex process to validate the use of the software for FDA. Robert George, director of industry analysis for Advanced Manufacturing Research (Boston), explains that "up to now, each user's validation of its software system has been an extremely tedious process. Each customer has been required to validate separately to FDA that its software supplier used the proper methodology in developing the software. For most companies, this is extremely onerous, and it has created a big bottleneck."

However, some relief may be in sight. "We've seen the rise of a number of third parties that are making a practice of validating software, and are auto- mating the process," says George. "And we're also seeing some ERP vendors take a fairly aggressive stance in validating their own software for FDA requirements."

Chris Jensen, corporate communications director for DataWorks, says that her company is one such vendor offering software that is functionally compliant with FDA regulations. Features of the software make traceability very easy. "We've found a lot of users that have one set of software for financial management and another for manufacturing, but there is a disconnect between the two," says Jensen. "Our ERP package brings them together."

One device manufacturer that uses the DataWorks software is Kaye Instruments, Inc. (Bedford, MA). Kaye is a pharmaceutical company that validates the sterilization process used in the manufacturing of intravenous drugs. "Our instrumentation and data collection systems require ultrahigh accuracy," says Tim Donaghey, the company's operations manager. "We run all aspects of our business on DataWorks, from accounting and engineering through sales and manufacturing."

Donaghey says his company's prior system was "a mixed batch of assorted third-party packages that required extensive maintenance and information-systems support, and were not integrated. By contrast, the DataWorks system works in fully integrated, real-time fashion. It has a lot of built-in administrative tools that allow us to tailor the software to effectively running our business."

Industrial & Financial Systems' Bridges reports that his company "is using the latest technologies for dealing with databases and document management, and integrating them with our strengths in manufacturing management software to create one large, all- encompassing enterprise system. We call it cradle-to-grave or complete product life-cycle management. The manufacturer can track and trace the entire process from product initiation to completion. And, when there is a failure in the field, the company can use the software to audit the production of the failed device. The system is even designed to handle the problem through closed-loop corrective action."

Ann Papenfuss, regional sales manager for ASI (St. Paul, MN), reports that her company's MES packages are capable of delivering a variety of functions, such as computer-integrated manufacturing (CIM), real-time statistics, loading and downloading of manufacturing data, G-code editing, preventive maintenance, tooling and gage management, document management, and several types of connectivity.

One aspect of ASI's package, Papenfuss explains, enables employees on the shop floor to view any kind of setup, testing, assembly, machining, or related procedures. "For medical device manufacturing, it's important to protect the integrity of all documentation," Papenfuss says. "With a lot of electronic viewing packages, someone can go in and edit a document, either purposely or inadvertently. Our package includes a view-only program that prevents that from happening."

ASI's package also comes with a redlining feature. If an employee sees a problem on the production floor, he or she can place a red line on a similar, but completely different document, along with corrections and additional information for anyone who wants to view it. "This way, people don't keep rediscovering a mistake," says Papenfuss. "The employee can E-mail the information back to the engineering or design department to make the appropriate correction." The documents can then be put on engineering hold, while production is either given or denied viewing access.


Consilium's Couling maintains that his company's MES system "provides excellent visibility for the manufacturing history of a medical device, including traceability from raw materials through to the finished product." Couling says that the company's product is designed to reduce time to market, decrease new product development costs, cut down manufacturing cycle time, optimize manufacturing performance, improve regulatory compliance, and automate product data management and change processing. "We've had several sites report that they've gone to completely paperless operations," he says.

Another vendor who maintains that his company's software has created a completely paperless environment is Bill Atwell, president of RealVision, Inc. (Pomona, CA). "In the environment of our software, drawings, work instructions, tool lists, processing instructions, and inspection sheets can all be made available instantly through network PCs," he says. "What this means is that a company can vastly improve its productivity by eliminating the need to pass folders and papers around the shop floor. This can be critical if a change occurs on the shop floor and the manufacturer needs to prevent future units from being made with a bad or nonfunctioning part.

"Also, for complying with FDA regulations, the manufacturer has complete control over its documentation. Documents that are processed electronically always include up-to-date revisions, and the system also provides for document approval by means of electronic signatures."

Atwell helps us visualize what an advanced computer system can do for documentation, by contrasting computerized operations with the way manufacturing has been done in the past. "In the recent past, a manufacturer would use an MRP or accounting system to build a router, or work order—an itemized list of the steps required to make a component for an assembly. Together, these steps—even simple ones such as cutting stock, turning it, polishing it, inspecting it, and shipping it—constitute the manufacturing process.

"To document the manufacturing process, the company has to have a description of each of these steps, as well as any paper documentation needed to supplement the descriptions, such as blueprints. Next, the manufacturer has to prepare a process sheet that shows an operator how to machine the material, or a computerized word address program for making the part on a CNC machine. In addition, the manufacturer may have an inspection report, statistical process control documentation, and shipping instructions.

"In other words," Atwell continues, "traditional manufacturing required a lot of paperwork to be retrieved from notebooks or file cabinets—and then someone had to bring it all to the people on the shop floor.

"Now, we've allowed computer technology to replace paper completely. Instead, documents that are created electronically are tied into the existing MRP or accounting system, making an electronic router. All of the documents are then immediately accessible to everyone who needs them, enabling the manufacturer to manage the manufacturing process out on the shop floor. The entire process of gathering documents—viewing, revising, releasing, dispersing, and controlling them—is handled by software."

In the past, says Atwell, MRP and accounting systems took into account such factors as time and labor, but never captured the work instructions necessary to run a job. Meanwhile, the engineering department would have an entirely different process for drawing up blueprints and manufacturing instructions, and manufacturing would require yet another set of more detailed, nuts-and-bolts information. "What all this has meant is that people have walked from department to department to convey information," says Atwell. "But when it's done electronically, everyone can work concurrently."

One user of the RealVision system is Nellcor Puritan Bennett (Carlsbad, CA). The company's senior production technician, Ray Bronoske, says the system is very helpful, "especially compared to what we had before. Now we can manage just about all our documentation. We use the system to keep track of inspection records, and also to load programs into the machinery. If someone does some editing on a machine's programming in order to adjust the process, this information can be uploaded to the software system for later reuse. There's no need to risk making a mistake when retyping changes, because they're already in the system. The software is also useful for managing the presetting of tooling prior to use, and for handling quick-change tooling. It cuts setup time approximately in half."


There's no reason to think that any of the competitive or regulatory pressures that have been driving advances in computerized manufacturing will go away any time soon. So device manufacturers will probably be faced with the need to streamline their processes even further in the years to come. Companies will continue to feel pressure to integrate their manufacturing systems with other business systems, and to make even greater use of the planning capabilities of such systems.

There will be payoffs for the use of such systems—if only in the fact that the company is able to remain in business and stay competitive. In the regulatory arena, manufacturers may find that computerized manufacturing systems are a key element in maintaining the documentation required by FDA's new design control regulations. Similarly, computerized systems can make it possible for a company to participate in FDA's new program for electronic product submissions.

In the meantime, the marketplace includes a wide range of computer systems designed to streamline manufacturing, and many include features intended especially for use in a regulated environment. Device companies may be challenged by the marketplace, but if they look hard enough they should have no trouble finding a manufacturing software package that will meet their needs.

Michael J. Major is a freelance contributor to MD&DI.

Advances in computerization are enabling manufacturers to routinely achieve high standards of performance. According to Phil Couling, product manager for Consilium, Inc. (Mountain View, CA), the following figures represent current best-practice benchmarks in key areas of medical device manufacturing.

Time to market: Indirect/direct labor ratio = 1:1

Manufacturing cycle time = 1.5 theoretical

Waste: Quality = 6 sigma

Customer service: >99%

Regulatory compliance = 100%

Safety compliance = 100%

Resource utilization: >70%

Return to article

Definitions for many of the terms used to describe manufacturing software tend to be somewhat flexible. Since not all systems contain equivalent features, a company that is seeking to purchase a system should be certain that the one selected includes everything the company needs. The following definitions should provide a starting point for discussions with suppliers.

Capacity Requirement Planning (CRP). A subsystem of MRP II (see below) that enables manufacturers to plan and schedule equipment use and production.

Computerized Numerical Control (CNC). A system that permits manufacturers to control manufacturing equipment by means of data loaded into their computerized memories.

Enterprise Resource Planning (ERP). A computerized system for integrating data from throughout a company in order to improve planning activities, provide better control of operations, and enable products to get to market more quickly.

Manufacturing Execution System (MES). A real-time system for coordinating all data relating to the manufacture of products and applying them directly to shop floor activities.

Manufacturing Resource Planning (MRP). An automated system for handling information directly related to manufacturing, including inventories, bills of materials, and orders from purchasing.

Manufacturing Resource Planning II. An expanded version of MRP that includes enhanced capacity for planning and scheduling the use of manufacturing resources.

Return to article

Copyright© 1996 Medical Device & Diagnostic Industry

Evaluating Sterilizer Performance as Part of Process Equivalency Determination

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

Originally published November 1996

Paul J. Sordellini and Vincent A. Caputo

Medical device manufacturers that use ethylene oxide (EtO) to sterilize their products, either in-house or at a contract sterilizer, often validate the cycle in one particular vessel and later find it no longer satisfies their requirements. A validated vessel becomes outgrown when the number of devices manufactured per week exceeds its volumetric sterilization capacity. Increased market demand for a particular product, the merger of one manufacturer with another, the expansion of a particular product line, or expansion into foreign markets may cause a manufacturer to go from comfortably processing 5 sterilizer loads a week to needing to process 10 loads, thus creating an ever- increasing backlog.

Consider, for example, a shallow-vacuum cycle engineered for pressure-sensitive products. Among the time-consuming attributes of such a cycle are multiple shallow vacuums and slow ramp rates. When these process steps are added to a 6-hour sterilant dwell, the total resident time per lot inside the vessel may reach 18–20 hours. Using a single vessel, a contract sterilizer could treat only 7 or 8 pressure-sensitive loads per week (assuming the contractor had no other customer using the same validated chamber—which is unlikely).

Cycle optimization is often considered as a means of reducing chamber dwell time to allow more process cycles per week. However, optimization of an already validated cycle is rarely a viable alternative since any modification of critical parameters will usually compromise the original validation. More often, solving the problem of insufficient capacity requires the use of additional sterilizers. If another vessel of identical size and configuration is available, then the protocol used in the validation of the first vessel may be followed to validate the second one. Another possibility is for the manufacturer to increase the size of each production lot and use a larger sterilizer. In this alternative, the volume of product treated per cycle increases while cycle time remains the same. The amount of related paperwork and QA review will also remain constant for the larger vessel.

In either case—using a second identical vessel or a larger nonidentical vessel—the same issue arises: If the sterilization cycle has already been fully validated and proven effective in the first vessel, and the second vessel performs equivalently, is it necessary to perform a complete validation in the second vessel? Or is there enough scientific support and regulatory tolerance to accept a situation in which a primary vessel is selected and fully validated (physically and microbiologically) while secondary vessels are determined equivalent after a reduced validation (full physical qualification but reduced microbiological qualification)? Both the manufacturer and contract sterilizer would like to avoid performing a full validation of the second vessel, which can take months to prepare. Obviously, the sooner the cycle can be run in the second vessel, the sooner more product can be processed and shipped to satisfy market demand. In addition, a number of sometimes-costly product samples are destroyed during validation-related testing, and the laboratory testing itself is expensive.

Unfortunately, no guidance documents have been available to enable manufacturers and contract sterilizers to determine when, after a full validation is completed for one vessel, additional vessels can be considered equivalent to the first in their ability to deliver the same set of process parameters and obtain the same level of sterility assurance. Probably because of this lack of a standard policy for determining vessel equivalency assurance (VEA), the sterilization industry has remained very conservative, limiting itself to demonstrating process equivalency only between vessels of identical size located within the same facility. In a nutshell, industry practice has been to commission every vessel, conduct empty-vessel test studies in every vessel, perform a full microbiological qualification in one vessel, and then perform a reduced microbiological qualification of identical vessels for the same product family and process. The contents of this reduced equivalency validation have been known to vary from a series of sublethal challenges to simply running one half cycle and one full cycle.

Any standard method of determining process equivalency among sterilizers must begin with a thorough assessment of their physical performance. Only after a physical comparison of the vessels has been successfully completed can a facility think of proceeding with a comparison of their microbicidal performance. Therefore, performing a VEA protocol requires a thorough understanding of how to physically evaluate a vessel. The next section of this article is intended to be a guide to the basic physical testing of industrial-size EtO sterilizers, thus enabling companies to take the first step toward VEA. The section after that discusses the reduced microbiological qualification requirements for validating additional vessels that are now being developed by a task group of the Association for the Advancement of Medical Instrumentation (AAMI).


The basic elements for physically qualifying individual sterilizer vessels are detailed in Medical Devices—Validation and Routine Control of Ethylene Oxide Sterilization (ANSI/AAMI/ISO 11135). Upon installation of a vessel, a full commissioning or installation qualification (IQ) is performed. Regardless of whether VEA is intended, no part of this commissioning may be excluded. The IQ process is followed by a performance qualification (PQ), which ANSI/AAMI/ISO 11135 divides into physical qualification and microbiological qualification. Both are fully carried out if no VEA is planned. However, if process equivalency is intended, all parties involved have a mutual interest in avoiding as many repetitious microbiological validation sequences as possible. Physical qualification includes a number of studies performed on empty vessels, including leak tests, wall profiles, and operational qualifications. Each of these test methods is described below, along with several enhancements that are useful to the validation engineers performing the comparative studies needed to determine VEA.

Leak Tests. Multiple leak tests, both under vacuum and under pressure, are conducted during physical qualification to attest to the correct fitting and sealing of all vessel doors, gaskets, valves, welds, and penetration points. The unintentional introduction of air into a sterilizer during operation can alter the efficiency of both the humidification and the sterilant phases of the process cycle. To determine the leak rate, the vessel is pressurized (to 20.0 psia, for example) and left for 12 hours at a steady temperature. During that time, measurements of pressure, jacket temperature, and the vessel's internal air temperature are logged at intervals (commonly every 5 minutes). Next, a vacuum is drawn inside the vessel (down to 1.0 psia, for example) and the same data-logging procedure is followed for another 12 hours.

The main concern of the validation engineer is the total variation in pressure during the respective 12 hours under pressure and under vacuum. Respect for the leak-test tolerances, which are defined in the IQ protocol, and repeatability of the results must be clearly demonstrated. Because the protocol will include such specifications as the minimum and maximum allowable operating pressures for the vessel, all vacuum and pressure leak tests should be engineered to reflect and validate those parameters.

Wall Profiles. The purpose of a wall profile is to map and evaluate the heat-circulation patterns throughout the vessel's system of jackets. To obtain the necessary data, temperature probes are fastened to the six internal surfaces of the sterilizer. The number of probes will vary with the size of the vessel. (Guidance in this area is offered by ANSI/AAMI/ISO 11135, Annex B, paragraph B.2.3.2.) A location diagram of the probe placement must be designed so as to ensure that all internal surface areas are monitored. This diagram becomes a permanent part of the vessel's performance data file.

If VEA is being sought, it is very effective at this point to also determine the temperatures of the vessel's various jacket contents (air, water, steam, or oil). One or more probes can be inserted directly into the fluid pathway within each jacket and the resulting temperature data can be sent to a recorder for later comparison with the temperature spread verified on the internal surfaces of the vessel. Analysis of these comparative data will provide an insight into the vessel's thermal conveyance efficiency. The closer the temperature spread of the internal walls is to the temperature of the fluid content of the jackets, the more efficient is the overall heating system of the vessel. Among other things, this efficiency analysis will indicate whether the vessel (including its feed and return heating lines) is properly insulated, so that heat loss to the surrounding work area is minimal. It will also confirm that the flow within the jacket system is adequate to heat all areas of the vessel to required temperatures.

Once the vessel and jackets are probed, the vessel is closed and a jacket temperature set point is programmed. The vessel is allowed to reach thermal equilibrium by, for instance, leaving the jacket system set at 110°F for 24 hours. Data collection then begins and continues for 12 hours, after which the jacket set point is raised (to 120°F, for example) and data collection continues for another 12 hours. After the test, all data are analyzed for two purposes. First, the range of temperatures from the probes is reviewed to determine the evenness of the temperature spread throughout the vessel and to identify any cold spots. Second, a graph is created and statistical analysis is performed on the data compiled when the jacket temperature was raised to determine the vessel's thermal response time (that is, the time lag between the sterilizer control system's call for additional heat and the actual arrival of that heat to the vessel walls as well as the time to attain a new state of equilibrium). A faster transfer of heat on the jackets' internal walls may contribute to reducing the heat-call response time of the vessel.

Operational Qualifications. Once acceptable results are achieved for the leak tests and wall profile, an operational qualification (OQ) study can be performed on the empty vessel. To prepare for this study the vessel is programmed for a cycle and conditions that closely emulate those expected to be encountered during routine procedures. The ability of the sterilization system to accurately and repeatedly attain all critical parameters is revealed during the OQ.

Following a geometric pattern defined in the OQ validation protocol outlined in ANSI/AAMI/ISO 11135, probes are suspended in the vessel so as to occupy space that would normally hold product. (Guidance in determining the correct number of probes is given in ANSI/AAMI/ISO 11135, Annex B, paragraph B.2.3.2.) According to the document, the object of the study is twofold: (1) to establish the correlation between the test probes and the permanent wall-mounted temperature probes used to monitor and control the vessel's heat supply, and (2) to confirm that heat is distributed evenly throughout the vessel. Thus, for a well-functioning vessel, the OQ will demonstrate that the average of the temperature-control probe data is within a tight range of the preprogrammed air-temperature set point, and that all the test probe readings fall within ±5.4°F of the air-temperature set point. These temperature-limit requirements may be altered provided the change is scientifically supported. If parametric release is planned, for example, requirements may need to be more demanding to demonstrate greater control of the process.

In most cases, an example of an acceptably functioning vessel would be one with a set point of 120.0°F, a verified average temperature-control probe range during the sterilant dwell phase of 118.0°–122.0°F, and a suspended probe temperature spread during sterilant dwell of 114.6°F–125.4°F. Successful completion of at least three consecutive OQ cycles, with consistent results, ensures the proper functioning of the sterilizer and ancillary equipment. If significant changes to the vessel's heating or recirculation systems are required to meet this goal, the wall profile and operational qualification will need to be redone to evaluate the consequences of the changes.

As was suggested during the discussion of wall profiles, OQ studies can be enhanced to provide more performance data for use in a VEA evaluation. The cycle's sterilant dwell phase should be allowed to proceed until a thermal equilibrium has been reached throughout the vessel. The air-temperature set point should then be increased at least 10°F and the dwell continued until attainment of a new state of equilibrium. The collection of these data will enable the valida- tion engineer to compare the heating char- acteristics of two or more vessels. A careful review of all ramp rates, set points, and temperatures achieved will confirm that the vessels perform each segment of the cycle in an identical fashion.

It is important to stress that the practical value of an OQ study is limited. During the initial vessel-commissioning process, an OQ will immediately reveal the failure of any system component (the recirculation blower, jacket feed pump, temperature-control probes, vacuum pump, gas valves and actuators, and so on). Also, if OQs are repeated semiannually, a comparison of the data collected over time can expose changes in vessel performance. For example, if the probes' temperature range around set point is expanding with each OQ study, it may indicate that some component affecting even heat circulation may require replacement. However, a genuine assessment of a sterilizer's performance can occur only when the vessel is operating with a product load of maximum density. Little heat and recirculation is needed to successfully run an OQ, but with a full vessel load of actual product, all the dynamics concerning heat control and distribution can be realistically studied.

When vessel equivalency is being evaluated, a comparison of leak rates, wall profiles, OQ data, and thermal response times will reveal which vessel should be the primary vessel (which is subject to full microbiological qualification) and which should be the secondary vessel or vessels (cross-validated for the same product family using a reduced microbiological validation). To offer the maximum challenge to the process validation, the vessel with the fastest leak rate, the widest temperature range during the wall profile and OQ studies, and the slowest thermal response times should be the primary vessel. Because the secondary vessels will have more-efficient results during the above-mentioned tests, the use of a less-stringent microbiological qualification will be scientifically justified.


Any standard protocol to determine equivalency among multiple vessels must include extensive physical testing. However, as mentioned above, once this testing is completed and does in fact present data attesting to the similarity of the vessels' performance, a reduced microbiological qualification, which saves both time and resources, can be performed for the secondary vessels following a guidance soon to be published by AAMI.

In June 1995 a task group within the association began working on a technical information report (TIR) to be titled Engineering Aspects of Industrial EO Sterilization. Since then, the guidance document has undergone three revisions. After each draft was circulated to all members of the group, written comments were received and discussed, resulting in modifications to the document. The fourth draft was scheduled to be presented and reviewed by the group this past September during a meeting in Washington, DC. It was expected that the final proposal would then be balloted and published in early 1997.

In its current form, this TIR contains a section dealing with demonstrating process equivalency for multiple preconditioning rooms, sterilization vessels, and aeration rooms. While it advocates a full commissioning and physical performance qualification for every vessel regardless of its size, it also presents a strategy for allowing reduced microbiological qualifications of secondary vessels targeted for process equivalency. The TIR places little importance on a mere structural comparison of the equipment; instead, it suggests that process equivalency be based on the equipment's ability to consistently deliver the same set of physical process parameters. Such equivalency can be demonstrated through a comparison of the commissioning and physical performance testing data. The TIR also defines those cases in which, after a complete microbiological qualification of one vessel or room, subsequent vessels and rooms may be declared equivalent and validated through reduced microbiological qualification.

Once published, the AAMI document will help gain industry and regulatory agency support for process equivalency protocols. The TIR targets every possible scenario: identical vessels and rooms in the same or in different locations, and nonidentical vessels and rooms in the same or in different locations. Suggestions for validation of process equivalency are made for each of the four possible situations. The TIR also offers guidance in using statistical process capability indices to further strengthen the validity of the results. Basic formulas to calculate process capability (Cp) and the process capability index (Cpk) are provided, along with guidance for interpreting the resulting standard deviation data.


Engineering Aspects of Industrial EO Sterilization, AAMI Technical Information Report, 4th working draft, Arlington, VA, Association for the Advancement of Medical Instrumentation (AAMI), June 1996.

Guideline for Industrial Ethylene Oxide Sterilization of Medical Devices, ANSI/AAMI ST27, Arlington, VA, AAMI, 1988.

Hoborn J, "Ethylene Oxide Sterilisation—A Proven Method," in Ethylene Oxide Sterilisation Conference Proceedings 1989, London, European Confederation of Medical Devices Associations (EUCOMED), pp 33–50, 1989.

Medical Devices—Validation and Routine Control of Ethylene Oxide Sterilization, ANSI/AAMI/ ISO 11135-1994, Arlington, VA, AAMI, 1994.

Perkins JJ, Principles and Methods of Sterilization in Health Sciences, Springfield, IL, Charles C. Thomas, 1983.

Winckels H, "Equipment and Process Validation," in Ethylene Oxide Sterilisation Conference Proceedings 1989, London, EUCOMED, pp 13–23, 1989.

Paul J. Sordellini and Vincent A. Caputo are industry consultants with Quality Solutions, Inc. (Annandale, NJ), which is a participating member of the Ethylene Oxide Sterilization Association.

Copyright© 1996 Medical Device & Diagnostic Industry

QUALITY SYSTEMSGetting Compliant for theGlobal Market

Medical Device & Diagnostic Industry Magazine | MDDI Article Index

Originally published November, 1996

Gunter Frey and Kimberly Hughes

[Related Links]

Although it is still more than a year and a half away, the date of June 15, 1998, is approaching very rapidly for device manufacturers seeking to begin or continue selling in the European Union (EU). That is the day the EU's Medical Devices Directive (MDD) will become fully effective throughout the member states of the union, and when device manufacturers will have to either demonstrate compliance with the directive or cease selling in the EU.

Among the most important terms of the MDD are specific requirements for quality systems that must be observed by manufacturers. The directive does not explicitly mention any of the quality systems standards commonly used in the United States or Europe, but it is clear that the EU expects manufacturers to adopt and comply with such standards as are appropriate to their particular industry.

With the growing importance of overseas markets and the rapid globalization of the device industry, most U.S. manufacturers are faced with the task of developing and implementing a quality system that complies with both international and domestic requirements. As a guide to those who are about to undertake this daunting process, this article describes how our company, Pryon Corp. (Menomonee Falls, WI), successfully accomplished certification to the ISO 9001 and EN 46001 quality systems standards (see Figure 1, below). The article poses many of the same questions we raised during our preparatory work, and provides some resources to help obtain the answers. Although our approach cannot be expected to work for every company, our experience may benefit those who will shortly be guiding their own companies toward a certification audit.


The first step for any company preparing for its initial audit is to become familiar with the relevant standards and available resources. Since standards are revised over time, companies should be sure to obtain the most recent versions. Copies of the EU directives, relevant standards, and guidance documents are available from a number of organizations, including the American Society for Quality Control (ASQC), the Office of European Community Affairs, and the National Institute of Standards and Technology. Numerous books, videos, and software packages are also available to provide either a general overview or answers to detailed questions. Information about these resources can be found in trade publications as well as in the publicity materials issued by their publishers.

Among the most helpful sources of information is the National ISO 9000 Support Group. This group provides an independent arena for discussion, advice, and information exchange. It promotes a better understanding of the standards, the steps involved in implementation, the benefits of certification, and the process of becoming certified. Members of the group represent a great deal of ISO 9000 experience. The support group provides members with general ISO information, ideas on where to go for additional information, tutorials, the Continuous Improvement Newsletter, and discounts on publications. In addition, the support group is accessible via the Internet. The group can be contacted at 616/891-9114.

While the goal in this step is to gather as much information as possible, it's equally important not to become overwhelmed. This phase is comparable to a first semester at college, and it takes roughly the same time--three to four months--to comprehend the unfamiliar terminology and get a general sense of what lies ahead.


The second step is to determine which standards apply to the firm and its products. For device manufacturers, the most widely used quality management and quality assurance standards are those of the ISO 9000 family, compiled by the International Organization for Standardization (ISO). A virtually identical series of standards have been published as European norms under the designation EN 29000, and the requirements of the ISO standards are also reflected in FDA's revised good manufacturing practices (GMP) regulation.

The ISO 9000 series includes five parts, the first and last of which (ISO 9000, "Guide to Selection and Use," and ISO 9004, "Guide to Quality Management Elements") assist manufacturers in determining which of the remaining three standards apply to them. Pryon Corp. conducts research, development, manufacturing, installation, and servicing of CO2 monitors, so our goal was to become certified to ISO 9001, which covers design, production, and installation. Some companies approach ISO 9001 certification by first attempting certification to ISO 9002 (production, installation, and servicing) and later expanding their quality system to include design and development.

To do business in Europe, most medical device manufacturers will also have to comply with another set of European norms, designated the EN 46000 series, which translate the general requirements of the ISO 9000 series for use specifically by manufacturers of medical products. Compiled by the European Committee for Standardization and the European Committee for Electrotechnical Standardization (CEN and CENELEC), the EN 46000 standards mandate additional requirements for medical products to be sold in the European market. For example, requirements of EN 46001 include or refer to postmarket surveillance activities, customer complaint handling, advisory notices, and recalls.


Once the company has determined which of the various types of quality certification will be required, it is important to ensure that executive management is fully involved in the process. Even though company executives may have identified the need for certification and may have recommended creation of an audit team or committee, they may still require some education about the complexities of the certification process.

Some of this education takes place gradually, as managers observe preparations and are given updates on the company's progress. But it is also necessary to conduct meetings with executive management apart from the training classes held for all staff. At Pryon, management was supplied with an outline of the audit and certification process, and was continuously updated on schedules, activities being performed, and tasks still remaining to be done.

At most companies, the individual or committee coordinating the certification process is also responsible for establishing the company's training requirements. But to ensure that executive managers are fully engaged in the process, it is a good idea to involve them in this process and to update them regularly. Executive management can provide valuable assistance in developing the company's training programs, and in supporting travel to training seminars for key employees.

Without executive management's continuing commitment and support, it will be difficult to achieve successful completion of the certification program. But since preparation for the certification audit is complicated and time-consuming, commitment may wane at times. To keep managers motivated, it may be helpful to remind them that registration is intended to ensure that the company's products will have ready access to important markets. At Pryon, management's commitment was evidenced by its support of all activities, and its active participation in the design and conduct of internal audits and employee training.


After completing the initial phase of information gathering described above, companies must undertake the more difficult task of understanding and interpreting the materials. Even with many books and videos in hand, company staff will still need help interpreting portions of the standards and answering questions from their fellow employees. A number of resources can help to resolve such difficulties.

One good approach is to ask other firms how they had addressed shortcomings in their quality systems. Peers and business contacts within the medical device industry can be invaluable sources of information. If they have already completed an audit, they speak the language, know the process, and may be willing to share their experiences, problem-solving methods, and advice. When making these contacts it is best to start with friends and then branch out to peer companies, and always ask for recommendations of additional contacts. Also, callers should have ready a list of detailed questions; people are rightfully reluctant to take time from their schedules to answer vague or open-ended questions.

Another resource is FDA's Division of Small Manufacturers Assistance (DSMA), which can supply copies of key agency documents related to quality systems. One such document is the "Working Draft of the Current Good Manufacturing Practices (CGMP) Final Rule," [Now superseded by the October 7, 1996 Final Rule--Ed.] which was released by the agency in July 1995. It includes a preface that summarizes comments from device industry experts about the proposed regulation as well as FDA's position toward those comments. Because it offers insight into the ways the agency has adapted elements of ISO 9001 and EN 46001, the July 1995 draft is still useful even now that the revised GMP regulation has been issued. Experts at DSMA are also available to answer questions regarding the revised GMPs.

Although our company elected not to retain a consultant, there are many such companies and individuals that can answer questions and provide information and assistance. Lists of such service providers are available through many sources, including the National ISO 9000 Support Group and the Internet. Before Pryon decided to conduct its own preparatory work, the company obtained quotes from several consulting services. Our research indicated that the average fee for conducting an ISO 9001 gap analysis for a single-site company of 100 employees was approximately $6000; the average fee for a certification audit was $19,000. Totaling $25,000 (excluding traveling and lodging expenses), these fees did not include certification to EN 46001 or the MDD. Our approach enabled us to limit expenses to about two-thirds of the prices quoted, and also to include certification to EN 46001 and annex II of the MDD.


Just as the regulatory regime in the EU differs from that of the United States, so also does the structure of authoritative bodies that administer it. Each member state of the union has a competent authority, usually a governmental agency, which is charged with carrying out the requirements of the EU directives. Competent authorities determine whether a certification organization is qualified to act as a notified body, and they also oversee the auditing of notified bodies to ensure that they remain qualified.

A notified body is a certification organization that has been designated to carry out one or more of the audits described in the annexes of the directives. There may be more than one notified body in any given member state. In many cases, independent European test agencies have achieved qualification to serve as notified bodies. However, an organization's designation as a notified body may be restricted to specified types of devices, so companies must be sure to choose only those that are notified for their products. The EU's Delegation of the European Commission maintains a list of all notified bodies (currently about 34 organizations), which includes their names and addresses; their identification numbers; and the products, procedures/modules, and annexes of the directives for which they are notified.

Notified bodies assume full and final responsibility for the performance of audits, but they may subcontract with a registrar to perform the actual work of the audit. Registrars are organizations that assess and certify companies to the appropriate ISO 9000 and EN 46000 standards. While there are many registrars in North America that can provide ISO 9000 certification, not all are recognized in Europe, and not all can certify companies to the EN 46000 series of standards. In the United States, registrars are accredited by the Registrar Accreditation Board (RAB) using criteria based on internationally recognized standards and guides. The RAB is a private, not-for-profit organization that can provide manufacturers with a list of registrars; it can be contacted at 800/248-1946.

To avoid the possibility of working with a registrar whose accreditation might prove unacceptable in Europe, our company determined to select a notified body that would be responsible for the competence and qualifications of its auditors. We first determined which notified bodies subcontracted with or accepted audits by registrars located in the United States. Then we again contacted our peers to ask them about the notified bodies they were working with. Companies that elect to use this approach should be sure to have a set of short and direct questions that will elicit the information they need, such as the following:

  • What is the company's impression of its notified body?
  • Is the company satisfied with the notified body's services?
  • Has the company encountered any difficulties with its notified body?
  • How responsive is the notified body?

Once the answers to these questions have been digested, the company should then prepare to contact various suitable notified bodies. Not every notified body is a candidate; a company may decide to interview only those located in the country where it does the most business, or only those in a country where it has an office. With the help of feedback from its peers, the company should be able to narrow its list to about 10 notified bodies of interest.

When the manufacturer begins to interview notified bodies, it is important to conduct information gathering according to a well-structured process so that each firm can be compared with the others. There are two reasons for contacting notified bodies. The first is to obtain a detailed quote and related information, including answers to the following questions:

  • Does the notified body have one or more U.S. offices? If so, do those offices have personnel knowledgeable about your company's type of products?
  • How many clients has the notified body audited? Does it publish a list of these clients?
  • How many people will the notified body send, and from where?
  • How long does the notified body anticipate its auditors will be on-site?
  • Will the notified body try to combine the auditors' visits with those to other customers? (If so, it may be possible to divide travel costs with them.)
  • How often does the notified body perform follow-up audits?
  • What are the costs for the initial audit and for follow-up audits?
  • What is the margin of error on the costs the notified body quotes?

Overall, the manufacturer should try to determine how well the notified body's personnel know and understand the directives and standards that relate to its products. It is also important to ask how backlogged the notified body is with audit requests; the closer we come to June 1998, the greater the likelihood that such firms will be inundated with requests.

The company should request an informational package and an application package from the notified body. The former should provide data regarding the certification process and some background about the notified body; the latter should include forms for the manufacturer to use in supplying general information about itself and its quality systems.

The second reason to contact notified bodies is to enable the manufacturer to get a feel for its potential business relationship with each firm. To determine this, the manufacturer should make careful observations of the body's professionalism in every contact. Some areas that might be worthy of note include the following:

  • How timely is the notified body in responding to inquiries?
  • How does it handle phone calls? Frequent transfers to staff members throughout the organization quickly become tiresome.
  • Does the notified body make a practice of assigning one contact person to handle a manufacturer's project? If so, is there a back-up in case that contact is ill or on vacation?
  • Does the notified body view questions as criticisms or honest inquiries?
  • How well does the notified body seem to understand your company, its products, and its operations?
  • Is the notified body genuinely interested in learning about your company and helping with its audit?
  • How well do the notified body personnel speak English?

    Feedback from peers can help a manufacturer to determine whether a given notified body has good depth of knowledge and expertise about its products, but if the manufacturer fully understands the directives and standards it should be able to reach its own conclusions early on. A notified body whose customary response sounds like "We'll get back to you," may be going through a learning process or may not have time for manufacturer questions. In either case, such a firm may not be the best choice, especially if the manufacturer has an aggressive audit schedule.

    When all of the above information has been gathered, the manufacturer should be in a position to make its final selection. At Pryon, the factors that were weighed in making that selection were ranked as follows: reputation, past relationship, expertise, cooperation, response time, and availability. Depending on the structure and size of the company, that decision may be made by one person or a committee. Whenever possible it's a good idea to use a committee, in part because it is another way of keeping executive management involved.

    U.S. manufacturers often select as their notified body a firm that previously served as their European test agency. Such a body's prior knowledge of the company and its products can be invaluable in expediting the certification process; in addition, the manufacturer is able to take advantage of the business relationships and good rapport that have already been formed. Ultimately, this was the course that Pryon elected to follow. The company's experience with its European test agency offered evidence that the firm's personnel intensely review products; carefully scrutinize documentation; and thoroughly evaluate product design, manufacturing processes, quality systems, and company facilities. Meeting that notified body's criteria would solidly underscore the company's confidence that its quality system was adequately designed, implemented, and controlled to meet ISO 9001 and EN 46001 requirements.

    Even after the manufacturer has made its selection, it should not rush into making a commitment to the notified body. Problems can arise unexpectedly that might make it im-possible to work with the selected firm. If the manufacturer must make a commitment at this point, it should be sure that the agreement between the two firms has an exit clause.


    The next step is for the manufacturer to begin working on an internal quality system audit called a gap analysis. A properly conducted gap analysis will identify areas in which the company's quality systems do not meet the requirements and where corrections must be implemented in order for the company to pass the certification audit. Gap analyses can be conducted by consultants, by an in-house committee, or by a peer company. Since Pryon wanted to speed the audit process and save money, it used an in-house committee to perform the gap analysis.

    If the manufacturer has sufficient expertise in-house, using a committee to perform the gap analysis can provide employees with a better understanding of the firm's internal systems. One risk in conducting a self-directed gap analysis is that members of the employee committee may overestimate their ability to interpret standards and their intent.

    To help design its gap analysis, the manufacturer should request certification audit questionnaires from several notified bodies. The company should ensure that it understands the questionnaires, and should contact the relevant notified body if it requires a more detailed interpretation. However, if it must frequently ask a notified body for more background or interpretation, it is probably not ready for the certification audit. If this is the case, perhaps attendance of appropriate personnel at an ISO 9000 seminar or auditor class would be helpful.

    The questionnaire from the company's notified body of first choice should be used as the basis of the gap analysis questionnaire, but it should be amplified with questions from other questionnaires and with the manufacturer's own questions. When the questionnaire has been developed, the manufacturer should appoint a gap analysis team to conduct the internal audit of its quality systems. At Pryon, this phase of the project was spearheaded by the directors of quality assurance and regulatory affairs. They provided relevant excerpts from the questionnaire to the department heads, and asked them to prepare for the internal audit.

    Throughout the internal audit, the gap analysis team should be objective, carefully scrutinize each area to establish compliance, take notes, review the results, and start corrective action immediately. A graph of the audit results can be a useful tool for identifying compliant areas, distinguishing areas in which further work is required, and communicating those results to executive management and appropriate staff. Notified bodies can supply the manufacturer with their grading systems, which manufacturers can apply to their audit results to determine their level of compliance. Companies shouldn't expect 100% compliance at this point, since the gap analysis is specifically intended to identify discrepancies.

    When the internal audit is completed, its results should be shared with executive management and with department heads, to ensure that all key personnel understand the gaps. Colleagues can help to determine the costs and timelines for correcting these gaps. If the company finds that discrepancies cannot be fixed within its projected timelines, it should reassess the schedule for the certification audit. When it believes its quality systems are fully in compliance, it can send the completed gap analysis questionnaire, application package, and supporting documentation, along with a commitment letter, to the notified body of its choice.


    The commitment letter should state that the firm has been chosen to serve as the manufacturer's notified body and should request a final, firm quote. Approximately four weeks after receiving and reviewing the commitment letter, supporting documentation, and the completed gap analysis questionnaire, the notified body should provide the manufacturer with several dates of its availability and an estimate as to how long the auditors will be on-site. It may be several months from the time of submission before a notified body can actually visit. It's a good idea to have a backup date, even though it might be months from the original date. Again, the closer the 1998 deadline approaches, the less scheduling flexibility notified bodies will have.

    Supporting Documentation. A general rule of thumb is that the better a submission is structured, the sooner a response can be expected. The company's quality manual must be included in the submission. It has embedded in it both quality procedures and policies. As supporting documentation, Pryon also furnished its notified body with a copy of the company's regulatory master. This contained a table of contents, procedures and policies that helped show compliance, and process and production flowcharts. Above all, the submission must prove that quality systems are in place that meet the requirements of the chosen standard.

    Submission Review by the Notified Body. The notified body will conduct an extensive review of the company's questionnaire and supporting documentation to determine if its quality systems meet the standards. It may provide feedback on many points of the questionnaire, and may also request clarification of some of the company's responses. Such clarification can be provided in writing or on the phone. At Pryon, we found that a combination of the two worked best. In a few instances, further documentation was required; in others, the notified body just had to be directed to a certain procedure or section of the quality manual. The manufacturer shouldn't attempt to create additional documents in response to every question, but should seek to understand the notified body's concerns and then clarify the needed information. The manufacturer should be honest and open, and should prevent discussions about interpretations from evolving into arguments.

    After completing its review, the notified body will create a summary report showing the company's compliance level point by point. From this report, it will establish a general rating. At Pryon, the summary determined that the company's quality systems were sufficient to undergo the certification audit.

    In addition to examining the issues raised by the notified body's summary, it is a good idea for the manufacturer to prepare its own list of issues that still need attention. The summary report and list can then be distributed to executive managers to enlist their help with final preparations for the certification audit.


    At Pryon, we started preparing for the certification audit by providing all employees with in-house courses on ISO 9000 and EN 46000. Staff were given a general introduction that included handouts explaining the upcoming audit and its purpose. We suggest that two such classes be held, one about three months before the scheduled visit, the other three to six weeks before. The latter serves as a refresher course to ensure that all employees understand the purpose, scope, and importance of the audit and are committed to its positive outcome.

    During this period, the manufacturer's quality systems should be monitored continuously. However, no major changes should be put into effect, because the systems should be operating a solid three months before the auditors arrive. If a significant change must be made, the manufacturer should be sure to inform the notified body. But it's best not to overwhelm it with numerous new or rewritten procedures, since this would make its initial review of the documentation null and void. At Pryon, we used the list of issues prepared earlier to correct minor discrepancies, and provided our notified body with corrected documentation when warranted.

    Preparing the Audit Team. As the date of the audit approaches, the manufacturer should ask the notified body to provide an audit agenda, which may or may not follow the same sequence as the audit questionnaire. The notified body should indicate how much time it intends to spend in particular areas and to whom the auditors expect to talk. Based on the agenda, the manufacturer can then identify the employees who must be present during each session.

    At Pryon, we conducted a dry run of the audit, which helped determine how much time would be spent in each area, gave employees a better understanding of the audit proceedings, and established a level of comfort for those areas not frequently audited. It also ensured that all necessary staff would be available and on standby during the entire audit or visit.

    Tips for the Actual Audit. As final preparations, the company should be sure a meeting room is set up and available throughout the audit. Writing materials, protective garments, and other supplies should all be ready. Such preparations enhance the auditors' confidence in the company.

    Essentially, the auditors are looking to see how well the company's documented quality control systems are followed. There are several common audit paths. The auditors may choose to follow any one of these or a combination of several. One path begins with an order for materials and follows those materials through the system to the point of their being shipped as a finished product. Another starts with a finished product and works backward. Or the auditors might evaluate how the company's quality systems work across the organization.