Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

PROCESS CONTROL

When doing its own product testing, how can a device company be sure that test results will lead to both a safe product and FDA approval?

Unlike some other forms of performance testing, strength testing usually results in the destruction of the element being tested. Testing an endoscope joint, for example, might involve pulling on it until it breaks to determine the degree of stress it will be able to withstand in use in the body. For this reason, strength testing cannot be performed on parts actually intended for sale. Rather, the device maker must select a sample from the total produced, test that sample, and extrapolate from the data obtained an estimation of the reliability of the (untested) parts that will actually be sold and used, often in life-critical situations.

How large should that sample be? Large enough to give a statistically valid picture of the safety of the parts to be sold. Too small a sample can lead to FDA scrutiny and with it the necessity of expensive retesting, or worse, to the sale of unsafe devices. On the other hand, particularly if the parts in question are expensive, the manufacturer does not want to throw money away by destroying a larger number of them than necessary. If sample size is allowed to increase until it destructively tests all the parts, the manufacturer will be certain about their strength but have nothing left to sell.

The Bionix 200 (MTS Systems Corp; Eden Prairie, MN) performs material testing. Photo courtesy Annex Medical.

Common sense tells us that the larger the sample size, the greater the certainty about the extrapolation of sample results to all parts. And the greater the cost. So we must determine the minimum sample size that satisfies both regulatory standards and our resolve to make and sell the safest devices at the highest possible level of confidence.

This leads to other questions. How high can our level of confidence be that the part is safe? And how confident can we be that all parts sold will be as safe as the ones we have tested?

These questions recently became matters of practical concern for Annex Medical, Inc. (Eden Prairie, MN), which had developed a new ureteral stone basket and needed to make a statistical analysis of product validation test results to ensure that the product would meet the minimum strength specifications set down in the company's product validation protocol. The analysis included selection of sample sizes and determination of confidence intervals and population percentages (all defined below). In trying to conform to industry standards and practices, however, Annex discovered that guidelines were difficult to locate.

Individuals contacted at FDA declined to provide specific guidance, explaining that the manufacturer, who is intimately acquainted with the details of a product and its use, is the one best qualified to determine levels of confidence, population percentages, and required sample sizes for validation testing. In Annex's case, however, these confidence levels and percentages were precisely what the company needed help with. The only advice FDA was able to offer Annex was that it hire a consultant.

Most small medical manufacturers lack staff statisticians but would like to do at least some of their own analysis, both to learn more about their business and to save money. The result of Annex's call to FDA was the development of step-by-step procedure for setting up and analyzing minimum-strength tests.

The procedure met Annex's needs and could be applied in full or in part to other companies' situations. The following article explains what that procedure entails.

STATISTICAL TOOLS

Several statistical concepts are involved in making good extrapolations. Initial decisions include determination of confidence intervals, population percentages, and sample size. The meaning of these terms and how they are determined are discussed below.

Confidence Intervals. To determine confidence intervals, measurements are first taken from a sample group of a few parts. The average (or sample mean) and sample standard deviation are two statistical measures calculated from the sample data. (The average is the central value of a data set, derived by dividing the sum of the values of the set by the number of terms in it. The standard deviation is the common measure of the dispersion of the data set around the average.)

These measures are then used to estimate the mean and standard deviation of all the parts, including those that are untested. In simple terms, a confidence interval is an upper and lower limit for how far the mean of all the parts might vary from the sample mean. For example, at a 90% confidence level, a company would be 90% sure that the mean of all the parts would fall between the upper and lower limits of the interval.

Traditionally, statistical analysis has been carried out at 95% confidence, with many peer-reviewed journals refusing to publish studies that use confidence intervals below that level. The 95% figure, however, is based more on the availability of published tables than on any solid rationale. Many large medical manufacturers use a sliding scale of confidence levels, usually ranging from 95 to 99.9%. Higher confidence levels are used for more-critical measurements or tests.

Population Percentage. Population percentage is the proportion of the product expected to exceed the minimum strength. Ideally, this percentage would be 100%, but such certainty could only be ensured by 100% destructive testing. Since that is not feasible, analysis is performed based on a high, but realistic, population percentage. Like confidence levels, more-critical measurements and tests are usually assigned higher population percentages than less-critical ones. For a critical joint, for example, 99.9% of the product might need to be expected to exceed a safe minimum, while for a less-critical item 95% might be acceptable. For a given data set, then, increasing the confidence level or the population percentage increases our assurance about the true minimum strength of a product but paradoxically--since we are in effect upping the ante--makes it less likely to be found acceptable.

Sample Size. Sample size should be as large as necessary to provide statistically valid information about the total population of (untested) parts, but no larger. How the appropriate sample size is determined for a particular test protocol will be further discussed below.

Endoscopic instruments from Annex Medical (left). Tensile testing of a urological endoscopic instrument (right). Photos courtesy Annex Medical.

K VALUES

K values are a convenient way of expressing the statistical interaction of confidence intervals, population percentages, and sample size.

Suppose the strengths of a sample group of parts have been tested, with the results providing an average strength and a standard deviation. Extrapolation from this test-sample information provides a range of strengths expected for all the parts. K value can be defined as the number of standard deviations that the upper and lower limits of this strength range are away from the sample mean. A larger K value corresponds to a wider strength range. Reversing this procedure, if we start by knowing the K value--which we do because past statistical analysis has provided K-value ranges (see Table I)--we can find a lower or minimum limit for a strength measurement by simply multiplying the standard deviation of the sample by the appropriate K value and subtracting the result from the sample mean. (Later in the article this procedure is illustrated with an example.)

The more critical the part, the more stringent the test criteria should be. Higher K values correspond to more-stringent test criteria. K values can be made higher by increasing the confidence level, the population percentage, or some mix of the two. K values become lower as sample size increases.

Table I. One-tailed K values for calculating the probability of exceeding the minimum. Highlighting and outlining refer to the example discussed in the text.

The top of each Table I column shows the one-tailed confidence level and population percentage used to develop it. (The table used here is one-tailed because we are interested in only one end, or tail, of the strength distribution curve--namely, minimum strength. If we were interested in both minimum and maximum strength we would use a two-tailed test.) Not all the confidence level/population percentage combinations listed in the table would have been specified in past (precomputer) statistical procedures, largely because intermediate values of confidence and population percent are not widely published.

The columns of Table I represent an increasing progression of both confidence level and portion of population from left to right, that is, from 95/95 to 99/99.9. Arranging the table in this fashion results in an orderly increase in K values as a product or part becomes more critical, while giving appropriate credit to the increased certainty obtained from testing larger sample sizes. Population percentages in this table are for one-tailed testing, and are specifically arranged to find the probabilities of exceeding a minimum (for example, a minimum strength).

The method presented in this article produces results similar to those resulting from more complicated statistical specifications.

Normally Distributed Data. Table I is based on the assumption that the measurements are normally distributed. This assumption can be tested without a computer using normal-probability graph paper. Data that are normally distributed will plot as a straight line.

Plot each data point (y-axis) against its median rank (x-axis). In the present case, each data point represents a break point in a sample subjected to strength testing. The median ranks are assigned by arranging the data in ascending order. The smallest value becomes rank order 1, the next smallest becomes rank order 2, and so on. The median rank for each data point is then found in the column of a median-rank table--such tables are readily available in reliability texts--headed with the appropriate sample size and the row associated with the rank order.

For sample sizes over 100 a common numerical method of determining normality of data distribution is the chi-square test. Examples of this test are found in most statistics books.

Data that, when graphed, form two distinct lines on normal-probability graph paper or two humps in a frequency histogram are said to be bimodal. If this pattern is observed the data are not normally distributed. (See the discussion on bimodality under Common Problems, below.)

FINDING MINIMUM STRENGTH VALUES

The procedure for evaluating a part's minimum strength values can be broken down into the following series of steps. Steps A­G set up the analysis by determining sample size and critical test values. Steps H and I cover analysis of the results. We use Annex Medical's testing of an endoscopic instrument joint is used as an ongoing concrete example of the procedure.

Table II. Severity-of-failure categories and their respective safety factors.

Step A. Use failure mode and effects analysis (FMEA) or simply brainstorm to identify the different ways the product to be tested might fail. Determine the severity of the failure for each failure mode. The following severity categories, which are listed in Table II, have proven useful for products such as medical instruments:

1.Nuisance: A failure the user (physician or patient) will become aware of but can tolerate without the patient being affected.

2.Decreased Device Performance: The device is functional but not at the intended level.

3.Lengthened or More Complex Procedure: The failure of the device creates the need for additional procedural steps. These may include simple recoveries of broken product components from introducer instruments or from the patient.

4.Surgical Intervention: Surgery and/or other serious steps must be taken to correct a problem caused by the device failure.

5.Serious Injury: A device failure in this category is likely to cause injury to the patient. The injury may result directly from the device or from subsequent intervention.

6.Death: The device failure could cause death either directly or as a result of subsequent intervention.

If the endoscopic joint being used in our example were to fail, a smooth, easily grasped piece of the instrument would remain in the patient, and an additional endoscopic procedure would be required to retrieve it. Therefore, the Lengthened or More Complex Procedure column in Table I would be used to measure the joint's severity of failure.

Step B. Devise a mechanical test to evaluate each failure mode for the part under consideration. The endoscopic joint, for example, was evaluated using a pull test to measure its strength.

Step C. Select a minimum acceptable result for the test performed in step B. Consider the conditions of actual use and answer the question, How strong must this part be to perform its function?

In the example in which the endoscopic instrument was tested, a small amount of experimental measurement was done. It was discovered that only 1 lb of force was required to actually operate the instrument.

Step D. Typically, the minimum result from step C is multiplied by a safety factor, normally ranging from 2 to 10, to arrive at the minimum acceptable value. Product knowledge and experience are the best guides in selecting this factor. Alternatively, the consequence of failure determined in step A can be used as a guideline, as shown in Table II.

Some products or parts of products are much stronger that needed for their intended function. In such cases, the minimum acceptable value may simply be set so it exceeds the maximum load that can possibly be applied to a part in use or abuse. This approach to assigning the minimum acceptable value is safest if the maximum load on one part is limited by the strength of some weaker part. For instance, if a steel screw is threaded into a plastic handle, the maximum load on the screw's threads will not equal the force required to break the threads themselves, but rather the force required to strip them out of the plastic.

In the example, as mentioned above, the endoscopic joint failure had been given a severity-of-failure described as a "lengthened or more complex procedure." The corresponding safety factor of six (per Table II) was multiplied by the operating force of 1 lb to establish a minimum acceptable value of 6 lb.

Step E. Test 10 samples and determine the sample mean and standard deviation of the test results. In the sample, include parts made by several different operators using different fixtures and machines; try to introduce as much variation as possible at this step. In addition, write a short explanation of how each sample failed, that is, the failure mode. Examples are "joint broke" or "wire broke."

In the endoscopic joint test procedure, 10 sample joints were pull tested. Their average strength was found to be 15.3 lb, with a standard deviation of 3.4 lb.

Step F. Determine the number of standard deviations (NSD) that the minimum acceptable value from step D is away from the sample mean by using this equation:

NSD = Mean ­ minimum acceptable value

---

Standard deviation

In the example:

NSD = (15.3­6) = 2.735

---

3.4

Step G. Using the K-value table (Table I), find the column associated with the severity of failure determined in step A. Trace downward in the column until you find a row with a K value equal to or just smaller than the NSD value found in step F, in this case, 2.706. Follow this row to the left to find the minimum sample size (n^_min), in this case 22, that should be used to validate the product on this test.

A minimum of three separate validation test batches (or runs) should be made to conform to good statistical practice (and minimize the possibility that a single run was a lucky result of an otherwise poor process). The sample size number, n^_min, can be used in two ways: For each batch to be statistically valid on its own, run at least n^_min pieces in each. This is the soundest procedure statistically and should be used if the consequence of failure is surgical intervention, serious injury, or death. Conversely, for a high-cost product with a lower severity of failure, use a minimum of three separate batches to test a total of at least n^_min samples. For instance, if n^_min is 22, as in the example above, then three batches of 8 each for a total (n^_actual) of 24 would be a good choice. Be aware that in this case the required confidence level will not be achieved until all three batches have been tested and the results pooled. If n^_actual is large enough, it is good practice to spread the testing over as many batches as possible. For example, if n^_actual is 100, 5 batches of 20 or 4 batches of 25 would be preferable to 3 larger batches. It can be advantageous (but not necessary) to select a batch size of at least 6, since this will allow some separate analysis of the runs using the information in the K-value table.

This experiment has now been set up. The next sequence of steps covers analysis of the results.

Step H. Test n^_actual samples. Find the mean and standard deviation for the results. If you are using a batch size of n^_actual, repeat this for each batch. As in step E, write a short explanation of how each sample failed. In the example, after testing 24 sample joints (three batches of 8), the overall average was found to be 15.21 and the standard deviation 2.67.

Step I. Multiply the standard deviation from step H by the K value corresponding to the n^_actual used. Subtract the resulting number from the average found in step H. This is the critical value.

Critical value = Mean ­(K value x standard deviation)

If the critical value exceeds the minimum acceptable value found in step D, the validation is complete. Write the report. If the critical value is less than the minimum acceptable value, see the following section on common problems. In the example, n^_actual was 24, so the K value (from the Lengthened or More Complex Procedure column) was 2.654.

Critical value = (15.21) ­ (2.654 x 2.67) = 8.124

Since the critical value (8.124) exceeds the minimum acceptable value (6 lb from step D), joint strength is adequate.

COMMON PROBLEMS

If the value of NSD found at step F is smaller than anything in the table, or extremely large sample sizes are indicated, there are two possible causes for these related occurrences.

In the first case, the variability of pieces from the process is large compared to the mean strength of the piece. If the standard deviation is more than about 30% of the mean, it would probably be a good idea to look at the process to see if some of the variables can be better controlled.

Operator-controlled variability is particularly suspect, especially if, as recommended, more than one operator made the samples. Study the operators' work processes to determine and correct the differences between their techniques. (If this is not done, the variability will occur when actual production begins and will be much more expensive to correct.) Some other sources of variation include the tolerances of incoming material, amount of adhesive or solder applied, times, and temperatures.

Another way to reduce variability is to limit the combinations of fixtures, machines, and operators used. For instance, always run fixture A on machine A and fixture B on machine B. Keep in mind, however, that if this is done only the combinations that have been tested will be valid for production, and subsequent QC records will need to confirm that only these validated combinations are being used for production.

The second possibility is that the average strength of the product is too close to the minimum acceptable value. The best fix here is to rethink the design or materials in the piece. Otherwise, the product will always be running on the ragged edge of not working.

If the value of NSD found at step F is larger than anything in the table, or step G indicates a sample size of less than 10, 10 will serve as an acceptable validation sample size for the initial test run. Running two more small batches (6 parts each, for example) to meet the three-batch minimum should be sufficient to validate the product when the 12 pieces are lumped with the first 10.

If the critical value found at step I is smaller than the minimum acceptable value, the most probable cause is that something in the process or materials has changed between the time of the pilot run of 10 and the time the three validation batches were run. Review everything! If the cause of this variation is not found, it could surface again unexpectedly, resulting in problems with production parts. Correct the changes and rerun this procedure starting at step E, taking 10 new samples. Continue sequentially through the steps. Do not use the old results in the rerun. If no changes can be identified and corrected, use the mean and standard deviation of the validation batches to find a new NSD. Go to step G with this NSD and continue sequentially determining a new sample size, n. In this case, the data that have already been collected can be included in the analysis, necessitating only a run of enough additional pieces to bring the total up to the new n.

Bimodality. If bimodality is evident, look at the failure mode for each individual sample (as in step E, above). Group the data so that similar failure modes are together. Does one failure mode seem to have a different mean and standard deviation than the other(s)? This could identify the cause of the bimodality. Try to understand why the bimodality is occurring. Perhaps one of the failure modes can be reduced or even eliminated.

Alternatively, divide the data into subgroups by failure mode and analyze each as if it were a separate test, starting at step H. The size of each subgroup becomes n^_actual for the analysis. It may become necessary to run more samples to accumulate enough of each subgroup to complete the analysis. Frequently each failure mode subgroup will meet the minimum requirements if it is analyzed separately.

Vaughan Weeks is the president of Novus Technical Service (Caledonia, WI).

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

CLINICAL TRIALS

For medical device manufacturers who must conduct clinical trials, extra effort early in the process, in the form of voluntary auditing, can prevent costly errors.

Clinical trials for establishing the safety and efficacy of a medical device can be among the most costly and time-consuming elements of product development. If procedural errors render unusable the information gathered in a series of clinical trials, the time and expense required to conduct the studies again can easily terminate the entire development effort. Thus, for the many medical device manufacturers who need to sponsor clinical trials to gain premarket approvals (PMAs) or submit 510(k) notifications, it is essential to maintain confidence throughout the clinical trial process that the data will be usable.

One way for sponsors to ensure that there will be no unpleasant surprises after clinical trials are finished is to plan for one or more independent investigator audits during the trials. An audit of a clinical trial provides the research sponsor with independent appraisal of the quality and completeness of the data generated by the trial. Although auditing alone cannot transform a poorly planned, executed, monitored, or analyzed trial into a credible one, an active clinical trial audit program will point out potential problem areas early, so solutions can be found before it is too late. Used effectively, therefore, audits can reduce costs, maintain project schedules, and ensure regulatory compliance.

FDA already requires that clinical trials be subject to monitoring. Current procedures at the Center for Devices and Radiological Health (CDRH) make it highly likely that at least two investigational sites of each major PMA clinical study will be inspected by the CDRH bioresearch monitoring program. Studies undertaken for 510(k) submissions are now also reviewed by FDA bioresearch monitoring staff.^1,2 By choosing to perform independent auditing of investigational sites, sponsors can increase their chances of performing well during an FDA inspection.

Auditing, which is typically conducted no more than once or twice during the course of a clinical trial, encompasses many aspects of the study. Auditors review a sample of data taken from a representative group of study sites. They also check regulatory documents such as the study protocol, institutional review board (IRB) communications and approvals, informed consent documentation, investigator agreements and curricula vitae, clinical laboratory certifications and normal values (if necessary), randomization documentation, and device accountability records to determine whether the documents conform to sponsor requirements. Auditors compare the data in the source documentation--most commonly, the medical record--against data entered on the case report form (CRF). Because of the vast amount of material involved, this comparison must often be limited to just a sample of patients and selected data fields. Auditors also review source documentation to ensure protocol compliance and full evaluation and reporting of adverse device effects (ADEs).

The most common reason that sponsors choose to perform independent auditing is to prepare for an FDA inspection. However, many device manufacturers are beginning to follow the lead of the pharmaceutical industry and incorporate this activity into all phases of major research programs. Sponsors are finding that clinical trials audits serve a variety of purposes.

Independent review of a clinical trial can bring a fresh perspective and new insights into the project. For example, the longstanding relationship between a clinical research associate (CRA) and the investigative site that he or she oversees can be a potential source of bias. A CRA typically visits a study site at regular intervals to make sure that procedures are being followed. An auditor, who does not have such ongoing contact with a site, can be more forceful in calling for any needed changes.

Audits provide the chance to review one clinical research system, such as control of investigational devices, across several studies. They can also determine how data generated outside the United States can be used in an FDA submission.

In rare instances, an audit can be used to investigate possible fraud. In such cases, data audits can show whether data in source documentation, such as medical records and laboratory reports, are consistent with entries on CRFs and data listings.

So, the benefits of auditing can be well worth the extra effort required. To make sure that these benefits will be realized, however, sponsors must develop a comprehensive auditing strategy. Sponsors must decide not only which sites and data will be audited and how often the audits will be performed, but also how the information generated from the audits will be used to improve procedures.

CHOOSING AN AUDITOR

One of the first, and most important, decisions the manufacturer faces is choosing an auditor. Because they are sometimes not part of the sponsors' organizations, contract auditors can promote a spirit of objectivity and encourage investigators to communicate problems openly. Auditors can also be chosen from the sponsor's quality assurance organization or from any group not associated with direct management of the clinical trial.

Auditors must be knowledgeable about the applicable regulations, the technical aspects of device operation, and the medical condition indicated for the device. This can represent a rather unusual combination of knowledge, which is acquired only by working in related fields in industry. Experience in good manufacturing or laboratory practices is certainly a good qualification for auditors, but a strong background in the relevant clinical area and experience reviewing medical records are also essential. Many potential problems can only be recognized by someone with clinical experience. For example, if a CRF indicated that a study subject was discharged from a hospital halfway through a course of antibiotics given to treat a postoperative infection, an auditor with clinical experience would be able to recognize that such a patient would be unlikely to be discharged at that point, and question the discharge date on the CRF.

Auditors must also be able to organize and write a report that accurately describes the audit observations. For the report to be useful, descriptions of deviations must be detailed enough that solutions can be found and implemented.

In addition to the well-defined qualifications mentioned above, there are some equally important, but less tangible, characteristics that are important in an auditor. The auditor must have an aptitude for details and logical thinking. A successful auditor is able to notice a minor inconsistency, deduce possible causes, and then investigate, uncovering significant issues. An auditor should have the patience to search for details and follow these leads, often under tight deadlines.

Auditors must also conduct themselves in a professional manner during all interactions with the study site personnel and corporate coworkers.

PLANNING THE AUDIT

Once the auditor is chosen, the manufacturer must plan where and when the audit will be performed as well as what data will be audited.

Confusion and schedule conflicts can often be avoided by preparing an audit agenda and sharing it first with internal management and then with site personnel. The agenda will tell all parties what information will be audited, who must be present, and when each portion of the audit will occur.

Audits are most frequently conducted when studies are at least half completed, so that the auditors can review a meaningful number of patients. Audits performed after a study site has been closed can present some special challenges when corrective actions are necessary. Some site personnel can become reluctant to invest time in the correction process once all research fees have been paid. Also, corrections become more difficult as time elapses. Patients are more difficult to track down, and medical records can become more difficult to access.

Neither research sponsors nor FDA have the resources to audit every site of every clinical trial. Site-selection criteria include the greatest number of patients enrolled, or the largest number of ADEs or protocol violations. The number of patients is often the most important factor in selecting audit sites because it is more cost-effective to review many patient records at a small number of sites than a smaller number of patient records at many sites. Geographic considerations can also be a factor in determining where to conduct an audit.

Of course, one of the most important aspects of planning the audit is determining what data to investigate. To decide this, auditors should become thoroughly familiar with the sponsor's study protocol and all amendments to it. In some cases, the sponsor may wish to add additional agenda items for an audit, and the auditor should become familiar with these items as well.

Auditors should also meet with the responsible CRAs and project managers, and review data listings, if available, for unusual patterns. The meetings should cover all relevant issues, such as ADE reporting, IRB and informed consent issues, investigational device accountability and failure issues, and patient-assessment questions. The review of the data listings also provides an opportunity for key safety and effectiveness CRFs and specific data fields to be identified.

CONDUCTING THE AUDIT

Good communication is essential to a good audit. Auditors who arrive at a site unprepared waste time and diminish goodwill between the sponsor and the researcher. Researchers who either do not understand or do not prepare for an audit can also adversely affect the process.

The first step in promoting this communication is for the auditor to contact the site to announce that the audit will be conducted and plan the details. In most cases in the United States, three or four weeks' notice is adequate. International audits often require longer notification periods.

The auditor must confirm that all necessary site personnel and the documentation that will be reviewed will be available during the audit. To make the initial communication as unambiguous as possible, any telephone agreements should be followed up by a fax or E-mail. Written documentation is useful for U.S. audits and essential for international ones.

During the audit, every opportunity to gather information should be taken, and auditors should be alert for communication nuances that may signal hidden problems. For example, an investigator who claims to have only a vague idea where the study protocol is stored may not be familiar enough with the protocol.

There is no single method that is appropriate for all medical device clinical research audits. Auditors should adapt their techniques to the clinical trial under review. For example, an audit of an in vitro diagnostic study will concentrate on sample handling and a data comparison between analytical methods, whereas an audit of a study for a life-sustaining device will include detailed review of medical records.

However, there are documentation tools that auditors can use with minor modifications for nearly any study.

Audit Checklists. Checklists are invaluable when conducting multisite audits, especially if more than one auditor is involved or if the study is international. Checklists can capture statistical summary data, such as the total number of informed consent forms reviewed, the number acceptable, and the most common deviations. They can be customized for different studies to list the key data points that must be reviewed during the audit. Although they are not substitutes for good observation and analysis, checklists can help ensure that each subject at each site during a complex audit is reviewed in a uniform manner.

Chronology Lists. A chronology of key events can simplify regulatory document review. To make such lists, the auditor constructs a table with columns for the date and for IRB, sponsor, and FDA actions. Each action, as described in the study site's documentation file, is listed in sequence on its own line. Once completed, chronologies can clearly show how various actions are interrelated and point out any conflicts between actions and exceeded time frames.

AREAS TO REVIEW

Although audits vary depending on the study, there are several aspects of clinical studies that should be routinely reviewed during an audit because they are potential sources of difficulty for many studies.

Regulatory Documentation. A variety of documents must be present before an investigator can begin a medical device clinical trial, and an auditor should make sure they are in order.

If the device poses a significant risk (SR) to study subjects, an investigational device exemption (IDE) must be filed and approved by FDA. If the device poses a nonsignificant risk (NSR), no formal IDE filing is necessary; however, the record-keeping requirements as stated in 21 CFR 812 continue to apply. In an NSR study, the reviewing IRB must make two determinations: first, that the study is NSR, and second, that the study protocol is approved. Some IRBs become focused on the second objective and fail to document the first decision.

Randomization. One item becoming more common as medical device clinical trials increase in complexity is randomization. In order to reduce bias, patients are randomly assigned to either the experimental treatment or to treatment with a comparison, or control, device. There are several valid methods for randomizing treatments. The chosen method must be carefully reviewed to ensure that it is functioning as intended. Clinical investigators should not be able to influence the treatment group assignment for any patient.

Clinical Documentation. Comparison of CRFs to source documentation is almost always the most time-consuming on-site task of the audit. CRFs are rarely arranged in the same order or format as the medical records. Also, the organization of medical records varies greatly from one site to another. An effective auditor can easily adapt to changing conditions, however.

Unlike a CRA, who reviews virtually all data recorded on each CRF, the auditor commonly reviews only a carefully chosen sample of data fields. Often, only a subset of patients at the site are reviewed. Selecting these patients for review is a key aspect of ensuring the validity of the entire audit process. In situations where all the patients appear to have been treated in the same way and with the same result, a simple technique, such as, can be employed to determine sample size, where n = number of patients enrolled. When using this method, care should be taken to ensure that patients are then selected from the beginning, middle, and end of the study enrollment period.

If there is greater variability among the study subjects, those with large numbers of ADEs and several protocol violations, early terminators, and those lost to followup often involve the greatest number of functional areas in the clinical research system and will demonstrate the degree of coordination between these systems.

CRF and Source Documentation Consistency. Comparing CRF entries with source documentation is a major component of all clinical trials audits. Source documents, or medical records, may have a wide variety of formats. For example, orthopedic implant studies will use x-ray films, while cardiology studies will use electrocardiographic traces, angiography films, and other testing data. Clinical laboratory data, pathology data, and even patient self-assessments and diaries may be the source documentation.

Outpatient studies can often present documentation challenges. An investigator may enter only a few lines of progress notes, and most of the study data may be entered directly on the CRFs. To ensure that there will be adequate backup data for the CRFs, some research sponsors will work with investigators to develop worksheets for the investigator's records where the data can be recorded.

No matter which form the source documentation takes, the auditor must assess it for credibility and consistency with the data recorded on the CRF.

Assessment of data for credibility is usually straightforward, unless fraud is suspected. In rare instances, investigators have manufactured data and even entire patients by repeatedly photocopying a laboratory report and adding a new name for each new study record generated.

One indicator of potential fraud is vastly different appearance of the signatures of either the study subject or the study site personnel on different documents. If a study subject signed the informed consent form the day before the use of an investigational device and the signature appears significantly different from the signature on the hospital's surgical consent form, then further investigation is justified.

Auditors need to review several aspects of the medical records. For example, the appropriate personnel must operate the device. Many studies of therapeutic devices require that the device be operated by the investigator or a subinvestigator. It is often hard to show that the device was operated according to the protocol, but this should be done whenever possible.

Another area that an auditor should review is whether all measurements generated consistent results. Inconsistencies in assays may be caused by technical faults in the test methods, or by how the methods are applied by investigators. If a site is performing a nonroutine assay in conjunction with the study, sufficient validation data should be available to demonstrate that the assay is capable of producing reproducible results.

CRF Corrections. It is inevitable that improper entries in CRFs will be identified during the data entry process. When data entry personnel are unsure of an entry or the data are clearly out of the expected range, a query is generated. Large, complex trials require special systems to document and track queries. Data entry errors may be handled in a less formal manner for most smaller medical device studies. However such problems are resolved, the copy of the CRF at the study site must be made to accurately reflect the sponsor's database. Audits conducted at the close of a study often review a sample of data queries to ensure that site documentation conforms with the sponsor database. Here, auditors need to distinguish between random lapses and persistent system failures.

ADE Documentation and Reporting. ADE reporting is one of the most difficult tasks in medical device clinical research. All members of the research team can become involved in reporting an ADE, and all activities are governed by time limits in the regulations. Chronology lists can make this part of an audit easier by showing when reports were made.

Investigational Device Accountability Records. The FDA IDE regulations require that supplies of investigational devices be controlled. When dealing with single-use devices, this requirement is usually simple to audit. Sponsor shipping records and study site receiving records show what devices should have arrived at the site. Use records can be found in the medical records, CRFs, or dispensing logs. It is essential that all these sources agree. Disposals, returns, and repairs of devices should also be recorded in site documentation and a final check made to determine the number of devices shipped to the site and the fate of each one.

FDA concerns in this area involve possible protocol violations stemming from using different quantities of the device per subject, diverting investigational supplies to nonstudy patients, or transferring investigational supplies to investigators who are unaffiliated with the study. Accountability for durable medical devices, both therapeutic and diagnostic, can become more complicated, especially if there are no consumables necessary for the use of the device. Auditors may have to gain access to data stored in device controller memory or use other methods to confirm accountability.

SPECIAL CHALLENGES FOR AUDITORS

One of the most difficult tasks the auditor will encounter is to ask the hard questions. For example, if it becomes clear that a study did not have IRB approval for a three-month period after the initial IRB approval expired and before the next approval took effect, the investigator or study coordinator must be asked about this situation. This could be a significant observation, or simply a misunderstanding. The auditor must resolve such situations so that the audit report will be accurate.

Although outright fraud is rarely encountered in medical device clinical research, its consequences are so severe that one must always be on the alert for it. If it appears that the site is involved in fraudulent activities such as the creation of study subjects or falsification of test data, the auditor needs to gather conclusive evidence, if possible, and quickly inform the sponsor. Frequently, the sponsor's response is to terminate the audit and conduct further site interactions through regulatory or legal representatives.

USING THE FINDINGS

When the audit is complete, a wrap-up meeting should be held to discuss the findings. This meeting is the most important part of the audit process for study site personnel and often one of the most difficult tasks for the auditor. The auditor's goal for the meeting is to summarize key findings and identify any misunderstandings before they become part of the audit report. The investigator's goal is to get an answer to the question: "Did we pass?"

A scheduled date for the wrap-up meeting should be agreed on at the beginning of the audit to make sure that as many of the key people as possible will be present. Rushed meetings in the hallway with the auditor's bags packed and the airport shuttle van waiting outside should be avoided. A little thought and even 10 or 15 minutes of organization on the part of the auditor can make a big difference in the quality of the presentation.

A good way to start the wrap-up meeting is to thank site personnel for their time and assistance. Next, the (hopefully) brief list of important audit findings should be reviewed. In some cases, auditors can be reluctant to share negative observations with the study staff. If there is a possibility of fraud, this concern may be justified, but in other cases, it helps everyone if findings are frankly discussed. Discussions should center on facts, not personalities. Photocopies of documents demonstrating the observations are very useful.

A somewhat controversial item that may also be included at wrap-up meetings is a review of corrective actions. Some firms encourage auditors to work with the sites to solve problems, while others feel that this type of interaction jeopardizes the objectivity of the auditor. Auditors who choose to suggest corrective actions need to be aware of the conduct of the entire study, not just one site, and should be careful not to commit their organization to promises that it cannot keep.

Many sponsors of clinical research also ask auditors to perform some form of brief training of study site personnel as a part of the audit. Distributing a package of relevant documentation including copies of 21 CFR 50, 56, and 812, the FDA compliance guidance manual for clinical investigators, and some of the FDA information sheets can be helpful for training.³ This information package can be especially useful if the audit is conducted after the last monitoring visit, because the next time the site study documentation is reviewed may be with an FDA investigator during a bioresearch monitoring inspection. Most clinical investigators realize that this type of training and information is generally applicable to all the studies they conduct and welcome the opportunity to gain a better understanding of the regulations.

The Audit Report. The observations discussed during the wrap-up meeting need to be documented in a report so that effective corrective actions can be taken. In addition, appropriate levels of management at the sponsor's firm must be made aware of the results of the audit. One advantage that auditors of medical device clinical trials have over auditors of manufacturing facilities is that nearly all the observations made during clinical study audits relate to documents, which can be copied and included in the report to support the report's conclusions.

There are many possible formats for audit reports. Reports for data audits generally contain less narrative and more tables that list individual deviations along with error rates for specific CRFs or even for specific data fields. Audit reports of the more conventional midstudy, or retrospective, audits usually include most of the following items.

Scope. This section delineates what was reviewed (and not reviewed) during the audit. It is important to inform management of any limitations in the scope of the audit. This section could include phrases such as, "Only study X was reviewed during this audit," or, "Time permitted review of only seven out of the nine patient records that were initially designated for review."

Introduction. This section gives the background for the audit, listing all the personnel involved and their roles, the reason for the audit, the audit methods, and a very brief description of the study that was audited.

Executive Summary. This section contains one or two paragraphs explaining in general terms the audit results and their significance.

Major Findings. The most significant findings are listed in this section as bulleted items, with the most significant ones first. References to more-detailed descriptions in the body of the report can also be included.

Regulatory Documents. Observations relating to informed consent, IRB activities, or any other nonmedical study issues should be described in this section. In most cases, it should be possible to photo- copy documents related to these items, which may be cross-referenced and included as an appendix to the report.

Clinical Data Review. Results of the comparison of source documentation against CRFs or data listings are described in this section, as well as protocol violations such as difficulties reporting ADEs, improper use of the device, failure to perform specified testing on patients, or failure to follow study schedules.

Conclusion. Opinions and impressions can be important results of the audit. They need to be documented, but clearly identified as opinions.

Follow-Up Actions. Completed audit reports must be carefully protected. A very limited number of copies should be circulated within the company, and review of the report should be completed according to a predetermined schedule.

Corrective actions must be agreed on, implemented, and documented. Once an organization invests time and money in an audit program, the follow-up activities become as important as the audit and should not be forgotten.

All corrective actions, including minutes of postaudit meetings and revised documents, should be reported to the auditor for inclusion in the audit file. Once all issues have been resolved, all copies of the report should be recalled and the report, along with supporting documentation, filed in secure storage. An audit log should be maintained listing studies and sites audited, dates of audits, and the identities of the auditors. Ordinarily, FDA investigators are entitled to review the audit log but not the audit reports.

Many medical device firms can address a specific list of problems. Ensuring they do not arise at other sites and with other studies is more difficult.

Quality assurance tools used in the manufacturing part of the organization, such as trend analysis and examination of the root causes of problems, are very useful in the clinical research environment as well. If issues arise because of lack of knowledge on the part of the responsible CRA or the study site staff, training can be modified to address them. One of the most important functions of the clinical audit is to look at systems and ask whether they continue to be adequate for the clinical research that is being conducted.

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

EDITOR'S PAGE

In the managed-care marketplace, medical device companies may increasingly find cooperation more beneficial han competition.

One of the benefits of making mistakes in print is learning from the readers who correct you. For instance, a misleading statement I made in last month's column ultimately gave me a renewed appreciation of device industry history and of the complex relationship between competition and cooperation.

What I had written was a summary of a remark made by Harvard law professor Arthur Miller in the course of a panel discussion at HIMA's annual meeting last March. In trying to capture the gist of it, I attributed the invention of the pacemaker to Medtronic cofounder Earl Bakken.

This statement caught the attention of two readers, both connected with the codevelopers of the implantable version of the pacemaker, Wilson Greatbatch, William Chardack, and Andrew Gage. As both pointed out, these men were the inventors of the first successful implantable pacemaker. A call to Medtronic confirmed this, but added the information that Bakken was the inventor of the first successful wearable pacemaker.

As I subsequently read in one of the references I consulted, "there is much controversy over who was the first to invent the artificial pacemaker." This statement refers not to Greatbatch et al., however, but to other inventors active several decades earlier.

Ultimately, it seems, the pacemaker was a team effort, whether intentionally or not. No doubt the pacemaker pioneers were driven by competition. But it succeeded in the end by virtue of cooperation, one advance building on another.

This conclusion brings me back to the HIMA panel discussion. In considering the implications of managed care, one of the panelists, Barbara McNeil of Harvard Medical School, described the quandary facing medical device companies. If managed-care organizations demand expensive cost-effectiveness studies of new products, who will fund them? she asked. If one company does so, its competitors will benefit without bearing any of the cost. If they jointly sponsor the studies, however, each loses the competitive advantage of getting into the market first. The result might be that no studies are done, and no one will benefit.

This scenario reminds me of the Prisoner's Dilemma, a philosophical problem that pits self-interest against cooperation. Two prisoners held in separate cells for a jointly committed crime are offered a deal: confess, implicating the other, and go free. If only one prisoner accepts this deal, the other will be sentenced to three years. If each implicates the other, each will receive two years. If neither takes the deal, both get one year in prison.

In this scenario, the best outcome is the last, in which the total time served jointly is two years. This result depends on cooperation. The worst outcome is the second, in which a total of four years is served. This result is driven by self-interest. Human nature being what it is, the latter result is the most likely.

In the real world, the relationship of self-interest and cooperation is usually more complex. As in the case of pacemaker development, the best results will come from a mix of competition and cooperation. But as the market changes and managed care becomes more dominant, dilemmas like McNeil's may become more common.

In such cases, is cooperation clearly the best choice? For medical device companies, the answer will increasingly be yes.

John Bethune

[email protected]

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

DESIGN

For medical device manufacturers, FDA's new design control requirements are just one more reason why the team approach to product development is gaining ground.

In the medical device industry, the practice of bringing together a multidisciplinary team of specialists early in the development of a product has been kicking around for at least a decade. Both large and small companies have adopted the practice in order to make better use of their resources. Now, the growing competitiveness of the device marketplace and new regulatory requirements for design control are making the team concept an essential part of doing business.

Increasingly, both in-house talent and outside consultants representing manufacturing, marketing, quality assurance, and regulatory affairs are being asked to offer their expertise at the earliest--some would say most critical--stages of product development. "There is no one person in a company experienced enough to understand the whole process," says Jim Sandberg, director of quality assurance and regulatory affairs at Protocol Systems (Beaverton, OR). "So you have to look to several people to get the job done."

The team approach to product development is gaining favor because companies are coming to realize that if the contributions of staff are not organized, the result can be deadly. Bill Wood, vice president for research and development at RELA (Boulder, CO), recalls performing a postmortem on a project completed five years ago. During the two-year development of the device, there were nine major changes. "What I discovered was that as different staff joined the project, they would each initiate a major change," Wood says. "This incremental addition of staff was pretty damning. The project ended up behind schedule and over budget."

ADVANTAGES OF TEAMWORK

Involving representatives of the key disciplines in a timely, coordinated fashion not only saves money and time, but also provides advance warning to the various departments about when they will have to switch into high gear. Regulatory affairs specialists, for instance, know when a submission must be made to keep a project on track. Their assessment of the time that will be required to get through regulatory review can be used to alert the manufacturing and marketing departments about when they must be ready to carry the ball. The team approach also adds predictability to a process that is inherently uncertain. "We check and double-check everything," says Dennis Tomisaka, director of R&D at Medex, Inc. (Dublin, OH).

The practice of making quality assurance personnel a key part of the development process has grown as more and more companies have embraced the ISO 9000 family of quality systems standards compiled by the International Organization for Standardization (ISO). And that practice will likely proliferate after June 1, 1997. On that date FDA will implement its new quality system standard, which emphasizes the use of quality systems throughout the design and development process. "This regulation is going to push companies to establish formal programs that ensure quality in the design process," says Kim Trautman, quality systems expert in the Office of Compliance at FDA's Center for Devices and Radiological Health.

For development projects to succeed, companies will need the support of all key players. "Companies will have to understand the relationships among their design, R&D, manufacturing, corrective and preventive actions, and management functions," says Trautman. The new standard, she says, will "force manufacturers to integrate quality into every part of their process and to think of quality more as a system than as a discrete function."

The new quality system regulation, including design control requirements, has simply made it more desirable to adopt what has already become a best practice at many device manufacturing firms. Companies not following the team approach can easily run into problems. "What tends to happen is that when R&D staff can't meet the specifications provided by the marketing department, they create something else," says Glen Freiberg, vice president for regulatory, clinical, and quality systems at Bard Diagnostic Sciences, Inc. (Redmond, WA), which focuses on the marketing of in vitro diagnostics. "That can lead to failures in the marketplace."

In such instances, R&D personnel may rationalize their actions by concluding that the expectations established by marketing are simply unrealistic. Applying the team concept provides Bard Diagnostic Sciences the means for hashing out such misgivings. "We can put marketing on the spot and say, 'You gave us your design input but you didn't justify it. Where is your marketing study? Where is your strategy?'" says Freiberg. If the design specs established by marketing can't be met, the project might be scrapped, rather than going to market with a noncompetitive product.

BEYOND DESIGN

But the team approach does not stop with product design. At many firms, it carries over to every step on the way to the end-user--testing, regulatory approval, packaging, sterilization, marketing, distribution, installation, and service. In the past, staff charged with these tasks were isolated from one another. Now they are being integrated. Team members versed in specific areas take lead responsibility for getting a job done, but key processes become a team effort.

Product development teams are typically formed in the same way they have been for decades--by and around R&D staff proposing technology within the state-of-the-art envelope. When the team concept is applied, however, this nucleus of engineers soon reaches out to involve others.

"The old way of developing a product was to throw it over the wall at each stage," says Tomisaka of Medex, which makes a wide range of medical devices from pressure monitors to disposables. "We used to do that, but it didn't work very well. We had a lot of hitches at the last minute. So we have formalized our development approach around what we call cross-functional teams."

Medex starts with a team leader, usually from R&D, and a senior "sponsor"--an executive who serves as liaison between the team and top management. To this core the company then adds other members, usually marketing and engineering experts, who put together a proposal and budget request. When completed, the request is sent to executive management for approval.

As the project unfolds, additional team members are selected from the various disciplines within the company. At a minimum, specialists from R&D, marketing, and quality assurance are involved early on. "You can't be just R&D driven, or marketing driven, or quality driven," Tomisaka says. "The whole company needs to be involved."

To ensure that product development remains on track, phased gate reviews are scheduled throughout the process, marking the end of such phases as the development of the prototype, design transfer to manufacturing, and process validation. Along the way, the team leader tries to build consensus. "The process is driven by the team leader and that is where management skill comes into play," says Tomisaka.

At Protocol Systems, multidisciplinary teams have been a cornerstone of the company's development process since its inception some 10 years ago. The company, whose 400 staff members design and manufacture patient-monitoring instruments, organizes the team around two lead personnel. One is an engineering project manager; the other is a marketing product manager.

Over the course of the project, the two leaders bring in different professionals as needed. "People can be in and out of the project, depending on the phase," explains Protocol's Sandberg.

The size of the team is proportional to the project. "The smaller the project, the fewer the people, and vice versa," says Sandberg.

MAKING TEAMS WORK

At Protocol Systems, a project typically begins with a nucleus of about a half dozen specialists who develop a project proposal and solicit input from others as needed to prepare the proposal for senior management. That nucleus expands in the next phase of development, when product specification requirements are established. At this point, team members try to define the product from the perspective of its customers. "There are a whole series of customers," Sandberg says. "The customer can be a regulatory agency, a distributor, or the end-user." The engineering project manager is responsible for controlling changes, as in the case of document specifications or the development schedule.

Marketing specialists must know what the end-user wants and what competitors will be offering. "If you're trying to compete against something head-on--a similar technology--you need to know how long that competing product will be on the market," says Freiberg of Bard Diagnostic Sciences. "If you're going to be competing against an evolving technology, you need to know how much longer until a similar competing product will show up."

Academic and major medical institutions provide an important source of expertise, particularly when determining a product's features. Marketing and clinical affairs specialists at Cardiovascular Dynamics (Irvine, CA) routinely survey health-care providers, conduct face-to-face visits with medical opinion leaders, and document hands-on experience by end-users to help the company fine-tune its products as they move toward full production.

Quality assurance staff are becoming increasingly important members of the team. The quality system regulation is providing the spark for including QA on the ground floor for new projects. "QA professionals aren't going to be isolated anymore; they will have to interface with management, with R&D professionals, and so on," says FDA's Trautman.

In some instances, they may be directly involved in the development of product specifications, even though the actual design is being developed by a product engineer. "In our organization, the quality team establishes the product requirements necessary to meet performance, safety, and environmental standards," says Protocol's Sandberg. "If you don't have those pieces laid out at the beginning of the project, it's going to be pretty hard to know whether a product's design will meet its objectives."

At Protocol Systems, quality assurance specialists are brought in at the very genesis of a project. "They help develop the product proposal," says Sandberg. This proposal briefly describes the product's opportunities in the market, the regulatory requirements it must meet, and any special manufacturing requirements.

Protocol Systems breaks its quality personnel into two groups. One identifies quality standards such as those developed by the government or by other standards-writing bodies. The other documents safety standards used by testing laboratories and regulatory bodies. At least one member of every quality team is assigned to look into requirements that have been specified by customers.

Manufacturing personnel must be involved to ensure that the final product can ultimately be mass-produced and that the correct equipment is available, not only for manufacture but for testing. Engineering and manufacturing personnel work together to develop the process needed to make the finished device. Specialists are assigned to establish the documentation requirements up front.

At Bard Diagnostic Sciences, all departments get involved early in planning. "Marketing should tell you what the product could sell for, and should also estimate what the reimbursement for its use may be," Freiberg says.

The company also brings manufacturing concepts to bear on the development process long before production begins. "Where we used to have a potential for certain things to pass from one department to another without an efficient transition, the quality systems department is becoming more and more relied upon to make that transition smooth," says Freiberg.

According to Freiberg, the company now uses R&D staff to develop the step-by-step procedures for manufacturing, and to set up a pilot run that will create products for clinical evaluation. A manufacturing engineer is then brought in to observe the pilot process. The R&D team documents the major steps involved; manufacturing--through its observation of the pilot--documents the necessary detailed information that correlates with those steps. "For example, if R&D has developed a 10-step procedure, by the time manufacturing is done we may have 30 steps," says Freiberg. The job of quality systems personnel is to manage the paperwork and ensure that all documentation is satisfactory, he explains.

PLANNING THE REGULATORY PROCESS

The most critical role of the regulatory affairs staff is to plan the strategy for gaining regulatory approval or clearance of products. The regulatory affairs department identifies the review process at FDA that must be used--premarket notification (510(k)), premarket approval (PMA), product license application (PLA), or a combination if the product is a hybrid between a drug and a device. The department also helps to estimate the path of least resistance to clearance or approval, and the shortest time to market. Such estimates, however, remain "more of an art than a science," says Freiberg.

With many companies looking to gain approval first in Europe, where the regulatory maze can be navigated more quickly, regulatory affairs specialists are often charged with developing a global strategy. The trick is to come up with a complementary process for European clearance that will support approval or clearance of the device by FDA, since the U.S. marketplace accounts for literally half of all medical device sales globally.

OUTSIDE ASSISTANCE

Consultants can also play a key role in the team approach to product development. Their advice may be very specific--relating to issues such as the type of electrical connector or sterilization process needed to meet regulatory requirements--or it may span a range of considerations. Consultants to the medical device industry include a varied array of specialists in R&D, design, process design and automation, manufacturing, sterilization, packaging, labeling, and regulatory affairs.

Consultants can--and often do--become intimately involved with a manufacturer's product development team through strategic relationships that make them essential to the design and development process. Consultants with Quality Solutions, Inc. (Annandale, NJ), a firm that specializes in providing advice regarding sterilization issues, routinely serve as team members for device development. In that role, they may actually function as project managers with authority for all decisions related to the selection, purchase, installation, and qualification of equipment required for product sterilization.

"On its team a company may have a wide variety of in-house staff such as engineers, quality systems experts, a registered nurse in charge of clinical trials, and an MBA handling company finances. But it might not have anyone with sterilization experience," says Paul Sordellini, vice president of Quality Solutions. "So the company hires us and turns over the entire project of planning the sterilization strategy for the device."

Quality Solutions staff often play an important and necessary role in designing a company's product. They review engineering drawings and plans for manufacturing and packaging in the context of materials and components to be used. "Certain raw materials may preclude the use of gamma radiation or E-beam sterilization, because of the way the devices would be bulk packaged," Sordellini says. "The physical configuration of a device may create other concerns."

The relationship between a company and its consultants--and their direct input into the team decision-making process--continues throughout the product development stage and into the manufacturing stage to ensure the quality of the finished device. Sterilization protocols developed by Quality Solutions and implemented either by in-house personnel or by contract sterilizers must be audited, Sordellini notes. They may also need to be adjusted if the company modifies the device. Quality Solutions, in the role of consultant, is often called in to provide that assistance.

MANAGEMENT CONSULTANTS

Also available are consulting firms that focus mainly on a company's management structures and internal processes--including those used for product development. Often such firms help companies prepare for certification to one of the ISO 9000 family of standards or for the CE marking of their products. When they play a role as both consultant and assessor, a certain detachment is expected.

"The extent to which we get involved really depends on how large the company is and whether it has resources on the inside to help," says Susan Reilly, director for quality assurance and compliance at Medical Device Consultants (North Attleboro, MA). Start-up companies, she says, tend to be looking "for somebody to do a little more than hand-holding."

For the most part, such management and systems consultants remain on the outside looking in. They provide advice when asked, but exert relatively little direct influence on the decision making for a particular project. "Companies typically hire us to lay some groundwork, conduct some training, and perhaps put together a plan or system that they can implement," says Reilly. "Then when they are done--or think they are done--we come back to see if they really are."

When such consultants do become more deeply involved in development, it is usually in the role of facilitator or catalyst. "I'll go in and get everybody on the same page," says Anita Thibeault, principal of Anita Thibeault & Associates (Rogers, AR), a consulting firm specializing in the device industry. "We'll go through the nomenclature and company procedures so everybody knows what has to be done and what the rules are for doing it."

The vehicle Thibeault uses for getting to that stage is the workshop--a two-day interactive session attended by about 20 professionals representing R&D, marketing, quality assurance, manufacturing, and regulatory affairs. "These folks typically represent the team that the company plans to put together," she says. "The workshop gets them started asking the right questions." Thibeault typically holds a follow-up workshop focused on reviewing the design of the device.

A workshop serves as an event that signals the true beginning of a project, says RELA's Wood. "It makes the project real. It's like somebody ringing a bell and saying, 'You'd better get on board because the train is leaving.'"

Workshops pull together all of the company staff that hold a stake in the final product--engineers in R&D and manufacturing, as well as marketing, QA, and regulatory affairs specialists. RELA promotes workshops as a catalyst for team building to advance the design and development process. What goes on at these workshops can be remarkable, says Wood. "Often you will see discoveries being made among department personnel, a unifying among the participants about the nature of the product itself."

RELA has facilitated 18 workshops over the past 24 months. The common challenge addressed in each has been keeping participants moving toward their goal. Wood uses tricks, such as allowing each participant two 'tangent tickets.' Spending a ticket allows the person to go off on a tangent for three minutes. Then he or she has to come back to the main discussion. "I time them," Wood says. "That tends to control where we go."

The result of the workshop is a clear sense of what the team members must do to make the project progress. Direction is clarified and missing information is identified. "We have brought these people together and they now feel like a team," Wood says. RELA follows up, often within 24 hours, with a flip-book containing the key points discussed and conclusions reached. "The book captures all the things that occurred, but it is not a lengthy document that you have to wade through," Wood says. "Here is the workshop we did yesterday and here are the conclusions. Bang." Quickly putting a summary in the hands of participants enables them to continue with the project while the excitement of the workshop is still with them, Wood explains.

GOING BEYOND WORKSHOPS

Following the workshops she facilitates, the role played by Thibeault usually shifts to one that is more often associated with consultants, namely, giving advice at critical points en route to full production, such as process validation. "Companies ask me to come in to make sure they're taking the right steps--for instance, to advise them about what should be documented and how they should put that documentation together," she says.

RELA also changes direction after the initial workshop, but often to a much different track. The company is actually more a device development firm than a consulting firm. "Consultants typically write a report and leave," Wood says. "They have no stake in how things eventually turn out. We want to be part of the development team. Our stake is a long-term relationship that leads eventually to manufacturing."

The company can and often does run the entire product development process, using the client's in-house talent in a review capacity. In those instances, RELA initiates the project with a workshop and uses periodic review sessions with the client "to ensure that we are on the right track," Wood says. Under those circumstances, the client is an extension of the team, bringing to bear an extended, multidisciplinary source of feedback.

PURCHASING AS A TEAM EFFORT

The use of product development teams not only shapes a company's finished devices but also exerts significant influence over its purchasing decisions for materials, components, manufacturing supplies, and services. In small companies, which most medical device manufacturers are, these decisions have traditionally been made by manufacturing personnel. But as a company's development process becomes more sophisticated, its manufacturing personnel become more likely to call in other specialists to help make these determinations.

At Protocol Systems, the key members of the team that develops a device are also responsible for selecting its materials, components, supplies, packaging, and labeling. "This effort is a combination of design, engineering, manufacturing, and quality assurance," Sandberg says. In some cases, the team brings in a materials manager to help.

Specialists in materials management and inventory control can sometimes exercise control over the specifying and ordering of raw materials, manufacturing supplies, packaging, labeling, and software for inventory control. But the actual decision-making process depends heavily on what is being purchased and on the risk involved in its use. For example, when purchasing an off-the-shelf component or buying from a vendor with a track record, the project team can act on its own. But if a subassembly must be custom-built, then the team makes a recommendation to senior management. "Usually the vice president for manufacturing reviews the proposal," says Sandberg, "because he's the person who will be working with it throughout the life of the product."

CONCLUSION

The team approach is itself a product of the modern health-care environment. To compete effectively, companies must waste nothing of value--not expertise, money, or time. "A multidisciplinary team approach is a lot better than R&D just making a product with a little feedback from marketing and then cramming it down manufacturing's throat," says Mark Siminuk, senior development engineer at Cardiovascular Dynamics. "You end up with a product that is more marketable."

Teams also reduce the risk of losing time and money in virtually every phase of product development, from design through production. Individual savings mount up as fewer unexpected problems arise along the way. But the biggest advantage of all, says Sandberg, is that less time is needed to commercialize a new technology. "The more things you do in parallel, the faster you can get to market."

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

FIRST PERSON

Associate vice president and special counsel for HIMA commends FDA for its recent reforms but urges the agency to continue change.

If FDA wants to promote rather than stifle innovation, it must continue to modify its inspection procedures. In the past, an FDA inspection of a manufacturer's facility was often an adversarial encounter: FDA's goal was to collect evidence for prosecution or another type of regulatory action. But in the mid-1990s, reduced congressional funding forced the agency to reevaluate its operation. Practical suggestions from industry convinced forward-thinking FDA officials to take action. FDA has since begun the process of changing its procedures. Even though both industry and the White House have acknowledged FDA's efforts, much work remains to be done.

INSPECTIONS WITHIN THE SCOPE OF FDA'S MISSION

FDA's mission, which was revised in July 1995, states that the agency is a team of dedicated professionals working to protect and promote the health of the American people. FDA is responsible for ensuring that . . . products are in compliance with the law and FDA regulations; noncompliance is identified and corrected; and any unsafe or unlawful products are removed from the marketplace.

For many years FDA maintained a different stance. To carry out its mission, it developed manuals explaining its regulatory philosophy. These manuals, available to the public from the National Technical Information Service, specifically define an investigator's responsibility during each step of an FDA inspection. Subchapter 150 of the Investigations Operations Manual (IOM) explained,

All regulatory procedures are designed to discover and develop evidence of violations with the ultimate aim of assuring compliance with the law and protection of the public.

According to David Haggard, director of FDA's Division of Compliance Policy, this statement was not in line with the agency's philosophy and has been deleted from the manual. FDA's current philosophy for inspections, which, according to Haggard, is more appropriately reflected in section 502 of the IOM, states,

An establishment inspection is a careful, critical, official examination of a facility to determine its compliance with the laws administered by FDA. Inspections may be used to obtain evidence to support legal action when violations are found, or they may be directed to obtaining specific information on new technologies, good commercial practices, or data for establishing . . . regulations.

This subtle shift in philosophy shows that the agency is learning it must work with device companies to improve regulation. It has also modified its thinking about the objectives of the inspection process. Perhaps, after the new quality system regulation is implemented, the description might be further modified to reflect a more positive purpose, such as verifying that a company is producing safe and effective products that comply with its internal quality procedures.

Unfortunately, the IOM still encourages investigators to be aggressive when requesting records to which they are not legally entitled. Section 514 states,

Refusals must be also discussed in the EIR [establishment inspection report] whether or not it is [sic] a refusal of legally required information. Provide comprehensive details on all refusals in your EIR under the heading captioned "A Refusal to Permit Inspection."

FDA should delete this section from the IOM. Industry regulatory experts report that responding to an investigator's request for data that FDA is not entitled to is one of the most terrifying challenges they face.

SUGGESTIONS FROM INDUSTRY FOR IMPROVEMENT

In 1994, the Health Industry Manufacturers Association (HIMA) polled industry on its concerns and developed recommendations to improve the inspection process. Using these suggestions, HIMA officials began meeting regularly with Ron Chesemore, Gary Dykstra, and Debbie Ralston from FDA's Office of Regulatory Affairs and Bruce Burlington and Lillian Gill from the Center for Devices and Radiological Health. The suggested items included:

• Conducting preannounced inspections.
• Annotating the FDA-483 with completed or promised corrective actions.
• Requiring that observations on the FDA-483 be put in context (e.g., the inspector examined 50 complaints and found that only 3 had not been reported as medical device reporting [MDR] items).
• Promoting and recognizing investigators for working with industry to achieve compliance rather than initiating regulatory actions.
• Having medical device specialists perform inspections.

After listening to these suggestions in meetings at FDA headquarters, in grassroots meetings around the country, and during FDA's teleconference on compliance, agency officials met with an industry group chaired by Wendell Gardner, senior vice president for COBE Laboratories, Inc. (Lakewood, CO). The agency, aware of its diminishing resources and the group's reasonable suggestions, decided to implement a pilot program that included the following features:

• Preannouncing inspections.
• Annotating the FDA-483 to include corrected items and putting observations in context.
• Issuing close-out letters after inspections were completed.

The program was so successful that this March these pilot program features became part of FDA's standard operating procedures (SOPs) for conducting medical device inspections.

One important feature of the pilot program that was not incorporated into the SOPs was the provision allowing the companies under inspection to evaluate investigators for following the spirit of the pilot and to judge whether the program improved communication. The fact that investigators knew they would be evaluated when the inspection ended appeared to improve communication between the investigators and the companies.

FDA should reinstate this provision. To preserve anonymity and quash industry fears of retribution, FDA should have companies send evaluation forms directly to headquarters or to a third party rather than to the investigator's district director.

Another initiative that FDA has implemented is a certification program for medical device investigators. Its primary objective is to educate investigators about the medical device industry and specific technologies so they can conduct better investigations. In addition, the certification program promotes uniformity in the enforcement of FDA's regulatory responsibilities. Industry supports this effort, and medical device manufacturers are eager to work with FDA to provide training on emerging technologies.

In response to industry's concern about linking awards and promotions to an investigator's success in detecting and supporting regulatory actions, FDA in 1995 initiated the compliance achievement reporting system (CARS). This system allows frontline regulators to enter compliance accomplishments into a database. The information can be used to reward investigators for working with industry to achieve voluntary compliance. Industry applauds this initiative but wants to work with FDA in an oversight capacity to ensure that corrective actions are voluntary, and that individual investigators are not encouraging companies to change their procedures just so the investigator can receive recognition.

To get more feedback on specific initiatives, FDA invited industry from Dallas and Boston to meetings on ways to facilitate effective communication between FDA and industry. FDA has also met with industry during grassroots meetings in Atlanta, Charlotte, and Orlando. During these meetings, attendees were polled on ways to improve the inspection process. Their suggestions included:

• Conduct joint training of industry and FDA investigators on the new requirements under the quality system regulation and MDR requirements. Rationale: Training FDA investigators and members of industry together minimizes the chance of interpreting specific requirements differently.
• Provide EIRs automatically to companies after inspection. Rationale: If the purpose of an inspection is to judge a company's compliance with regulations, the company should have access to FDA's conclusions about its compliance status as soon as possible after inspection.
• Mandate that FDA investigators be permited to request only that information to which they are specifically entitled under the law. Rationale: By having FDA investigators request only what they are specifically entitled to under the law, industry officials will be less intimidated during FDA inspections.
• Issue warning letters from headquarters rather than from the field. Rationale: The industry needs to operate in an environment that is predictable and consistent throughout the country. Empowering individual districts to issue warning letters allows for various interpretations of the regulatory requirements and reduces predictability for companies that have plants in several districts. Having a centralized office at FDA issue or approve the issuance of warning letters will help ensure that similar conditions receive similar sanctions.
• Do not include in a warning letter items that have been corrected or that the company has promised to correct. Rationale: Investigators annotate FDA-483 items with information about performed or promised corrective actions. The purpose of a warning letter is to put companies on notice that, if they fail to correct the listed conditions, FDA will take regulatory action against them. Because the FDA-483 already lists the conditions, repeating the items that companies have corrected or have promised to correct serves no useful purpose.
• Give companies 5 to 10 days to respond to an FDA-483, include the response in the warning letter, and allow FDA 20 days to send out a warning letter. Rationale: By increasing time frames, companies will have more time to provide a thorough response to the FDA-483. Then FDA can read the response and include it in the warning letter if necessary.

In response to the suggestion to provide joint training, FDA's Southwest Region conducted joint training for FDA and industry personnel on how to comply with the MDR requirements. FDA also conducted joint training on how to comply with the design control portion of the new quality system regulation. Moreover, in response to the suggestion to provide EIRs automatically to companies after an inspection, the agency set up a pilot program to do so temporarily. Industry is optimistic that FDA will continue to conduct grassroots meetings and implement these suggestions and others that arise from those meetings.

OTHER STIMULI TO CHANGE

Quality System Regulation. Although it mandates the requirements for quality systems, FDA's new regulation allows manufacturers to develop systems that meet their individual needs. Because the regulation is written in a nonprescriptive manner, investigators should not be prescriptive during inspections.

The National Performance Review. The National Performance Review program to reinvent government shows that the government realizes that the majority of firms in any industry, including the medical device manufacturing industry, want to comply with clear, reasonable, and predictable regulatory requirements. To keep in line with this philosophy, FDA should enter into productive partnerships with the majority of firms to allow innovation and promote voluntary compliance. By doing so, inspection time should not be focused on the majority, which are essentially in compliance with the law, but rather on the minority that seek to get around the rules.

In 1996, Jean Logan from the National Performance Review and Marie Urban from FDA's Office of Regulatory Affairs began to look for occasions on which FDA entered into productive partnerships with the medical device industry. On five occasions, FDA personnel from the districts and from headquarters were recognized for partnering with industry and received the National Performance Review's Hammer Award. On three of these occasions the industry partners of FDA--HIMA, FDLI, and Storz Instrument Co. (St. Louis, MO), respectively--were also presented with the award.

FDA Reform. The slowdown in product approval in the 1990s, competition from overseas companies, the drying up of venture capital, the movement of research and development overseas, and the Republican Congress have prompted industry to work to achieve legislative reform of FDA.

One goal of such reform is the establishment of a reasonable and predictable regulatory environment that does not depend on the incumbent FDA commissioner or center directors, but rather on a statute about which FDA officials have limited discretion. FDA has consistently contended that it needs unlimited power, and that industry should trust it to use that power judiciously. Conversely, industry's position has been that FDA should be given power judiciously.

Industry's efforts to achieve legislative reform may have motivated FDA to implement many of industry's suggestions. Although industry is pleased by FDA's actions, without legislative reform there is no guarantee that the changes will be permanent.

FDA, industry, and Congress need to work together to develop regulation that allows companies to operate in a reasonable and predictable regulatory environment based on law rather than on the discretion of the officials who are charged with carrying out the law.

PROSPECTS FOR THE FUTURE

In any regulatory system, abuses can occur when the government operates in isolation. For the past few years, the White House, Congress, and the press have focused on the challenges brought on by FDA regulation. On the positive side, FDA has worked with industry on many new programs to produce a regulatory environment that is more reasonable and predictable than before. Industry's continuous assistance will help the agency build on its recent successes.

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

SNAPSHOT

Robert Thompson on developing medical packaging.

It started with the complaint that gastroenterologists couldn't read labels in their darkened labs. The physicians, who view the internal organs of a patient on a video monitor, couldn't turn the lights on to read, because it took too long for their eyes to readjust to the dark when they turned them off again. To help these physicians use medical products in this setting, Robert Thompson, manager of the package engineering department at Boston Scientific Corp. (Watertown, MA), created packaging with phosphorescent labeling. Not only did the physicians welcome this feature, but the Institute of Packaging Professionals (IoPP) recognized Thompson's efforts with an AmeriStar award and the World Packaging Organization gave him a WorldStar award.

"Listening to a customer is often the best way to design a package," explains Thompson, who has been designing packages for medical devices and diagnostics for a decade. "Customers know what they're looking for and can give you the best information to help you develop the package design." The phosphorescent labeling was the result of such communication.

Before focusing solely on medical devices, Thompson, who holds several U.S. and international patents, designed packaging and packaging systems and equipment for pharmaceutical, food, medical device, and industrial products for various clients. Such experience prepared him for his position as senior packaging engineer at American Cyanamid Co., Davis & Geck Div. (Danbury, CT), where he developed packaging for sutures, wound-closure devices, bioabsorbable bone screws and pins, and operating room products. To design such packaging, he had to understand how extreme heat and high doses of gamma, E-beam, and EtO sterilization affect packaging.

Over the years Thompson has come to believe that a good package does more than just hold a product. "A good package meets end-user preferences, offers convenience, protects a product, meets regulatory requirements, addresses environmental concerns, and satisfies manufacturability and cost issues," he explains.

In order to heed the environmental call to reduce, reuse, and recycle, Thompson says, designers are producing packages that use less material. "Due to health issues, a large percentage of medical waste is not recycled. Source reduction is the best way to lessen the impact of medical packaging on the environment," he says. "Whenever we can, we do try to make packages recyclable." The thermoformed-tray-and-lid system designed to protect Boston Scientific's Symphony nitinol stent, for instance, was made from recyclable materials. Thompson's efforts secured him an IoPP merit award in 1996.

Influencing how Thompson develops packages is the industry's move toward a global market. In order to obtain the CE mark, necessary for selling devices in the member states of the European Union, manufacturers must label their devices with instructions in the languages of the targeted markets. "To deal with this trend, most of the industry is using a multilingual labeling format," Thompson observes. "This challenges the packaging developers to deal with limited real estate--it's very difficult to fit all the necessary languages on the package."

Another trend, driven by changes in regulatory and quality control requirements, is the push for designers and engineers to validate the process as well as the package. "In the past, packages were developed and then tested for applicability and performance," explains Thompson. "Today, most companies are also validating the systems that develop the package. You therefore have a greater assurance that the package is going to be effective."

This trend has fostered a relationship in which packaging component suppliers and developers work together to validate the entire packaging supply chain, including the processes for making packaging components. "This helps ensure that the components the suppliers provide meet the developer's needs," Thompson says.

Thompson also sees a continued move toward using less-invasive medical devices and early interventional procedures. "Ailments that once required surgical repair are often being corrected at an earlier stage with minimum trauma to the patient. These procedures are being performed by radiologists, cardiologists, and others. The packaging designer must understand the various needs of these physicians and the environments in which they work."

To keep up with these trends, the packaging designers of the future must wear many hats--"engineering, regulatory, and marketing hats," explains Thompson.

Daphne Allen is associate editor for MD&DI.

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

VERBATIM

An interview with Michael Baker, director-general, European Confederation of Medical Devices Associations (EUCOMED)

With the increasing globalization of the medical device industry, it might be tempting to think that the importance of national boundaries is on the wane. But that notion is soon put to rest by even a short conversation with Michael Baker, who was recently named director-general of the European Confederation of Medical Devices Associations (EUCOMED).

A veteran of the medical device industry, Baker was a cofounder and the first chairman of the UK Sterilization Packaging Materials Association. Before joining EUCOMED, he served as secretary-general of the European Sterilization Packaging Association (ESPA).

Now Baker is confronting the inherent challenges facing a trade association whose membership comprises a variety of national associations. In this interview with Steve Halasey, editor of MD&DI's sister publication IVD Technology, Baker discusses the issues of importance to the European medical device industry, and how the European approach might influence U.S. manufacturers.

Q.  As we speak, U.S. and European officials are very involved in talks to establish mutual recognition agreements (MRAs) between the two markets. What is Europe's position in those discussions, and what is EUCOMED's role in advancing that position?

A.   President Clinton's hope was that there would be an MRA on medical devices by the end of January, but that hasn't happened. There have been a number of talks in both Brussels and Washington, and those are to be furthered with more discussion later this year.

We have probably reached a situation where FDA is unable, for whatever reason, to give any more than it has already. There is an agreement to recognize quality systems inspections and to allow a limited number of devices to undergo a pilot program for premarket approval by a third party. But obviously, Europe wanted far more than that.

Europe wanted the two markets to agree to a complete recognition of each other's systems and procedures but, unfortunately, FDA is not willing to relinquish its rights as the sole custodian of the health of U.S. citizens. Nor, at present, does it have a legal basis for relinquishing those rights. As a result, Europe is unlikely to get precisely what it wanted at the outset.

From a EUCOMED perspective, it would certainly be my hope that European industry and the European Commission will be able to accept FDA's response. Provided we have an understanding, this could be merely a stopping point in the progress toward mutual recognition. And we could then agree on a transition period, maybe 18 months or so.

Q.   What would you see developing in that subsequent period?

A.   I see the interim agreement as the basis for a transitional, confidence-building period of up to three years, leading to an extended operational agreement. It would be my hope that by the end of that period, we would get to a situation where FDA is prepared to throw more into the pot. But if Europe agrees now to move ahead on the basis of such an interim agreement, it must retain the right to opt out if, at the end of the transitional period, it appears that the United States is not prepared to move any further.

Q.   Does it make the European medical device associations nervous that the United States doesn't have a system for qualifying notified bodies or third-party inspectors?

A.   The fact that FDA does its own product approvals eliminates the need for notified bodies. But it would be a lot easier to attain complete recognition between the two systems if in the course of time FDA were to recognize third-party assessment bodies in the United States. Notified bodies make life a lot easier for Europeans, and I think they would make life a lot easier for FDA, because one of its problems at the moment seems to be its workload.

Q.   The biggest stumbling block in the MRA talks seems to be in mutual recognition of product approvals. If some kind of interim agreement is reached, how long do you think it will take to evolve a system where there is really complete mutual recognition? That may be asking, "How quickly can FDA give in?"

A.   I think that is probably what it boils down to. Because Europe is ready for that now, and that is what we would welcome now. We recognize that FDA has legal restraints that will not allow it to relinquish its final veto, but we hope that there is commitment to seek the legislative changes necessary to achieve this.

It is very creditable that the agency takes its responsibilities so very seriously, and obviously it has to do so. But there has to come a time--in the interests of free trade, open competition, and global market access--when FDA relaxes its hold and gives other bodies the credibility they deserve. Sometimes FDA gives one the impression that it believes there's nobody in Europe who is as concerned about the health of their own people as FDA is about the health of American citizens--and that is just blatantly not true.

So I think FDA needs a period of confidence building to appreciate and understand that we do take these matters very seriously in Europe and that we are equally intent on ensuring that our peoples' health is not impaired by a lack of enforcement or regulation.

In terms of how long it's likely to take, I think the answer is, "How long is a piece of string?" We would hope that if there is an agreed transitional period for this confidence-building exercise, it would be as short as practically possible. Having said that, it's likely to be three years. Even then, we may not see complete mutual recognition for all devices. But we would hope to see at least progress toward full mutual recognition on Class I and Class II devices via a change in legislation, followed by initiation of a program to build confidence with the higher-risk, Class III devices. It may take more than five years to actually get a level playing field and complete mutual recognition.

Q.   FDA has said that guidances are essential to ensure that all device reviewers are on the same page when they review a product submission. Do you think that might be a stumbling block?

A.   I think that is probably a very major practical stumbling block. But one needs to be pragmatic about this sort of situation. Europeans need to recognize that FDA does have a few practical problems, and this is one of them.

Obviously if the doors were to be opened to specific devices, then FDA would have to have inspection guidelines for those devices. Otherwise, how would the European test houses and notified bodies know what they're testing for and what criteria are to be used? I have heard that the list of some 30 devices that FDA has agreed to release for free interchange was in part dictated by the fact that it has the necessary documentation for those devices.

But FDA doesn't have guidance for the hundreds of other products that we would ultimately want to see, and it would be a hell of a task for the agency to produce the required documentation in a very short space of time. So, it's a very practical problem, one that in Europe we have to try to understand. It is a much more practical and pragmatic approach to rely on the use of international standards.

Q.   FDA has suggested that industry begin the process of writing guidances and submit them to FDA as a starting point. Would European device associations be interested in contributing to that process?

A.   No. European industry is interested in using international performance and safety standards--not in the production of guidances on a device-by-device basis.

Q.   In the international standards-writing process, U.S. organizations sometimes feel that they are inappropriately outvoted. How does one organize such an effort to ensure that the greatest weight is given to the proper entities?

A.   To my knowledge, this is not true. FDA needs to assign greater resources to standardization work, and fewer to the production of internal guidance documents that essentially become mandatory.

But I would question whether the United States is not one of the dominant forces in ISO [International Organization for Standardization] already. From the European perspective, the United States is certainly perceived to be dominating ISO, and I am not personally aware of any instances where standards adopted by ISO did not first meet with U.S. favor or approval.

Q.   But in some instances--such as the ISO 10993 standard on biocompatibility testing--FDA has refused to accept the standard "as passed" by the international committee. And the result is that U.S. manufacturers still have to satisfy two sets of standards, however similar.

A.   n that situation, of course, it's the governing body for U.S. industry that's the stumbling block. It's not European industry or Korean industry that's telling the United States, "No, you can't have your way." It's the United States itself saying, "We are not prepared to accept this standard." And presumably this is because the procedures are not up to the standards that FDA would require.

Q.   So, essentially, U.S. manufacturers have to put more pressure on FDA to accept those standards?

A.   If that's the situation, yes. The rest of the world recognizes ISO as being primarily U.S.-controlled. But if U.S. industry is having a problem getting approved standards through its own controlling body, then that is very much a U.S. problem. So U.S. industry needs to exert a little more pressure to make sure that its views are well understood at FDA. Or, vice versa, FDA needs to explain to U.S. industry why the approved standards are not acceptable.

&Theoretically, since FDA is usually involved in the ISO committees that generate standards, the agency should have ample opportunity to provide input. When FDA has had a chance to be involved in the discussions from the word go, it shouldn't then effectively veto industry use of an approved standard. I find that absolutely amazing.

Q.   EUCOMED must have to face these questions of procedure and structure all the time in dealing with its various national associations and corporate members. How do you manage to deal with all of them?

A.   Basically all the decision-making processes within Europe are based upon consensus. There are occasions when one of the big five states, which have the power of veto, will actually exercise that power. But that doesn't necessarily stop legislation from happening. For example, the United Kingdom can be against a particular piece of legislation, but that legislation will nevertheless go through by consensus view.

We try to operate EUCOMED in much the same way. The association membership includes at least 15 different nationalities, so it is difficult to reach consensus. But we try to ensure that issues of importance are discussed very openly, and as a result of that discussion we hope to arrive at a consensus.

When it is empowered by its members, EUCOMED seeks to take a leading position on certain issues. If there is no consensus view, then there is a structure within EUCOMED that determines what view will be adopted.

Q.   Could you explain EUCOMED's request for an extension of the June 14, 1998, effective date for the Medical Devices Directive?

A.   EUCOMED is not really requesting an extension. It is trying to make it clear that for medical devices legally placed on the market without CE marking prior to the end of the transitional period, there is no legal basis to force their removal or withdrawal after that date.

The point at issue really is the question of CE marking. The directive currently says that no product without the CE mark can be placed on the market in the EU after the June 14, 1998, deadline. We are suggesting to the European Commission that this would cost the medical device industry an inordinate sum of money, because it would likely require manufacturers to withdraw a lot of stock that was legally placed on the market prior to the deadline.

Instead, we are recommending that the commission agree to a transitional period of five years, up to June 2003. During that period, non-CE-marked products already placed on the European market prior to June 1998 could continue to be sold and used. June 2003 would be an agreed deadline for their removal.

We think that this proposal will minimize the cost and inconvenience of CE marking not only to industry but also to hospitals. Otherwise, it's going to be a fairly massive task for hospitals to go through all their inventory on June 13, 1998, to discover what's got to be removed and what can continue to be used, and then to replace their old, unmarked stock. The hospitals aren't really equipped to do this and shouldn't have to spend time doing it. Our proposal will benefit both health-service providers and the medical device industry.

Q.   Has EUCOMED estimated what it would cost companies to withdraw, relabel, and redistribute their products?

A.   We estimate that if 10% of products do not have the CE mark on June 14, 1998, withdrawing them from the marketplace will probably cost industry in the region of half a billion dollars.

But it's not only cost that's at issue. The inconvenience and logistical difficulties involved in such a massive withdrawal are quite horrendous. I maintain that EUCOMED's proposal is of benefit not only to industry, but to the hospital end-user as well. It would save an awful lot of cost and inconvenience. In all honesty, the commission probably recognizes this.

Q.   Would this recommendation apply to both U.S. and European companies?

A.   Yes. It would apply across the board.

Q.   In Europe, the term multinational tends to be read as "U.S.-owned." Does it create problems for Europeans to have such a heavy penetration by U.S. companies in the European market?

A.   I don't think it causes any practical problems. Obviously, it may impinge upon one's sense of national or European pride, but I don't think it generates specific problems.

There are a number of multinationals that are not U.S.-owned and that are very strongly represented in the U.S. market. There are some from Germany, and many from Scandinavia and the UK. So it does work in reverse.

Generally speaking, the biggest device companies tend to be U.S.-owned, and that presumably arises from the fact that the United States is the largest market for medical devices. So U.S. companies have grown to a size that reflects the market's importance and strength.

But because of the length of time it takes to get a premarket approval (PMA) in the United States, many U.S. companies are now going to Europe first to get their new products on the market. In this sense, FDA might be described as not only American industry's best defense, but also its worst enemy.

Even so, Europeans prefer to buy European if they can, and I think this is why U.S.-owned multinationals don't openly portray themselves as being from the United States. When they are in Europe, they set up local operations, they employ local people, and they are represented by local people. That's a very sound move that helps to alleviate problems, perceived or otherwise.

Q.   How might the completion of MRAs between Europe and the United States change that picture?

A.   Right now, the major factor that is causing U.S. companies to adopt a Europe-first strategy is the short time it takes to get a product to market in Europe compared with the United States. Hopefully MRAs will even out the availability of products in the United States and Europe. That will certainly be part of the benefit of having completed MRAs so far as U.S. industry is concerned. Whether the agreements so far completed will have that effect for the 90 devices that FDA has agreed to include in the first three years, I don't know.

Q.   Are there other crisis points that EUCOMED is concerned about--particular market issues that have the potential to affect device manufacturers harshly?

A.   I don't want to be too specific, because that could arouse interest where there is not much at the moment. But the whole problem of cost containment is an increasingly important issue. The German situation is just one example. Health-care budgets are under pressure and are being cut, whether through DRG reductions or simply through sweeping cuts of a certain percentage, and this is a very general problem that industry has to address in some way.

For the most part, companies did not need to be too clever to cover the initial round of cost-containment pressures. So long as the decreases were in the range of 3 to 5%, companies could consider them a straight reduction in pricing--and therefore also a straight reduction in margin--but they were of a magnitude that could be handled. Companies got rid of their fat, reduced their prices a bit, and they covered the decreases. But when demands go beyond that stage--that's when companies have to start thinking cleverly.

Q.   Are current budget pressures forcing manufacturers to trim essentials such as their R&D budgets--or is that still to come?

A.   That is probably still to come. But let us hope it does not arrive at that. R&D is the long-term lifeblood not just of industry but of patient health. If we want to improve our overall standard of health care, we have to ensure that industry continues with its R&D programs.

Q.   Where else in the world is EUCOMED involved?

A.   We work in the overall framework of global harmonization with Australia and Japan, both of which have very strong manufacturer associations. We also get involved in seminars and promotional work in parts of Eastern Europe and Southeast Asia, where we are endeavoring to promote the European style of device regulation. Eastern European countries are keen to become part of the EU. These countries, therefore, need to begin developing the infrastructure and technical expertise that will be required for their participation at the competent authority and notified body levels of the EU regulatory structure.

So, EUCOMED is involved quite heavily in parts of the world that are theoretically not within our immediate scope--in other words, not within Western Europe.

I think it is relevant to stress the importance that EUCOMED attaches to the development of a good working relationship with other associations. In the United States, we work with the Health Industry Manufacturers Association as well as with the National Electrical Manufacturers Association. We also enjoy good rapport with MEDEC in Canada, MIAA [Medical Industry Association of Australia] in Australia, and JFMDA [Japan Federation of Medical Devices Associations] in Japan. International cooperation is an increasingly important element as our industry becomes increasingly global.

Q.   Is EUCOMED also involved in supporting its member associations in technical areas?

A.   Yes. One issue that has become very important in Europe is the propensity for end-users to reuse single-use devices. This is a practice that we are very much against, and we are aggressively publicizing the fact that it is highly dangerous for patient health. We have issued a publication on this topic that is available in six languages, and we are hopeful that hospital end-users will read the document and understand why it is so important not to pursue this policy of reusing single-use products.

It is very understandable why it happens, particularly in light of current cost-containment pressures. Hospitals have demands placed upon them which they feel obliged to meet, and if they can reduce some of their costs by reusing a product, then they will obviously do so. I am quite sure that no hospital would adopt the practice unless it felt that it was reprocessing the device correctly. But the fact remains that this is a highly dangerous practice, and end-users need to be aware of that. The practice should be forbidden.

---

Copyright ©1997 Medical Device & Diagnostic Industry

BIOCOMPATIBILITY

Biomaterials and medical devices constitute an extremely diverse, heterogeneous category of items. Because the use of these products normally entails their direct or indirect contact with patients, there is an obligation on the part of manufacturers to establish the safety of their products before they are marketed. Medical device safety evaluation assesses the risk of adverse health effects due to normal use and likely misuse of a device. Since adverse health effects could result from exposure to the materials from which a device is made, preclinical assessment of the toxic potential of such materials or components is needed to minimize the potential hazard to the patient.

A thorough biocompatibility safety testing program will typically comprise in vivo studies supplemented by select in vitro assays. Photo: Northview Biosciences, Inc.

Until recently, the regulations governing the manufacture and sale of medical devices varied greatly among countries. Since January 1995, medical devices to be marketed in the European Union (EU) have been required to comply with EU Medical Devices Directive 93/42/EEC, which specifies requirements for safety assessment issues. The purpose of the directive is to promote a single European market for trade in medical devices, while ensuring that users and patients are not exposed to unnecessary risks.

At present, safety assessments of medical devices are guided by the toxicological and other studies recommended in the International Organization for Standardization (ISO) 10993-1/EN 30993-1 standard. At present, 17 parts of the standard are either accepted or under preparation. Tests that may be used in an evaluation of medical device biocompatibility include procedures for cytotoxicity, skin sensitization, dermal irritation and intracutaneous reactivity, acute systemic toxicity, subchronic toxicity, mutagenicity, implantation, hemocompatibility, chronic toxicity, and carcinogenicity.

This article is an introduction to a relatively new and rather complicated field in toxicology—the toxicological testing of medical devices. The guidelines for testing such products are discussed and a general description of the various test procedures given. Future developments in the field of biocompatibility regarding international harmonization and the potential for new methodologies are also addressed.

INTERNATIONAL REGULATORY EFFORTS

ISO is in the process of publishing a series of standards on the biological evaluation of medical devices—ISO 10993.¹ Many parts of this series have been accepted as international standards, while the rest are under development (see Table I). The subject of the first part, ISO 10993-1, is the categorizing and performance of safety testing. Part two of the standard, ISO 10993-2, is concerned with animal welfare requirements; another section, ISO 10993-12, deals with sample preparation and reference materials. Most of the remaining parts of the standard treat the individual tests.

Table I. Listing of individual parts of ISO 10993, Biological Evaluation of Medical Devices.

The EU has issued a council directive—93/42/EEC, 1993—concerning medical devices.² All medical devices to be sold on the EU market must comply with this directive after June 14, 1998. The European Committee for Standardization (CEN) is currently in the process of adopting the ISO 10993 standard as the European standard.

In 1986 the responsible authorities in the United Kingdom, United States, and Canada issued the Tripartite document, which was a guidance on the selection of toxicological tests for medical device safety testing.³ This document has now been replaced by ISO 10993-1 as a first step in the process of international harmonization. In 1995 FDA chose to accept the ISO 10993-1 standard, with a modification of the matrix listing (see sidebar below).

Japanese authorities have also issued a guideline for toxicological testing of medical devices. This document is available in an unofficial translation as Guidelines for Basic Biological Tests of Medical Materials and Devices.⁴ It resembles ISO 10993 in structure and content, but recommends modified tests and sample preparations.

Figure 1. Flowchart illustrating steps in the biological evaluation of medical devices according to the ISO 10993-1 standard.

The procedure for using the ISO 10993-1 standard is illustrated by the flowchart in Figure 1. The standard is applicable only for devices that are directly or indirectly in contact with the body or body fluids. If a device is to be subjected to the standard, the first step is to characterize the material. Such characterization need not always be followed by biological evaluation, because there may be sufficient historical data to verify that the device meets the requirements of the standard. If the material and/or the intended use of the device is different from any historical safe device, biological evaluation has to be performed. By following the standard, a suitable test program can be chosen depending on the type and duration of body contact.

Within the EU, all new medical devices must carry the CE mark from June 14, 1998. This should ensure the availability of relevant documentation regarding biocompatibility and the lack of health problems associated with the use of a device. It is noteworthy that the approval of such documentation is not, as it was previously, accorded by the national health authorities, but rather by the so-called notified bodies, whose experts review the products and production facilities of medical device manufacturers.

DEVICE CATEGORY AND CHOICE OF TEST PROGRAM

The need to evaluate a medical device biologically depends on the material used in the device, the intended body contact, and the duration of that contact. A device designed for surface contact for a limited time is not as likely to be bioincompatible as a permanent-exposure implant device made of the same material. The ISO 10993-1 standard divides medical devices into three main categories: surface devices, externally communicating devices, and implant devices. Each category is further divided into subcategories according to the type of contact to which the patient is exposed (see Table II).

Table II. Device categories and examples according to ISO 10993-1.

The choice of test program for a device in a given category depends on the duration of the contact. Three different time spans are given: limited contact (<24 hours), prolonged contact (24 hours­30 days), and permanent contact (>30 days). ISO 10993-1 lists the tests that must be considered for each category (note: the accompanying table could not be reproduced. Please contact the editors for more info).

The ISO test matrix should not be considered as a checklist for the different tests that have to be performed, but rather as a guide for qualified toxicologists who also take into consideration material information and historical data from similar devices. The certifying authorities in most countries (e.g., notified bodies, FDA, Japanese authorities) are generally cooperative when a company must decide on a test program for a device. It is therefore advisable to maintain close contact with the relevant authorities during the entire process. However, testing should not be performed simply to meet regulatory requirements. This is important not only to lessen the risk of overtesting and excessive use of experimental animals, but also because a strict regulatory approach may mask potential negative health effects that might be identified via optional or nonroutine testing procedures.

As regards CE marking of existing products on the market or safety evaluation of medical devices already in clinical use, appropriate historical or clinical data should be employed whenever possible to avoid unnecessary testing.

PREPARATION OF EXTRACTS

ISO 10993-12 describes how samples for biological evaluation should be selected, prepared, and extracted. Other guidelines provide similar descriptions, which differ slightly in the specifics of the extraction procedures.

The device to be tested (the test article) should be a representative specimen of the mass-produced device. It should also be finished or treated (e.g., coated or sterilized) in the same way as the mass-produced device.

Because the toxic potential of materials and devices depends to a substantial degree on the leachability and toxicity of soluble components, extracts of the device are normally used in the tests. In some tests, however, an evaluation under normal-use conditions is mimicked by using the device or a piece of the device directly. Ideally, extraction media should constitute a series of media with decreasing polarity to ensure the extraction of components of widely different solubility properties. The most commonly used extraction media are physiological saline, vegetable oil, dimethylsulfoxide, and ethanol. Other extraction media such as polyethylene glycol or aqueous dilutions of ethanol may be selected in certain cases. For in vitro cytotoxicity testing, complete cell-culture medium is most often employed.

The various guidelines also differ somewhat with respect to the temperature at which the extraction is conducted. Some leachable compounds may be chemically altered at high temperatures, and it is now generally recommended that extraction be conducted at 37°C—simulating body temperature—for 72 hours. This procedure will probably become increasingly accepted as the most appropriate extraction method. For in vitro cytotoxicity tests, extraction at 37°C for 24 hours is usually recommended, since certain constituents of the media are relatively labile.

The amount of leachable substances released to the extraction media is related to the surface area and thickness of the product to be extracted. Recommendations vary from 1.25 to 6 cm² of product per milliliter of extraction medium, depending on the size and shape of the product, or from 0.1 to 0.2g of product per milliliter of extraction medium when a surface area cannot readily be estimated (e.g., for powders or granulates). In any case, the specific properties of the product must be taken into account in order to make usable extracts.

For cases in which a medical device comprises several components made from different materials, the ideal procedure from a toxicological point of view would be to test extracts of the components separately. However, in some situations this is not practical, and extracts of the whole device may be used instead.

BIOLOGICAL CONTROL TESTS

Biological control tests are not described in the ISO 10993 standard for biological evaluation of medical devices, since these particular tests are designed primarily for batch-control purposes. Such tests are also used during the product development phase to identify sources of contamination and to establish procedures that ensure the intended quality of the end product.

Microbiological Control Tests. Microbiological control tests are necessary to establish the microbiological status of an end product—factors such as sterility, absence of pathological bacteria, or limits for microbial counts. Furthermore, it is often necessary to monitor the microbiological load of raw materials and intermediary products, or to check the efficiency of production and sterilization processes. The tests are performed by rinsing the materials or products in physiological saline and assessing the rinsing medium for microbes, or by directly incubating the products in growth media. Although the presence of pathological microbes on medical devices can represent a potential health problem, the subject is not within the scope of this article.

Tests for Endotoxins. Even sterile medical devices may contain cell-wall lipopolysaccharides originating from gram-negative bacteria. Such so-called endotoxins or pyrogens can cause an abrupt fever reaction after entering directly into the body from sources such as venous catheters, syringes, or implant components.

Two different biological assays can be used to measure the presence of endotoxins: the rabbit pyrogen test and the Limulus test. In both cases, an eluate is prepared—normally by rinsing the surfaces of the product with water—and then tested for endotoxins. In the rabbit pyrogen test, the eluate is injected intravenously and the rectal temperature of the animal is measured after the injection. In the Limulus test, the eluate is incubated together with lysate from the blood of the horseshoe crab (Limulus polyphemus), which contains a substance that forms a gel in the presence of endotoxins.

Test for Nonspecific Toxicity. This test is designed to assess any nonspecific adverse effect that occurs following intravenous injection of a device eluate in mice. The test is often performed with the same eluate used for the pyrogen test. The mice are inspected regularly for any signs of ill health, which can indicate the presence of toxic substances leaching from the product.

In May 1995, FDA issued Blue Book Memorandum G95-1, Use of International Standard ISO-10993, "Biological Evaluation of Medical Devices Part 1: Evaluation and Testing." This memo was a huge step toward international harmonization of device biocompatibility testing, although there are still some significant differences between FDA and European requirements. The Blue Book supersedes the Tripartite Guidance as a guideline for planning biocompatibility testing of medical devices for the U.S. market. The most significant change from the Tripartite is FDA's increased emphasis on a case-by-case testing strategy for individual devices. The biocompatibility of a device depends on several factors, especially the type of patient tissue that will be exposed to device materials and the duration of the exposure. Neither the Blue Book memo nor ISO 10993 prescribes a specific battery of tests for any particular medical device. Rather, they provide a framework that can be used to design a biocompatibility testing program. As with the Tripartite and the ISO standard, the core of the Blue Book memo is a materials biocompatibility matrix. The matrix categorizes devices based on the type and duration of tissue contact. It also presents a list of potential biological effects. For each device category, certain effects must be considered. In preparing a regulatory submission, manufacturers must address each of the biological effects pertinent to their device. The matrix from the Blue Book memo incorporates several ISO features: Contact-Duration Scheme—This is potentially one of the most important changes for device manufacturers. The upper limit for duration category A ("limited") was raised from 30 minutes to 24 hours. For some types of devices, such as surgical devices and accessories, this has significantly reduced testing requirements. Device Categories—Those in the Blue Book matrix are substantially the same as in the ISO standard. Test Selection—The Blue Book matrix moves in the direction of what is recommended in the ISO standard. However, FDA often goes beyond ISO, especially in tests for systemic toxicity, subchronic toxicity, and implantation. The Office of Device Evaluation (ODE) emphasizes that the matrix is "only a framework for the selection of tests and not a checklist of every required test." In some cases (e.g., subchronic testing of balloon catheters), FDA has already said it will require testing beyond what is suggested in the matrix. The test-selection matrix applies to most premarket approval submissions. The Blue Book memo also contains a flowchart to assist in selecting toxicity tests for 510(k) devices. If there is no significant change between the new device and the predicate device in materials, manufacturing, sterilization methods, and body contact, no further biocompatibility testing may be necessary. Otherwise, manufacturers are referred to "Device Specific Tox Profiles" (to be issued by FDA at some yet-to-be-determined date) or to the Blue Book matrix. When designing a biocompatibility testing program, manufacturers should also consult other applicable FDA publications, such as Biocompatibility of Medical Devices (from the Center for Devices and Radiological Health), the Guidelines for the Intraarticular Prosthetic Knee Ligament, or the PTCA Catheter System Testing Guideline. Other useful documents include ASTM F 748, Practice for Selecting Generic Biological Test Methods for Materials and Devices, and the AAMI Standards and Recommended Practices, Volume 4: Biological Evaluation of Medical Devices. (These and other publications are available from CDRH Facts-On-Demand at 800/899-0381.) For most projects, companies should review testing plans with FDA before beginning the actual studies. Since the Blue Book memo was issued, a few trends have emerged: Test-Method Harmonization— FDA has not adopted the ISO standards for specific test methods. However, these standards are generally very similar to what has been standard industry practice in the United States, and FDA has accepted ISO-compliant test procedures. Increased Emphasis on Analytical Characterization—Most reviewers will now insist on at least basic analytical characterization of device materials, usually physicochemical tests, infrared spectroscopy, and cytotoxicity. Often a more thorough characterization is appropriate, using techniques such as extract studies, chromatography, and/or determination of physical properties of the materials. More-Sophisticated Genotoxicity Testing—Under the Tripartite Guidance, manufacturers often simply used an Ames test to satisfy genotoxicity requirements. The ISO standard on genotoxicity (10993-3) requires at least three assays to look at the different levels of genotoxic effects—gene mutations, chromosomal aberrations, and DNA effects. While FDA has not adopted this test standard, its review of genotox data is becoming more rigorous—moving closer to that of European regulators even as the preferred test batteries continue to differ. So what's the bottom line for manufacturers planning biocompatibility testing programs? Work with an experienced toxicologist to apply the Blue Book matrix and the ISO standard to your device. Avoid cookie-cutter solutions by carefully evaluating the intended use of the device. And if at all possible, consult your FDA reviewer before beginning your biocompatibility testing program. RELATION TO U.S. PHARMACOPEIA To test medical device biocompatibility, manufacturers often use USP procedures such as the USP in vivo biological reactivity tests (Class I–VI plastics tests). While class plastics tests have some value in a biocompatibility testing program, a full Class VI test is rarely needed for a medical device. As a general rule, the Blue Book memo and ISO documents take a broader and more thorough view of biocompatibility than does the U.S. Pharmacopeia, and they supersede the USP for evaluating which studies to submit to FDA in support of product registrations. SOURCES OF BIOCOMPATIBILITY TEST DATA Data to satisfy biocompatibility test requirements may come from any of several sources. Most commonly, companies arrange for their own biocompatibility studies. Material vendors are often willing to share test data they have generated. If vendor data are used, manufacturers should obtain copies of the original study reports, and they should conduct at least some confirmatory testing of their own (e.g., cytotoxicity and hemocompatibility studies). If available, clinical data can be used to satisfy some biocompatibility biological-effect categories in the matrix. Manufacturers may use analytical data (e.g., extraction studies) to eliminate the need for biological testing in a particular category. Note on Contributors: Timothy V. Doherty, DVM, is director of in vivo services and Jeffrey Wallace is manager of toxicology services at the Northview Pacific Laboratories (Berkeley, CA) division of Northview Biosciences, Inc. (Northbrook, IL). The information is derived from the company's publication entitled Assessing Biocompatibility: A Guide for Medical Device Manufacturers.

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI May 1997 Column

DESIGNER'S TOOLBOX

A new way to make 3-D objects can help designers communicate better early in the product design and development cycle.

Engineers and designers conceptualize ideas in three dimensions, develop designs in three dimensions on computer-aided design (CAD) systems, but then print out a two-dimensional paper representation to communicate the design to others. In most medical device companies, the use of drawings is still the primary means of communicating early design concepts to other members of the product development team--marketing, engineering, and manufacturing. Most designers have experienced the communication failures that result from someone trying to visualize a 3-D design from a 2-D drawing, leaving much of the design intent to interpretation. Because of this, many design decisions are made and significant dollars are committed early in the product development process based on limited product and performance knowledge.

Because of the cost of rapid prototyping, device manufacturers historically have not built 3-D prototypes until the verification and test phase. Often, it is at this point that misconceptions and misinterpretations in the product development cycle come to light. For example, a designer of a socket for a prosthetic leg might use a CAD program to prepare the design, but print a 2-D drawing to present the concept to the design team. The team makes their recommendations based on their observations of these drawings. Many changes made to the original design result from misinterpretations of the drawings. Such a process is useful for a range of medical devices, including hip replacements, prostheses, catheters, and even diagnostic equipment.

The 3-D printer deposits plastic droplets in layers to produce a model of the device. Photo courtesy of BPM Technology.

To eliminate possible costly errors, companies are exploring methods for printing 3-D models instead of 2-D drawings. One technology that allows creation of 3-D concept models is known as 3-D object printing. This ink-jet technology, which jets ballistic particles, is more closely related to CAD printing than to other low-cost rapid prototyping systems. It uses a five-axis piezoelectric ink-jet head to deposit thermoplastic at up to 12,000 droplets per second. Each droplet begins to solidify on its way to the previous layer so that it adheres quickly, replicating the 3-D CAD image created by the designer (see photo above).

Designed to be a computer peripheral, 3-D object printers are similar to standard laser printers. No special training is necessary to use them. Users can scale or reorient the image, if needed. The printer produces a scaled object that fits within 10 x 8 x 6 in. BPM Technology (Greenville, SC), which has developed a system, says the printer has an accuracy of 0.015 in. and a surface finish of 250 µin. rms. Many devices cannot be printed actual size. Devices that are larger than the build envelope can be scaled. Objects can be constructed as hatch-fill or as a shell. Shell versions are usually produced for visual design verification only because they are not as durable as solid versions. Such 3-D object printing may represent a natural evolution in communication of design intent because it shifts design communication to the earliest stages of the concept.

With 3-D object printers, designers can create 3-D objects for about the same cost as a 2-D CAD color print or plot. Because 3-D object printing is less expensive than rapid prototyping, engineers and designers can create unlimited iterations of designs for just a few dollars per object. Producing a 3-D object instead of 2-D drawings provides definitive communication of design intent early in the design cycle.

However, 3-D printing is designed to produce a representation for visual verification and should not be used as a substitute for rapid prototyping. Once a company is ready to begin product testing, standard prototyping methods should be used to construct a physical model. As a result, 3-D object printing can provide a viable alternative to building prototypes late in the development process.

---

Copyright ©1997 Medical Device & Diagnostic Industry

Medical Plastics and Biomaterials Magazine
MPB Article Index

Originally published May 1997

TUBING DESIGN

Microbore medical tubing is becoming ever smaller and more complex. Photo: NDH Medical

The use of vacuum as a sizing technique for tubular extrusions has been a common and highly successful practice since the 1950s. In the late 1980s, a modified vacuum sizing principle--generally known as free extrusion with vacuum assist--was developed to enhance the processing of sticky or low-durometer materials such as flexible PVC and urethanes. However, when these same techniques have been tried in the processing of microbore tubing, results have been mediocre at best. One reason is, of course, the size of the tubing: a microbore tube can have a diameter as small as 0.030 in., with a wall as thin as 0.005 in. Given dimensions of this scale, tubing will set up almost immediately upon contact with the cooling water, which makes it very difficult to use vacuum to help with the sizing process.

In this article, we review recent modifications in equipment and processing methods undertaken in order to effectively apply vacuum sizing principles to the production of microbore tubing. Changes such as smaller tank cross sections, specialized low-turbulence features, split plate­type vacuum tooling, and the necessity of increased vacuum levels are among the items discussed. The article also explains how--when properly applied--the use of vacuum can enhance the manufacturing process, offering the potential for on-line cutting with no effect on dimensional stability and even minimizing the need for excessive drawdown to ensure roundness on multilumen tubing. We will begin by analyzing current sizing principles used in the processing of both single- and multilumen tubing.

FREE EXTRUSION

In the free-extrusion process, precision internal air­regulation systems are used to induce a specific air pressure/flow through individual or multiple holes in the die and into the molten tube. The belt or wheel puller then pulls the tube from the die and through the cooling medium at a consistent speed. As the polymer exits the die, the process of heat transfer begins, with air as the initial cooling medium, followed by the water within the cooling tank. The speed of the puller and the temperature of the cooling mediums determine how much drawdown occurs. By increasing or decreasing air pressure, product designers can vary the tube diameter along with individual lumen sizes and wall thicknesses. The cooling mediums remove heat from the tube, while the diameter and/or individual lumen configurations are maintained by the internal air pressure.

Given proper die design and precise control of air pressure, water temperature, and belt-puller speed, free extrusion can provide the means to hold extremely tight diameter tolerances. One drawback of free extrusion is that roundness is difficult to maintain. Mutilumen tubing can vary in roundness because of uneven wall thickness and the size and shape of the individual lumens, which cool at different rates. Many single-lumen tubes vary in roundness because of the weight of the water used in the free-extrusion process.

Figure 1. Typical single-lumen microbore tube (left) and a multilumen microbore tube with the design known as double-D (right).

Figure 2. The image on the left shows how ovality can occur when vacuum or internal air pressure is used on a nonconcentric tube. On the right, pressure on a tube with concentric wall thickness results in an even outward force and good roundness.

Figure 3. Irregularities that can arise during free extrusion of a multilumen double-D tube (left) and a three-lumen microbore tube (right).

PHYSICAL PROPERTIES OF TUBE DESIGN

A brief review of the physical properties of single- and multilumen tubing enables one to better understand issues that can affect a processor's ability to make the tube round, especially with free extrusion (see Figure 1).

Single-Lumen Tubing. When a tube is designed as a single lumen, its wall is generally concentric. Roundness is relatively easy to maintain, since any internal pressure will exert an equal force pushing out on all sides. If any section of the wall is thin, however, the same internal pressure can cause an unequal force on the thin side, causing ovality (see Figure 2). With some microbore tubing featuring wall thicknesses as small as 0.003 to 0.005 in., ovality can result when a product is only a few ten-thousandths of an inch out of concentricity.

Multilumen Tubing. If a tube is multilumen in design, it can potentially have a variety of shapes and sizes for the different lumens within the tube, as well as several different wall thicknesses between or surrounding the lumens. Each of these walls will then go through the heat-transfer process at different rates, which can lead to shrinkage variations and ovality. When internal air is used to size the lumens individually, the air pressure will again push out with equal force on all sides. But if the wall thickness varies or the shape of the lumen is something other than round, an unequal force can once more cause ovality.

Because of these problems associated with multilumen tubing, processors have resorted to excessive die drawdown and to the use of minimal air pressure/flow through each lumen. Given the inability to control or limit the growth of the outside diameter with free extrusion, the amount of trial and error involved in fine-tuning the process of multilumen die design can be tedious and expensive (see Figure 3).

FREE EXTRUSION AND THE COOLING PROCESS

Now that we have defined the two basic types of tubes, let's look at the free-extrusion process and discuss how roundness can be affected by the cooling process itself. When a tube enters the cooling tank, the water used as a heat-transfer cooling medium presents several problems that can lead to ovality or variations in wall concentricity.

Water Turbulence. The initial contact of the tube and the cooling water produces the most dramatic effect on tube diameter, ovality, and wall concentricity. The reason for this is the dramatic temperature differential between the tube and the water. If water is pouring out of the cooling tank--typically in a downward and forward motion--this initial form of water turbulence can cause uneven or variable heat transfer. This drooling of entrance water will generally cause the bottom of the tube to cool faster, with potentially a heavier wall; the top wall will be thinner because it will keep drawing-down longer. With multilumen tubing, this condition can make it almost impossible to fine-tune the individual lumens.

It should be noted that this drooling is not consistent. It generally surges back and forth as much as 0.125 to 0.250 in., depending on the size of the tube and of the clearance hole. Vibration or movement in the tank itself, caused by water pumps or poor mechanical design, can further exaggerate this surging, which in some of the more sensitive materials can result in water marks or even a slight flattening of the tube. Drooling of water can also cause actual movement of the tube--in some cases all the way back to the die. Wall concentricity, especially in very small, thin-walled tubing, will be directly affected by any movement at the die.

Split-Plate Tooling. In horizontal free extrusion of small-diameter tubing, the entrance tooling--which is generally of a split-plate variety--has a clearance hole for the tube to pass through. The only physical contact with the tube is made by the guide rollers, which limit flotation. For this reason, any movement or turbulence can cause the tube to move, allowing the internal air to vary its effect on the tube size and roundness.

The split plate serves as a dam, which enables the tube to enter the tank at a particular level under the water. When properly adjusted, the tube should enter the tank through the clearance hole and never touch the actual plate or tool. The hole should be as small as possible to minimize unwanted drooling; some processors simply drape strips of wet paper or fiber cloth around the hole to reduce the clearance size. Extreme care must be taken not to touch the tube with the paper or cloth so as to avoid blemishes or scratches on the outside.

In certain cases, processors intent on further enhancing free extrusion of their tubing have used an adjustable iris to replace the split plate. They begin with the hole greatly oversized in order to simplify the start-up procedure, then adjust the iris to minimize the clearance hole and consequently the drool.

Both movements of the tube and variations in the tank-temperature gradient can affect tube sizing. It has been observed that when there is no movement of the cooling water--and thus of the tube--sizing tolerances improve by as much as ± 0.001 in.

Heat Exchange. The heat-exchange process itself can potentially be a source of product instability. As heat is pulled from the tube, the surrounding water warms up; this heated water tends to "follow" the tube, like a physical layer. If water turbulence then occurs--whether from water introduction, water removal, rotating product guides, or even vibration--this heated layer is no longer controlled, introducing another variable into the process.

Water Weight. During free extrusion, the actual weight of the water can influence the roundness of the tube. This is especially true with respect to small-diameter, thin-wall tubing. In some cases, the level of the water is kept as little as 0.062 in. above the tube to help maintain roundness. However, the use of this practice to limit flattening can lead to concentricity problems because of water-temperature differentials above and below the tube.

Effects of In-Line Cutting and Coiling. Another occasional complication of free extrusion is the temporary blockage of internally regulated air caused by in-line cutting or flattening in the coiling process. This condition can result in variable flotation and even slight expansion in the outside diameter of the tube as the extrudate first exits the die. Very small annular rings directly related to in-line cutting can sometimes be seen on the tube surface.

VACUUM SIZING

With strict roundness, precise dimensional tolerances, and superior appearance becoming ever more critical in both single- and multilumen medical tubing, many processors are now investigating new sizing methods that feature the application of vacuum. The vacuum sizing technique is growing in popularity and offers several recent enhancements designed to facilitate processing difficult materials and smaller-sized tubes that were previously achieved with free extrusion.

Vacuum Calibration. Vacuum calibration refers to the use of differential pressure (vacuum) to force the extrudate against a calibration/sizing tool, during which operation sufficient heat transfer occurs to maintain a specific outside diameter. Once the tube has exited the sizing tool but is still within the vacuum chamber, vacuum is applied to maintain an equal force on all sides of the extrudate to ensure roundness.

Water Rings. In many cases, water rings have been employed both to help create a seal and to lubricate the extrudate so as to minimize sticking problems during initial contact with the calibration tooling. The use of a water ring is called preskinning, which is the process of cooling a very thin (0.001­0.005-in.) layer on the outside of the extrudate before it enters the vacuum chamber.

Vacuum Calibration Tooling. Many styles of vacuum sizing tools have been developed to process different kinds of polymers, but most can be divided into plate-type or sleeve-type units. Common to all vacuum calibration is the actual contact of the extrudate with the vacuum sizing tool. Generally, the tool is made slightly larger than the product, which shrinks after exiting the tool according to polymer formulation, cooling-water temperature, and line speed. The force of the vacuum must be kept at a sufficient level to maintain contact between the extrudate and the tool. This process has proven highly successful in sizing most tubular extrusions, with the key issue being the assurance of roundness.

Figure 4. Microbore noncontact vacuum-assist system.

FREE EXTRUSION WITH VACUUM ASSIST

Materials such as flexible PVC, urethanes, and certain thermoplastic rubbers have created problems during vacuum sizing because of their tendency to stick to the sizing tool on contact. Consequently, a new form of sizing has been developed with these materials in mind.

The technique--free extrusion with vacuum assist--is increasingly being used to improve tube roundness and attain higher line speeds with many of these difficult-to-process materials. Typically, a short cooling trough known as a preskinning chamber (from 2 to 8 in. long) is situated in front of the entrance to the vacuum chamber (see Figure 4). Ambient air is introduced as needed through the die, as with vacuum sizing. The short trough is designed to work in concert with the drawdown of the polymer to preskin the extrudate and initiate the sizing process before the tube enters the vacuum chamber.

Vacuum-Assist Tooling. The nature of the materials processed via free extrusion with vacuum assist also dictates a change in the vacuum tooling employed. A glass-filled Teflon or similar material is commonly used to make bushings that are located on both ends of the vacuum chamber and that act as a water and vacuum seal. Typically, from 0.015 to 0.070 in. is allowed for tube clearance through these bushings. In this case, the vacuum is no longer being used to push the tube against the tool, but rather to maintain roundness.

A type of a water ring may also be used in the process, keeping the tube centered in the bushing to avoid contact. Most of the actual sizing is completed within the preskinning chamber, with the vacuum section primarily serving to maintain size and roundness during secondary heat-transfer stages. These vacuum sections generally range from 2 to 6 feet in length, with, in some cases, multiple sections for extruding materials--like nylon--that require differential-temperature processing. Generally, very low vacuum levels--from approximately 0.1 to 135.0 in. water gauge--are maintained in the primary section. Levels must be kept within 0.1 in. to ensure consistent sizing and roundness.

Contoured Rollers and Belt Pullers. It is also necessary to use contoured rollers within the vacuum section to offset flotation and potential ovality. These rollers must be made of a low-coefficient-of-friction material, such as PTFE, to minimize sticking.

It is essential that the belt or wheel puller have very precise speed control. The speed of the tube being drawn from the die and through the cooling medium will directly affect both wall thickness and diameter.

Vacuum Tanks. When processors first looked at the potential of vacuum as a tool to assist in the sizing of microbore tubing, it was assumed that the technique would be an immediate success. But early on, problems surfaced when the only vacuum tanks available to the industry were huge in proportion to the tubing to be sized. The need to minimize water turbulence or vibration was scarcely considered in their design; even roller positions were limited to every one or two feet, which is inadequate for microbore tubing. For these reasons, many medical tubing processors did not attempt the process, but continued to struggle with free extrusion to make their products. Today, however, very rigid small vacuum tanks are being built with features specifically for the processing of microbore tubing. These new tanks are quite similar in size to conventional open free-extrusion tanks, but are far more rigid in design. In some cases, heat-exchange systems are built into the walls of the tank itself to minimize water turbulence.

OPTIMIZING VACUUM-ASSIST EQUIPMENT AND TECHNIQUES

For single-lumen microbore tubing, free extrusion with vacuum assist was the first process to be tried other than free extrusion. It was assumed, given the uniform wall thickness of the single-lumen design, that noncontact-type tooling would work on the majority of materials. A problem soon arose, however, with the length of the preskinning chamber before the vacuum section. Even with the shortest available preskinning chamber, prior to entering the vacuum section a tube was preskinned from 0.002 to 0.005 in., which in many cases represented the majority of the wall thickness. For this reason, vacuum had little or no effect on tube size or roundness.

Specialized Vacuum Tooling. New tooling has now been developed specifically for the process of free extrusion with vacuum assist of single-lumen microbore tubing. The preskinning chamber is now built into the vacuum tank and used as a low-turbulence area measuring approximately 1.25 in. long. Split plates are used on both ends of the chamber, with clearance holes gauged according to the tube material and size and the line speed. The main innovation in this specialized preskinning chamber is that it is under vacuum. In this system, the water is independently controlled within the preskinning chamber, and is used along with the vacuum to control any water drool and maintain roundness. Depending on the material, the water temperature within the system can be used to vary the effect of vacuum on the actual sizing of the tube. In this way, the vacuum can be employed primarily to maintain roundness of the single-lumen tube.

In the case of multilumen tubing, where internal air would normally be required to size the lumens individually, the die design used for microbore product is such that almost no internal air is needed. If vacuum is applied with the free-extrusion vacuum-assist process, the tube will be affected in the same way as if excess internal air was applied to the lumens: the vacuum will exert an equal force on all sides of the tube and pull where the walls are thinnest, resulting in ovality.

Figure 5. Hybrid vacuum calibrator (patent pending).

For this reason, special vacuum tooling has been developed for multilumen tubing (see Figure 5). To simplify its use, it was designed similar to a split plate, with top and bottom sections to enhance the string-up procedure. As with vacuum calibration, a water ring is used to apply a thin layer of water--in many cases, only 0.001 to 0.005 in. thick--around the tube as it enters the tool. This serves to create a water seal, a lubricant, and an initial skinning of the outer wall of the tube.

Figure 6. Hybrid vacuum calibration tool with water ring.

Because many multilumen tubes are processed from potentially sticky, flexible materials, it was desirable not to have the tube touch the tool. Depending again on the material, product size, and line speed, two or more additional water rings can be installed within the calibrator to maintain this thin water barrier between the tube and the tool (see Figure 6). The tool itself is made of glass-filled Teflon to minimize sticking in case of contact. As vacuum is applied to the vacuum chamber, the tube is actually pushed out against the water barrier, limiting its expansion. Through adjustments in water temperature, water volume within the vacuum tool, and vacuum level, tube roundness can be maintained even with the most difficult multilumen designs (such as the "double D").

Water Weight. An additional benefit of vacuum systems is that, because the vacuum decreases the weight of the cooling water, the water level above the tube becomes less critical than with free extrusion alone. In particular, water turbulence has much less effect on tube size when the multilumen vacuum calibration tool is used, since most of the sizing is completed within the tool itself.

CONCLUSION

The increasing popularity of specialized vacuum sizing means that many of the unwanted aspects of the use of internal air common to free extrusion can be eliminated, as can most of the negative effects of in-line cutting and coiling. As in the past, die design will continue to be very important in the processing of microbore tubing. However, with the application of vacuum, fine-tuning of the die can be minimized, saving time and money.

As we gain a better understanding of vacuum sizing and the processes are further improved, the production of even smaller microbore tubing, with more lumens and tighter tolerances, will become possible. This will doubtless contribute to the development of increasingly effective miniaturized medical devices in the future.

Robert H. Bessemer is director of development at Conair Jetro/Gatto (Bay City, MI), where his current responsibilities involve developing products for downstream extrusion applications in addition to direct sales and training. His areas of expertise include extrusion technology, vacuum sizing techniques, and tapered-tubing production. David Czarnik is product development manager at Conair Jetro/Gatto. He is also involved in creating downstream extrusion equipment and developing tooling and process improvements. He holds several patents on dies, calibration tooling, and vacuum systems.

---

Copyright ©1997 Medical Plastics and Biomaterials