As a consultant, I am often asked to improve quality at a company and sometimes the work product takes the form of writing and/or revising SOPs and training people to follow those SOPs.  In many cases, no prior SOP exists within the company and we are asked to create the SOP document from scratch. 

Now anyone who has written one of these SOP documents knows, this is not exactly an easy task.  The trick with good SOPs is to write them at a high enough level so they are ‘Standards’ and not detailed ‘Work Instructions’ which will be more prone to lots of deviations and potential non-compliant activities when different people perform the procedure in slightly different settings.

Anyway, when I was writing SOPs some years ago, I picked up a book of template SOPs to make the job a little easier.  The title of the book (a 3-ring binder actually) which is still on my bookshelf is: A Handbook of SOPs for Good Clinical Practice by Donald E Maynard and B Joyce Baird (Interpharm Press, Inc, Buffalo Grove, IL, 1996).  These generic SOPs were available in electronic format on CD-ROM and were really quite helpful and easy to use when laying the infrastructure for a new business conducting clinical trials for pharmaceutical research.  Unfortunately, as I was asked to write this column about SOPs, I was disappointed to learn this Handbook was out of print.

As I am a really lucky person, I was not surprised to find today a book review discussing a new book with SOP templates published just this year and titled “Clinical Research Quality System Manual” by Nancy J. Stark, 2009, 355 pages, Clinical Device Group.  And, how convenient, this book was written for the device sector!  Unfortunately, the price tag is $3,000.00.  Wow.

So, I thought I would share some thoughts from our common and collective e-bookshelf - the internet.

The reason I had not looked at my Handbook of SOPs book for so many years is that I have started using the web rather than the book templates to draft new SOPs for my clients.  I have discovered the web holds hundreds of SOPs which can be used to create the SOP you need for your company.  The web has templates and forms and literature and training materials to help your people follow the SOPs as well.  For example, here is just one website with an interesting style and content: http://www.jefferson.edu/ohr/SOPTOC.cfm.  The information found at this website could be used as examples to help you develop an SOP manual to cover your initial SOP needs for Good Clinical Practice.  You can start with the SOP on SOPs and work all the way through the documents to see which are the most helpful to you.

Before I can recommend buying the new $3000.00 book of SOP documents, I think I would advise you to look to the web for the SOP you need to develop, and perhaps hire a contractor to help you get the job done quickly and efficiently.  SOP writing doesn’t get any easier, but the longer you wait to get those SOPs written, the more likely you will find yourself on the wrong side of a 483 detailing how you need to have SOPs in place to control the quality of your clinical trial activities.

Dr. Frestedt is president and CEO for Frestedt, Incorporated and has over 30 years of scientific, clinical and regulatory experience in the pharmaceutical and medical device industries. Dr. Frestedt has a Ph.D. in Pathobiology from the University of Minnesota Medical School and she currently serves as Chair of the Ethics Committee for the Regulatory Affairs Professionals Society.

Back to Top

FDA's authority to regulate medical devices is really all about labeling.  The simple go/no-go decision for approval (or clearance in a 510(k) situation) gets most of the attention, but labeling can have a substantial impact on the acceptance and marketability of a new device.

Among other things, the device label tells for whom the device is intended, the purpose for which it is to be used and what it can do.  "What it can do" often comes in the form of labeling claims.  Unlike, say, indications for use, for which FDA has a definition in 21 CFR 814, the term "labeling claim" is not specifically defined.  Nevertheless, to paraphrase a famous U.S. Supreme Court ruling, FDA knows one when it sees one.

From a statistical standpoint the question is simple: when can labeling claims be made?

This issue arises especially when a study has evaluated several endpoints, which almost all do.  In a pivotal study, one or more of the endpoints will be primary and the rest will be secondary (some may even be "tertiary" or "ancillary," depending on your taste in study design vocabulary).

Although it is possible to get a device approved with a failed primary endpoint, the most common situation for approval is that the primary endpoint was met -- by which we mean that the statistical test came back with a p-value below the magical cutoff of 0.05.  Labeling claims can then be made for that endpoint.

Secondary endpoints are a different matter, and FDA's attitude toward labeling claims for secondaries has crystallized over the past several years.  The problem with secondary endpoints is that there are often a lot of them, and it's well understood that the practice of performing statistical tests on many different endpoints tends to produce false positives: findings of effect which look "real" but actually arise by chance.

This means that simply showing a p-value less than 0.05 on a secondary endpoint does not mean a labeling claim can be made; stronger statistical controls are required.

These are called multiplicity adjustments, and there is no sweeping FDA guidance that covers all situations.  However, several draft guidance documents mention the issue of statistical multiplicity.  For instance, FDA's recent draft guidance for atrial fibrillation states that "[i]f you intend to present comparisons between groups for a secondary effectiveness endpoint in your labeling, your protocol should include a prespecified hypothesis and an adjustment for multiplicity, as appropriate."

Last year's draft guidance for urinary incontinence is even more detailed, listing examples of specific techniques for multiplicity adjustment.  There are a wealth of such things, but they basically fall into two categories: "hierarchical" or "gatekeeping" methods and methods that adjust p-values.

Hierarchical methods mean that the study analysis plan defines in advance the order in which secondary endpoints are to be tested statistically, and that labeling claims can be only be made for secondary endpoints as long as the sequence of p-values stays below 0.05.  So if you have four secondary endpoints you hope to label and the prespecified sequence of p-values goes 0.015, 0.043, 0.353 and then 0.011, only the first two qualify for consideration for labeling.

Notice what happened there: even though the fourth and final p-value was below 0.05 as well, you don't get it in the labeling.  That's the price of the hierarchical approach -- the sequence matters.

If that seems arbitrary and unscientific (and sometimes it does to me, too) then the alternative is p-value adjustment, which either inflates the p-values or reduces the cutoff below 0.05 -- they're the same mathematically.  There's no sequence; instead, the statistical bar is raised for each endpoint.

The easiest such method is the Bonferroni approach, which says that the cutoff is 0.05 divided by the number of tests.  In the example above, that makes the cutoff 0.0125 instead of 0.05, meaning only the fourth endpoint (the one with a p-value of 0.011) would qualify.

Other p-value adjustment methods have been developed that are "smarter" than Bonferroni adjustment.  I won't bore you with the mathematical details, but one such approach, called Hochberg's method, would allow both the 0.011 and 0.015 p-values to qualify.  Again, though, you still have a p-value below 0.05 (the one at 0.043) that doesn't count for labeling.  That is, there's a price to be paid here too.

Which approach is the best?  There's no simple answer, since a statistical strategy for multiplicity adjustment has to take into account the clinical importance of the endpoints and the perceived likelihood of meeting each endpoint.

To reiterate, though, these decisions need to be made in advance to make the statistical analysis valid in FDA's eyes.  As always, when in doubt contact your friendly neighborhood statistician for advice.

Scott Brown is a Principal Statistical Consultant with The Integra Group, providing statistical expertise to the medical device and biotechnology industry. Contact him at 763-951-7461.

Back to Top

ARKRAY, Inc. Receives FDA Clearance for New GLUCOCARD(R) Vital(TM) Blood Glucose Monitoring System

ARKRAY, Inc., the world's 5th largest manufacturer of diabetes self-monitoring systems, today announced that it has received 510(k) clearance from the US Food and Drug Administration for its new GLUCOCARD Vital blood glucose monitoring system.

The new meter and strip platform, which ARKRAY will make available for distribution in early December, is expected to be a significant complement to the existing GLUCOCARD line of products, and will be ideal for the home medical equipment (HME), durable medical equipment (DME), mail order and managed care markets.

"We're very excited about the latest addition to our GLUCOCARD brand," said Jonathan Chapman, President of ARKRAY, USA. "The Vital meets all of ARKRAY's rigorous design and performance standards. And we can offer it at a price point that will make it an ideal solution for our customers to use in competitive bidding."

The new GLUCOCARD Vital meter and strip platform is highly accurate, auto-coded, requires only 0.5 microliter of blood, holds 250 dowloadable tests in its memory and has a glucose oxydase (GO) strip chemistry that is not affected by interferences that cause testing errors in other common test strips.

The Vital, along with all GLUCOCARD brands, are backed by ARKRAY's unique YouChoose((TM)) wellness program, a comprehensive education and support system that includes printed tools, videos, recipes and an integrated website designed to help patients manage their diabetes

BridgePoint Medical receives European approval for devices

BridgePoint Medical Inc. announced Monday that it has received approval to sell and use its new catheter devices in Europe. The Plymouth-based medical-technology company earned the European CE Mark, the European equivalent of the U.S. Food and Drug Administration’s (FDA) seal of approval, to market its CrossBoss CTO Crossing Catheter and the Stingray CTO Re-Entry System.

A recent medical trial showed the products resulted in a greater success rate in allowing doctors to access heavily diseased arteries than the standard guidewires that are now in use. The BridgePoint devices are under clinical investigation in the United States and have yet to receive FDA approval.

Entellus secures $30 million in capital

Entellus Medical Inc. will use $30 million in new financing to expand sales and marketing of its chronic sinusitis treatment, the company announced Tuesday.

In June, Maple Grove-based Entellus announced a new treatment called FinESS Sinus Treatment, which it said is less invasive than the surgery that sinusitis patients sometimes undergo to reduce inflammation of the sinus cavity and allow normal mucus drainage to resume. The Maple Grove ear, nose and throat medical company will use the new funding to hire more sales and customer support staff and market FinESS nationwide in 2010.

Essex Woodlands Health Ventures of Palo Alto, Calif., a new investor in the company, led the financing, while existing investors Split Rock Partners of Eden Prairie and San Francisco-based SV Life Sciences supported it. Essex Woodlands managing director Guido Neels will join the Entellus board of directors.

Vascular Solutions gets FDA approval for catheter product

Vascular Solutions Inc. announced Monday that it has received clearance from the U.S. Food & Drug Administration to launch its GuideLiner catheter product in the United States.

The GuideLiner is used to help guide catheters into place during especially tricky coronary interventions — a procedure in which a balloon-like device is inserted into the body and used to remove blockage in the arteries leading from the heart.

The Minneapolis-based medical device company (Nasdaq: VASC) had already received approval for sale of its product in Europe, and plans to begin selling the product to U.S. heath care providers starting in November.

The Medical Device Alliance includes press releases from local medical device companies every month. Please feel free to send all PR to kirbys@mfrall.com for potential future inclusion.

Back to Top