April 17, 2000
Dear Friends:
INCLUSIVENESS:
This message is being sent to all addresses in Connie Conrad's most recent email to HCFA/MCAC panel speakers. For convenience, "hcfa.gov" addresses have been moved to the "CC:" line, but everyone has been included. Some of the addresses are not the actual speakers, but groups that arranged speakers. They may wish to supply the relevant individual's email addresses, or take responsibility for forwarding mail to them.
EXCLUSIVENESS:
Ideal: Initially I thought of excluding the HCFA names, but my intention is to create a free and open dialog in the incontinence community, and you don't get openness by being closed. (Practical: Besides, surely someone would attempt to curry favor by forwarding anything interesting, anyway.)
REMOVAL:
If you do not wish to be included in this effort, just "REPLY TO ALL" and tell people to remove your name from the "INCONTINENCE AT HCFA" list.
REPRESENTATION:
We will assume that, absent formal declarations to the contrary, everyone speaks only for themselves, and any organizational or institutional affiliations are listed for identification purposes only.
WHAT DO WE HAVE TO TALK ABOUT?
I came away from Baltimore with several strong impressions about issues that were raised. I wonder if others share my concerns, or want to raise new issues? Here are some of mine.
1. The Relevance of "Evidence-based Medicine". One issue is whether EBM can even be used to analyze a behavioral intervention like biofeedback. As an educational activity, biofeedback depends on creating changes WITHIN the person of the research subject. In traditional ANOVA, such changes are ascribed to "error variance" and discarded.
Detrol had to prove to the FDA that it produced changes apart from and in spite of personal characteristics of the patients and clinicians. But biofeedback only works if it can produce changes in the personal characteristics of the patients. They must develop better neurological control over their muscles, AND they must remember to invoke these new skills in daily life.
Of course behavioral interventions CAN be analyzed as if they were pharmaceutical interventions, in the same way that, for example, category data "can" be analyzed as if it met the basic assumptions for parametrical statistical testing --- but graduates of Statistics 101 know that is not valid.
2. On "Treatment Effects" a.k.a. "Intermediate Results". Many weeks ago I sent John Whyte an extensive email explaining how, since 1986, it has become the standard in biofeedback research to look for -- and demand -- evidence of "treatment effects" BEFORE looking at outcome effects [e.g., Symptom Improvement]. If you claim, for instance, that using biofeedback to reduce muscle tension will lead to headache relief, researchers are now required to show that they did, in fact, reduce the muscle tension, before reporting any improvement in the headaches.
This is an important scientific check, because if there was no change in muscle tension but the headaches improved, it establishes that some other mechanism, such as placebo or attention, was responsible, instead of the hypothesized intervention. That is important information.
Lefevre demonstrates an incredible ignorance of biofeedback research when he writes, of pelvic muscle evaluations, "These physiologic measures are intermediate outcomes that may or may not be directly related to health outcomes of interest to this assessment, and are not analyzed in this review of evidence."
"May or May Not" is simply FALSE. The theory behind Kegel's biofeedback training was that his method increased the effective strength of the pelvic muscles, and that, in turn, led to an improvement in incontinence symptoms. If there had a symptom improvement without a change in pelvic muscle strength, that would shoot a big hole in the Kegel theory.
[On closer examination, there was apparently a good reason why Lefevre avoided discussing the topic, since he included urodynamic testing as well as pelvic muscle testing in that section, and it is well known that urodynamic tests often do not correlate with symptom changes. This is very embarrassing to the urological theory, and would have required extensive discussion.]
3. On Winning the Argument in the Premises. HCFA and/or BCBS defined the biofeedback problem as focusing on the "addition" of biofeedback to "PME alone". As I tried to point out in my 8 minutes of fame, this represents a very naive "20-year-old" perspective on incontinence. If we adapt a more learned 50-year historical perspective, going back to Arnold Kegel's original work, the question should have been: "What is the effect of SUBTRACTING biofeedback from Kegel's program?" That is, after all, what really happened historically when Kegel died and his biofeedback device was no longer available.
4. On Categorical Confusion. One of the problems inherent in the TEC/TAR analysis is the arbitrary bifurcation of incontinence research into two supposedly exclusive and exhaustive categories, "PME alone" and "PME+BFB". The primary problem is that the title given to the first category does NOT match the category contents, so when reported to the public the results will surely be misunderstood and misinterpreted. (There was even some confusion among the speakers!)
The category "PME Alone" does not include just PME Alone, but in fact included one or more of the following therapeutic interventions: (1) written handouts, (2) oral descriptions of exercise, (3) external coaching at monthly, weekly, or thrice weekly clinic sessions, (4) internal pelvic muscle examinations (initially, monthly, weekly, or twice weekly), (5) with or without an unspecified amount of verbal (relayed) biofeedback from (6) clinicians of various professional orientations and with varying levels of experience.
So, if HCFA decides NOT to cover "biofeedback" with electronic instruments, how many sessions of, say, internal vaginal examination by experienced nurses and/or PTs will they cover under the category "PME Alone"? How many times a week?
I would NOT want to put my biofeedback instruments into a head-to-head (if you'll pardon the misleading metaphor) efficacy competition with Carol Brink, RN. She MIGHT win. (I hope she has her fingers insured like a pianist.)
On the other hand, how much would 10 1-hour sessions with Carol Brink cost Medicare? How many patients can Carol Brink see in a week? In how many states?
And there are very few, if any, clinicians as experienced as Carol Brink. What about the rest? Are they statistically the same as biofeedback? <g>
[One of the primary advantages of objective measurement with EMG biofeedback instruments is that it provides even inexperienced clinicians with accurate, objective measures of pelvic muscle strength. It doesn't take 25 years to become an expert.]
5. Congratulations to Cyndy Feldt, PT. Cyndy, representing the APTA, was the only person I heard who openly expressed anger at HCFA's dramatic change in the substance of the Baltimore hearings with only two weeks notice. I expect that she was speaking for all presenters who had spent months of preparation time on one set of assumptions, only to have the ground rules totally revised just two weeks before the hearing.
Likewise, the posting of the "real questions" a few hours before the hearing caught almost everyone unprepared. I made a few PowerPoint word changes at 6 am Wednesday morning when I downloaded the official questions from the HCFA site. Not everyone was able to do that.
6. The agenda was as poorly organized as the overall process. In retrospect, speakers had to sign up by March 22 to be eligible -- but the TEC report, which was the only discussable topic -- wasn't posted until March 28th. So everyone had to commit (and buy those damned non-refundable airline tickets) before they knew exactly what they would be allowed to discuss when they arrived.
The same sequence reversal took place on April 12th. In the morning we had 120 minutes of public commentary --- followed, after lunch, by the actual presentation of the TEC report by Frank Lefevre. Now I'm not a JD, but I have seen a few Perry Masons in my day, and I thought it was always the government's duty to present its case, followed by the defense's argument. It isn't sufficient to say that we all had two weeks to read the TEC report on the internet; there was no way of knowing ahead of time which of the 30-odd pages (each) would form the basis for Frank's presentation. Some of us made good guesses, but I hope in the future a more logical sequence will be used.
This is not to suggest, in the slightest, that HCFA staff didn't do a good job; under the circumstances, they did an EXCELLENT job, handling a nearly impossible situation. And old-timers told me that this hearing was infinitely better than previous ones. That's probably true.
These are just a few thoughts upon returning from Baltimore. Hopefully others will join in and raise issues that really need to be discussed. If anyone feels intimidated, I will gladly post anonymously any messages sent to me in private with that request. I realize not everyone is able to say what they really think due to employment circumstances, etc.
FREE PUBLICATION OFFER
I have already posted my own PowerPoint presentations on my "IncontiNet" website, at
http://www.incontinet.com/AAPBBIO.ppt for biofeedback, and
http://www.incontinet.com/AAPBSTIM.ppt for the Electrical Stimulation presentation. (If you don't have PowerPoint and can't view it in your browser, you can download a FREE PowerPoint Viewer from the Microsoft.com website.)
I would be happy to post, without charge, ANY other individual's or organization's HCFA presentation on the same website. Just send it as an email attachment to "webmaster@incontinet.com".
The legal transcript is supposed to be available in a week, and it is predicted that HCFA will post that to their website in another week, but it is very unlikely that the text of your slides will be included in the transcript, and certainly NOT in color! In the spirit of fairness, this free offer is also extended to HCFA and BCBS for their April 12-13 presentations.
Hope this helps,
John
John D. Perry, PhD, MDiv, BCIA, FAACS
AAPB representative