1. Berghmans et al Abstract on Urge Incontinence (February 2000)
2. Commentary on Berghmans et al on Urge Incontinence
3. Comments on Electrical Stimulation for Stress Incontinence (Sand 1995, Brubaker 1997)
4. Comments on SANS (leg nerve) stimulation for Incontinence.
5. Katherine Moore on E-stim for Men (JWOCN) Jan. 2000
6. Minimalist Therapy for Elderly Incontinence (Weinberger, Goodman, and Carnes, 1999)
Written, complied and edited by John D. Perry, PhD
With HCFA/MCAC public hearings less than a month away, you should be aware of a new article by Berghmans et al that isn't yet indexed on Medline, but IS available in full (see below) on the internet.
In order to get this information out quickly, I resist the temptation to editorialize about "another European academic hatchet job" on incontinence treatments. Suffice it to say that the authors obtain their results, as usual, by carefully selecting the research that was included.
In addition, their review is ENTIRELY based on "external" factors, such as randomization and data reporting. There is absolutely no consideration about the quality of the interventions, as if it did not matter at all.
But I can't help but point out that Burgio's JAMA article received the highest rating on "methodological" form of all the RCTs.
Note especially that these authors consider behavioral treatment to be a form of "physical therapy", not behavioral medicine or nursing. (Note the wording of their conclusion below.)
It will take some time to prepare a critique, because the paper is so badly conceived and cryptically presented. But don't take my word for it.... here's the abstract, and ordering instructions follow.
British Journal of Urology International, Feb 2000 (85:254-263)
Conservative treatment of urge urinary incontinence in women: a systematic review of randomized clinical trials.
L.C.M. BERGHMANS, H.J.M. HENDRIKS, R.A. DE BIE, E.S.C. VAN WAALWIJK VAN DOORN, K. Bo and PH.E.V. VAN KERREBROECK Department of Urology, University Hospital Maastricht, The Netherlands, Dutch National Institute of Health Professions, Amersfoort, The Netherlands, Department of Epidemiology, University of Maastricht, Maastricht, The Netherlands, Norwegian University of Sport & Physical Education, Oslo, Norway Objective: To assess the efficacy of physical therapies for first-line use in the treatment of urge urinary incontinence (UUI) in women, using a systematic review of randomized clinical trials (RCTs).
Materials and methods: A computer-aided and manual search was carried out for RCTs published between 1980 and 1999 investigating the treatment of UUI defined by the keywords `physical therapies', e.g. bladder (re)training (including `behavioural' treat- ment), pelvic floor muscle (PFM) exercises, with or without biofeedback and/or electrical stimulation. The methodological quality of the included trials was assessed using methodological criteria, based on generally accepted principles of interventional research.
Results: Fifteen RCTs were identified; the methodological quality of the studies was moderate, with a median (range) score of 6 (3±8.5) (maximum possible 10).
Eight RCTs were considered of sufficient quality, i.e. an internal validity score of >5.5 points on a scale of 0-10, and were included in a further analysis. Based on levels-of-evidence criteria, there is weak evidence to suggest that bladder (re)training is more effective than no treatment (controls), and that bladder (re)training is better than drug therapy. Stimulation types and parameters in the studies of electrical stimulation were heterogeneous.
There is insufficient evidence that electrical stimulation is more effective than sham electrical simulation. To date there are too few studies to evaluate effects of PFM exercise with or without biofeedback, and of toilet training for women with UUI.
Conclusion: Although almost all studies included reported positive results in favour of physical therapies for the treatment of UUI, more research of high methodological quality is required to evaluate the effects of each method in the range of physical therapies.
Keywords: Lower urinary tract, systematic review, urinary urge incontinence, physical therapy, over- active, biofeedback, electrical stimulation
Complete text is available from Healthgate at http://www.blackwell-synergy.com/Journals for $16.00.
(BJU Int Feb. 2000) [see previous post]
1. Confusing Form and Substance
The 81 studies retrieved by computer and friends yielded 15 RCTs that that met their "inclusion criteria" of (1) RCT design, (2) Urge results reported separately and (3) interventions included bladder (re)training and/or pelvic muscle exercises (with or without biofeedback and/or stim), (4) reliable outcome measures were used, and (5) the report was published in English, German or Dutch. (Sorry, "others")
At first glance these criteria seem reasonable, except perhaps for the unwarranted assumption that "urge" can be easily separated from "stress" incontinence.
This apparently led to the exclusion of McDowell et al 1999, with 85 subjects, of which 74 had urge accidents, which decreased by a statistically significant amount. McDowell apparently did not think the distinction was that "real". Most of her subjects had "mixed" incontinence.
The 15 included studies were rated on their "form"; e.g., how well the study presented their "inclusion criteria", described randomization procedures, and presented outcome data. Oh, yes; they had to have 50 subjects in each group. And less than 10% dropouts.
Where does the magic number "50" come from?
Why not "48" (four dozen) or "64" (8^2)?
And why is 10% the cutoff rate? Wouldn't it help to know WHY people dropped out?
It should be obvious that these criteria are entirely arbitrary and prejudicial. They do, however, allow the researchers to derive conclusions without any indication that they understand what actually happened within the various studies.
In contrast to their rigid "form" analysis, the researchers do not even consider "substance" in their evaluation. If a study claims to use "bladder retraining", then it does -- regardless of whether the particular details of that training conform at any known or accepted or proven standard.
The result of this external analysis of "form" is to end up comparing apples AND oranges to drugs and sugar pills.
In their zeal to find bladder retraining studies, the authors mis-categorize Burgio's JAMA study as a "retraining" project, despite the fact that the study's authors describe it as "biofeedback".
This represents severe category confusion, since biofeedback and bladder retraining are both forms of behavioral treatment. For reasons which fairly ooze from the text, these authors prefer to conceptualize "retraining" as the primary classification.
2. Painting by the numbers
Berghmans et al established a rigid and formal set of criteria for evaluating urge incontinence research. To be labeled as "strong", the evidence had to supported by three high-quality RCTs.
"Weak" evidence consisted for three or more low-quality studies, or three high quality studies that had equivocal results. (Excuse me?)
"Insufficient" evidence was concluded when there were less than three studies in the literature, that met the inclusion criteria, regardless of the outcome.
At first glance this "painting by the numbers" approach to evaluation of scientific research might appear to be "logical" and "fair", but is it really a substitute for careful, thoughtful analysis?
This is a "bean counter's approach" to science; not surprising, since Berghmans is an epidemiologist, not a clinician.
But it does lead to internal contradictions and silly results. For example, Berghmans rates Burgio's 1998 JAMA study as the best of the entire batch of 15 RCTs reviewed -- it scores 8.5, fully 1.5 points higher, on a scale of 1-10, than any other.
Yet they conclude that because there is only ONE such study, albeit the best and, at N=197 also the largest, since no other biofeedback study made the cut, it was ruled "insufficient evidence" (when comparing biofeedback to placebo).
The very same Burgio study, which also compared biofeedback to a drug treatment, is one of three such "behavioral treatment vs. drug" studies, and since there are THREE, the evidence for BT over drugs is "weak", but not "insufficient".
There is an inconsistency here that should be obvious to anyone who UNDERSTANDS the literature. In well-documented clinical trials for the FDA, the drug in question has been shown to be superior to placebos -- it would not be on the market if it wasn't.
So how can we conclude that BT is superior to drugs but not to placebos? It just doesn't make sense, but that's the kind of quandary that Berghmans gets in when "painting by the numbers".
Looking at the numbers from a different perspective, Berghmans et al find "weak (but sufficient) evidence for bladder retraining over no treatment, based on 168 patients in four studies, but "insufficient" evidence for biofeedback over placebo, based on 197 patients in one large study. Does that make sense? If Burgio had written up her results in three batches, that would have made three high-quality studies, leading to a verdict of "strong" evidence in support.
3. Re-defining the professions
Perhaps the most offensive part of the Berghmans et al study is the attempt to classify all behavioral interventions as forms of "physical therapy".
The term (PT) is used five times in the abstract alone. Since one of the authors, Kari Bo, is a world renowned physical therapist, this is not surprising. It is, however, a rather blatant "turf grab" that is sure to offend many other professionals.
Bladder retraining, for example, is NOT a form of Physical Therapy; it is an off-shoot of behavior modification therapies developed by psychologists in the 1950-1970s.
Likewise, pelvic muscle exercise was developed not by PTs, but by a gynecologist working closely with psychologists in California. More recently it has become a nursing intervention as well as an activity of physical therapists, but that does NOT make it a form of physical therapy.
Pelvic muscle exercise never got off the ground (Davies, 1932) until (in 1948) Kegel linked it with what later become known as "biofeedback"; the use of external instruments to monitor internal activity.
Indeed, the only CPT code that alludes to pelvic muscles and incontinence, 90911, is a "physician" code that psychologists and medical doctors use, and nurses under supervision (or law as in NPs). It is NOT a code that PTs can use; they are limited to generic "muscle rehab" codes under the PT sections.
Yet Berghmans and Bo continually attempt to reclassify these treatments as a form of "physical therapy". They are treatments that PTs, as well as nurses, doctors and biofeedback technicians use. But when a PT uses bladder retraining, that doesn't make her a psychologist, and when a nurse uses biofeedback that doesn't make him a PT. We can only imagine that the conceptual confusion was both intentional and polemic.
Knowledgeable readers will recall that the first article in this series, dealing with Stress U.I., concluded that biofeedback was worthless.
Mark Weatherall of New Zealand re-evaluated the five cited studies, and concluded that the evidence actually favored biofeedback, in contradiction to Berghmans' conclusion.
This year's Urge U.I. conclusion is much more cautiously worded. The actual conclusion is "almost all studies included reported positive results in favor of physical therapies [sic] for the treatment of U.U.I...."
They also call for "more research", but can anyone cite a field of study for which that is NOT true?
In fact, ALL of the these studies were positive for behavioral treatments of UUI, with two exceptions.
One e-stim vs. sham stim study found no differences (are we going to "accept" the null hypothesis????).
Even more interesting was that one study compared bladder training with and without prior urodynamic evaluation, and found that urodynamics evaluation made no difference in the outcome.
Excuse me? Since when is urodynamic testing a "treatment"? This has got to be the ultimate "Painting by the numbers" blunder.
Hope you find this helpful.
Why would one study find a significant impact on STRESS urinary incontinence with E-stim, while a later study finds no effect at all?
As the TUMS lady used to say, "GOOD Question!"
Sand, Richardson, Staskin et al, in "Pelvic floor electrical stimulation in the treatment of genuine stress incontinence: A multicenter, placebo- controlled trial" (Am J Obstet Gynecol 1995; 173:72-9) reported a modest but significant improvement in weekly leakage reports (14>10), pad use (6.2>4.1) and pad weight (45>15) after 14 weeks of therapy.
Brubaker, Benson, Bent et al, in "Transvaginal electrical stimulation for female urinary incontinence", writing in the same journal two years later (177:536-540) found NO significant change in stress incontinence rates after 8 weeks of stim therapy.
Brubaker et al DID find a significant change in detrusor hyperactivity (49% cured, i.e., stable on provocation). However, there was no corresponding change in urinary frequency or in urinary accident rates, which certainly raises questions about the predictive validity of urodynamic testing without pelvic muscle evaluation.
At first glance, the substantial difference in number of therapy sessions -- Sand's patients got 75% more therapy -- might account for the difference.
Unfortunately, Sand took only "before" and "after" leakage reports, so we don't know if the extra six weeks made the difference, or if Sand's patients were also improved at 8 weeks.
One might be tempted to look at the stimulation devices themselves for effectiveness clues. Sand used the EMPI Innova system, while Brubaker used the Hollister/Incare Microgyn II.
The Innova device delivered two channels of pulses, at 50 Hz and 12.5 Hz. The Microgyn II has only one channel, at 20 Hz.
The Innova has a 0.3 second pulse duration, compared with 0.1 second for the Microgyn (i.e., Innova gives three times more stim in each cycle).
The Innova program was ramped up during therapy, from 5 on/10 off for 15 minutes (2X daily) progessing to 5/5 for 30 minutes in the final 6 weeks.
The Microgyn used a 2 on/4 off for 15 minutes throughout the 8 weeks of the study.
In other words, the Innova group got 15 minutes of current per session for their final six weeks, while the Microgyn group had already terminated.
The Microgyn group got only 5 minutes of current in each of their sessions.
So we have 3 times longer pulses for 3 times longer sessions, or NINE TIMES as much electrical stimulation for the Innova group.
Is that significant? Since the Innova group got statistically significant reductions in pads and leaks, it would appear so. Of course the reductions were quite small (6.2 pads > 4.1 per week) compared to typical biofeedback comes; a 33% reduction vs. Burgio's typical 85%.
Interestingly, Brubaker collected pad use information in patient diaries, but that data was not published. There is a comment that diaries were incomplete, which is blamed on the "lack of compensation" for research subjects.
A more plausible explanation is that the research staff failed to stress the importance of the diaries. In our own clinical trials, we had 100% compliance with pad reports -- because we insisted on it. And our patients were all paying top dollar in a fee-for-service clinic.
In both Sand and Brubaker there is a lack of detail about the precise patient instructions, either explicit or implicit, with respect to concomitant voluntary pelvic muscle exercises. In other studies, such augmentation has often played a significant role.
One obvious difference here is that in the Sand project, patients were subject to pelvic muscle evaluation before and after electrical stimulation; in Brubaker they were not. Sand used the relatively crude Gynex perineometer, but nevertheless was able to report a statistically significant increase in vaginal muscle strength, from 10.6 mmHg before to 15.2 after. [This is only a modest result compared with biofeedback improvements (e.g., Sussett, 18 > 72 after)].
The fact that voluntary squeezes were measured and recorded in Sand but not in Brubaker certainly invites speculation about the expectations held by the research subjects.
The substantial differences in the amount of stimulation provided by the two devices might lead to suggestions of technological superiority.
However, prior to using the Innova product, Sand and Wheeler reported on "Transvaginal Electrical Stimulation for Incontinence" in a monograph that was privately printed by Hollister in 1989. [I haven't found that this was ever published in a peer-reviewed journal; anyone know?] In that report, using an early prototype of their Microgyn II, they found a significant reduction in leakage on pad test (24 gms before > 5 gms after), as well as a modest improvement in vaginal muscle strength (12 > 16), after only SIX weeks of 2 X 15 minutes therapy. So maybe device differences are NOT at fault.
It is worth noting that these 84 subjects were also subjected to before and after measurements of their voluntary vaginal muscle contractions, using a Contimed II biofeedback device.
The issue of voluntary contractions (Kegels) during e-stim treatment is extremely important, ever since Lamhut, Jackson and Wall (1992) discovered that among elderly and demented patients -- who were presumably unable to do covert contractions -- a substantial minority were actually made WORSE by the same Microgyn II device. They concluded: "it was ineffective in improving urinary incontinence."
In fact, there was a tendency for the treatment to worsen the incontinence. [abstract]
The issue of what is really effective about e-stim needs to be addressed in future research. It would be valuable to compare levels of patient cooperation, ranging from implied Kegels to explicit Kegels to biofeedback-trained Kegels as levels of treatment along with the stimulator device. Until such research is done, we will have a difficult time evaluating the very contradictory research results discussed above.
A distinguished senior member of one of these lists sent the following private reply to my RCT comments on e-stim studies:
My own gut feeling is clinical results are mostly limited by the non-specific nature of all intravaginal stims, i.e. they hit ALL effector mechanisms. By way of contrast, the Medtronic InterStim (an implant) and the UroSurge SANS (needle electrodes) are more "targeted" and appear to be more effective, at least for urge/OAB.
I was interested to hear that the SANS device, developed just down the Freeway in SF, is "more effective" than the non-specific stimulators using patch or vaginal/anal electrodes. So I put "UroSurge" into www.google.com, and instantly found their website, along with press releases, clinical reports, and even a "real player" video of their needle insertion technique.
Sure enough, the SANS was approved by the FDA on February 8, 2000.
BUT IS IT TRUE that the SANS method (e-stim in the leg) is "more effective" than the prior alternatives? Here's what UroSurge claims.
Initially, Dr. Stoller at UCSF reported an 80% "success" rate [i.e., greater than 50% improvement in symptoms], based on 98 patients.
Apparently this data has not been published, however.
In a "collaborative study", unnamed investigators reported a 71% "success" rate [at a much lower rate,
25% or more, improvement in symptoms.] This data is based on only 29 to 42 subjects. Let's look at the details:
Before After Median % change Daytime frequency 13.5 10.0 -29.2% For starters, remember that Burgio (JAMA, 1998)
obtained an (-)81% mean change using only a couple of sessions of biofeedback treatment. By what calculus is 29% "more effective" than 81%?
The more commonly reported stat for these figures is the "mean" % change, but that's only 25.9%; no wonder they use the median!
Even more to the point, after 12 weeks of SANS therapy, these patients are still having 10 voids a day -- three more than the ICS definition of urge incontinent!
Nighttime voiding went from 4.0 to 3.0, a 15% improvement, and still "incontinent" by ICS >2 standards.
24 hour voiding frequency went from 14.9 to 11.3, a paltry 22% median (24.1% mean) change.
Daytime incontinence episodes (leaks) went from 4.4 to 2.5, a 40% median (57% mean) improvement, but still not even close to biofeedback's 80%. [Data from: http://www.urosurge.com/study/sans_clinical_results.html.] So why, we might ask, would anyone want to have a needle inserted into their leg 10 times for 20-30 minutes each week when they can obtain far better results with biofeedback?
"More effective" than previous stim? Here's what some others claim:
Sand, Richardson, et al (1995)
62% improved >50%
Sand & Wheeler (1989)
81-85% improved (at all)
Neotonus website (ICS)
77% improved (at all)
Neotonus (AUA 99)
66% improved (at all)
75% improved > 50% (includes "selected" patients)
60-90% effective (whatever that means)
The latest SANS report:
based on 107 patients, 12 wks:
Litwiller et al (AUA 2000)
ca. 75% improved (at all)
Burgio (JAMA 1998)
90.5% improved > 50%
It is difficult to understand where the idea came from that SANS was "more effective" that other forms of stim; based on the data above, it seems right smack in the middle of the e-stim world. Perhaps EMPI should send them a brochure.
Based on the same physiology.
AND, when we stop to think about it, why would we expect anything different? For years conventional (perineal, perianal, vaginal and anal) electrodes have been used to stimulate the sacral nerves at S2-S3 and cause reflex inhibition of bladder contractions. It works. Not as well as voluntary contractions of the same muscles, but it works moderately well.
The SANS device does the same thing, except it uses the posterior tibial nerve -- a 3 foot longer pathway that arrives at the same spinal junction.
Since both conventional and SANS stimulation use the S2-S3 nerves, why would we expect the results to be any different? And they aren't.
About the only advantage of the SANS system is that it allows urologists to treat pelvic problems without having the patient undress. But since when has that been a problem for urologists?
Many years ago there was a Swedish study that tried to use forehead EMG electrodes to treat pelvic cramps (dysmenorrhea). The explanation was that it was "more convenient" to attach electrodes to the forehead than to the genitals. At the time, prior to the invention of vaginal/anal inserted electrodes, that was, of course true. But now we treat the pelvic muscle problem directly.
Somehow I think there's a parallel here.<g> On the down side, we have to break the skin a dozen times with very fine needles to stick the posterior tibial nerve. We are not told what the risk of infection or complications are for this procedure; it is probably slight, but as the old Vermont saying goes, "It depends whose ox is being gored!"
UroSurge boasts that an advantage of SANS is that it doesn't require "vaginal or rectal probes which can be uncomfortable or painful". Uncomfortable, yes. But never "painful", at least not among urge patients. This is a bit of a cheap shot, but perhaps it is the best they can offer. Still, as an empirical claim, it really should be based on collected data, not salesman's jargon.
Dishonesty in advertising.
The UroSurge website contains a patient handout that describes the alternatives to SANS as "bladder training and pelvic muscle exercises" only -- neither conventional electrical stimulation nor biofeedback are even mentioned. Misleading? You bet.
The Litwiller et al paper that will be presented at AUA by Dr. Govier states, in its Conclusion:
"Percutaneous peripheral afferent nerve stimulation with the PercSANS treatment presents a minimally invasive and potentially therapeutic alternative to other current treatment options (open-surgery or pharmacological intervention, and the associated riskd [sic] for patients diagnosed with documented urgency/frequency due to pelvic floor dysfunction."
[Quoted in its entirety from their website.] Note that members of the AUA will be told that "surgery or drugs" are the only alternative treatments for urge incontinence. That's not likely to go over any more.
On the other hand, sticking thin needles into extremities does LOOK a lot like "electro-acupuncture", which is a an up and coming "alternative" medicine. If Medicare doesn't cover it, probably consumers will pay out of pocket anyway.
Since SANS -- unlike conventional stim -- cannot be done at home, the procedure allows for much higher per patient charges. 10-12 office MD visits to achieve the same results as one office visit and ten weeks of home stim. That's the definition of a profit center!
Unlike biofeedback, which requires intensive labor, the SANS system requires only a few seconds time to insert the needle, adjust the stimulator so the toes don't flex, and leave the patient alone for half an hour. A good technician ought to be able to "treat" 5 to 10 patients simultaneously, in rotation, and bill Medicare for each one.
NOW I understand why SANS is "more effective" than conventional stim. I've got to stop thinking like a clinician, and start thinking like a business person. [But I'm probably too old to learn.] Happy Zapping!
Has Dr. Katherine Moore changed her mind about the efficacy of electrical stimulation? An article in January 2000 issue of the Journal of WOCN suggests that she has.
In the document "Electrical Stimulation Utilization Parameters", which Moore drafted and the Continence Coalition sent to HCFA last fall, Moore was enthusiastic about the role of stim in treating *Stress* Incontinence.
"Electrical stimulation has been reported to be effective for stress incontinence with cure rates ranging from 30-50% and improvement form [sic] 6 to 90% (Yamanishi, 1998)"
This was noteworthy, since no comparable endorsement was even given in the U.P. for Urge Incontinence.
In fact, "Stress incontinence" was listed at the top of the list for "Medical Necessity and Indications of Use", ahead of "Urge incontinence".
And "Post radical prostatectomy (moore 1999) [sic]" was also on the "indications" list recommended to HCFA.
But in the article just published in JWOCN, "Treatment of Urinary Incontinence in Men with Electrical Stimulation: Is Practice Evidence-Based?" Moore takes a much more sanguine view of the stim-stress connection.
After an enthusiastic discussion of the role of stim in treating "detrusor instability (urge incontinence)", Moore dismisses the application for Stress incontinence in two paragraphs. She mentions six studies -- none of them by name -- and notes that only two (actually, Sand 1995 and Yamanishi 1997) found any statistically significant improvement, while four found no significant differences.
And, while Sand et al did find a "statistically significant" difference between stim and a sham device, the "improvement" failed to reach *clinical* significance; i.e., stim patients were still "incontinent" by I.C.S. standards. Yamanishi likewise reported "statistical", but not "clinical" significance.
Sand's project involved 52 women but NO men; while Yamanishi's subjects included only six men among their 50 subjects.
Moore concludes that "...the evidence for electrical stimulation in treatment of stress incontinence is weak."
She goes on to say "The equivocal results indicate that practitioners need to consider carefully before recommending electrical stimulation as the first line treatment for stress urinary incontinence."
(I just LOVE understatement!)
Similar conclusions follow her analysis of the use of electrical stimulation for post-prostate incontinence.
The first such study (in 1976) found only three of 16 subjects improved. The second (1981) found 6 out of 8 improved, but apparently they were also doing Pelvic Muscle Exercises, so it is impossible to give stim all the credit. One unpublished study (a thesis) showed "positive results" for stim when combined with PMEs, but two others -- including Moore's own 1999 report -- found "no differences" to support stim.
Thus Moore concludes, with characteristic understatement:
"...stimulation alone or combined with (PME) does not significantly enhance treatment outcomes in men with incontinence after radical prostatectomy." (p. 29)
Moore remains convinced that "The primary role of electrical stimulation appears to be in the treatment of urge incontinence in both men and women" (p. 29) but even that conclusion is apparently a bit shaky.
Hasan 1996 studied 30 men (and 41 women) and found "(statistically) significant improvement" in voids and pad use (presumably equal for both sexes).
But Bower, Moore et al in a 1998 "replication" that included 19 men, apparently only found urodynamic improvements.
Likewise Brubaker 1997 found big urodynamic improvements -- but no symptom improvement.
Primus 1996 found symptom improvement, and all cases of "sensory urge incontinence" were cured. In contrast, Bower, Moore et al found NO sensory subjects improved.
Urodynamic test improvements are probably relatively meaningless to our clinical patients. I have had many patients complain of urgency, frequency, and urinary accidents; I've never had a patient complain of an abnormal urodynamics test.
The entire discussion of "Evidence-Based Practice" (like so many other such discussions) has been based on finding STATISTICAL significance, NOT clinical significance. In virtually all studies, electrical stimulation has only been shown superior to "no treatment".
Before subjecting patients to stim, the clinician really should address the question "Am I offering my patients the best possible treatment?"
When you compare Burgio's 81% symptom reduction rate for biofeedback with the results of stim studies, the answer is clear.
When pads used "after" treatment are still more than one a day, the patient has NOT been given an effective treatment.
The responsible clinician will spend the time and money needed to get proper training, so as to offer patients the best possible treatment.
Finally, there are some minor errors in the report that you should be aware of when reading it:
1. In Table 4, page 27, the footnote to Primus should be #24, not #36, which is a different Primus report.
2. On page 29, the reference to a study of 63 post prostatectomy men should be #1, Moore and Estey, 1999; not #21.
3. On page 24, Moore says "... a placebo effect of incontinence therapy as high as 39% has been reported. [footnote to Burgio, JAMA, 1998]".
But Burgio, in her JAMA paper, clearly says:
"The control group was intended to control not only for the placebo effect but also for the effects of clinic visits, self-monitoring (bladder diary), and therapist contact."
"Self-monitoring" has a long history and favorable status as an independent therapy in its own right as a form of "Behavior Modification", and should not be confused with "placebo effects".
A recent report underscores the hazards of conducting only "Medline" abstract-level research.
According to the abstract, elderly subjects received "pelvic muscle exercises, bladder retraining, estrogen replacement, biofeedback, functional electrical stimulation and pharmacologic therapy" -- but "subjects considered pelvic muscle exercises, delayed voiding, and caffeine restriction most effective in reducing incontinence severity".
Taken out of context, these results are not only misleading but erroneous. One has to read the whole text to find out that 64% of the 53 patients received only "instruction" in pelvic muscle exercise, and only 42% received instruction in bladder retraining.
In fact, only five patients received any electrical stimulation, and only one (1) received biofeedback.
Data was not presented which would allow comparisons between treatments, presumably an acknowledgement of the small sample size in these groups.
It does not appear, therefore, that "hard" behavioral therapies played much of a role at the University of Wisconsin's Geriatric Continence Center. This in spite of having access to e-stim and biofeedback on an InCare office computerized system, one of the most expensive made.
The paper under discussion is Weinberger MW, Goodman BM, and Carnes M. "Long-Term Efficacy of Nonsurgical Urinary Incontinence Treatment in Elderly Women". Journal of Gerontology, 1999, 54a:3, M117-M121.
In fairness to the authors, the article is not about comparisons between treatments per se, but about geriatric patient's one-year recall and satisfaction with their treatment. But on the other hand, the poor wording of the abstract will undoubtedly lead to this work being misquoted for years to come.
In contrast to their minimalist treatment, patients at the Clinic appear to have received an excellent diagnostic workup (but, then, isn't that always the way?)
A team consisting of a geriatrician, and a urogynecologist, as well as unnamed residents, fellows and nurse practitioners "evaluated and treated...all patients", but it seems likely that not all team members saw all patients -- which raises questions about the uniformity of the therapy.
The Evaluation included history and physical, 72-hour voiding diary, urine culture, Q-tip test, and a "simplified" urodynamics evaluation, consisting of "eye-ball cystometry" and a standing stress test.
The report states that "Biofeedback, using a perineometer connected to a video graphic display (InCare, Libertyville, IL) was used to teach women [note plural] with weak pelvic muscle strength how to perform pelvic muscle exercises." (p. M118)
But in Table 3 we learn that only one woman (singular) received biofeedback therapy (and she was so unimpressed with it that a year later she forgot the fact.)
Likewise, it is not clear how women "with weak pelvic muscle strength" were defined; presumably by the usually unreliable "finger" test. And it isn't clear why 64% of the patients were given verbal instruction in pelvic muscle exercise if they did NOT have "weak pelvic muscle strength". Perhaps they meant to say that only women (or a woman) with VERY weak muscles got biofeedback, the rest got verbal instruction.
Interestingly, only "refractory incontinence" was later treated with electrical stimulation. It was not a first-line treatment.
The interventions were not long (median visits only 3) nor terribly successful. Only 12 patients ("22%", but actually 23%) became "dry", and only another 25 (47%) "improved" at all. Fully 8 ("16%", but actually 15%) were "unchanged" by the treatment.
It is important to note, however, that these figures represent "patient reports", not objective (pad) or even diary reports.
Apparently another 8 patients (15%) failed to HAVE a final clinic visit, since the reported numbers don't add up to 53 (nor 100%). We don't know if they no longer had the initial problem, or no longer trusted the treatment.
The ineffectiveness of the Clinic's "verbal instruction and bladder retraining" treatment is actually acknowledged by the authors, who note that, based on *subjective opinions* of the patients, "improvement was significant (p=.005; see Figure 1)"
But in the very next sentence we read:
"There was, however, no statistically significant reduction in the mean number of daily incontinence episodes."
Excuse me? What was the purpose of their clinic visits? To gather baseline data for a future study of patient recall?
An examination of "Figure 1 (and 2)" shows what really happened over "12 to 25 months"; the patients had the same number of leaks, but they all rated their problem as "less severe" than initially; i.e., they learned to live with it. Now that's creative therapeutics!
It is worth noting that the patient population had somewhat "moderate" incontinence; 2.7 leaks per day (+/- 3) and 2.0 (+/- 2.3) pads per day.
I'm going to remember this study every time I hear the band play "On, Wisconsin..."
In the "Discussion" section, the authors make two major errors in discussing the work of Baigis-Smith, Smith, Rose & Newman, 1989.
They state that Baigis-Smith found that those "with the mildest incontinence at presentation experienced the greatest improvement following treatment".
Actually, what Baigis-Smith et al reported was this:
Those people with the mildest incontinence at the baseline visit had the fewest or no urinary accidents after the end..." (p.232)
That is a very different result. Obviously those with the fewest symptoms before treatment would have fewer, or even NO accidents, at the end.
Conversely, those with a large number of accidents at the start, even with the 78% overall improvement rate, might still have a measurable number of accidents at the end.
(E.g., a 78% reduction in accidents starting at 2 per day would be only 0.4 per day after, whereas 20 per day with a 78% reduction would be 4 leaks a day at the end.)
On the second point, these authors stated that Baigis-Smith et al provided "treatment with pelvic muscle exercises and habit training" (just like they did).
In fact, Baigis-Smith used EMG biofeedback in the office with 100% of their patients, using the "Clinical Perineometer from Farrall Instruments. In what they called a "focused treatment" model, patients were allowed to continue with the biofeedback so long as they were showing demonstrable improvement.
In addition, although not documented in the text, patients were allowed to use an EMG home trainer for two or more weeks; the project leased and used eight "Personal Perineometers" also made by Farrall.
All patients received full interventions including formal biofeedback at all visits; the mean duration of treatment was 3 visits (with two weeks between), plus formal EMG evaluations at 3, 6, and 12 month follow up, for an average of six clinic visits (and six biofeedback sessions.)
In contrast, the Wisconsin project averaged only three visits and no biofeedback.
The New Jersey biofeedback protocol resulted in a 78% symptom reduction, with 76% of patients improved > 60%, and 98% improved > 30%.
The Wisconsin verbal protocol resulted in NO measurable reduction in leaks, only a change in patient complaints.
It should be further noted that the Baigis-Smith project -- unlike many less successful programs, actually saw their symptom improvement rates go UP after termination of "focused treatment". Cured, for instance, increased from 37% to 59% a year later.
Whether this can be attributed to more effective teaching of at-home skills, or the reinforcing properties of biofeedback sessions at 3, 6 and 12 months is unclear -- but it is still impressive.
The project is all the more impressive since the express purpose was NOT to determine how effective biofeedback could be, but how many senior citizens could receive significant personal benefit from a time- and cost-limited intervention.
Finally, the nurse that directed the NJ clinic, Diane Smith, had received extensive personal training in the lab of Kathy Burgio, and several consultations and visits over two years from this author (JDP).
Wisconsin did not have either of these advantages.
Copyright © 2000 by Incontinet.com