Podiatry Arena members do not see these ads

Here's the latest from Foot and Ankle showing foot orthoses are effective in the treatment of plantar fasciitis. Patients showed better compliance and fewer side effects when using foot orthoses than when using night splints. This study correlates well with my 21 years of clinical experience of treating thousands of patients with plantar fasciitis.

__________________
Sincerely,

Kevin

**************************************************
Kevin A. Kirby, DPM
Adjunct Associate Professor
Department of Applied Biomechanics
California School of Podiatric Medicine at Samuel Merritt College

e-mail: kevinakirby@comcast.net

Private Practice:
107 Scripps Drive, Suite 200
Sacramento, CA 95825 USA
My location

Voice: (916) 925-8111 Fax: (916) 925-8136
**************************************************

Related threads:
Plantar Fasciitis Discussions

__________________
Are you addicted to Podiatry Arena?
Have you added your location to the map? (Map FAQ's)
Earn CPD Points

OMG ... possibly underpowered; didn't use 'intention to treat analysis'; etc. I agree the results do also match my clinical expereince, but it difficult to accept how the results in the study support their conclusion given these issues.

Like the thread on ESWT therapy, we should not blindly accept research that matches our clinical experience and do everything to rip to shreads research that does not. .... they all should be held up to the same level of scrutiny .... its not possible to have it both ways.

__________________
Craig Payne
Department of Podiatry
La Trobe University
Melbourne, Australia
http://www.latrobe.edu.au/podiatry
__________________________________________________ ___________________________________
God put me on this earth to accomplish a certain number of things - right now I am so far behind, I will never die.
The views expressed above are those of the author and not that of La Trobe University
This is where I am, where are you?

Craig, please tell me why I can not accept this new research as evidence of the benefit of foot orthoses. Are you saying that what these researchers did was in error or that the researchers were biased or possibly had a hidden agenda? I feel that we should not blindly dismiss research just because it doesn't have the statistical power or methodological prowess that other research studies have and then rip to shreds the research that does not live up to "research perfection". As a clinician, I tend to believe what I have seen with my own eyes and heard with my own ears in successfully treating thousands of patients with plantar fasciitis with custom foot orthoses over the past two decades...that custom foot orthoses are well tolerated and highly effective treatment modalities at treating these patients.

Are you then saying that custom foot orthoses are not effective and well tolerated in treating plantar fasciitis as the study published in Foot and Ankle International has concluded?? Or are you saying that this research doesn't meet the standards that you have established for foot orthosis research? If this study doesn't meet the standards that you have established as being acceptable for foot orthosis research, then please provide all of us here on Podiatry Arena a list of the research studies that have been done worldwide on foot orthoses in the last three decades that do meet your research standards and you feel are acceptable.

__________________
Sincerely,

Kevin

**************************************************
Kevin A. Kirby, DPM
Adjunct Associate Professor
Department of Applied Biomechanics
California School of Podiatric Medicine at Samuel Merritt College

e-mail: kevinakirby@comcast.net

Private Practice:
107 Scripps Drive, Suite 200
Sacramento, CA 95825 USA
My location

Voice: (916) 925-8111 Fax: (916) 925-8136
**************************************************

What I am saying is that because they did not use the 'gold standards' of analysis for RCT's -- ie no a priori power analysis and did not use 'intention to treat analysis', then there is significant potential that they could have overestimated the the effect of the treatment ..... that says to me that they could have got it wrong and journal editors have an ethical responsibility to ensure that these things do not happen. Hence the CONSORT statement on the reporting of clinical trials that researchers and, more importantly, journal editors should be adhering to.

In the ESWT thread, its obvious that the studies that support the use of EWST do not come up to the CONSORT standard and those that support it do come up to the standard. Those with vested financial interest in EWST go to great lengths to discredit those that do not support its use and yet are not prepared to hold the studies that support its use to teh same standard.

We should not blindly accept any research that matches our clinical experience nor should we blindly dismiss research that does not match our clinical expereince --- the CONSORT standard is there to try and prevent those kinds of bias creeping in.

__________________
Craig Payne
Department of Podiatry
La Trobe University
Melbourne, Australia
http://www.latrobe.edu.au/podiatry
__________________________________________________ ___________________________________
God put me on this earth to accomplish a certain number of things - right now I am so far behind, I will never die.
The views expressed above are those of the author and not that of La Trobe University
This is where I am, where are you?

Craig,

Can you explain this component?

As regards power analysis - it is a subject that has intrigued me. Ask a statistician how many subjects you should include in a study, invariably the answer is 'as many as possible'. Can you point to a source that can provide guidance with this? Whenever I have brushed against this issue the numbers required to produce a good robust study are oftentimes quite forbidding.

Is it possible to achieve all of the gold standards? - double blind, RCT etc. great for drugs, not so good for mechanical intervention. As a surgeon, how can I design a blind study and not know which treatment I am providing?

If we do accept study limitations and yield to the strength of our own observation, assuming we can accept it is possible to be sufficiently objective, and cast aside other possible confounding issues e.g. those patients who feel under pressure to tell us what they think we need/want to know/hear because they ‘like’ their clinician (? a version of Stockholm syndrome?) is it not necessary then also to accept the evidence from the patients of Brian Rothbart, Ed Glaser etc. who also will tell us of subjectively observed 'good' outcomes over many years?

With most of my professional time dedicated to the provision of socialized health care, I have been impressed how treatment outcomes can vary if the patient/client is fee paying instead, as opposed to those who do not i.e. National Health Service, where treatments are free. Not of course the o nly decisive factor, there is, it seems, variation in the patients’ perception of the value of treatment when it is not otherwise influenced / complicated with payment. There is a contrast with the fee-paying patient who, one could speculate, has a greater stake in finding success in their treatment and their ‘choice’ of health care provider. I am not sure if anyone has explored this socio-economic health issue but it is a variable and an object of my observation.

Thanks

Dieter

I'm no expert but this is my answer to your question, hope I understood it correctly and I'm not teaching you to suck eggs.

Statistics are educated guesstimates of the real data patterns

The optimum number of subjects for a study is determined by the researcher and what level of confidence they require from their analysis.

The reader should then be able to judge for themselves how accurately or how reliably the statistics characterise the parameters of the population.

If a study is done of the entire population of a group of interest then strictly speaking this is a census and the results are not statistical (ie estimates) but are parameters (real representitive data)

So ideally a pilot study or studies are done with a small number of subjects and the characteristics of composition of the data and the number of people in the group population will give an indication of the optimum number of subjects for the full research project.

The data characteristics of the research project will then be analysed so that a confidence level can be found the confidence level is usuaully expressed something like 95% within 2 x standard deviations of the mean.
This does not neccessarily mean that 95% of the population characteristics of interest will fall within 2 x SD of the data mean.
It does mean that you would expect that, if you repeated the experiment several times again, the statistics should show that 95% of subjects in each study have characteristics of interest that fall with 2 x SD of the mean.
It is down to the reader to decide if they think the confidence level is adequate and reliably quoted for their purposes.

If you required a higher confidence level say 99% within 1 x SD of Mean then usually your sample population or number of subjects in the study group would have to be much higher.

There are many books on statistical analysis I particularly like Statistics for Dummies, which is about my level.

Cheers Dave

Is this a good explanation Craig?

Dave, you seem to be talking about alpha levels here, statistical power is related to beta, more specifically 1-beta. In simplified other words, what is the probability that the conclusion is wrong? Sample size is not a random number dreamed up by the researcher as you seem to suggest, sample size is calculated either prospectively or retrospecitvely. Much work done on this by Cohen (whole book don't have reference to hand). Method of calculation is in part dependent on the analyses to be performed. Rule of thumb- minimum 80% power required. Spooner's rule of thumb- if an RCT doesn't disclose power, question why. All trials should disclose power.

__________________
Science is the antidote to the poison of enthusiasm and superstition

My location

Perhaps this reference will be of use:
DJ Pratt: A critical review of the literature on foot orthoses
J Am Podiatr Med Assoc 2000 90: 339-341

A few years old now, but the conclusions drawn should take account of all but the last 6 years , and are quite important to this debate.

Please don't shoot the reference supplier

__________________
Science is the antidote to the poison of enthusiasm and superstition

My location

Intention to treat analysis is how you deal with compliance and dropouts. Using it is the gold standard as advocated by the CONSORT statement. See this BMJ article on it: http://bmj.bmjjournals.com/cgi/conte...l/319/7211/670

__________________
Craig Payne
Department of Podiatry
La Trobe University
Melbourne, Australia
http://www.latrobe.edu.au/podiatry
__________________________________________________ ___________________________________
God put me on this earth to accomplish a certain number of things - right now I am so far behind, I will never die.
The views expressed above are those of the author and not that of La Trobe University
This is where I am, where are you?

Simon

I have not come across the alpha beta terms before, is beta the same as P value and Z or T test which test the probability that the null hypothesis is rejected or not.

I was not saying that the researcher randomly selects the sample number. The number the researcher chooses is subjective though. By looking at pilot study or past research the researcher makes a decision upon the numbers required for their study. Since the statistical output is not a true characterisation of the population parameters but merely an estimate of those parameters this decision based on the probable level of error they will accept when considered against logistics and cost.

Bigger samples don’t always mean better samples since reliability also depends on the quality of the method. Also much larger samples don’t necessarily give correspondingly much lower margins of error.

Here’s an example

N=500 P= 0.44 Confidence level = 95%

= 1.96 x square root of (0.44 x 0.56) / 500 = 4.4% error margin

N= 2000

= 1.96 x square root of (0.44 x 0.56) / 2000 = 2.2% error margin

So there is four times the work and cost for half the margin of error. Is the decreased error worth the extra time and expense?

In the same way Confidence Levels also exhibit this negative exponential of diminishing returns and therefore the reader of the research must then decide if the Confidence level and margin of error is acceptable for their purposes.

If not, then using the original research statistics as a baseline, they to can calculate the extra number of subject they will require for a follow up experiment that gives lower margin of error and higher confidence levels.

I’m asking for your opinion here Simon since statistics is quite new to me so I might be a bit mixed up.


Cheers Dave

Dave sent you a private mail cause I don't have time to convert word docs containing two tiny tables to allow me to upload here without going over size limit.

__________________
Science is the antidote to the poison of enthusiasm and superstition

My location

Simon:

My point in asking Craig this question is that there are so few "high quality" research articles on foot orthoses available, that we have no choice but to use all the lower quality research articles on foot orthoses along with our theoretical understanding of the mechanical function of foot orthoses and also along with our clinical experience to provide the evidence for their therapeutic effectiveness.

I strongly believe that just because orthosis research is published that has sufficient numbers of subjects to confidently draw conclusions from the research results, that it may not be research that is valuable for the practicing podiatrist. For example, if the prescription foot orthoses used in the study are not designed properly or are, for example, compared to "sham orthoses" or "placebo orthoses" that are in fact custom molded prefab orthoses, then that research may be less helpful to the practicing podiatrist than that research which has less statistical power, but has been carefully designed and executed to answer a question that has important clinical relevance to the practicing podiatrist.

__________________
Sincerely,

Kevin

**************************************************
Kevin A. Kirby, DPM
Adjunct Associate Professor
Department of Applied Biomechanics
California School of Podiatric Medicine at Samuel Merritt College

e-mail: kevinakirby@comcast.net

Private Practice:
107 Scripps Drive, Suite 200
Sacramento, CA 95825 USA
My location

Voice: (916) 925-8111 Fax: (916) 925-8136
**************************************************

Agreed, sample size and statistical power is only one aspect of research design. However, the problem of lack of statistical power should not be under-estimated. I could turn your argument around and say I strongly believe that under-powered research is less helpful to the practicing podiatrist because it is misleading and potentially harmful to patients, has the potential to make them mistrust their own experiences etc etc. But the real problems are the strong beliefs themselves. These cloud objectivity. I'm sure Brian Rothbart strongly believes in what he is doing, as does Ed Glaser, as do you and I.

As Craig has said, what is required is critical engagement with the literature, not blind acceptance of it. We do have a choice: I choose critical reading and view reported research with a sceptical eye, I don't believe what the authors are saying just because it fits with my mindset. The problem is that most podiatrists don't have the research skills to do this. Let me try to explain some of what we have been talking about in the hope that we can engage more podiatrists in critical review:

Lets take a research study:

The null hypothesis (Ho) = that there is no significant difference between treatment X and treatment Y

Lets say the "true" situation is that the Ho is true, i.e. in reallity there really is no difference in the outcomes obtained using either treatment.

Despite this the researchers could draw 2 conclusions from their study data:

1. Ho is accepted- this is the correct conclusion: probability= 1-alpha
2. Ho is rejected- this is the wrong conclusion, a false positive or type 1 error: probability= alpha


Now lets say the true situation is that the Ho is false, i.e. in reallity treatment X is superior to treatment Y

Again, the researchers can draw 2 possible conclusions based on their observations:

1. Ho is accepted- this is the wrong conclusion, a false negative or type 2 error: probability= beta
2. Ho is rejected- this is the correct conclusion: probability= 1-beta (statistical power)

The aim of the researcher is to keep alpha and beta as close to zero as possible. But:

If alpha is reduced without changing the sample size (n), beta increases
If beta is reduced without changing the sample size (n), alpha increases
Only by increasing n can both beta and alpha be reduced.

Most podiatrists are familiar with the concept of alpha, although they may not be familiar with the language here. Normally we set a level for alpha at 0.05, 1-alpha = 0.95 or 95% (seem familiar?). But most are unaware of beta, commonly we look for a statistical power (1-beta) of >0.8 or 80% so beta <0.2

When you read a paper, one of the first things you probably want to know is the P (alpha) value right? My question then becomes, why trust one probability related to the data and ignore the other? The first thing I want to know is what is the probability that the authors came to the wrong (my glass being half empty as it is) conclusion? To know this we need to look at both alpha and beta.

Hope this helps.

__________________
Science is the antidote to the poison of enthusiasm and superstition

My location

Simon:

Good discussion and thanks for the review of statistics.

Here is what I believe it all comes down to. Once one has become trained as a podiatrist, has seen thousands of feet with many years of practice, has tried many different treatment protocols, and has studied during that time to become an expert in the function of the human foot and lower extremity, how much should that podiatrist change what they do because of a published research study that is contrary to their practice habits?

In my case, I do things differently now than when I was first graduated from podiatry school 23 years ago in nearly all aspects of my practice. I change what I do because of trial and error experimentation, because of my continuing academic study that allows changes in theoretical understanding of a subject, because of communicating with other physicians and patients, because of attending seminars, because of participating in online podiatry forums, because of observing the responses of hundreds to thousands of patients to certain treatment, and because of reading scientific research.

Therefore, reading scientific research critically does not form the sole basis of why I practice the way I do; it is only one of many of them. In fact, scientific research is something that I am very critical of, especially if it is not designed well and especially if it is contrary to the other many sources that help determine my decision-making in my practice. I do change what I do because of research, if I feel the research was done well, it makes sense and it is supported by other research that was also done well. However, no matter how well one research study was done, if it is contrary to my knowledge and experiences and is contrary to existing research studies, I have a very hard time believing it, and will certainly not change what I do, until other research studies comes out to support it.

Finally you said "I choose critical reading and view reported research with a sceptical eye, I don't believe what the authors are saying just because it fits with my mindset." I would agree with the above. However, I think the more important question for you and other practicing podiatrists is this: If you were to critically read and view a research study, and that study has excellent research and statistical methods, but it tells you something that is very contrary to your professional education, very contrary to what you have seen hundreds and/or thousands of times, and is very contrary to methods that you have personally seen helped hundreds and/or thousands of patients, would you then change what you do just because of that one excellent statistical and methodological research study?

For me, research is just one avenue of informing my practice, it is not the end all. Personally, I would much rather have a competent, experienced, well-trained podiatrist that has never read research or has never taken statistics in his/her life be my podiatrist and make my foot orthoses or do my foot surgery than having a research-oreinted podiatrist that has great book learning and research skills but has poor motor skills and poor clinical judgement to do my foot orthoses or foot surgery. Of course, that does not mean that the formerly described podiatrist could not improve their techniques and knowledge by starting to become educated more in research methods and statistics. However, some of the most gifted physicians that I have known wouldn't know an alpha from a beta from a p value in research statistics and methodology, and this did not seem to affect their ability to heal people thousands of people over many years very effectively from their painful pathologies. Research has a proper place in the scheme of clinical medicine, but it does not, by itself, heal people.

__________________
Sincerely,

Kevin

**************************************************
Kevin A. Kirby, DPM
Adjunct Associate Professor
Department of Applied Biomechanics
California School of Podiatric Medicine at Samuel Merritt College

e-mail: kevinakirby@comcast.net

Private Practice:
107 Scripps Drive, Suite 200
Sacramento, CA 95825 USA
My location

Voice: (916) 925-8111 Fax: (916) 925-8136
**************************************************

I would probably experiment with the new information and review the outcomes I obtained through its application. As I know you would too prof. Kirby.:)

For me too, but we were talking about research.

As a friend and mentor to me over the years, I am not surprised to hear you say that, although I suspect many who are not so fortunate to know you as I do may be somewhat surprised. Here's what I think as of today:

"you can't teach old dogs new tricks"

Research is, in the main, the realm of the academic. If you are working in academia and teaching undergrads you should be actively researching. The fruits of this research are then "absorbed" into the curriculum and the teaching of new students. Hence progression occurs predominantly at the undergraduate level. By the time of graduation, the "way of doing things" is usually set and little changes there after. This is clearly not applicable to all, but in my view, the vast majority. Hence the graduates are only as good as their teachers. Those students with good teachers (actively researching and publishing), think yourselves lucky, those with bad teachers (turn up late, go home early, not interested in the subject, doing the job but nothing more) switch podiatry schools.

__________________
Science is the antidote to the poison of enthusiasm and superstition

My location

Good reply, Simon. Unfortunately, with my 50th birthday only five short months away, it is becoming more of a struggle every year to resist the temptation and just become a "comfortable old dog" in my views and clinical techniques. That is one of the reasons why I enjoy keeping myself as sharp as possible with continued reading, writing and having discussions with the likes of you, Simon, and the others here on Podiatry Arena who are "brave enough" to enter into an academic discussion with me on a subject.

The bottom line is we need more and better research on all the therapeutic modalities which we use on our patients to allow us to be better informed as to the efficacy and mechanism of action of those modalities for our patients. The better the research, the more it helps us and our patients. However, as you know, better theory and practical application of better theory is my bag. I hope to be able to have time in the near future to contribute more to these academic pursuits in the hopes that my work will better enable podiatrists to be able to think more like engineers when they are treating mechanically-based pathologies of the foot and lower extremity.

__________________
Sincerely,

Kevin

**************************************************
Kevin A. Kirby, DPM
Adjunct Associate Professor
Department of Applied Biomechanics
California School of Podiatric Medicine at Samuel Merritt College

e-mail: kevinakirby@comcast.net

Private Practice:
107 Scripps Drive, Suite 200
Sacramento, CA 95825 USA
My location

Voice: (916) 925-8111 Fax: (916) 925-8136
**************************************************

This is great! Your findings correlate with what I have seen in my Pedorthic Practice. I have found that the orthotics made by Birkenstock called birkobalance have worked sometimes better than custom made. The reasons are as follows:

1. The custom functional device is not always tollerated by the client and is not modifiable as needed post fitting.

2. Many times the top cover used presents too much force or the ground reaction forces are not spread out as much as they should be to help the patient feel more comfort in its use.

3. The birkobalance orthotic is a 3/4 length device made of thermoplastic cork. This allows one to heat up the problem areas of pressure and using a tool depress that spot only. The rest of the orthotic is uneffected and so is the support over the rest of the device. The device should feel cradled to the client before the device is out the door. If in a few days the pressure points are still too much for the patient then some slight modifications can be done. The most common areas I have seen are in the navicular,talus, cuneiform area of the foot. The next area is the longitudinal top line of the high flanged device which needs to be just thinned out some to accommodate most pronated flat foot. Also the area on the transvers plane on the plantar surface interface of the calcaneocuboid area. The one area that does not need much attention is the distal end of the device which has a slight met pad built into it. I have also been able to glue pieces to the orthotic and give the orthotic more inversion if it is indicated.

4. Minor grinding is also easily done to lessen the feel of the met pad that is placed in the device.The device is fit to the longitudinal arch and not to the shoe size as many other places will fit the other devices most commonly scripted. This allows true support of the arch only and is not placing too much pressure to the met heads or is not too short to the longitudinal arch.

5. I have also found that the type of shoe that is to be used must also be supportive. A mild rocker sole is indicated or even a larger amount of toe spring is called for. Flats , slip ons or other soft thin soled shoes will not do the patient the most good in the treatment of this condition. How tight the laces are on a lace up shoe or the velcro closure on a velcro shoe will also affect the way the foot is held in the shoe to the orthotic.

6. The stretches and icing and the NSAI modalities that are scripted all help. Pt and other interventions assist.

I have many patients wanting the stylish shoes for looks. Pathology will dictate to the modality scripted by the doctor as to what the treatment plan compliance should be. Even after the pain has subsided and no more pain is prevelant, the support system should still be maintained as often as possible to maintain the support to the plantar fascia.

I am sure there are other ways of working with this problem. Combining modalities like night splints is also indicated. I have had many patients with the orthotic superfeet who have not had much success with them. I have found that many times it is the person despensiing the device who is not able to properly fit the device to the client. The birkobalance orthotic has been replacing the superfeet orthotics by many of the clients who compalin of foot pain still when they are trying hard to use the superfeet device. I have had clients keep the birkobalance orthotics in their shoes when it is first placed in their shoes for thier use. WHen clients come back and get multiple units because they are pleased with the outcome, I know I am onto something good. I thought I would pass it along.

Rick Woodland
C.Ped 0810

408 617-0066 w.
408 617-9110 fax
rickwccfc@yahoo.com