Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Topical Applied Progesterone Effective for Cyclic Mastalgia
- Samantha McCormick (6 October 2001)
-
Insufficient Evidence of Current Prescribing Patterns
- David Lewis (8 October 2001)
-
A flaw in systematic review of pharmeceutical treatments
- Deborah C McDowell (16 October 2001)
-
The need for individual trial results in reports of systematic reviews
- Douglas G Altman, Christopher Cates (25 October 2001)
-
Re: Topical Applied Progesterone Effective for Cyclic Mastalgia
- Paul Dimmock (26 October 2001)
-
Re: Insufficient Evidence of Current Prescribing Patterns
- Paul Dimmock (26 October 2001)
-
Re: A flaw in systematic review of pharmeceutical treatments
- Paul Dimmock (26 October 2001)
-
Re: The need for individual trial results in reports of systematic reviews
- K M Wyatt, P W Dimmock, P W Jones, P M S O'Brien (8 November 2001)
-
Data on luteal progesterone inadequate
- Joseph B Stanford (8 November 2001)
-
Repeat request for individual trial results in reports of systematic reviews
- Douglas G Altman, Christopher Cates (22 November 2001)
|
|
|||
|
Samantha McCormick, Private Practice LaSalle, Illinois, USA
Send response to journal:
|
As a Nurse-Midwife in the US, I routinely perform gynecologic exams. Through my own personal struggle with one particularly common PMS symptom, cyclic mastaglia, I have discovered topically applied micronized progesterone cream to be highly effective for myself and my clients who suffer from this. It is equally effective for peri-menopausal cyclic mastaglia as well, in my clinical experience. I find it very confusing that topical treatment is apparently effective (at least in my clinical practice) when the literature on progesterone supplementation clearly suggests no positive effect on PMS. I would suggest a study exploring the topical route of supplementation for PMS before abandoning a possible treatment for a common disorder, for which we have so few acceptable treatments. |
|||
|
|
|||
|
David Lewis, General Practitioner Ty Bryn Surgery, Trethomas, Gwent
Send response to journal:
|
Editor, Wyatt et al demonstrate convincingly that there is little evidence to support the use of progesterone and progestogens in premenstrual syndrome.1 They offer less convincing evidence that this remains a commonly prescribed treatment. They cite references from the 1980's that apparently describe prescribing patterns in the 1990s.2,3 Progesterone and progestogens have been widely discredited in the treatment of premenstrual syndrome (and dysfunctional uterine bleeding) for many years. To assume that prescribing habits have not changed since the 1980s is highly subjective and almost certainly wrong. It assumes General Practitioners and gynaecologists take no interest in clinical studies and do not adjust their prescribing accordingly. This article has been publicised widely in the media. It has been portrayed as demonstrating that many doctors are out of date and prescribing inappropriately. In a week when another study in the same issue describes the negative portrayal of doctors in the national press it is ironic that an otherwise impressive systematic review contributes to this trend.4 David Lewis general practitioner Ty Bryn Surgery, Trethomas, Gwent CF83 8GL dailewis@doctors.org.uk 1. Wyatt K, Dimmock P, Jones P, Obhrai M, O'Brien Shaughn. Efficacy of progesterone and progestogens in management of premenstrual syndrome:systematic review. BMJ 2001;323:776-780 (6 October) 2. Stewart A. A rational approach to treating the premenstrual syndrome. Seal, Kent: National Association of Premenstrual Syndrome, 1989 3. Lyon KE, Lyon MA. The Premenstrual syndrome. J. Reprod Med 1984:29:705-11 4. Ali NY, Lo TYS, Auvache VL, White PD. Bad press for doctors:21 year survey of three national papers. BMJ 2001;323:776-780 (6 October) |
|||
|
|
|||
|
Deborah C McDowell Home
Send response to journal:
|
I read with dismay the Reuter's condensation of this article on the systematic review of progesterone administration for PMS. It concludes that there is no benefit in administering Progesterone or progestins in the treatment of PMS. I strongly disagree with this conclusion. and using a systematic review as the only method for reaching it. It has been shown in the 1960's and 70's that the timing and duration of treatment must be synchronized with each woman's menstrual cycle for the treatment to work. If the timing is missed, a cascade of reactions will occur and treatment will be ineffective. These results were published in a book "Once A Month" by Katerina Dalton, MD (England) (publisher Hunter Press). I have had PMS in the past and was subjected to several treatments which provided no therapeutic benefits. Before trying the only treatment that worked, I did a literature search in the US journals to look for a treatment. I found the same results as this systematic analysis of the data found. Also,my personal experience with progestins caused suffering from other side effects. However, a psychologist, Dr. Art Schimelfenig, recommended reading Dr. Dalton's book. This book described that the treatment had to be done in sync with a woman's menstrual cycle. No other published studies in my literature search had bothered to administer the progesterone or progestins in sync with each woman's cycle and for the prescribed portion of the cycle only. My severe headaches were immediately controlled. My general comment: Based on my own personal literature search, I would surmise that most of the studies in this review did not coordinate administration with a woman's menstrual cycle and therefore showed little response overall. If one gives equal weight to a study that shows little positive result with those that show measureable good results, then a review may reach a faulty conclusion. As a synthetic organic chemist, as an RN, and as a cook, I realize that the timing of addition of reactants, steps in a recipe, or administration of medication is extremely important as are the reagents, ingredients, or compound administered. In synthetic chemistry, we don't average results. We find the best "recipe" for using a reagent to produce the desired product. Then we try another compound and compare the best techniques and results for the second one with the best developed for the first. |
|||
|
|
|||
|
Douglas G Altman, (1) professor of statistics in medicine, (2) general practitioner (1) ICRF/NHS Centre for Statistics in Medicine, Institute of Health Sciences, (2) Bushey, Herts, Christopher Cates
Send response to journal:
|
We have commented previously about systematic reviews that do not include the summary data from each study. We wrote: ‘…we can see no valid excuse for not including the data in the report or making them available electronically. The missed opportunity is even clearer in those cases where an extended version of a paper appears on the web page…’.[1] We also observed that the inclusion of such data was recommended in the QUOROM statement for reporting systematic reviews.[2] Wyatt et al have just published a systematic review of trials of progesterone and progestogens as treatments for premenstrual syndrome.[3] Although an extended version appears on the BMJ’s web page, the actual results of each trial are not presented (as summary data for each arm of each trial) and treatment effect estimates and confidence intervals are shown only graphically. By contrast we are told whether each study was or was not statistically significant (table 1), information that is useless and irrelevant. There are several aspects of this review that readers cannot assess without summary data from each study. For example, we would wish 1) to assess the strange heterogeneity P values for Figures 1 and 2 (the quoted P values of 0.999 are implausible given the clear graphical heterogeneity); 2) to examine the consistency of SDs across studies, for example to be satisfied that SD and SE have not been confused at all; 3) to gain some insight into how the cross-over trials were included in the analysis (about which the authors say nothing at all) and whether the crossover and parallel group trials differed in their findings; 4) to seek an explanation for the apparent discrepancy for three trials (references 19, 31 and 32) between the ‘reported results’ in Table 1 and the results shown in Figures 1 and 2; 5) to see whether the correct pooled results from 4 trials of oral micronised progesterone are as shown in the text or as given in Figure 1; 6) to assess the claim that random and fixed models give the same answer in the face of graphical heterogeneity; 7) to confirm or otherwise their remarks about subgroup differences (that look significant in figure 1); 8) to see how the trials are weighted in the analyses; 9) to check the standardised mean difference (SMD) results against the raw figures for mean and SD (as SMD can conceal a great deal). In addition, we note that the authors make no comment about the varied nature of the outcome measures in these trials, nor do they say which outcome was used for those trials that presented more than one. It is hard to believe that all of the scales can be considered equally valid assessments of symptoms. Also, we wonder if they can clarify the meaning in the figure legends of ‘standardised mean difference … for proportion of patients who showed improvement …’. We are puzzled by this terminology as the SMD gives no direct information about proportions of patients improving. For the reasons given we ask Wyatt and colleagues to make available the summary data from all of the trials, preferably as an electronic annex to the extended version of the paper. There can be no valid reason for not doing so. Lastly, we feel that this case strengthens our argument [1] that the BMJ should insist that in future summary information from each study is included in published systematic reviews. Douglas G Altman
Christopher Cates
[1] Altman DG, Cates C. Publishing the results of studies included in systematic reviews. eBMJ http://www.bmj.com/cgi/eletters/323/7305/166#EL1 (20 July 2001) [2] Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup D for the QUOROM Group. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Lancet 1999;354:1896- 900. [3] Wyatt K, Dimmock P, Jones P, Obhrai M, O'Brien S. Efficacy of progesterone and progestogens in management of premenstrual syndrome: systematic review. BMJ 2001;323:776-80. |
|||
|
|
|||
|
Paul Dimmock, Research Fellow Keele University
Send response to journal:
|
Dear Nurse McCormick To our knowledge, there are no randomised controlled trials on the use topical progesterone for cyclic mastalgia so we can only comment that the treatment remains of unknown effectiveness. P. Dimmock, K.M. Wyatt, P.M.S. O'Brien |
|||
|
|
|||
|
Paul Dimmock, Research Fellow Keele University
Send response to journal:
|
Dear Dr. Lewis
We reference up-to-date prescribing pattern data in the longer web version of the paper (http://www.bmj.com/cgi/content/full/323/7316/776). These data will, we hope, be published in the near future and give details of UK prescribing patterns up to 1998. We regret the manner in which the paper has been reported in the media, but we have no control over this. P. Dimmock, K.M. Wyatt, P.M.S. O'Brien |
|||
|
|
|||
|
Paul Dimmock, Research Fellow Keele University
Send response to journal:
|
Dear Ms. McDowell We are unaware of any randomised controlled trials on the use progesterone using Dr. Dalton's methods, so we cannot comment on their efficacy. The trials we looked at gave progesterone in a variety of doses, forms and at different times of the month so we very much doubt that the relatively subtle changes suggested would have any difference on its efficacy P. Dimmock, K.M. Wyatt, P.M.S. O'Brien |
|||
|
|
|||
|
K M Wyatt, Lecturer in Health Servivces Research, Research Fellow, Prof. of Medical Statistics, Prof. of O&G Keele University and Exeter University, P W Dimmock, P W Jones, P M S O'Brien
Send response to journal:
|
Firstly we would like to state how saddened we were by the rapid response of Altman and Cates. Our research is undertaken for the sole purpose of producing evidence -based guidelines for premenstrual syndrome treatments. PMS is a serious and debilitating condition and for which there is a myriad of possible treatments. We are systematically reviewing the benefit, or otherwise, of the most commonly used or prescribed treatments, and where appropriate, submitting these results for publication to try and achieve the widest possible readership. We are not interested in a point scoring exercise over the idiosyncrasies of the various possible statistical techniques used in meta-analysis. All of the statistical methods we have used we have referenced and have used in our previous meta-analyses [1,2]. To suggest that we are confused over what a standard error and what a standard deviation is, is insulting to the Journal’s statistical reviewer and us. It is ironic that one of Prof Altman’s questions is actually answered in the extended web version of our paper (ref 13 relating to the conversion of ranges to standard deviations), when he is extolling the virtues of additional information on the web. We have considerable experience in PMS (clinically as well as though our research) and believe ourselves competent to judge the clinical appropriateness of combining trial data. We have found this personal attack unpleasant and upsetting and have to question the use of unsupported attacks in the Rapid Response forum. It would appear that correspondents are increasingly using Rapid Responses as a method of personal attack and self-publicity rather than as a forum for discussion of current medical issues. [1] Wyatt KM, Dimmock PW, Jones P, O’Brien PMS (1999) Vitamin B6 therapy - a systematic review of its efficacy in premenstrual syndrome BMJ 318: 1375-1381 [2] Dimmock PW, Wyatt KM, Jones P, O'Brien PMS. (2000) Selective serotonin re-uptake inhibitors - a systematic review of their efficacy in premenstrual syndrome. Lancet 356: 1131-1136 |
|||
|
|
|||
|
Joseph B Stanford, Assistant Professor University of Utah
Send response to journal:
|
To the Editor: The meta-analysis of Wyatt et al on progesterone for premenstrual syndrome is comprehensive, but it does not address a critical point. The physiologic rationale for progesterone use is only met when the progesterone is administered during the luteal phase. Studies of progesterone administered during the luteal phase for PMS have not used a marker of ovulation (such as "peak day" of vaginal discharge or basal body temperature shift), but instead have relied on fixed calendar calcuations, which are inaccurate for many women. Recently the BMJ published a study [Wilcox AJ, Dunson D, Baird DD. The timing of the "fertile window" in the menstrual cycle: day specific estimates from a prospective study. Brit Med J 2000; 321:1259-62] demonstrating the highly variable timing of ovulation within the menstrual cycle in a population of normally fertile women. The hypothesis that progesterone administered consistently during the postovulatory phase of the menstrual cycle unfortunately remains untested and is not disproved by the current meta-analysis. |
|||
|
|
|||
|
Douglas G Altman, (1) professor of statistics in medicine, (2) general practitioner (1) ICRF/NHS Centre for Statistics in Medicine, Institute of Health Sciences, (2) Bushey, Herts, Christopher Cates
Send response to journal:
|
Our goal was to illustrate why the summary results of each study should always be included in the report of a systematic review. Among our many reasons were some specific concerns about the review by Wyatt et al and also some points which would interest us in any report of a systematic review. It would indeed have been an insult to suggest that the authors do not understand the difference between a standard error and a standard mean, but our own experience of systematic reviewing is that these are not infrequently misreported in the original papers, and we are aware of cases of this passing unnoticed into reviews. The unpleasant tone of the authors’ response to our correspondence suggests that they feel we are challenging the methods used in the production of their systematic review. Not so. We are simply asking them to show good faith by posting the raw data used to generate the figures in their review, both as a matter of record and also so that we and other readers can set our minds at rest over some apparent inconsistencies within their paper. We note that the authors used a quality score that downgrades trials if they do not provide explanation of the terminology 'randomised' and 'double blind'. The same concerns apply more widely - readers should not have to accept authors' assurances. We regret that none of our specific concerns about this review has been addressed. Curiously, though, the authors have responded to a question (about converting ranges) that we did not ask! We remain disappointed that the review was published without the raw data. We hope that the authors and editors will now work together to post the data. Douglas G Altman
Christopher Cates
|
|||