JARG-18-2-comment
2
pages
English
Documents
Le téléchargement nécessite un accès à la bibliothèque YouScribe Tout savoir sur nos offres
2
pages
English
Documents
Le téléchargement nécessite un accès à la bibliothèque YouScribe Tout savoir sur nos offres
Publié par
Langue
English
P1: FLF/FLK P2: FMNJournal of Assisted Reproduction and Genetics PP014-290112 December 19, 2000 8:21 Style file version Oct. 14, 2000Journal of Assisted Reproduction and Genetics, Vol. 18, No. 2, 2001EDITORIAL1Assessing Outcome2NORBERT GLEICHERIn this issue of this Journal, Deonandan et al. pointout the limitations of life-table analyses in the evalua- See article on page 73tion of infertility treatment outcomes (1). And, whileone can continue to argue about the best statisticalmethodologies applied when assessing IVF outcome Simply stated, there is no way to compare IVF out-versus outcome of other fertility treatments, this pa- comes between programs, unless we use supercom-per should remind us that the average infertile couple puters with discriminant analyses, based on innumer-probably does not care very much what a program’s able patient variables. But is there really a need forspecific success rates for any particular treatment are. such a comparison?In fact, if patients do care about these kind of statis- In my opinion, there isn’t! Infertility patients dotical data, then they should be advised that they are not appear interested in outcome statistics based onbarking up the wrong tree! individual treatment steps. What they really want toknow when seeing a fertility specialist is, “What areWe all know and accept by now that outcome datafor fertility treatments are greatly biased by patient se- my chances to conceive in this program?” “How ...
Voir
Publié par
Langue
English
Journal of Assisted Reproduction and Genetics, Vol. 18, No. 2, 2001
EDITORIAL
Assessing Outcome1
NORBERT GLEICHER2
In this issue of this Journal, Deonandanet al.point out the limitations of life-table analyses in the evalua-tion of infertility treatment outcomes (1). And, while one can continue to argue about the best statistical methodologies applied when assessing IVF outcome versus outcome of other fertility treatments, this pa-per should remind us that the average infertile couple probably does not care very much what a program’s specific success rates for any particular treatment are. In fact, if patients do care about these kind of statis-tical data, then they should be advised that they are barking up the wrong tree! We all know and accept by now that outcome data for fertility treatments are greatly biased by patient se-lection and that women with poorer ovarian reserves have poorer pregnancy chances with IVF. Selection bias is, however, even further aggravated since some centers treat women with excellent ovarian function immediately with IVF, while others enter only those who have failed a number of ovarian stimulation cy-cles (often accompanied by intrauterine insemina-tions). Of course, the former can expect to have higher pregnancy rates than the latter and this kind of selec-tion bias cannot be as easily controlled for as age-related selection biases.
1The opinions expressed in this Editorial are the author’s and do not necessarily represent the opinion of the Editorial Board or e ublisher. 2fretuHrohTneCepthrOelnas57N0rohtduction,manRepro Street, Chicago, Illinois 60610. Fax: (312) 397-8396. e-mail: chrjournal@aol.com.
71
See article on page 73
Simply stated, there is no way to compare IVF out-comes between programs, unless we use supercom-puters with discriminant analyses, based on innumer-able patient variables. But is there really a need for such a comparison? In my opinion, there isn’t! Infertility patients do not appear interested in outcome statistics based on individual treatment steps. What they really want to know when seeing a fertility specialist is, “What are my chances to conceive in this program?” “How long will it take?” and “What will it cost?” Which brings me back to the paper by Deonadau et al.(1). Life-table analyses could represent an en-tirely acceptable methodology to, in fact, compare the performance of infertility programs as long as such statistical methodology not only is utilized to determine treatment-specific outcomes but evaluates program-specific pregnancy rates, independent of the treatment options chosen. The real questions that should be asked in comparing infertility programs should therefore be: What are a specific couple’s chances to conceive atthisfertility center (indepen-dent of the treatments utilized)? What are the chances to deliver a healthy child? And how long will it take, and at what cost? These questions could be easily answered by estab-lishing age-based life tables for individual programs and by comparing those between infertility centers with regard to pregnancy outcome and the respective cost a life birth generates.
1058-0468/01/0200-0071$19.50/0°C2001 Plenum Publishing Corporation
72
Cost is an important factor because of the obvious incentive to increase pregnancy rates at the expense of very costly multiple births (2) but also because there are high cost differences between programs and their respective commitments to cost management. The treatment of infertility lends itself better than most other areas in medicine to quality, outcome, and cost controls. It seems surprising that we have done so little so far.
REFERENCES
EDITORIAL
1. Deonandan R, Campbell MK, Ostbye T, Tyummon I, Robertson J: IVF births and pregnancies: An exploration of two methods of assessment using life-table analysis. J Assist Reprod Genet 2001;18 2. Gleicher N, Oleske D, Tur-Kaspa I, Vidali A, Karande V: Reducing the risk of high-order multiple pregnancy after ovar-ian stimulation with gonadotropins. N Engl J Med 2000;343: 2–7
Journal of Assisted Reproduction and Genetics, Vol. 18, No. 2, 2001
