Gaps in the evidence for fertility treatmentðan analysis of the Cochrane Menstrual Disorders and Subfertility Group database

Transcription

1 Human Reproduction Vol.18, 5 pp. 947±954, DOI: /humrep/deg260 Gaps in the evidence for fertility treatmentðan analysis of the Cochrane Menstrual Disorders and Subfertility Group database N.P.Johnson 1,2,3,4, M.Proctor 1,2 and C.M.Farquhar 1,2,3 1 Cochrane Menstrual Disorders & Subfertility Group Editorial ase, 2 University of uckland, and 3 Fertility Plus, Obstetrics & Gynaecology Department, National Women's Hospital, uckland, New Zealand 4 To whom correspondence should be addressed at: The University Department of Obstetrics & Gynaecology, National Women's Hospital, Claude Road, Epsom, uckland, New Zealand. n.johnson@auckland.ac.nz CKGROUND: The randomized controlled trial is considered the best approach to assess the effectiveness of treatments. The aim was to summarize the available evidence and determine gaps in the evidence for clinical decision making in subfertility. METHODS: search of the Cochrane Library for Menstrual Disorders and Subfertility Group reviews was undertaken and, where the reviews were related to subfertility, the authors' conclusions were appraised and correlated with the results and meta-analysis sections of the reviews. Each review was then categorized as to what extent it had answered the clinical question posed by the reviewers. RESULTS: Of 38 subfertility reviews currently or previously published on the Cochrane Library from the Menstrual Disorders and Subfertility Group, 12 reviews concluded that there was evidence of effectiveness of the interventions studied. There was insuf cient evidence of effectiveness in 26 reviews, from which the authors of 23 reviews called for further research. tabulated summary of the review conclusions is presented. CONCLUSION: Cochrane subfertility reviews have eliminated some gaps in the evidence and highlighted others. Future clinical trial design should focus on adequate power and reporting the major outcome of live-births per couple as well as adverse events. Key words: Cochrane systematic review/evidence-based/gaps/infertility/subfertility Introduction The randomized controlled trial (RCT) is considered the best approach for ascertaining the effect of a health care intervention. In 1979, rchie Cochrane criticized the medical profession ``that we have not organised a critical summary, by specialty or subspecialty, updated periodically, of all relevant randomised controlled trials'' (Cochrane, 1979). The challenge to create a comprehensive register of all was initially taken up in perinatal medicine. This ultimately led to the publication of a `Guide to Effective Care in Pregnancy and Childbirth' and the electronic publication `The Oxford Database of Perinatal Trials'. Following the launch of the Cochrane Collaboration in 1992, these publications became the prototype for the present day Cochrane Library, which publishes more than 1500 systematic reviews covering most areas of health care represented by 50 collaborative review groups. The goal of the Cochrane Collaboration is to improve decision making in health care by ``preparing, maintaining and promoting the accessibility of systematic reviews of the effects of health care interventions''. In addition to providing up to date, unbiased evidence on health care, systematic reviews can also identify `gaps' where there is insuf cient or no evidence. Gaps in the evidence for clinical decision making may arise for a number of reasonsðwhere no research has been performed; where trials of insuf cient power to determine an important treatment effect have been undertaken, but no meta-analysis has been performed; and where there is insuf cient evidence from trials even after a thorough systematic review and metaanalysis. The rst comprehensive review of robust evidence for fertility treatments (Vandekerckhove et al., 1993) was published a decade ago. This report aims to summarize the ndings of the Menstrual Disorders and Subfertility Group (MDSG) reviews in the eld of subfertility, to assess qualitatively how successful MDSG membership has been in the `coverage' of this eld and to assess where gaps in the evidence remain. Methods Each of the 38 reviews addressing subfertility from the 70 MDSG reviews currently or previously published on the Cochrane Library (Issue 1, 2003, Oxford Update Software) were sub-grouped as unexplained infertility, tubal infertility, anovulatory infertility, other female infertility, male infertility or assisted reproductive technology (RT). The following information was collected from each review: ã European Society of Human Reproduction and Embryology 947

2 N.P.Johnson, M.Proctor and C.M.Farquhar 22 Table I. Cochrane reviews of unexplained infertility Review title (uthors) included romocriptine for unexplained subfertility in women (Hughes et al., 2003b) Clomiphene citrate for unexplained subfertility in women (Hughes et al., 2003a) Danazol for unexplained infertility (Hughes et al., 2003c) Oral ovulation induction agents versus injectable ovulation induction agents (gonadotrophins) for couples with unexplained subfertility or male-factor subfertility (thaullah et al., 2003) Tubal ushing for infertility (Johnson et al., 2003a) Kinin-enhancing drugs for unexplained infertility in men (Vandekerckhove et al., 2003a) w Single versus double intrauterine insemination in stimulated cycles for subfertile couples (Cantineau et al., 2003) IVF for unexplained infertility (Pandian et al., 2003) the number of trials available for meta-analysis, the total number of trial available for meta-analysis and whether there was an answer to the primary clinical questionðinto which category from () to (C), below, the review fell. () Where there is evidence of effectiveness or harm from a metaanalysis of trial data. The term `relative effectiveness' was used when two interventions were compared and the term `effectiveness' was used when the treatment was compared with either placebo or no treatment. () Where there is insuf cient evidence of effectiveness and the review authors have called for further research. (C) Where there is insuf cient evidence of effectiveness and the review authors have not called for further research. Those reviews withdrawn from publication by virtue of not having been updated for at least 2 years have been highlighted in Tables I±VI by suf x w. Results The ndings of the MDSG subfertility reviews are summarized in Tables I±VI. Overall, from the 38 MDSG subfertility No signi cant difference versus placebo Signi cant increase in odds of pregnancy versus no treatment or placebo 2 68 No signi cant difference versus placebo No signi cant difference in odds of live birth 8 (2) (6) (224) (1241) Signi cant increase in the odds of pregnancy for tubal ushing with oil-based media versus no treatment. No signi cant difference in the odds of pregnancy but a signi cant increase in the odds of live birth for tubal ushing with oil-based media versus water-based media No signi cant difference in pregnancy rates No signi cant difference in odds of pregnancy 4 No signi cant difference in odds of: (35) Pregnancy for IVF versus expectant management (113) Live birth for IVF versus IUI (118) Live birth for IVF versus COS/IUI (146) Pregnancy for IVF versus GIFT reviews, there was suf cient evidence of effectiveness of the interventions from primary trial data or from meta-analysis of trial data in 12 reviews and insuf cient evidence of effectiveness in 26 reviews (from which the authors of 23 reviews called for further research). Two of the three reviews, which did not nd evidence of effectiveness and where the reviewers did not feel that further research was warranted, have now been withdrawn from publication on the Cochrane Library because they had not been updated for >2 years. Unexplained infertility From eight reviews of unexplained infertility, two metaanalyses provided evidence to answer the clinical question and there was insuf cient evidence in six reviews, four of which called for further research (Table I). In unexplained infertility, there was evidence of effectiveness of clomiphene citrate (CC) for women (Hughes et al., 2003a) and of tubal ushing with oil-soluble contrast media, although there was insuf cient evidence for the relative ef cacy of oil-soluble versus water- C C

3 23 The Cochrane menstrual disorders and subfertility group database Table II. Cochrane reviews of tubal infertility Review title (uthors) included Techniques for pelvic surgery in subfertility (Watson et al., 2003a) arrier agents for preventing adhesions after surgery for subfertility (Farquhar et al., 2003a) Liquid and uid agents for preventing adhesions after surgery for subfertility (Watson et al., 2003a) soluble media (Johnson et al., 2003a). There was insuf cient evidence of the effectiveness of bromocriptine (Hughes et al., 2003b) and danazol (Hughes et al., 2003c) for women and kinin-enhancing drugs for men (Vandekerckhove et al., 2003a). There was also insuf cient evidence for oral versus injectable ovulation induction agents for controlled ovarian stimulation (thaullah et al., 2003), for double versus single intrauterine insemination (IUI) (Cantineau et al., 2003) and for the relative effectiveness of IVF versus expectant management, IUI with or without ovarian stimulation and gamete intrafallopian transfer (GIFT) (Pandian et al., 2003). Tubal infertility In the ve reviews of tubal infertility, one meta-analysis provided evidence of effectiveness and there was insuf cient evidence in four reviews (Table II). In tubal infertility, there was evidence of effectiveness of laparoscopic salpingectomy for hydrosalpinges prior to IVF (Johnson et al., 2003b). lthough there was evidence of effectiveness of Interceed, 7 No signi cant difference in pregnancy rates for: (1) (72) Loupes spectacles versus operating microscope (1) (18) Microsurgery versus macrosurgery with prosthesis (2) (197) Tubal surgery with prosthesis versus no prosthesis (2) (135) Carbon dioxide laser adhesiolysis versus standard adhesiolysis (1) (24) No signi cant difference in postoperative adhesion formation with brin sealant versus standard technique for sterilization reversal 15 lthough there is a signi cant reduction in pelvic adhesion formation, there is no evidence of a signi cant difference in odds of pregnancy for: (5) (175) Interceed versus no treatment at laparoscopy (both reformation and de novo) (6) (554) Interceed versus no treatment at laparotomy (1) (42) Gore-Tex versus no treatment (1) (127) Sepra lm versus no treatment for myomectomy (2) (61) Gore-Tex versus Interceed 13 No evidence of signi cant bene t in terms of pregnancy or subsequent adhesion formation from: (6) (633) Steroids: systemic, IP or HT (4) (310) 32% Dextran 70 i.p (1) (92) Heparin solution i.p. (1) (93) Systemic antihistamines (1) (126) Noxytioline i.p. Postoperative procedures for improving 5 No evidence of effectiveness for: fertility following pelvic reproductive (3) (494) Postoperative hydrotubation surgery (Johnson and Watson, 2003) (2) (114) Second-look laparoscopy with adhesiolysis Surgical treatment for tubal disease in women due to undergo IVF (Johnson et al., 2003b) Laparoscopic salpingectomy prior to IVF signi cantly increases the odds of pregnancy and live birth Gore-tex and Sepra lm adhesion barriers and of superiority of Gore-tex over Interceed in terms of adhesion reduction, there was no evidence these interventions were associated with an increased chance of pregnancy (Farquhar et al., 2003a). There was insuf cient evidence for various microsurgical techniques (Watson et al., 2003a), various intra-operative liquid and uid agents for adhesion prevention following subfertility surgery (Watson et al., 2003b) and post-operative hydrotubation or laparoscopic adhesiolysis following pelvic reproductive surgery (Johnson and Watson, 2003). novulatory infertility From ve reviews of anovulatory infertility, one provided evidence of effectiveness and there was insuf cient evidence in four reviews (Table III). In anovulatory infertility, there was evidence of effectiveness of CC for oligo-amenorrhoea (Hughes et al., 2003d). In polycystic ovary syndrome (PCOS), there was insuf cient evidence of: relative effectiveness of FSH versus hmg (Nugent et al., 2003); effectiveness of 949

4 N.P.Johnson, M.Proctor and C.M.Farquhar 24 Table III. Cochrane reviews of anovulatory infertility Review title (uthors) included CC for ovulation induction in women with oligo-amenorrhoea (Hughes et al., 2003d) Gonadotrophin therapy for ovulation induction in subfertility associated with PCOS (Nugent et al., 2003) GnRH analogue as an adjunct to gonadotrophins for clomiphene-resistant PCOS (Hughes et al., 2003e) Pulsatile LH releasing hormone for ovulation induction in subfertility associated with PCOS (ayram et al., 2003a) Recombinant FSH versus urinary FSH or recombinant FSH for ovulation induction in subfertility associated with PCOS (ayram et al., 2003b) Laparoscopic drilling by diathermy or laser for ovulation induction in anovulatory PCOS (Farquhar et al., 2003b) addition of GnRH analogue to gonadotrophins (Hughes et al., 2003e); effectiveness of pulsatile LH releasing hormone (LHrH) (ayram et al., 2003a); relative effectiveness of recombinant FSH (rfsh) versus urinary FSH (ufsh) nor of a preferential dose regime for rfsh (ayram et al., 2003b). In clomiphene-resistant PCOS, there was no evidence of a difference in treatment pregnancy outcomes between laparoscopic ovarian drilling and gonadotrophin injections (Farquhar et al., 2003b). Other female causes of infertility The one review in this category (Table IV), provided evidence of effectiveness of laparoscopic surgical treatment of endometriosis for increasing the chance of pregnancy (Jacobson et al., 2003). Male infertility From seven reviews of male infertility, two meta-analyses provided evidence of effectiveness and there was insuf cient evidence in ve reviews, four of which called for further research (Table V). In male infertility, there was evidence of increased effectiveness of IUI over timed intercourse (Cohlen et al., 2003) and IUI versus cervical insemination (O'rien and Vandekerckhove, 2003). There was insuf cient evidence of Signi cant increase in the odds of ovulation and pregnancy versus placebo Reduced odds of OHSS with u-fsh versus hmg in cycles without concomitant GnRHa. Increased odds of overstimulation when GnRHa added to gonadotrophins. No signi cant difference in pregnancy rates No signi cant difference in odds of pregnancy (trials too small) 3 29 InconclusiveÐtrials too small, assessing different interventions 5 (4) (1) Table IV. Cochrane reviews of other causes of female infertility Review title (uthors) Laparoscopic surgery for subfertility associated with endometriosis (Jacobson et al., 2003) 950 (445) (103) No signi cant difference in odds of pregnancy outcomes or complications for: ufsh versus rfsh rfsh standard dose versus chronic low dose No signi cant difference in pregnancy rates between laparoscopic ovarian drilling (6±12 months follow-up) and gonadotrophin injections (3±6 cycles); signi cantly fewer multiple pregnancies with ovarian drilling included Laparoscopic surgery signi cantly increases the odds live birth plus ongoing pregnancy effectiveness of anti-estrogens (Vandekerckhove et al., 2003b), androgens (Vandekerckhove et al., 2003c), bromocriptine (Vandekerckhove et al., 2003d) or of surgery or embolization for varicocele (Evers et al., 2003) for subfertile men. There was insuf cient evidence to support any particular technique of surgical retrieval of sperm for men with azoospermia undergoing ICSI (Van Peperstraten et al., 2003). RT For the sub-group RT, of the 12 reviews, six found evidence of effectiveness and there was insuf cient evidence of effectiveness in six reviews (Table VI). In RT cycles, there was evidence of increased effectiveness in terms of pregnancy outcomes of recombinant over urinary FSH (Daya, 2003a; Daya and Gunby, 2003), long course over short course pituitary desensitization with GnRH agonists (Daya, 2003b), and long course GnRH agonists over GnRH antagonists (l-inany and boulghar, 2003). There was insuf cient evidence of effectiveness of growth hormone in IVF protocols (Kotarba et al., 2003) and of depot versus daily administration of GnRH agonists (lbuquerque et al., 2003). There was evidence of increased effectiveness of ICSI over standard IVF in the case of borderline semen but not with normal semen (Van Rumste et al., 2003). There was insuf cient evidence of effectiveness

5 25 The Cochrane menstrual disorders and subfertility group database Table V. Cochrane reviews of male infertility Review title (uthors) included Clomiphene or tamoxifen for idiopathic oligo/asthenospermia (Vandekerckhove et al., 2003b) w ndrogens versus placebo or no treatment for idiopathic oligo/asthenospermia (Vandekerckhove et al., 2003c) w of blastocyst versus cleavage-stage embryo transfer (lake et al., 2003). In the prevention of ovarian hyperstimulation syndrome (OHSS), there was evidence of effectiveness of i.v. albumin (boulghar et al., 2003), but insuf cient evidence of effectiveness for cryopreservation (D'ngelo and mso, 2003a) or coasting versus early unilateral follicular aspiration (D'ngelo and mso, 2003b). Discussion In spite of increasing recognition of the need for trials to have suf cient power to answer a clinical question, many of the included trials in Cochrane subfertility reviews, even those performed in recent times, have been under-powered. There are many reasons for this, including dif culty recruiting subfertility patients to clinical trials and using surrogate or intermediate outcomes to calculate power. It is therefore unusual in subfertility for a single RCT to provide suf cient evidence to answer a clinical question. y comparison, large multi-centred trials of women with pre-eclampsia and breech presentation have now been undertaken, by adopting a multi-centred approach throughout both the developed and developing world. lthough further trials are planned using this approach in obstetrics, no similar networks of clinical trials exist on such a scale in subfertility and there has been a paucity of multicentre trials in subfertility. Therefore meta-analysis of the results of randomized trials, which involves pooling of trial results in a statistically appropriate manner, has been, and will continue to be, a valuable technique. There is a perception that Cochrane reviews rarely give a clear answer to a clinical question. This is clearly not the case since almost one third (12 out of 38) of the subfertility reviews did provide evidence of effectiveness of the interventions lthough a suggestion of a bene cial effect on sperm count and motility, no evidence of a signi cant effect on pregnancy rate No evidence of a signi cant effect on sperm parameters or pregnancy rates No evidence of a signi cant effect on sperm parameters romocriptine for oligo/asthenospermia (Vandekerckhove et al., 2003a) w or pregnancy rates Surgery or embolisation for varicocele in subfertile men (Evers et al., 2003) No evidence of a signi cant effect on pregnancy rates Intrauterine versus cervical insemination of donor sperm for subfertility (O'rien and Vandekerckhove, 2003) Timed intercourse versus IUI with or without ovarian hyperstimulation for subfertility in men (Cohlen et al., 2003) Techniques for surgical retrieval of sperm prior to ICSI for azoospermia (Van Peperstraten et al., 2003) IUI gives signi cantly higher chance of pregnancy than cervical insemination with cryopreserved donor sperm, but no evidence of a signi cant difference with fresh sperm completed cycles IUI gives signi cantly higher pregnancy rates per cycle than timed intercourse 1 59 lthough in this RCT there was a signi cantly higher pregnancy rate from a micropuncture technique including neurovascular stimulation compared with microsurgical epididymal sperm aspiration, there are insuf cient data from to recommend any particular technique for surgical retrieval of sperm over another studied. Furthermore, the conclusion `not enough evidence' is important to emphasise the need for further research before adopting a new treatment that might be more invasive or expensive. common mistake of authors of trials and systematic reviews is to conclude that there is `no difference' or `no effect' of a treatment (lderson and Chalmers, 2003). The preferred terminology is `no evidence of effectiveness' (lderson and Chalmers, 2003) since the null hypothesis can never be proved, only disproved, and there is always some uncertainty around the estimates of a treatment effect. In this report, where reviewers have concluded no evidence of effectiveness but suggested that no further research is required, this implies that the review has ruled out a pre-speci ed clinically important effect of the treatment based on the numbers included in the meta-analysis. Reviewers should only draw this conclusion if the individual trials are adequately powered or if clinically important differences between treatments have been excluded. The width of the con dence intervals around the point estimate may also give some indication of whether or not there is a need for further research. However the methodology for power calculations in systematic reviews (unlike power calculations in ) is not well established. There are particular methodological challenges for subfertility trials and in particular, many trials do not report livebirth outcomes, clearly the most important outcome to a couple with a fertility problem (Vail and Gardener, 2003). Whilst awareness of the likelihood of conceiving after one cycle of treatment is of some interest, the one piece of information that a woman or a couple really want is the likelihood of having a baby at the end of a course of treatment, not necessarily after each cycle. Yet there has been a preference to report C 951

6 N.P.Johnson, M.Proctor and C.M.Farquhar 26 Table VI. Cochrane reviews of RT Review title (uthors) included Recombinant versus urinary FSH for ovarian started stimulation in RT cycles (Daya and Gunby, 2003) w cycles implantation rates or `per cycle' data in subfertility trials. `Per woman' or `per couple' data are not always reported but reporting and pooling of `per cycle' data is statistically inappropriate when it is women or couples who are randomized in trials, as many of the women will have undergone more than one cycle. Furthermore, few trials report by intention-to-treat; the inappropriate use of cross-over trial design and the failure to use a secure method of randomization is also common (Vail and Gardener, 2003). The design of future clinical trials needs to consider these issues in order to improve the utility of the research undertaken. Whilst this paper may serve as a reference for the best available evidence for infertility treatments, indiscriminate use of the summary tables is not appropriate. The strength of Signi cant increase in the odds of clinical pregnancy and live birth or ongoing pregnancy for r-fsh versus ufsh Signi cant increase in the odds of clinical pregnancy for FSH versus hmg FSH and hmg for ovarian stimulation in RT cycles (Daya, 2003a) w started cycles HMG versus recombinant FSH for ovarian Insuf cicient evidence of a difference in the odds of stimnulation in RT cycles (van Wely et al., 2003) pregnancy or live birth GnRH agonist protocols for pituitary desensitization started Signi cant increase in the odds of pregnancy for long in IVF and GIFT cycles (Daya, 2003b) cycles course versus short course GnRH agonist and long Depot versus daily administration of GnRH agonist protocols for pituitary desensitization in RT cycles (lbuquerque et al., 2003) GnRH antagonist protocols for assisted conception (l-inany et al., 2003) course versus ultra-short course GnRH agonist No signi cant difference in odds of pregnancy. Length of stimulation phase and number of gonadotrophin ampoules used signi cantly increased with depot GnRH agonist Compared to standard long course GnRH agonist protocols, GnRH antagonist protocols carry decreased odds of pregnancy Growth hormone for IVF (Kotarba et al., 2003) No evidence of a signi cant difference in pregnancy rates in either standard IVF or in previous poor responders Embryo freezing for preventing OHSS (D'ngelo and mso, 2003a) i.v. albumin for preventing severe OHSS (boulghar et al., 2003) Coasting for preventing severe OHSS (D'ngelo and mso, 2003b) ICSI versus partial zona dissection, subzonal insemination and conventional techniques for insemination in IVF (Van Rumste et al., 2003) Cleavage stage versus blastocyst stage embryo transfer in assisted conception (lake et al., 2003) 3 (1) (125) No signi cant difference in pregnancy outcomes or OHSS for cryopreservation versus fresh embryo transfer (2) (26) No signi cant difference in OHSS but signi cantly more pregnancies for cryopreservation versus i.v. albumin Signi cant reduction in the odds of severe OHSS with i.v. albumin with no signi cant difference in the odds of pregnancy 1 30 No signi cant difference in odds of moderate or severe OHSS fpr coasting versus early unilateral follicular aspiration ICSI versus IVF (8, n = 405) no signi cant difference in fertilization or pregnancy rate with normal semen; ICSI signi cantly higher fertilization rate with borderline semen. Fertilization signi cantly better in context of poor semen with ICSI versus SUZI (1 RCT, n = 12) and ICSI versus additional IVF (1 RCT, n = 28) No signi cant differences in odds of pregnancy, live-birth or multiple pregnancy Key to Tables I±VI. Figures in brackets refer to individual comparisons. = Evidence of effectiveness from meta-analysis = Insuf cient evidence and review authors called for further research C = Insuf cient evidence and review authors have not called for further research bbreviations: RT = assisted reproductive technology; CC = clomiphene citrate; COS = controlled ovarian stimulation; GIFT = gamete intra-fallopian transfer; HT = hydrotubation; ICSI = intracytoplasmic sperm injection; IUI = intra-uterine insemination; OHSS = ovarian hyperstimulation syndrome; PCOS = polycystic ovary syndrome; RCT = randomized controlled trial; rfsh = recombinant follicle stimulating hormone; SUZI = sub-zonal insemination; ufsh = urinary follicle stimulating hormone. w = Withdrawn from full publication on Cochrane Library. 952 systematic review evidence inevitably re ects the quality of the original trials within the review. Vandekerckhove et al. (1993a) noted the poor quality, or even lack, of true randomization, a major potential source of bias. Thus each trial contributing to a systematic review must be scrutinized in detail. Cochrane review work can help to `plug' gaps in the evidence. Firstly, a meta-analysis may demonstrate effectiveness of an intervention where the relevant primary trials had not shown a signi cant effect. Secondly, conclusions of insuf cient evidence assist with the de nition of a research agenda and may lead to the commissioning of trials. There are numerous examples of trials being commenced after a Cochrane systematic review has highlighted de ciencies in

7 27 The Cochrane menstrual disorders and subfertility group database Table VII. Subfertility reviews in progress Review title (stage of review) uthors Unexplained IUI for unexplained subfertility (P). Cohlen et al. infertility IUI versus Fallopian tube sperm perfusion for non-tubal infertility (P) Cohlen et al. Tubal infertility novulatory Oral agents for ovulation induction in subfertility associated with PCOS (P) oothroyd et al. infertility Gonadotrophins and GnRH agonists for Rt in subfertility associated with PCOS (P) Nugent et al. Insulin-sensitising drugs for PCOS (P) Norman et al. Other female Surgical treatment of broids in sub-fertility (P) Grif ths et al. infertility Immune modulation therapy for autoimmune ovarian failure (P) Kalantaridou et al. Ovulation induction in young women with karyotypically normal spontaneous premature ovarian failure (P) Kalantaridou et al. IUI versus timed intercourse for cervical hostility in subfertile couples (P) Helmerhorst et al. Lifestyle changes for improving fertility outcomes in overweight women (T) Farquhar et al. Chinese medicinal herbs for subfertility (T) Haojie et al. Laparoscopic versus open myomectomy for the treatment of uterine broids (pt) Taylor et al. Male infertility ssisted Recombinant hcg (rhcg) for ovulation induction in assisted conception (P) l-inany et al. reproductive Techniques for intrauterine embryo transfer (P) Wilson et al. technologies ssisted hatching for assisted conception (IVF and ICSI) (P) Seif et al. Cycle regimes for frozen-thawed embryo transfer (P) Ghobara et al. Number of embryos for transfer in IVF or ICSI (P) Ozturk et al. IVF versus tubal re-anastomosis (sterilization reversal) after tubal sterilization (T) Yossry et al. Interventions for controlled ovarian hyperstimulation of poor responders in the context of IVF (T) Shanbhag et al. Follicular ushing for oocyte retrieval prior to RT (T) Wongtra-ngan et al. Day 3 versus day 2 embryo transfer following IVF or ICSI (T) Oatway et al. Luteal phase support in treatment cycles with assisted reproduction (T) Daya et al. P = Protocol published on Cochrane Library.; T = Title registered with MDSG; pt = Proposed title under consideration by MDSG. the evidence. Indeed, there is a compelling argument in favour of performing a systematic review prior to the design of any new trial. Thirdly, a strategic appraisal of the `coverage' of the scope of a collaborative review group enables reviews to plug other gapsðthis is the principle of registering titles through the editorial base of a Cochrane Collaborative Review Group to minimize the likelihood of duplication and overlap within reviews. Table VII summarizes the published protocols, registered titles and proposed titles on subfertility within the MDSG. Those interested in registering a review title for a clinical question in subfertility not covered by the titles, protocols and reviews outlined in this paper should contact Michelle Proctor, Collaborative Review Group Co-ordinator, on m.proctor@auckland.ac.nz. In conclusion, there are many gaps in the evidence for clinical decision making in subfertility and Cochrane review work has eliminated some gaps in the evidence and highlighted others. cknowledgements Carolyn ilbrough assisted with the initial extraction of data from subfertility reviews in the Cochrane Library. We are grateful to all current and previous members of staff at the of ce of the MDSG Editorial ase in uckland and to those who have contributed to MDSG subfertility reviews. References boulghar, M., Evers, J.H. and l-inany, H. (2003) Intra-venous albumin for preventing severe ovarian hyperstimulation syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update lderson, P. and Chalmers, I. (2003) Survey of claims of no effect in abstracts of Cochrane reviews. rit. Med. J., 326, 475. lbuquerque, L.E., Saconato, H. and Maciel, M.C. (2003) Depot versus daily administration of gonadotrophin releasing hormone agonist protocols for pituitary desensitisation in assisted reproduction cycles (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update l-inany, H. and boulghar, M. (2003) Gonadotrophin-releasing hormone antagonists for assisted conception (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update thaullah, N., Proctor, M. and Johnson N.P. (2003) Oral versus injectable ovulation induction agents for unexplained subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update ayram, N., van Wely, M., Vandekerckhove, P., Lilford, R. and van der Veen, F. (2003a) Pulsatile luteinising hormone releasing hormone for ovulation induction in subfertility associated with polycystic ovary syndrome ayram, N., van Wely, M. and van der Veen, F. (2003b) Recombinant FSH versus urinary gonadotrophins or recombinant FSH for ovulation induction in subfertility associated with polycystic ovary syndrome lake, D., Proctor, M., Johnson, N. and Olive, D. (2003) Cleavage stage versus blastocyst stage embryo transfer in assisted conception (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Cantineau,.E.P., Heineman, M.J. and Cohlen,.J. (2003) Single versus double insemination (IUI) in stimulated cycles for subfertile couples Cohlen,.J., Vandekerckhove, P., te Velde, E.R. and Habbema, J.D.F. (2003) Timed intercourse versus intra-uterine insemination with or without ovarian hyperstimulation for subfertility in men (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Cochrane,.L. (1979) 1931±1971: a critical review with particular reference to the medical profession. In: Teeliz Smith, G. (ed) Medicines for the Year Of ce of Health Economics, London, UK. pp. 1±11. D'ngelo,. and mso, N. (2003a) Embryo freezing for preventing ovarian hyperstimulation syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update D'ngelo,. and mso, N. (2003b) `Coasting' (withholding gonadotrophins) for preventing ovarian hyperstimulation syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Daya, S. (2003a) Follicle-stimulating hormone and human menopausal gonadotropin for ovarian stimulation in assisted reproduction cycles 953

8 N.P.Johnson, M.Proctor and C.M.Farquhar 28 Daya, S. (2003b) Gonadotrophin-releasing hormone agonist protocols for pituitary desensitization in in vitro fertilization and gamete intrafallopian transfer cycles (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Daya, S. and Gunby, J. (2003) Recombinant versus urinary follicle stimulating hormone for ovarian stimulation in assisted reproduction cycles (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Evers, J.L.H., Collins, J.. and Vandekerckhove, P. (2003) Surgery or embolisation for varicocele in subfertile men (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Farquhar, C., Vandekerckhove, P., Watson,., Vail,. and Wiseman, D. (2003a) arrier agents for preventing adhesions after surgery for subfertility Farquhar, C., Vandekerckhove, P. and Lilford, R. (2003b) Laparoscopic drilling by diathermy or laser for ovulation induction in anovulatory polycystic ovary syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Hughes, E., Collins, J. and Vandekerckhove, P. (2003a) Clomiphene citrate for unexplained subfertility in women (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Hughes, E., Collins, J. and Vandekerckhove, P. (2003b) romocriptine for unexplained subfertility in women (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Hughes, E., Tif n, G. and Vandekerckhove, P. (2003c) Danazol for unexplained infertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Hughes, E., Collins, J. and Vandekerckhove, P. (2003d) Clomiphene citrate for ovulation induction in women with oligo-amenorrhoea Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Hughes, E., Collins, J. and Vandekerckhove, P. (2003e) Gonadotrophinreleasing hormone analogue as an adjunct to gonadotropin therapy for clomiphene-resistant polycystic ovarian syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Jacobson, T.Z., arlow, D.H., Koninckx, P.R., Olive, D. and Farquhar, C. (2003) Laparoscopic surgery for subfertility associated with endometriosis Johnson, N.P. and Watson,. (2003) Postoperative procedures for improving fertility following pelvic reproductive surgery (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Johnson, N., Vandekerckhove, P., Watson,., Lilford, R., Harada, T. and Hughes E. (2003a) Tubal ushing for subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Johnson, N.P., Mak, W. and Sowter, M.C. (2003b) Surgical treatment for tubal disease in women due to undergo in vitro fertilisation (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Kotarba, D., Kotarba, J. and Hughes, E. (2003) Growth hormone for in vitro fertilization (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Nugent, D., Vandekerckhove, P., Hughes, E., rnot. M. and Lilford, R. (2003) Gonadotrophin therapy for ovulation induction in subfertility associated with polycystic ovary syndrome (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update O'rien, P. and Vandekerckhove, P. (2003) Intra-uterine versus cervical insemination of donor sperm for subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Pandian, Z., hattacharya, S., Nikolaou, D., Vale, L. and Templeton,. (2003) In vitro fertilisation for unexplained subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Vail,. and Gardener, E. (2003) Common statistical errors in the design and analysis of subfertility trials. Hum. Reprod., 18, in press. Vandekerckhove, P., O'Donovan, P.., Lilford, R.J. and Harada, T.W. (1993) Infertility treatment: from cookery to science. The epidemiology of randomised controlled trials. rit. J. Obstet. Gynecol., 100, 1005±1036. Vandekerckhove, P., Lilford, R., Vail,. and Hughes, E. (2003a) Kininenhancing drugs for unexplained subfertility in men (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Vandekerckhove, P., Lilford, R., Vail,. and Hughes, E. (2003b) Clomiphene or tamoxifen for idiopathic oligo/asthenospermia (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Vandekerckhove, P., Lilford, R., Vail,. and Hughes, E. (2003c) ndrogens versus placebo or no treatment for idiopathic oligo/asthenospermia Vandekerckhove, P., Lilford, R., Vail,. and Hughes, E. (2003d) romocriptine for idiopathic oligo/asthenospermia (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Van Perperstraten,.M., Proctor, M.L., Phillipson, G. and Johnson, N.P. (2003) Techniques for surgical retrieval of sperm prior to ICSI for azoospermia (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Van Rumste, M.M.E., Evers, J.L.H., Farquhar, C.M. and lake, D.. (2003) Intra-cytoplasmic sperm injection versus partial zona dissection, subzonal insemination and conventional techniques for oocyte insemination during in vitro fertilisation (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Van Wely, M., Westergaard, L.G., ossuyt, P.M.M. and Van der Veen, F. (2003) Human menopausal gonadotrophin versus recombinant follicle stimulating hormone for ovarian stimulation in assisted reproductive cycles Watson,., Vandekerckhove, P. and Lilford, R. (2003a) Techniques for pelvic surgery in subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Watson,., Vandekerckhove, P. and Lilford, R. (2003b) Liquid and uid agents for preventing adhesions after surgery for subfertility (Cochrane Review). In The Cochrane Library, Issue 1. Oxford Update Submitted on January 10, 2003; accepted on March 3,