Types of studies

Randomised controlled trials (RCTs) will be included in this review. Only trials that are either clearly randomised or claim to be randomised and do not have evidence of inadequate sequence generation, such as date of birth or hospital number, will be included. Cross‐over trials will be included in the review for completeness but only data from the first phase will be included in meta‐analyses.

Types of participants

Women who are part of a couple with subfertility and undertaking IVF or ICSI treatment cycles with a GnRH antagonist or agonist protocol will be included.

Types of interventions

Trials comparing long‐acting FSH versus daily FSH will be eligible for inclusion. Any dose will be included.

Types of outcome measures

Electronic searches

The following electronic databases, trial registers and websites will be searched using Ovid software.

• Cochrane Central Register of Controlled Trials (CENTRAL), see Appendix 2.

• The Menstrual Disorders and Subfertility Group (MDSG) Specialised Register of Controlled Trial, see Appendix 3.

• MEDLINE, see Appendix 4.

• EMBASE, see Appendix 5.

• PsycINFO, see Appendix 6.

• CINAHL.

The MEDLINE search will be combined with the Cochrane highly sensitive search strategy for identifying randomised trials that appears in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The EMBASE search is combined with trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN).

Other electronic sources of trials will include the following.

• Trial registers for ongoing and registered trials: 'Current Controlled Trials' (http://www.controlled‐trials.com/); 'ClinicalTrials.gov', a service of the US National Institutes of Health (http://clinicaltrials.gov/ct2/home).

• The World Health Organization International Trials Registry Platform search portal (http://www.who.int/trialsearch/Default.aspx).

• Citation indexes (http://scientific.thomson.com/products/sci/).

• Conference abstracts in the ISI Web of Knowledge (http://isiwebofknowledge.com/).

• LILACS database, as a source of trials from the Portuguese and Spanish speaking world (htpp://regional.bvsalud.org/php/index.php?lang=en) (choose 'LILACS' in 'all sources' drop‐down box).

• Clinical Study Results, for clinical trials of marketed pharmaceuticals (http://www.clinicalstudyresults.org/).

• PubMed (http://www.ncbi.nlm.nih.gov/pubmed/), the random control filter for PubMed will be taken from the searching chapter of the Cochrane Handbook for Systematic Reviews of Interventions.

• OpenSIGLE database for grey literature from Europe (http://opensigle.inist.fr/).

Searching other resources

The reference lists of articles retrieved by the search will be handsearched and personal contact will be made with experts in the field and with the manufacturers of long‐acting FSH to obtain any additional, relevant data.

Selection of studies

Two review authors will independently scan the titles and abstracts of articles retrieved by the search and remove those that are very clearly irrelevant. Full texts of all potentially eligible studies will be retrieved. Two review authors will independently examine the full text articles for compliance with the inclusion criteria and will select studies eligible for inclusion in the review. We will discuss any disagreement or doubt, whether a study is eligible for inclusion or not, with a third review author in order to achieve consensus.
There will be a list of excluded studies and the reasons for exclusion will be provided in the 'Characteristics of excluded studies' table.

Data extraction and management

Data will be extracted from eligible studies using a data extraction form designed and pilot‐tested by the authors. Where studies have multiple publications, the main trial report will be used as the reference and additional details will be supplemented from secondary papers. The review authors will correspond with study investigators in order to resolve any data queries, as required. Two review authors will independently extract the data. Any disagreement between these review authors will be resolved by a third review author.

Assessment of risk of bias in included studies

The included studies will be assessed for risk of bias using the Cochrane risk of bias assessment tool, which recommends the explicit reporting of the following domains.

1. Random sequence generation (selection bias)

   • Adequate: use of central computer randomisation, independent central randomisation office, on‐site computer from which assignment can only be determined after entering patient data, random number table or serially numbered and sealed opaque envelopes

   • Inadequate: use of non‐opaque envelopes or systematic methods (e.g. date of birth, medical record number, day of the week presenting)

   • Unclear: insufficient information about the process of sequence generation

2. Allocation concealment (selection bias)

   • Adequate: sequentially numbered and identical drug containers are used

   • Inadequate: use of open random allocation (e.g. date of birth, medical record number, day of the week presenting)

   • Unclear: insufficient information about the process of allocation concealment

3. Blinding of participants, researchers and care providers (performance bias)

   • Adequate: blinding of the participants, researchers and the care providers, or incomplete or no blinding was used but it was not likely to influence outcomes

   • Inadequate: no blinding or incomplete blinding was used and likely to influence the outcomes

   • Unclear: insufficient information about the process of blinding the participants, researchers and care providers

4. Blinding of the outcome assessor (detection bias)

   • Adequate: blinding of the researchers or incomplete blinding had no effect on the outcome measurement

   • Inadequate: no blinding of the researchers, or incomplete blinding had influence on the outcomes

   • Unclear: insufficient information about the process of blinding the outcome assessor

5. Incomplete outcome data (attrition bias)

   • Adequate: there are no missing data, or reasons for missing data may not influence the outcomes

   • Inadequate: reasons for missing data may influence the outcomes

   • Unclear: insufficient information about the completeness of outcome data

6. Selective outcome reporting (reporting bias)

   • Adequate: all pre‐specified outcomes in the protocol have been published, or no protocol available but it is clear all pre‐specified outcomes are reported.

   • Inadequate: not all pre‐specified outcomes in the protocol are reported

   • Unclear: insufficient information about the process of outcome reporting

7. Other potential sources of bias

   • Adequate: the study was free of other biases

   • Inadequate: other biases are present

   • Unclear: insufficient information about the other sources of bias

Two authors will assess these seven domains as ' low risk of bias' (adequate), 'high risk of bias' (inadequate), or 'unclear risk of bias' (unclear). The assessments made by the two authors will be compared and any disagreements resolved by consensus or by discussion with a third author. The conclusion will be presented in the 'Risk of bias' table and will be incorporated into the interpretation of review findings by means of sensitivity analyses.

Measures of treatment effect

We will use the dichotomous data measures and express the results in the control and intervention groups of each study as Peto odds ratios (OR) with 95% confidence intervals (CI).

Unit of analysis issues

The primary analysis will be per woman randomised. Reported data that does not allow valid analysis (for example 'per cycle' rather than 'per woman' where women contribute more than one cycle) will be briefly summarised in an additional table and will not be meta‐analysed. Multiple live births (for example twins or triplets) will be counted as one live birth event.  

Dealing with missing data

In the case of missing data from the included studies, the original investigators will be contacted to request the relevant missing data. Where these are unobtainable, imputation of individual values will be undertaken for the primary outcomes only. Live births will be assumed not to have occurred in participants without a reported outcome. For other outcomes, only the available data will be analysed. Any imputation that is undertaken will be subjected to sensitivity analysis.
The data will be analysed on an intention‐to‐treat (ITT) basis, as far as possible.

Assessment of heterogeneity

We will consider whether the clinical and methodological characteristics of the included studies are sufficiently similar for meta‐analysis to provide a meaningful summary. We will use the I² statistic to assess the impact of the heterogeneity on the meta‐analysis. We will interpret the result of the I² statistic as follow:

• 0% to 40%, might not be important;

• 30% to 60%, may represent moderate heterogeneity;

• 50% to 90%, may represent substantial heterogeneity;

• 75% to 100%, considerable heterogeneity (Higgins 2011).

An I² statistic measurement greater than 50% will be taken to indicate substantial heterogeneity (Higgins 2011). If substantial heterogeneity is detected, possible explanations will be explored in sensitivity analyses.

Assessment of reporting biases

We will take care to search for within‐trial selective reporting, such as trials failing to report obvious outcomes or reporting them in insufficient detail to allow inclusion. We will seek published protocols to look for any pre‐planned outcomes that may not have been reported and compare the outcomes between the protocol and the final published study.

Where identified studies fail to report the primary outcome of live birth, but do report interim outcomes such as pregnancy, we will undertake informal assessment as to whether the interim values (for example pregnancy rates) are similar to those reported in studies that also report live birth.

In view of the difficulty of detecting and correcting for publication bias and other reporting biases, we will aim to minimise the potential impact by ensuring a comprehensive search for eligible studies and by being alert for duplication of data. If there are 10 or more studies in an analysis, we will use a funnel plot to investigate the potential for publication bias and other reporting biases.

Data synthesis

The data from primary studies will be combined using a fixed‐effect model. If substantial heterogeneity is found that cannot be explained by sub‐grouping, then we will employ a random‐effects model to investigate this. The statistical analysis will be carried out using Review Manager 5.1.

Subgroup analysis and investigation of heterogeneity

Where data are available, subgroup analyses will be conducted to determine the evidence within the following subgroups:

• age;

• weight;

• body mass index (BMI);

• dose of long‐acting FSH;

1. low dose (60‐120 μg);
2. medium dose (150‐180 μg);
3. high dose (240 μg).

• day of starting GnRH antagonist;

• poor responders to ovarian stimulation.

Sensitivity analysis

We will conduct sensitivity analyses for the primary outcomes to determine whether the conclusions are robust to arbitrary decisions made regarding the eligibility and analysis of studies. These analyses will include consideration of whether the review conclusions would have differed if:  

• eligibility was restricted to studies without high risk of bias;  

• alternative imputation strategies were adopted;

• studies with outlying results were excluded from the meta‐analyses.