Home > Procedures used during in vitro...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Fertility: Assessment and Treatment for People with Fertility Problems. London (UK): RCOG Press; 2004 Feb. (NICE Clinical Guidelines, No. 11.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

12Procedures used during in vitro fertilisation treatment

The Human Fertilisation and Embryology Act 1990 (HFE Act) requires that any fertility clinic in the UK offering licensed treatment services, such as IVF or use of donated gametes, must take account of the welfare of the potential child (including the determination of who will have parental responsibility for the child) and of any other existing children who may be affected by the birth, before treatment. Details on the issues of assessment of people seeking treatment, confidentiality, information, consent and counselling are referred to the HFEA Code of Practice.218

12.1. Medical assessment and screening

In addition to a detailed clinical assessment involving history taking and physical examination, careful screening before assisted reproduction aims to protect recipients and offspring from transmission of infections and genetic diseases. The welfare of children resulting from assisted reproduction should be considered in relation to screening.

A case series study showed that among patients seeking infertility treatment at an IVF clinic, 0.06% were seropositive for HIV, 0.5% were seropostive for the hepatitis B virus and 0.54% were seropositive for the hepatitis C virus.819 A cross-sectional study with 409 patients (248 women and 161 men) attending an infertility clinic reported a prevalence of anti-hepatitis C virus positivity of 3.2 % among women and 3.7% among men.820 Hepatitis C virus was detected in 5% of semen samples from men (n = 39) entering an IVF programme. Consideration needs to be given to the risk of hepatitis C virus transmission not only to the mother and child, but also through laboratory contamination of other non-infected couples’ gametes and of technicians, and even through storage and manipulation of cryopreserved semen.821 [Evidence level 3]

Screening for C. trachomatis infection before uterine instrumentation is discussed in Chapter 5, Section 5.3.


People undergoing in vitro fertilisation treatment should be offered screening for HIV, hepatitis B virus and hepatitis C virus; people found to test positive should be managed and counselled appropriately. [B]

12.2. Management of couples with viral infections

Where patients have chronic infections such as hepatitis B, hepatitis C and/or HIV, this should be taken into account when considering them for fertility treatment. If the treatment proposed is within the remit of the HFEA, then a ‘welfare of the child’ assessment is mandatory. Patients should be counselled thoroughly and given information about the potential risks and implications for themselves and their children in a manner that is sensitive to and specific for their own situation.

Where a positive case of these infectious agents is detected, fertility diagnosis and treatment must be carried out in facilities and using procedures which are appropriate for the handling of known positive specimens of the appropriate classification. Not all centres currently have such facilities available. Particular considerations apply to the use of cryopreservation, where there may be some risk of cross-contamination between samples.822–824 Such risks cannot be quantified and relate to the specific methodology used and the viral load of the specimen.

Whether fertility treatment is appropriate and the options available will vary depending upon the viral status of the male partner and/or the female partner, the particular infectious agent, the stage of their disease, their compliance with medication, and their fertility status. A strategy for the management of patients seeking fertility treatment and who are infected with HIV, hepatitis B and hepatitis C has been suggested.825 [Evidence level 3–4]

HIV infection

Current debates have focused on the welfare of the child perspective relating to vertical transmission or loss of a parent at a young age, and the improved treatment outcomes of antiretroviral drugs.826,827

Serodiscordant couples in which the man is HIV-1 positive and the woman is negative have limited options if they wish to have children safely because of the risk of transmitting HIV virus in semen to the female partner and offspring. One option is insemination with sperm from seronegative donors.

Sperm washing828,829 has been used as a risk-reduction option in which infected sperm are washed to reduce the titre of virus830,831 before insemination into the female partner at the time of ovulation, resulting in healthy live births and no reported seroconversions in either partners or children.828,832,833 [Evidence level 3] However, the risk of transmission still exists, as shown by the persistence of virus in washed sperm.834,835

In serodiscordant couples where the female partner is HIV positive and has no overt fertility problems, timed self-insemination with the man’s sperm can be considered. When assisted reproduction treatment is indicated (ovulation induction, IUI, IVF or ICSI), steps should be taken to minimise any risk of multiple pregnancy because of the increased risk to mother and fetus, perinatal morbidity and burden of caring for two or three babies at the same time when women infected with HIV are prone to ill health. Antiretroviral medication should be discussed with the treating physician. Little is known of the effect of invasive procedures involved in IVF treatment and ICSI (such as oocyte retrieval) on vertical transmission, or the long-term effects of antiretroviral treatments upon offspring.

Hepatitis B infection

Partners of individuals with hepatitis B should be vaccinated before fertility treatments begin and sperm washing will not be necessary. The normal course of pregnancy is not affected by hepatitis B infection and vertical transmission to neonates can be minimised with hepatitis B vaccination within 24 hours of birth and at six months.

Hepatitis C infection

As there is no vaccine for hepatitis C infection, risk-reduction measures such as sperm washing in assisted reproduction may be considered if the male partner is infected.830 [Evidence level 3] The normal course of pregnancy is not affected by hepatitis C infection. Both vertical transmission and nosocomial transmission (transmission within a health care setting) can be minimised by medical treatment to reduce viral load before fertility treatment or assisted reproduction. No specific vaccine is available to protect neonates.

The decision whether to provide fertility treatment in these patients should include an assessment of the welfare of the child. The patients’ own health, any associated high-risk behaviour, existence of a (homo-or heterosexual) couple etc. are all relevant to the decision-making process. Couples carrying HIV, hepatitis B and hepatitis C infections and who have fertility problems should be referred to centres having the appropriate expertise and facilities to provide safe risk-reduction treatment.


In considering the decision to provide fertility treatment for couples with HIV, hepatitis B or hepatitis C infections the implications of these infections for potential children should be taken into account. [D]

12.3. Ovulation induction during in vitro fertilisation treatment

IVF ovulation induction techniques are based on the use of the same drugs that are used in ovulation induction for ovulatory disorders. However, there are specific aspects of the use of these drugs that will be different in the IVF context. The more generic aspects of drug use (and their risks), especially in relation to gonadotrophins and GnRH analogues, are discussed in Chapter 7, whereas those drug techniques that are more specific to IVF are discussed below.

Natural cycle in vitro fertilisation

A literature review of studies involving 1800 cycles, 819 embryo transfers and 129 ongoing pregnancies reported an embryo transfer rate of 45.5% per cycle, an ongoing pregnancy rate of 7.2% per cycle and a cycle cancellation rate of 29% in natural cycle IVF.836 [Evidence level 2b–3] Natural cycle IVF was associated with no risk of OHSS or multiple pregnancy rate when a single embryo was transferred.836 [Evidence level 2b–3]

Natural cycle versus clomifene-stimulated cycle

An RCT showed no significant difference in clinical pregnancy rate between clomifene citrate cycle and natural cycle IVF (18% with clomifene citrate cycle versus 0% in natural cycle) but cycle cancellation rate was significantly higher in natural cycle IVF (10 cycles versus none).837 [Evidence level 1b]

Another RCT found a significantly higher pregnancy rate per cycle in patients undergoing clomifene citrate cycle IVF compared with natural cycle IVF (18% with clomifene citrate cycle versus 4% with natural cycle; RR 5.14, 95% CI 1.81 to 14.55).838 [Evidence level 1b] Modest side effects were reported following clomifene.

Natural cycle versus gonadotrophins

A crossover RCT found a significant improved clinical pregnancy rate per cycle with hMG cycle IVF versus natural cycle IVF (23% with hMG cycle versus 0% with natural cycle). There were no data on side effects or multiple pregnancy rate.839

Natural versus stimulated cycles with frozen embryos

The replacement of frozen-thawed embryos can take place in either a natural cycle or in an artificial cycle where exogenous hormones with or without GnRH analogue are used to prepare the endometrium. Patients with anovulatory or irregular cycles will be easier to manage with a programmed cycle such as a GnRHa-hormone replacement therapy protocol.840,841

A partly randomised controlled trial (n = 162) assessed the relative efficacy of two strategies of patient management for the replacement of frozen-thawed embryos. One group (n = 84) were treated with a GnRH analogue before receiving hormone replacement therapy (oral oestradiol valerate and intramuscular progesterone) for endometrial priming. The second group (n = 78) had their frozen-thawed embryos replaced during their natural cycles. Women with regular menstrual cycles were randomised to either group, but some categories of patients were allocated to the GnRH-hormone replacement therapy group without randomisation. These included women with amenorrhoea, oligomenorrhoea, inadequate luteal function or previously unsuccessful frozen embryo replacement in a natural cycle. There was no difference between groups in terms of age, obstetric history, duration of infertility, number of oocytes retrieved or fertilised or the numbers of embryos frozen following ovarian stimulation in the initial cycle. Eighty embryos were replaced in the first group and 16 (20%) clinical pregnancies occurred. A similar pregnancy rate was achieved in the second group with 14 clinical pregnancies (20%) occurring after replacement of 70 embryos.840 [Evidence level 2a]

In replacing frozen-thawed embryos, pregnancy rates were reported to be similar between natural cycle and programmed cycles;737 between natural cycle and GnRHa/hormone replacement therapy cycles in women undergoing replacement after elective embryo cryopreservation to minimise the risk of developing OHSS (RR 0.65, 95% CI 0.29, 1.42),842 between natural cycles and cycles controlled with exogenous oestradiol and progesterone administration,843 between natural cycle and GnRHa administration followed by oestradiol valerate supplementation,844 between spontaneous cycles, artificial preparation and ovarian stimulation cycles845 and between spontaneous cycles, an ovarian stimulation and oestrogen/progesterone replacement therapy.846 [Evidence level 3]

An RCT (n = 106) compared the outcome of frozen-thawed embryo transfer cycles using micronised 17 beta-oestradiol and micronised vaginal progesterone preparations with and without the concomitant use of a GnRH analogue and found comparable pregnancy rate per embryo transfer in both groups (26.4% with GnRHa versus 21.1% with no GnRHa).847 [Evidence level 1b]

Drugs for cycle control

In assisted reproduction, ovarian stimulation protocols enable the production and collection of multiple oocytes, which are fertilised in vitro and the resulting embryos then transferred into the uterus. IVF treatment is based predominantly on superovulation induced using gonadotrophins (such as hMG, uFSH and rFSH) in order that multiple follicles develop. In IVF treatment, gonadotrophins are most commonly used alongside gonadotrophin-releasing hormone (GnRH) agonists (such as goserelin, nafarelin and luprolide) and sometime antagonists (cetrorelix and ganirelix). Since many aspects of gonadotrophin and GnRHa use overlap with their uses in ovulation induction for ovulatory disorders, the evidence relating to these agents in IVF treatment is discussed in Chapter 7.

Management of women with a poor ovarian response

The lack of a consistent definition of poor ovarian response makes it difficult to develop or assess any protocol to improve the outcome. Women with poor ovarian response have lower pregnancy rates characterised by fewer follicles and number of oocytes retrieved, likely to be associated with diminished ovarian reserve.848,849 [Evidence level 3I]

A systematic review of available studies including RCTs found limited data that assessed the effectiveness of different management strategies in women with poor ovarian response.849 There is minimal or no benefit with the use of increased dose of gonadotrophins. There is insufficient evidence that the use of rFSH improved pregnancy rates when compared with uFSH in poor responders. Flare-up GnRH agonist protocols were reported to produce better results than standard long luteal protocols. Luteal initiation of FSH has not been shown to improve pregnancy outcome. The use of GnRH antagonists did not show any benefits. There were no studies reporting the use of corticosteroids involving poor responders. Data were limited on the use of nitrous oxide donors such as L-arginine in improving pregnancy rate in poor responders. Pre-treatment with combined oral contraceptives before ovarian stimulation may be beneficial. No benefit was shown with standard use of ICSI or assisted hatching of zona pellucida. Comparable pregnancy rates were reported between natural and stimulated cycles in poor responders. Further evaluation with large-scale and well-designed RCTs is needed to verify the role of these different approaches.849 [Evidence level 1b–2b]

Adjuvant growth hormone therapy

A systematic review of six RCTs found no significant difference between growth hormone augmented ovulation induction versus non growth hormone augmented ovulation induction in pregnancy rate per cycle in women with no previous poor response (OR 0.97, 95% CI 0.34 to 2.76) or in poor IVF responders (OR 2.55, 9%% CI 0.64 to 10.12).850 [Evidence level 1a]

Three additional RCTs were found. One small RCT showed no significant difference between adjuvant growth hormone GH 4 IU versus growth hormone GH 12 IU versus no growth hormone in downregulated ovulation induction in pregnancy rate per embryo transfer (0% versus 29% versus 0%).851 [Evidence level 1b] Another RCT showed no significant difference between growth hormone-releasing factor versus placebo in clinical pregnancy rate (8.3% versus 8%) and live birth rate (5.2% versus 4%) in poor responders.852 [Evidence level 1b] One quasirandomised trial showed no significant difference between growth hormone versus no growth hormone in downregulated ovulation induction in pregnancy rate (0% versus 7.7%) in poor responders.853 [Evidence level 2a]


Natural-cycle in vitro fertilisation has lower pregnancy rates per cycle of treatment than clomifene citrate-stimulated and gonadotrophin-stimulated in vitro fertilisation and is therefore not recommended, except in the rare circumstances where gonadotrophin use is contraindicated. [A]

For women who have regular ovulatory cycles, the likelihood of a live birth after replacement of frozen-thawed embryos is similar whether the embryos are replaced during natural or stimulated cycles. [B]

The use of adjuvant growth hormone with gonadotrophins during in vitro fertilisation cycles does not improve pregnancy rates and is therefore not recommended. [A]

12.4. Oocyte maturation – human chorionic gonadotrophin

Human chorionic gonadotrophin has been used as a surrogate LH surge to induce final oocyte maturation before oocyte retrieval in assisted reproduction.

An RCT found no significant differences between rhCG and uhCG in clinical pregnancy rate (33% with rhCG versus 24.7% with uhCG) and live birth rate (27% with rhCG versus 23% with uhCG) and OHSS incidence (7.2% with rhCG versus 6.4% with uhCG).854 [Evidence level 1b]

Another RCT showed no significant differences between 250 micrograms and 500 micrograms of rhCG and uhCG in clinical pregnancy rate (35.1% versus 36% versus 35.9%), live births (87.9% versus 84.4% versus 84.8%) or OHSS incidence (3.25% versus 9% versus 3.1%).855 [Evidence level 1b]


Couples should be informed that, in effecting oocyte maturation, recombinant human chorionic gonadotrophin achieves similar results to urinary human chorionic gonadotrophin in terms of pregnancy rates and incidence of ovarian hyperstimulation syndrome. Consideration should be given to minimising cost when prescribing. [A]

12.5. Monitoring of stimulated cycles

In assisted reproduction, the purpose of monitoring ovarian response is to ensure safe practice in reducing the incidence and severity of OHSS, and to optimise the timing of luteinisation before oocyte retrieval.

An average number of three-ultrasound-scan monitoring is commonly practiced: at the start of ovarian stimulation in GnRH agonist-controlled cycle, to assess at day seven to nine and to determine timing of hCG administration at days 11 to 14. The extent of monitoring is reduced in GnRH antagonist controlled cycles.856 [Evidence level 3]

One RCT (n = 114) found no significant differences between ultrasonic ovulation control with hormone determination versus ultrasound alone in pregnancy rate per embryo transfer (27.2% versus 29.5%) and OHSS rate (5.3% versus 7%) in women undergoing GnRHa-hMG during IVF-embryo-transfer for the first time.857 [Evidence level 1b]

One RCT (n = 279) found no significant differences between cycle monitoring using both serum oestradiol and ultrasound versus ultrasound alone in clinical pregnancy rate (34.3% versus 31.4%) and OHSS rates (4.9% versus 4.1%) in normal responders undergoing GnRHa-rFSH during IVF-embryo-transfer.858 [Evidence level 1b]

A non-RCT (n = 206) found no significant differences between ultrasound with hormonal determination versus ultrasound alone in clinical pregnancy rate (22.9% versus 23.4%), live birth rate (14.3% versus 14.8%) and OHSS rate (1.04% versus 0.9%) in women undergoing GnRHa-hMG/hCG during IVF-embryo-transfer.859 [Evidence level 2a]


Ultrasound monitoring of ovarian response should form an integral part of the in vitro fertilisation treatment cycle. [C]

Monitoring oestrogen levels during ovulation induction as part of in vitro fertilisation treatment is not recommended as a means of improving in vitro fertilisation treatment success rates because it does not give additional information with regard to live birth rates or pregnancy rates compared with ultrasound monitoring. [A]

12.6. Ovarian hyperstimulation syndrome

OHSS is an iatrogenic and potentially life-threatening complication of superovulation. The incidence of OHSS varies between 0.6% and 10% in IVF cycles. The severe form of the condition occurs in 0.5–2% of IVF cycles860 (see also Section 7.11).

Several risk factors have been associated with the development of OHSS:861

  • Young age (less than 30 years)
  • Lean physique
  • Polycystic ovary syndrome
  • High serum oestradiol (greater than 2500 pg/ml or 9000 pmol/l)
  • Rapidly increasing oestradiol levels (greater than 75% from previous day)
  • Size and number of follicles and ultrasonographic ovarian ‘necklace sign’ of multiple small follicles
  • hCG administration
  • Number of oocytes retrieved (greater than or equal to 20)
  • Multiple pregnancy.

Criteria for classifying the severity of OHSS are:

  • Mild:
    • abdominal bloating, mild pain
    • ovarian size usually less than 8 cm*
  • Moderate:
    • increased abdominal discomfort accompanied by nausea, vomiting and/or diarrhoea
    • ultrasound evidence of ascites
    • ovarian size usually 8–12 cm*
  • Severe:
    • clinical ascites, sometimes hydrothorax
    • haemoconcentration (haematocrit greater than 45%, white blood cell count greater than 15000/ml)
    • oliguria with normal serum creatinine
    • liver dysfunction
    • anasarca
    • ovarian size usually greater than 12 cm*
  • Critical:
    • tense ascites
    • haematocrit greater than 55%, white blood cell count greater than 25000/ml
    • oliguria with elevated serum creatinine
    • renal failure
    • thromboembolic phenomenon
    • ovarian size usually greater than 12 cm.*


There is no evidence to support the superiority of either hMG or rFSH517 (OR 1.60, 95% CI 0.60 to 4.3) or urinary preparations518 (OR 1.36, 95% CI 0.79 to 2.33) in preventing OHSS. [Evidence level 1a]

Cycle cancellation

Cancellation of a treatment cycle is a strategy that has been considered if ovarian ultrasound reveals a large number of developing follicles and/or serum oestradiol levels are excessively high. The principle behind this decision is to withhold the ovulatory trigger (hCG). In cycles where GnRH agonists have not been used, this may not completely prevent early-onset OHSS as a natural LH surge may still occur.862


Coasting involves discontinuation of gonadotrophins in cycles with an excessive response and delaying hCG administration, while continuing GnRH agonist administration in the presence of ultrasound and endocrine monitoring.863 It is an alternative to cycle cancellation in situations where there is a substantial risk of OHSS associated with high serum oestradiol levels above 2500 pg/ml (9000 pmol/l). The aim is to allow FSH levels to drop, thus inhibiting granulosa-cell proliferation and subsequent availability for luteinisation. The patient is monitored until the oestradiol level falls below a safe limit (< 2500 pg/ml or 9000 pmol/l). Although shown to be effective in observational studies, there is insufficient evidence to advocate the use of coasting in routine practice. It can potentially reduce the number of oocytes recovered and may even compromise pregnancy rates. A systematic review on the role of coasting for the prevention of OHSS identified only one RCT. Compared with elective unilateral follicular aspiration (elective aspiration of excess ovarian follicles), there was no convincing benefit associated with the use of coasting (OR 0.76, 95% CI 0.18 to 3.24).864 [Evidence level 1a]

Elective cryopreservation of all embryos

Following oocyte recovery in assisted reproductive treatments, fresh embryo transfer may be deferred if there are excessive numbers of follicles and oocytes recovered (for example, more than 20). All embryos are cryopreserved and electively replaced at a later date. The idea is to prevent a conception cycle and, hence, late-onset OHSS. A systematic review has found that there is insufficient evidence to support routine cryopreservation in cases with a high risk of OHSS (OR 5.33, 95% CI 0.51 to 56.24 for elective cryopreservation versus intravenous albumin; OR 0.12, 95% CI 0.01 to 2.29 for elective cryopreservation versus fresh embryo transfer).865 [Evidence level 1a]

Luteal-phase support

A systematic review has confirmed the effectiveness of routine luteal phase support after embryo transfer in IVF cycles involving the use of gonadotrophin-releasing hormone agonists.866 [Evidence level 1a] The use of hCG in this situation can aggravate OHSS and progesterone should be the preparation of choice in high-risk women.867

Prophylactic albumin administration

It has been suggested that administration of intravenous albumin around the time of oocyte recovery could be used as a preventative measure in the high-risk woman. The exact mode of action of albumin is unknown but it is thought to bind to vasoactive substances involved in the pathogenesis of OHSS. It also increases the intravascular oncotic pressure, thereby preventing loss of water from the intravascular compartment.861 A systematic review868 reported that the use of intravenous albumin at the time of oocyte retrieval significantly reduced the incidence of severe OHSS in high-risk women undergoing IVF (OR 0.28, 95% CI 0.11 to 0.73). [Evidence level 1a] However, the optimal timing and dose of albumin are unclear, as is its effect on implantation. There are also growing concerns about the possibility of febrile reactions, anaphylactic shock and the potential risk of virus and prion transmission.869 The systematic review,868 suggested that 18 women at risk would need to be treated with albumin infusion in order to prevent a single case of severe OHSS. This needs to be taken into account in the context of clinical decision making.

The alternative to albumin is infusion of hydroxyethyl starch solution, which is a plasma colloidal substitute. It may be a safer, cheaper and effective method that needs evaluation in an RCT, and there are concerns about its interaction with the blood-coagulation system.870

Role of follicular aspiration

Recovery of immature oocytes (which can then be cultured in vitro and subsequently used for IVF) has been suggested as a means of preventing OHSS when hCG is withheld.871 Follicular aspiration alone cannot be relied on to avert the development of OHSS or to arrest clinical deterioration in a pre-existing case. Despite this, practitioners are known to attempt meticulous puncture and aspiration of all stimulated follicles at time of oocyte recovery in the belief that this interferes with the mechanisms leading to production of the ovarian mediators of OHSS.861

Other methods of prevention

A number of other methods of preventing OHSS have been advocated. These include the use of recombinant LH872 and GnRH antagonists such as ganirelix or cetrorelix.873,874 A meta-analysis of five RCTs561 suggested that treatment with GnRH antagonists did not significantly reduce the incidence of severe OHSS in comparison with those treated with agonists (OR 0.51, 95% CI 0.22 to 1.18). [Evidence level 1a]

In a prospective randomised trial,875 ovarian electro diathermy in women with polycystic ovaries before IVF was compared with IVF alone. There was no significant difference in the incidence of OHSS in women treated by ovarian diathermy or not. [Evidence level 1b]


Treatment of OHSS is mainly supportive.862 Multidisciplinary local protocols involving gynaecologists, anaesthetists and haematologists should be generated and strictly followed. The condition is self-limiting and resolution parallels the decline in serum hCG levels (about seven days in nonpregnant women and 10–20 days in pregnant women). Mild OHSS is usually benign and resolves with the onset of the first period. Moderate to severe cases need hospital admission and monitoring. The principles of care include appropriate specialist involvement, circulatory support using intravenous fluids, maintenance of renal function, thromboprophylaxis and drainage of third space accumulation.


Clinics providing ovarian stimulation with gonadotrophins should have protocols in place for preventing, diagnosing and managing ovarian hyperstimulation syndrome. [GPP]

Women who have a significant risk of developing ovarian hyperstimulation syndrome should not be offered oocyte maturation (or luteal support) using human chorionic gonadotrophin. [A]


Further research is needed to determine whether prophylactic albumin treatment administered at the time of egg collection is effective. This research should include issues related to timing and dose.

12.7. Oocyte retrieval

Conscious sedation and anaesthesia or analgesia

It is accepted that transvaginal oocyte retrieval is unpleasant and painful. It is therefore important to provide effective anaesthesia or analgesia to minimise adverse effects and to minimise toxic effects on embryo cleavage rates and pregnancy rates. No technique of anaesthesia, analgesia or sedation is free from adverse effects. Whatever technique is used, it is essential that it should conform to the recognised standards of practice and guidance on the safe use of sedative drugs for patients undergoing health procedures as published by the Academy of Royal Medical Colleges.876 [Evidence level 4]

A narrative review of anaesthesia methods used for transvaginal retrieval of oocytes found that general anaesthetics can traverse into the follicular fluid and may be detrimental to cleavage rates of embryo and pregnancy rate. Epidural anaesthesia avoids many of the adverse effects of general anaesthesia and it may shorten recovery time. However, it requires the expertise of an anaesthetist. Local anaesthesia (paracervical block) or no anaesthesia can cause unnecessary discomfort. Conscious sedation requires less-specialised equipment, causes relatively few complications and is well-tolerated, although there is a theoretical risk of agents contaminating the follicular fluid.877 [Evidence level 2b–3]

Conscious sedation versus placebo

An RCT showed significantly higher median vaginal pain and abdominal pain levels in women given paracervical block and placebo as compared with paracervical block and conscious sedation. However, there was no significant difference in pregnancy rates per cycle.878 [Evidence level 1b]

Another RCT found significantly higher anxiety levels and vaginal and abdominal pain levels in women given placebo when compared with women given premedication with anxiolytic during oocyte retrieval.879 [Evidence level 1b]

Patient-controlled analgesia

An RCT showed no significant difference in mean pain score and patient satisfaction rate between fentanyl administration via a patient-controlled analgesia delivery system versus administration by a physician. However, significantly more fentanyl was used in the patient-controlled analgesia group.880 [Evidence level 1b] Another RCT reported no difference in patient satisfaction with conventional intravenous analgesia compared with patient-controlled inhalational isodesox during oocyte recovery, although the mean pain score was higher in the group receiving isodesox. There was no difference in fertility outcomes between the two groups.881 [Evidence level 1b] Patient-controlled sedation using propofol or alfentanil was also reported to provide less pain relief for patients than physician-administered sedation using diazepam and pethidine during transvaginal ultrasound-guided oocyte retrieval. Fertility outcomes were similar in the two groups.882 [Evidence level 1b]

Conscious sedation versus general anaesthesia

An RCT found significantly higher mean pain score with conscious sedation using midazolam and ketamine when compared with general anaesthesia using fentanyl and propofol, although the higher pain score with sedation was not sufficiently high to render it unacceptable to women. There was no difference between the two groups in pregnancy rate per embryo transfer (22.7% with sedation versus 23.8% with general anaesthesia). The mean number of embryos transferred was significantly higher in the sedation group (2.8 versus 1.9). Patient satisfaction did not differ between the two groups.883 [Evidence level 1b]

Intravenous midazolam and remifentanil and intravenous propofo and fentanyl were reported to be similar in providing effective sedation during oocyte retrieval for IVF procedures. However, a significant proportion of women (13%) given intravenous midazolam and remifentanil found the experience unpleasant due to awareness during the surgical procedure and said they would not accept conscious sedation for the same procedure in the future. All of the women given propofol and fentanyl were satisfied and said they would accept conscious sedation again.884 [Evidence level 1b]

Exposure to the intravenous anaesthetic drug propofol was not reported to have a detrimental effect on oocyte quality.885 [Evidence level 3]

A cohort study (n = 202) compared the effects of general anaesthesia with conscious sedation on oocyte retrieval and IVF outcome. This study found that significantly more oocytes were collected in the general anaesthesia group compared with the sedation group but there were no differences in cleavage and pregnancy rates between the two groups (23.6% with general anaesthesia versus 31.3% with conscious sedation).886 [Evidence level 2b] Epidural anaesthesia was reported to be effective in pain control when compared with intravenous sedation in an IVF programme. The pregnancy rates were similar in the two groups.887 [Evidence level 2b] Clinical pregnancy rates and delivery rates were lower following oocyte retrieval performed under general anaesthesia using nitrous oxide compared to epidural and local anaesthesia.888 [Evidence level 2b] A meta-analysis of three RCTs and one case–control study reported no difference in pregnancy rates (pooled OR 0.71, 95% CI 0.47 to 1.08) between general anaesthesia and locoregional anaesthesia in patients undergoing laparoscopic oocyte retrieval.889 Meta-analysis of the three RCTs showed similar results (OR 0.84, 95% CI 0.28 to 2.56) [Evidence level 1a]

A 1997 survey of UK fertility centres found that many different techniques were used for anaesthesia in IVF programmes.890 [Evidence level 3] A recent survey reported that 84% and 16% of IVF clinics used intravenous sedation and general anaesthesia, respectively, for transvaginal oocyte retrieval.891 [Evidence level 3] There was wide variation in personnel present during the procedure, the use of drugs, the degree of monitoring and the availability of emergency drugs. This wide variation in current practice within the UK highlighted the need for adoption of national guidelines for safe use of sedation in women undergoing IVF treatment. A set of guidelines with recommendations for good practice for sedation in assisted reproduction procedures has since been developed.892 [Evidence level 4]

Follicle flushing

Follicle flushing is traditionally employed during transvaginal ultrasound-directed oocyte recovery for IVF in the belief that flushing allows a larger number of oocytes to be collected that would otherwise be missed if aspiration alone were used.

An RCT (n = 36) reported similar oocyte recovery rate using a single-lumen needle without flushing or a double-lumen needle with flushing at ovum pick up. Administration of hCG occurred when the dominant follicle reached 18 mm in diameter in the presence of an appropriate oestradiol level. The number of follicles at the time of hCG administration was not reported. Operating time may be longer with follicle flushing.893 [Evidence level 1b]

Another RCT (n = 34) showed no significant differences between follicular aspiration with flushing and follicular aspiration only in mean number of oocytes retrieved (7.0 versus 8.5), fertilisation rate (64% versus 60%) and ongoing pregnancy rate (17% versus 19%). This trial included women who had developed at least three follicles that had attained a diameter of 18 mm with corresponding oestradiol levels at the time of hCG administration. Significantly longer time was required for the procedure of flushing.894 [Evidence level 1b]

A further RCT found no significant differences between follicular aspiration with flushing and follicular aspiration only in mean number of oocytes retrieved (9 versus 11), fertilisation rate (60% versus 55.6%) and clinical pregnancy rate per woman (26% versus 24%; RR 0.92, 95% CI 0.47 to 1.82). This trial excluded women who had developed less than four or more than 25 follicles that were wider than 14 mm on the day of hCG administration. Significantly longer time and higher doses of pethidine were required for the procedure of flushing.895 [Evidence level 1b]

The use of follicle flushing in women with less than three follicles has not been evaluated but it may be useful for ensuring that oocyte yield is maximised.


Women undergoing transvaginal retrieval of oocytes should be offered conscious sedation because it is a safe and acceptable method of providing analgesia. [A]

The safe practice of administering sedative drugs published by the Academy of Medical Royal Colleges should be followed. [D]

Women who have developed at least three follicles before oocyte retrieval should not be offered follicle flushing because this procedure does not increase the numbers of oocytes retrieved or pregnancy rates, and it increases the duration of oocyte retrieval and associated pain. [A]


Further research is needed to evaluate the effect of general anaesthesia on oocyte retrieval and outcome of in vitro fertilisation treatment, taking into account patient preference.

12.8. Sperm recovery

Spermatozoa can be retrieved from both the epididymis and the testis using a variety of techniques with the intention of achieving pregnancies for couples where the male partner has obstructive or nonobstructive azoospermia. Sperm recovery is also used in ejaculatory failure and where only non-motile spermatozoa are present in the ejaculate (see Section 6.3) Ejaculatory failure is not uncommon on the day of egg collection and is usually caused by anxiety.

Surgically collected sperm in azoospermia are immature (because they have not traversed the epididymus) and have low fertilising ability with standard IVF. It is therefore necessary to use ICSI. Sperm recovery for ICSI has made it possible for infertile men to father children who are genetically their own.

Surgical techniques for sperm retrieval from the epididymis or the testis include:

  • percutaneous epididymal sperm aspiration (PESA)
  • testicular sperm aspiration (TESA), which is also described as testicular fine needle aspiration (TEFNA)
  • testicular sperm extraction (TESE) from a testicular biopsy
  • microsurgical epididymal sperm aspiration (MESA).

In obstructive azoospermia, sperm can usually be obtained from the epididymis (PESA or MESA) and from the testis (TESA or TESE). In some men, sperm can be recovered from naturally occurring spermatoceles by percutaneous puncture.

In nonobstructive azoospermia, sperm needs to be obtained directly from the testis by aspiration (TESA) or biopsy (TESE). The chance of finding sperm is reduced. PESA and TESA can be performed under local anaesthesia in an outpatient clinic.896,897 PESA does not jeopardise future epididymal sperm retrieval.898

A systematic review that includes one RCT (n = 59) compared MESA to epididymal micropuncture with perivascular nerve stimulation techniques and aspiration in men with obstructive azoopsermia such as CBAVD. MESA achieved lower pregnancy (OR 0.19, 95% CI 0.04 to 0.83) and fertilisation rates (OR 0.16, 95% CI 0.05 to 0.48). Caution is required in the interpretation of this trial as the method of randomisation used was not reported clearly, nor was there any dropout or loss to follow-up reported.899 [Evidence level 1a]

PESA and TESA are two alternatives to MESA. MESA is more invasive, costly and technically more difficult but may be performed at the same time as correction of epididymal obstruction. In order to avoid subsequent scrotal surgery, cryopreservation of supernumerary spermatozoa during MESA should be undertaken.900 Facilities for genetic screening with a view to referral to preimplantation genetic diagnosis should be available in any sperm retrieval programme.901

The best method of extracting spermatozoa from the testicular tissue in nonobstructive azoospermia is uncertain. The relative merits of TESA and TESE using small (5-mm), multiple or large (10–15-mm) diameter biopsies is unknown.902–906 Compared with TESE, TESA has a reduced rate of sperm recovery but is less invasive.907–910 [Evidence level 3]

Failure rates of retrieval

Reported failure rates of sperm retrieval vary with study and with technique (see Table 12.1). A further complication is added by the inconsistent method of reporting (for example, per attempt, per patient, or per couple).

Table 12.1. Failure rates of sperm retrieval.

Table 12.1

Failure rates of sperm retrieval.

In nonobstructive azoospermia, testicular size, plasma FSH levels and testicular histology are related to spermatogenesis but they cannot be relied upon to exclude the presence of any spermatozoa within the testis.901,903,911–919 The quality of the sperm retrieved vary widely among aetiological groups, but are of no value in predicting fertilisation or pregnancy rates, or the embryo cleavage rate following PESA/ICSI cycles.920

Clinical outcomes of using surgically recovered sperm (success rates of epididymal, testicular, or ejaculate spermatozoa)

Epididymal and testicular spermatozoa yield similar fertilisation, cleavage and ongoing pregnancy rates using ICSI927,928 and are both successful for establishing pregnancies.915,922 Some authors report these success rates as being lower than those achieved by spermatozoa from the ejaculate. One study929 found that the normal fertilisation rate was significantly higher with ejaculated spermatozoa than with epididymal or testicular spermatozoa but no differences were observed with regard to embryo quality, the percentages of transfer after ICSI and the clinical pregnancy rates in the three groups of women. However, another study898 showed that the outcome of PESA–ICSI treatment compares favourably with that of ICSI using ejaculated spermatozoa. One study896 also found that the results of PESA–TESA were similar to ejaculate sperm. [Evidence level 3]

Another study930 found that the normal fertilisation rates with testicular and MESA spermatozoa did not differ significantly from each other but, with testicular spermatozoa, the rate was significantly lower than that obtained with ejaculated spermatozoa and ICSI in matched couples. [Evidence level 3] Spermatozoa can be retrieved from the testis in couples in whom epididymal aspiration failed.901,928,931 When spermatozoa cannot be recovered by one technique another one can be employed, for example, TESE after MESA.922 Testicular spermatozoa can be successful in achieving fertilisation and pregnancies for couples in whom epididymal aspiration failed.901,916 However, some studies report fertilisation or pregnancy rates lower than those achieved with epididymal spermatozoa. For example, one study901 found a transfer rate lower with TESE than with epididymal spermatozoa but there was little difference in pregnancy rate using epididymal or testicular spermatozoa. Also, the spermatozoa could not be frozen and saved for use in future cycles. PESA, MESA or TESE and ICSI are effective in men with CBAVD and in those with failed reversal of vasectomy.923,928,932 [Evidence level 3]

Variation in outcome using testicular sperm in nonobstructive azoospermia compared with obstructive azoospermia has been demonstrated by various studies.933–935 Results in nonobstructive azoospermia are generally inferior.

Testicular sperm cryostorage

Cryopreservation of spermatozoa does not negatively influence the outcome. Various studies have shown that the fecundity rate, clinical pregnancy rate, overall rate of clinical pregnancy rate per embryo transfer or clinical abortions after ICSI using cryopreserved or fresh surgically retrieved spermatozoa are not significantly different.901,927,936 In one study,901 the only significant factor appeared to be the age of the woman. [Evidence level 3] Using cryopreserved testicular sperm (cryo-TESE) for ICSI is an effective and successful approach for the treatment of severe testicular insufficiency.921 Because cryopreservation of spermatozoa has many additional advantages (for example, in comparison to the use of native testicular sperm with the necessity of repetitive testicular biopsies), it is routine in the performance of MESA–ICSI and TESE–ICSI.921,927 Testicular tissue which is intentionally obtained well before any planned ICSI cycle and cryopreserved could then serve as an efficacious sperm source in a subsequent ICSI cycle. This approach should be an alternative to repeated testicular tissue sampling and the availability of spermatozoa is assured before the initiation of ovulation induction. This tissue can be harvested at the same time as diagnostic biopsy, thereby minimising the number of surgical procedures.937

A retrospective consecutive case series938 compared the results of ICSI with fresh and with frozen-thawed epididymal spermatozoa obtained after MESA in 162 couples suffering from infertility because of CBAVD, failed microsurgical reversal for vasectomy or postinfectious epididymal obstruction, irreparable epididymal obstruction, ejaculatory duct obstruction or anejaculation. Overall, 176 MESA procedures were performed in the male partners, followed by 275 ICSI procedures with either fresh (n = 157) or frozen-thawed (n = 118) epididymal spermatozoa. The overall pregnancy rate (as indicated by raised hCG levels) per ICSI cycle was significantly lower when frozen-thawed epididymal spermatozoa were used (26.3% versus 39.5%). However, no significant differences were found either in clinical or ongoing pregnancy rates, or in implantation rates, and there were no differences in pregnancy outcome. [Evidence level 3] In men suspected of having obstructive azoospermia with no work-up or an incomplete one, MESA was preferred as a method for sperm recovery because a full scrotal exploration can be performed and, whenever indicated, a vasoepididymostomy may be performed concomitantly. Recovery of epididymal spermatozoa for cryopreservation during a diagnostic procedure is a valid option in these patients since ICSI may be performed later or even in another centre using the frozen-thawed epididymal spermatozoa without jeopardising the ICSI success rate. In a retrospective study939 the authors aimed to determine whether fertilisation and implantation rates after ICSI with fresh or frozen-thawed testicular spermatozoa were comparable. They found that the fertilisation rate after ICSI with frozen-thawed testicular spermatozoa was significantly lower than with fresh testicular spermatozoa (71% versus 79%), the pregnancy rate was similar for both groups (38% and 27%), the implantation rate per transferred embryo was significantly lower in the frozen-thawed rather than in the fresh testicular sperm group (9% versus 25%), and the live birth rate per transferred embryo was higher in the group in which fresh testicular spermatozoa were used (19% versus 8%). [Evidence level 3]

A retrospective analysis of consecutive ICSI cycles940 compared the outcome of ICSI with fresh and frozen-thawed testicular spermatozoa in patients with nonobstructive azoospermia. No statistically significant differences were noted in any parameters examined between ICSI cycles with fresh or cryopreserved testicular spermatozoa from the same nine men and comparing all ICSI cycles performed (two-pronuclear fertilisation, embryo cleavage rates, implantation rates and clinical pregnancy rate). The delivery or ongoing pregnancy rate using fresh sperm was better but the difference was not statistically significant. Cumulative clinical pregnancy rates and ongoing pregnancy rates per testicular sperm extraction procedure were 36% and 24%, respectively. [Evidence level 3]


Surgical sperm recovery before intracytoplasmic sperm injection may be performed using several different techniques depending on the pathology and wishes of the patient. In all cases, facilities for cryopreservation of spermatozoa should be available. [C]

12.9. Assisted hatching

Assisted hatching has been proposed as a method to disrupt the zona pellucida, which may facilitate and enhance implantation and pregnancy rates. A narrative review of four RCTs and three non-randomised controlled trials found considerable heterogeneity in study methodology, populations selected, indications and techniques of assisted hatching. It reported that assisted hatching might be suggested for women aged over 38 years, those with elevated day-three serum FSH and repeated IVF failures. Data from this review did not support generalised assisted hatching for all patients.941 [Evidence level 1b–2a]

The four RCTs from the previous review941 were included in a systematic review of 23 RCTs (2572 women) assessing the impact of assisted hatching on live birth, clinical pregnancy and implantation rates.942 [Evidence level 1a] This review showed that assisted hatching had no significant effect on live birth rate (OR 1.21, 95% CI 0.82 to 1.78; based on six RCTs, n = 523 women). However, there was an increase in clinical pregnancy rate with assisted hatching (OR 1.63, 95% CI 1.27 to 2.09, based on 19 RCTs, n =2175 women). This effect may be increased in a subgroup of women who had previously had one or more cycles of IVF or ICSI that did not result in a live birth (OR 2.33, 95% CI 1.63 to 3.34, based on four RCTs, n = 666 women). However, these results should be interpreted with caution because of the poor methodological quality of the included trials, with unclear methods of randomisation in 13 trials and inadequate concealment of allocation in 23 trials.

Multiple gestation

Monoamniotic multiple gestation may be increased in zona-manipulated cycles. The potential obstetric risks and complications of zona manipulation should be discussed with couples. In an anonymous survey of 42 IVF centres in the USA,943 143 pregnancies were ascertained from zona-manipulated cycles (ICSI, subzonal sperm injection, zona drilling and mechanical assisted hatching). A multiple gestation frequency of 16.1% was reported. There were five monoamniotic twin gestations (all of which resulted in live births), four being from manipulated cycles and one being from a non-manipulated cycle. There has also been one case report of conjoined twins in a triplet pregnancy after IVF and assisted hatching.944 [Evidence level 3]


Assisted hatching is not recommended because it has not been shown to improve pregnancy rates. [A]


Further research is needed to evaluate the effects of assisted hatching on live birth rates and long-term consequences for children born as a result of assisted hatching.

12.10. Embryo transfer techniques

Use of ultrasound

Ultrasound-guided embryo transfer is a complex intervention. Four RCTs945–948 and four quasi-RCTs949–952 comparing ultrasound-guided embryo transfer versus clinical touch embryo transfer were identified. [Evidence level 1b–2a]

We performed a meta-analysis using data from all eight studies. This showed a significant increase in pregnancy rates with routine ultrasound-guided embryo transfer (pooled OR 1.46, 95% CI 1.25 to 1.70, n = 3358 embryo transfers). When the quasi-RCTs were excluded, there was still a significant increase in pregnancy rates with routine ultrasound-guided embryo transfer (pooled OR 1.42, 95% CI 1.17 to 1.73, n = 2051 embryo transfers). Overall, the meta-analyses suggest that use of ultrasound at the time of embryo transfer increases pregnancy rates. However, there was clinical heterogeneity among different groups of women and in the specific role of ultrasound in each trial. [Evidence level 1a]

Day two to three versus day five to six transfers

This has been the subject of a systematic review.953 A single quasi-randomised trial showed no difference in live birth rates between day 2/3 transfer and blastocyst transfer on days 5/6 (OR 1.59, 95% CI 0.80 to 3.15). A meta-analysis of the results of four trials also failed to show any advantage associated with day 5/6 transfers (combined OR 0.86, 95% CI 0.57 to 1.29). It is not possible to perform an intention-to-treat analysis for blastocyst transfer and so the results of these studies may be biased. [Evidence level 1a]

Four new RCTs were identified.954–957 Results from these trials were combined with those from the earlier studies. A new meta-analysis showed the following results. [Evidence level 1a]

  • Pregnancy and live birth rates per ovum pick up (that is, intention to treat analysis) (OR 1.08, 95% CI 0.94 to 1.25) and embryo transfer (OR 0.92, 95% CI 0.64 to 1.32) are similar in the two groups, suggesting no difference between the groups.
  • Pregnancy rate per embryo transfer (combined OR 1.20, 95% CI 1.04 to 1.38, based on 14 RCTs) and live birth rate per embryo transfer (combined OR 1.41, 95% CI 1.0 to 1.98, based on five RCTs) are higher in the day 5/6 transfer group.
  • Some caution should be exercised in interpreting the results of these meta-analyses as combining cycles as opposed to women can affect the precision of the results and widen the confidence intervals.
  • Day 5/6 transfers appears to offer no advantage over day 2/3 transfers in terms of increased pregnancy and live birth rates per cycle started. The apparent advantage in terms of pregnancy/live birth rate per embryo transfer at 5/6 days may be achieved at the cost of a number of women who do not proceed to embryo transfer.

Type of catheter

Seven RCTs have been identified comparing a number of different catheters.958–964 The results of these trials suggest that the choice of embryo transfer catheter can affect pregnancy rates. In particular, data from large trials suggest that certain types of soft catheter are more effective that other types of catheter. [Evidence level 1b] Data from the various studies could not be aggregated due to significant clinical heterogeneity and differences between individual catheters.

Endometrial thickness

Endometrial thickness and endometrial pattern are the two anatomical parameters suggested to evaluate the endometrium by ultrasound. The role of endometrial thickness as a single factor in predicting pregnancy following IVF is controversial. A narrative review of 27 cohort and observational studies found insufficient data for an association between endometrial thickness and the probability of conception during IVF cycles. The mean endometrial thickness for conception and non-conception cycles were similar, ranging from 8.6 mm to 12.0 mm. There was also no case in which the endometrial thickness was less than 5 mm which resulted in pregnancy (based on 1605 cycles in 13 studies).965 [Evidence level 2b–3] In such circumstances, the IVF cycle should be abandoned and consideration given to preparing the endometrium with exogenous hormones before a frozen embryo replacement cycle. Implantation and pregnancy rates were reported to be significantly reduced in women with an endometrial thickness of greater than 14 mm on the day of hCG administration in an IVF programme.966 [Evidence level 2b] One study reported that reduced endometrial thickness had only a marginal effect on the probability of achieving a pregnancy rates with assisted reproduction.967 [Evidence level 2b]. However, no significant correlation was found between endometrial volume and thickness and occurrence of pregnancy during IVF treatment in two studies.968 [Evidence level 3] 969 [Evidence level 2b]

Bed rest versus no bed rest

One RCT (n = 182) found no significant difference in pregnancy rate per embryo transfer between 20 minutes of bed rest versus 24-hours of bed rest following embryo transfer (24% versus 23.6%), spontaneous miscarriage rate (19% versus 18%) and multiple pregnancy rate (14% versus 13.6%).970 [Evidence level 1b] Another RCT (n = 211) assessed the role of fibrin sealant for embryo transfer and found no significant difference in implantation and pregnancy rates when both study and control groups were instructed to routine activities without any bed rest after embryo transfer. There was no group that was assigned to bed rest.971 [Evidence level 3]


Women undergoing in vitro fertilisation treatment should be offered ultrasound guided embryo transfer because this improves pregnancy rates. [A]

Embryo transfers on day 2 or 3 and day 5 or 6 appear to be equally effective in terms of increased pregnancy and live birth rates per cycle started. [B]

Replacement of embryos into a uterine cavity with an endometrium of less than 5 mm thickness is unlikely to result in a pregnancy and is therefore not recommended. [B]

Women should be informed that bed rest of more than 20 minutes’ duration following embryo transfer does not improve the outcome of in vitro fertilisation treatment. [A]


Further research is needed to evaluate the effect of cleavage (day 2 or 3) and blastocyst (day 4 or 5) stage methods of embryo transfer on live birth rates.

Further research is needed to evaluate the effects of different types of embryo transfer catheters on pregnancy rates.

12.11. Luteal support

Progesterone versus no support in non-downregulated cycles

A 1988 meta-analysis of five RCTs found no significant difference between luteal-phase progesterone support in non-downregulated IVF cycles and no such support in pregnancy rate (OR 1.25, 95% CI 0.93 to 1.66) in women undergoing IVF or GIFT after ovarian stimulation with clomifene and hMG.972 [Evidence level 1a]

Human chorionic gonadotrophin versus no treatment/human chorionic gonadotrophin versus progesterone in downregulated cycles

A meta-analysis of 18 RCTs showed significantly higher pregnancy rate per cycle in women treated with hCG compared with no treatment (OR 1.9, 95% CI 1.3 to 3.1, based on five RCTs) when used with GnRH agonist.866 [Evidence level 1a] A significantly higher pregnancy rate per cycle was also found in groups treated with intramuscular or oral progesterone (progestagen) compared with no treatment (OR 1.2, 95% CI 1.0 to 1.7, based on eight RCTs). In three RCTs that compared hCG luteal support with intramuscular or oral progesterone, pregnancy rate per cycle was significantly higher in women treated with hCG compared with progesterone (OR 2.0, 95% CI 1.1 to 3.9). However, this effect was to due a difference in the effectiveness of hCG and oral (rather than intramuscular) progesterone. There was no significant difference in spontaneous abortion rate between women given luteal support or no support (OR 0.8, 95% CI 0.4 to 1.7, based on seven RCTs). The overall incidence of OHSS with hCG was 5% (n = 220) versus 0% (n = 193) with progesterone or no treatment.866 [Evidence level 1a]

Another meta-analysis973 of 30 RCTs showed that intramuscular hCG significantly improved clinical pregnancy rate when compared with no treatment (RR 2.72, 95% CI 1.56 to 4.90, based on four RCTs). Intramuscular progesterone significantly improved clinical pregnancy rate (RR 2.38, 95% CI 1.36 to 4.27, based on three RCTs), ongoing pregnancy rate (RR 3.8, 95% CI 1.42 to 11.38, based on three RCTs) and delivery rate (RR 5.50, 95% CI 1.25 to 35.53, based on one RCT) when used with long GnRH agonist protocol. Intramuscular hCG significantly improved clinical pregnancy rate (RR 8.36, 95% CI 1.44 to 173.74, based on four RCTs) and ongoing pregnancy rate (RR 7.43, 95% CI 1.22 to 156.64, based on four RCTs) when compared with oral progesterone used in a short GnRH agonist protocol.973 [Evidence level 1a]

The same meta-analysis reported that intramuscular progesterone significantly improved clinical pregnancy rate (RR 1.33, 95% CI 1.02 to 1.75, based on five RCTs) and delivery rate (RR 2.06, 95% CI 1.48 to 2.88, based on two RCTs) when compared with vaginal progesterone. There were no significant differences in fertility outcomes when comparing: vaginal progesterone with no treatment; different doses of progesterone; intramuscular progesterone with oral progesterone; intramuscular hCG with oral progesterone in both long and short GnRH agonist protocols; intramuscular hCG with intramuscular progesterone; oestrogen plus progesterone with progesterone only in long GnRH agonist protocols; hCG plus progesterone with vaginal progesterone in long and short GnRH agonist protocols; intramuscular progesterone plus oestrogen with hCG. Given the increased risk of OHSS associated with hCG use, progesterone was favoured for luteal-phase supplementation with addition of oestrogen.973 [Evidence level 1a]

The review did not consider patient satisfaction. However in one of the RCTs, 4/30 women discontinued treatment because of their inability to administer intramuscular progesterone.

The two meta-analyses show inconsistency in the relative effectiveness of the different drugs and routes of administration for luteal support. Although the meta-analyses involved a total of 18 and 30 RCTs, respectively, most of the detailed comparisons were based on meta-analyses of very few RCTs.

Patient satisfaction was assessed as part of a non-randomised multicentre study conducted in the USA.974 [Evidence level 3] Women were asked to report their preferences between vaginal progesterone and intramuscular progesterone; 94% of the women found vaginal progesterone easier to use, and 84% preferred vaginal progesterone to intramuscular progesterone.


Women who are undergoing in vitro fertilisation treatment using gonadotrophin-releasing hormone agonists for pituitary downregulation should be informed that luteal support using human chorionic gonadotrophin or progesterone improves pregnancy rates. [A]

The routine use of human chorionic gonadotrophin for luteal support is not recommended because of the increased likelihood of ovarian hyperstimulation syndrome. [A]


Further research is needed to compare the effectiveness (including patient satisfaction) of different drugs and routes of administration for luteal support during in vitro fertilisation using gonadotrophin-releasing hormone agonist cycles.



Ovarian size may not correlate with severity of OHSS in cases of assisted reproduction because of the effect of follicular aspiration. Nevertheless, recording ovarian measurements of ovarian size is not currently considered useful as a prognostic indicator nor as an indicator of the stage of the disease.861

Copyright © 2004, National Collaborating Centre for Women’s and Children’s Health.

Apart from any fair dealing for the purposes of research or private study, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Cover of Fertility
Fertility: Assessment and Treatment for People with Fertility Problems.
NICE Clinical Guidelines, No. 11.
National Collaborating Centre for Women's and Children's Health (UK).
London (UK): RCOG Press; 2004 Feb.

NICE (National Institute for Health and Care Excellence)

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...