Discussion and conclusions

Summary of main findings

A total of 1578 couples consented, 619 of whom were randomised (310 to the fresh-embryo transfer arm and 309 to the freeze-all arm). Most non-randomisations (n = 959) were owing to the non-availability of three good-quality embryos (n = 476). Of those randomised, 117 (19%) couples did not adhere to their allocated intervention: 96 (31.3%) in the freeze-all arm and 21 (6.8%) in the fresh-embryo transfer arm. Non-adherence was higher in the freeze-all arm than in the fresh-embryo transfer arm, with the most common reason being patient choice.

The ITT analysis showed that the healthy baby rate was 20.3% in the freeze-all arm and 24.4% in the fresh-embryo transfer arm (RR 0.84, 95% CI 0.62 to 1.15). Similar results were obtained from CACE analysis (RR 0.77, 95% CI 0.44 to 1.10), per-protocol analysis (RR 0.87, 95% CI 0.59 to 1.26) and as-treated analysis (RR 0.91, 95% CI 0.64 to 1.29). There was no statistical difference in the healthy baby rate across age groups (< 35, 35 to < 40 and > 40 years), the number of previous embryo transfers (0 or ≥ 1), whether it was cleavage or blastocyst transfer, or whether one or two embryos were transferred.

There was no evidence of a difference in live birth rate (28.3% vs. 34.3%; RR 0.83, 99% CI 0.65 to 1.06) or clinical pregnancy rate (33.9% vs. 40.1%; RR 0.85, 99% CI 0.65 to 1.11) between the freeze-all arm and the fresh-embryo transfer arm.

There were no statistical differences between the two arms in any of the obstetrics and perinatal outcomes (i.e. hypertensive disorders of pregnancy, antepartum haemorrhage, preterm delivery, very preterm delivery, onset of labour, mode of delivery, low birthweight, high birthweight, low birthweight for gestational age, high birthweight for gestational age and congenital anomalies).

The risk of ovarian hyperstimulation was 3.6% in the freeze-all arm and 8.1% in the fresh-embryo transfer arm (RR 0.44, 99% CI 0.15 to 1.30). There were 30 reported AEs, but these were not related to the intervention.

A total of 88.6% embryos survived the freezing–thawing process.

There was no statistical difference in STAI scores for male participants (MD 0.1, 99% CI –2.4 to 2.6) and female participants (MD 0.0, 99% CI –2.2 to 2.2) between the arms.

Following adjustment for minimisation criteria, the mean post-randomisation treatment cost (inclusive of OHSS) per woman randomised was £1395 (95% CI £1294 to £1505) in the fresh-embryo transfer arm and £1576 (95% CI 1514 to £1642) in the freeze-all arm. The mean between-group difference was £181 (95% CI £60 to £292). Based on the estimated difference in the healthy live birth rate (–0.039, 95% CI –0.101 to 0.027), fresh-embryo transfer was found to dominate frozen-embryo transfer, being, on average, less costly and more effective. Considering the joint uncertainty surrounding the estimated differences in costs and effects, the probability of fresh-embryo transfer being preferred on the grounds of cost-effectiveness was > 89% across all willingness-to-pay thresholds per additional healthy live birth.

When ANC and delivery costs were included in the cost-effectiveness analysis, the freeze-all strategy was, on average, less costly, owing to a smaller number of pregnancies and live births (MD –£75, 95% CI –£623 to £461). However, fresh-embryo transfer retained a higher probability of being cost-effective compared with frozen-embryo transfer above a willingness-to-pay threshold of £1921 per additional healthy live birth. Furthermore, when cumulative costs and outcomes associated with the transfer of the remaining frozen embryos were simulated using a Markov model, fresh-embryo transfer was found to be less costly and more effective than frozen-embryo transfer, even when including the ANC and delivery costs. The same pattern of results was observed when using live births as the measure of effectiveness.

The difference in treatment costs was found to be sensitive to the application of more conservative costs for monitoring ultrasound scans prior to frozen-embryo transfer, but the overall cost-effectiveness findings remained stable, with fresh-embryo transfer retaining a substantially higher probability of being cost-effective than frozen-embryo transfer.

Update and comparison with existing literature

The E-Freeze trial was planned in 2014, awarded funding in 2015 and started recruitment in 2016. Several trials40–46 across the world were conducted and published during this time (i.e. 2016–20) comparing freezing all embryos, followed by frozen-embryo transfer, with fresh-embryo transfer. Two trials report on hyper-responders (i.e. those who are at high risk of OHSS),40,41 and five reported on those who were predicted as normal responders (i.e. not at high risk for OHSS).42–46

Table 29 summarises the trials reporting on normal responders that were published after the E-Freeze trial began. We present the aggregated meta-analysis on these trials on key clinical outcomes with and without incorporating data from E-Freeze trial. For the comparison of the results and data from the E-Freeze trial with the existing literature, we have included only those trials reporting on normal responders, the populations for which were similar to the population of the E-Freeze trial.

The two trials on hyper-responders40,41 reported that the freeze-all approach improves the live birth rate and reduces the risk of OHSS in those who are hyper-responders.

The five trials on normal responders reported from Viet Nam,42 China, 43,44 Denmark45 and the Netherlands.46 Our results are consistent with three of these trials,42,43,45 but are in contrast with the others.44,46 Wei et al.44 suggest that the singleton live birth rate is higher with freeze-all, followed by frozen-embryo transfer, and Wong et al.46 showed that the live birth rate was significantly lower when freezing all embryos rather than using fresh embryos. There are differences in population, outcome measures and the timing of randomisation in each of the trials (Table 30), which may account for these differences.

The outcome healthy baby rate was not reported by any other trial. The closest comparison was singleton live birth, reported by Wei et al.44 Hence, it is not possible to compare the primary outcome measure reported by the E-Freeze trial with any other studies in the literature.

Live birth rate

The live birth rate in our trial was 28.3% in the frozen-embryo transfer arm and 34.3% in the fresh-embryo transfer arm. Although there were no statistically significant differences between the two arms, these figures were similar to the live birth rates reported by the two trials from Europe45,46 and the trial from Viet Nam.42 However, the rates are much lower than those of both of the trials reported from China.43,44 This could be because the trials reporting from China had an upper age limit of 35 years and, therefore, included patients with a better prognosis.

The combined data from these five trials showed no difference in the outcome of live birth between the two arms, which is similar to the results of the E-Freeze trial (Figure 16).

FIGURE 16 Meta-analysis of existing trials,42–46 with and without the E-Freeze trial, for live birth rate. (Figure not reproduced here; see the original article.)

Based on nearly 5246 women randomised to fresh-embryo transfer compared with frozen-embryo transfer, an aggregate-data meta-analysis did not seem to favour either fresh-embryo transfer or a strategy of freezeing all embryos, followed by frozen-embryo transfer, with a combined RR of 1.01 (99% CI 0.78 to 1.31). An updated meta-analysis, including randomised data from an additional 616 women from the E-Freeze trial, does not appear to result in a convincing change in the direction, size or precision of the overall effect, with a combined RR of 0.97 (95% CI 0.77 to 1.24).

Ovarian hyperstimulation syndrome

The risk of OHSS reported in the E-Freeze trial was 3.6% in the freeze-all arm and 8.1% in the fresh-embryo transfer arm. Most cases of OHSS were mild, with moderate to severe OHSS in only 1.6% in the freeze-all arm and 5.8% in the fresh-embryo transfer arm. This is higher than the rate quoted for moderate to severe OHSS in national statistics.6 This difference is because of better ascertainment in the trial setting; it is well known that cases of OHSS are not reported in clinical practice.49

The total OHSS figures were higher than those reported by other trials. This could be because we also reported mild OHSS, whereas other trials reported only moderate and severe OHSS. Combined data from the five trials reporting on OHSS42–46 showed a statistically significant reduction in OHSS when all embryos were frozen compared with fresh-embryo transfer. The addition of data from the E-Freeze trial did not change the direction or magnitude of these figures, but increased the precision (Figure 17).

FIGURE 17 Meta-analysis of existing trials,42–46 with and without the E-Freeze trial, for OHSS. (Figure not reproduced here; see the original article.)

Miscarriage

The risk of miscarriage per couple randomised was 14.3% in the frozen-embryo transfer arm and 12.9% in the fresh-embryo transfer arm. The corresponding figures for risk of miscarriage per pregnancy were 31.7% and 26.0%, respectively. Miscarriage rates were slightly higher in the E-Freeze trial than in other studies (ranging from 9.9% to just over 25% per pregnancy42–45). We were unable to explain this higher rate of miscarriage. To explore the higher risk of miscarriage, we undertook a post hoc analysis of the miscarriage rate per centre; there were no significant differences between the five clinics that contributed most data. The combined data from other trials showed no difference in miscarriage rates between trial arms, which is similar to the results reported in the E-Freeze trial.42–45 Wong et al.46 did not report on miscarriage per pregnancy; hence, their data are not included in the meta-analysis graph. The addition of E-Freeze to existing data does not change the direction, magnitude or precision of effect (Figure 18).

FIGURE 18 Meta-analysis of existing trials,42–45 with and without the E-Freeze trial, for miscarriage rate. (Figure not reproduced here; see the original article.)

Obstetric and perinatal complications

There was no statistical difference in obstetric and perinatal complication between arms in the E-Freeze trial. Observational studies and meta-analysis of published data50–52 suggested that the risk of pre-eclampsia and large for gestational age (LGA) babies is increased in pregnancies that are a result of frozen-embryo transfer. The numbers of each individual complication were too small to draw any definite conclusions from this trial alone, or from any of the other existing trials individually.42–46 In comparison with the risk in the general population (see Table 30), not all risks were higher in IVF pregnancies, irrespective of whether they were the result of fresh-embryo transfer or frozen-embryo transfer. Although this is reassuring, it is in contrast to the previous findings from systematic review of observational data7 and could be owing to small numbers of each complication in our trial.

Natural compared with hormone replacement therapy frozen-embryo transfer

Most of the frozen-embryo transfers in the E-Freeze trial were undertaken in hormonally mediated cycles. By contrast, worldwide,53 45% of cycles are natural cycles. This can be explained by the fact that participants in the E-Freeze trial were eager to receive treatment and wanted a planned date for their frozen-embryo transfer, after having their treatment delayed by participation in the trial. By contrast, the worldwide data are based on observational, non-randomised data and are more likely to be from patients undergoing transfer of surplus frozen embryos after fresh-embryo transfer has failed, rather than frozen-embryo transfer as their first treatment.

The two Chinese trials43,44 had a large proportion of participants who underwent endometrial preparation for frozen-embryo transfer using natural cycles. The two European trials45,46 and the Vietnamese trial42 had the largest number of patients undergoing hormonally mediated hormone replacement therapy (HRT) cycles.

Hormonally mediated frozen-embryo transfer is more convenient for the clinic and patients, as the date of thaw and transfer can be planned in accordance with the workload of the clinic and at the convenience of the staff and patients. Recently, some concerns have been raised that pregnancies following HRT-mediated frozen-embryo transfer may be at a higher risk of complications than those following a natural cycle.54

Non-adherence

Non-adherence in the freeze-all arm was very high in this trial. Other trials have also shown non-adherence to the allocated intervention, but at much lower levels than those seen in the E-Freeze trial. The highest level was reported by Shi et al.43 (freeze-all, 18%; fresh-embryo transfer, 15%). However, non-adherence was reported in both arms, whereas in the E-Freeze trial non-adherence was predominantly seen in the freeze-all arm (31.3%). This level of non-adherence occurred despite the fact that the trial was specifically designed to reduce non-adherence, with consent being reconfirmed just before randomisation and information being provided throughout the process. There could be two reasons for this non-adherence. The funding of IVF treatment is limited across the UK, especially in England, where most of the participating centres were based, and > 60% of treatments in England are funded by the patients themselves. Where funding is available, only one cycle of treatment is funded by most CCGs. Hence, there will be apprehension about freezing all embryos among clinicians, scientists and patients, especially when there are only one or two embryos. There is always a fear that the embryos may not survive the freezing–thawing process, leading to loss of funding for the only funded treatment. Our data showed that, on average, 86% of embryos survived the freezing–thawing process across the participating clinics; hence, the survival rate is far below 100%.

Although there was an intention and acceptance from patients to be randomised to either arm at the time of consent, this did not translate to real practice, with a noticeable preference seen for fresh-embryo transfer. Studies from two different parts of the world55,56 have shown that patients would accept the intervention of freeze-all, followed by frozen-embryo transfer, if freeze-all reduces the side effects and has at least equal, if not better, success rates. Both conditions must be fulfilled to accept the delay.

Economic analysis

The few published economic evaluations of freeze-all compared with fresh-embryo transfer have produced mixed findings. Roque et al.57 used observational data from a private centre in Brazil to compare the cost per ongoing pregnancy among patients receiving treatment with each strategy. Although the total cost of treatment was higher per patient undergoing the freeze-all approach, the average cost per pregnancy was lower, given the substantially higher pregnancy rate in the non-randomised cohort (39.7% vs. 31.1%). Thus, Roque et al.57 concluded that a freeze-all strategy was a cost-effective option compared with fresh-embryo transfer.

In an economic evaluation carried out using data from a RCT conducted in Viet Nam, Le et al.58 reported higher mean costs in the freeze-all arm than in the fresh-embryo transfer arm over a full cycle, including the transfer of all embryos obtained from a single, controlled, ovarian hyperstimulation cycle. The live birth rate was also slightly higher in the freeze-all arm than in the fresh-embryo transfer arm (48.6% vs. 47.3%), but the average cost per live birth was higher. In an incremental analysis, the additional cost per additional live birth was estimated to be €30,997 per additional live birth (freeze-all vs. fresh-embryo transfer). The probability of cost-effectiveness for the freeze-all approach remained < 60% irrespective of the willingness to pay per additional live birth. Based on these results, Le et al.58 concluded that the freeze-all approach did not constitute a cost-effective use of resources.

Our findings of increased treatment costs with the freeze-all approach, compared with the fresh-embryo transfer approach, are consistent with those in the published studies.58 However, directionally, both the health baby rate and the live birth rate favoured fresh-embryo transfer in the E-Freeze trial, leading to a lower probability of the freeze-all approach being considered cost-effective. This pattern remained when the subsequent transfer of the remaining embryos was simulated, both with and without the inclusion of pregnancy and delivery costs.

A further economic analysis was carried out using data from a multinational RCT comparing a personalised embryo transfer strategy, guided by endometrial receptivity, with frozen-embryo transfer and fresh-embryo transfer.59 The personalised embryo transfer strategy involved the freezing of all viable blastocysts and the use of an array test to predict each individual’s optimum window for implantation in a subsequent frozen-embryo transfer with hormonal replacement. Similarly, this study found the cost per embryo transfer to be higher in the freeze-all arm than in the fresh-embryo transfer arm, and reported a slightly lower live birth rate per first frozen-embryo transfer than for single fresh-embryo transfer. The personalised embryo transfer strategy generated the highest live birth rate, but incurred higher treatment costs than both the fresh-embryo transfer and the freeze-all arm. No such treatment arm was included in E-Freeze for comparison.

Strengths

To the best of our knowledge, this is the first and only trial in the UK comparing fresh-embryo transfer with a policy of electively freezing all embryos, followed by subsequent frozen-embryo transfer. In total, 18 clinics participated, of which 13 recruited participants.

The recruited trial population is consistent with what would be expected given the NHS-funded treatment criteria (i.e. age < 40 years, BMI < 30 kg/m², non-smoker, no previous children).3,60 The occurrence rate of aetiological causes of infertility in this trial were similar to those that have been reported in previous population-based studies for other aetiologies of infertility,61,62 except unexplained infertility, the proportion of which was higher in this trial. The duration of infertility was > 2 years for most participants, which is the minimum duration criterion for unexplained infertility used for accessing NHS-funded IVF, as per national guidance.3,60

The E-Freeze trial was a pragmatic trial (i.e. except for randomisation, all other process pre and post randomisation were as per local protocols) involving multiple clinics across the UK. The participants were recruited from both the NHS and private clinics, given that > 60% of IVF in the UK is privately funded by couples themselves. The pragmatic nature of the trial provides a true reflection of what happens in clinical practice.

A healthy baby is a unique outcome chosen for this trial because it encompasses efficacy and safety together. All other trials on this topic have reported on live birth rate or ongoing pregnancy rates.40–46 As all complications in pregnancy and delivery have an impact on the short- and long-term health of an individual, the E-Freeze trial was unique in taking a holistic view of the infant, rather than just focusing on a live birth.

A detailed economic analysis is presented, based on all costs incurred to the NHS as well as participants, from post randomisation to the delivery of the baby. Cost analysis also included modelling to incorporate the longer-term costs in each of the trial arms of a heathy baby if all remaining embryos were used. Real-time data from clinical record forms and participant questionnaires were used for cost analysis. The response rate for the questionnaires was > 70% from both partners. Of the seven existing trials across the world on this topic,40–46 only one Vietnamese trial has reported on an economic analysis conducted alongside the trial,58 stating similar conclusions to ours; however, the Vietnamese trial did not have a societal perspective and was not performed in an NHS setting.

The E-Freeze trial is the only trial on this topic exploring the emotions of both the male and the female partners. It is becoming clear that, along with the clinical outcome, patient perceptions are equally important for any process or procedure to be put in place. This is especially important when one is recommending radical changes to the treatment, such as how IVF treatment is delivered, with delay and uncertainty associated with it. As with the economic questionnaire, the emotions questionnaire had a good return rate of > 70% at both of the time points at which they were administered (i.e. at consent and the actual intervention). Most literature on emotions relates to women only; this trial was unique in exploring the emotions of both the male and the female partner’s STAI scores separately. None of the other trials on this topic has evaluated this aspect.

Although a lot of assumptions were made in the sample size calculations because of the unique nature of the primary outcome, the results suggested that most were correct. We anticipated that the healthy baby rate in our trial population would be towards the lower end of the CI, around 25%, taking into account the higher risk of preterm delivery and babies with a low birthweight for their gestational age following IVF. The healthy baby rate was 24.4% in the trial, which was in line with this assumption. We also anticipated that, of those couples who consented, 50% would not obtain three good-quality embryos. In the trial, 49.6% of those consented did not have at least three good-quality embryos and, hence, were not randomised. Therefore, the trial design and sample size calculation were robust, based on the assumptions, which were very close to what was seen in reality.

Limitations

The trial did not reach its predetermined sample size of 1086 and, therefore, lacked the power to provide a definitive answer to the research questions. This was compounded by the bias introduced from non-adherence to the allocated intervention, especially in the freeze-all arm, of up to 31%. The additional analysis by per protocol, as treated and CACE did not change any results. There was no difference in the baseline characteristics (i.e. demographic and clinical, pre and post randomisation) or primary outcome rate between those who adhered to the allocated intervention and those who did not adhere to the allocated intervention. The consistent result across all analysis types indicates that it is likely that non-adherence did not alter the result. However, this cannot be definitively concluded.

There was a significant drop in the number of participants from consent to randomisation, with most participants not reaching three good-quality embryos. This was a result of the broad inclusion criteria, which allowed patients to take part in the study regardless of ovarian reserve, rather than including only those with good ovarian reserve. This was agreed by the co-investigator group at the outset, after much debate, acknowledging that ovarian reserve tests predict only the number of eggs, not their quality. A total of 30 participants had fewer than six eggs; they would not have been included if we had strict criteria based on ovarian reserve tests.

There was a change in practice during the trial, as clinics moved from embryo transfer on day 3 to embryo transfer on day 5. The reason for randomisation at day 3 in this trial was that embryo transfer would be as close to randomisation as possible. When the trial protocol was written, most embryo transfers occurred at day 3. However, as practice changed to transfer on day 5, there was a gap between randomisation and the actual intervention, which provided an opportunity for clinicians and participants to change their mind, contributing to non-adherence.

It has been suggested that freeze-all will benefit those with a large number of eggs, but we did not plan a subgroup analysis based on the number of eggs obtained a priori, as a large number of eggs are at a higher risk of OHSS, which was an exclusion criterion. There are now data available from other trials, published after the E-Freeze trial started recruiting, reporting that the strategy of freeze-all benefits those with a large number of eggs.40,41 This was revisited by the co-investigator group; however, it was agreed not to undertake a post hoc analysis because the number of patients with more than 15 eggs was small (45/616) and, therefore, it was unlikely to provide a clinically meaningful answer.

Some adverse birth outcomes (e.g. preterm delivery and low birthweight) can have a far-reaching impact on costs and child health outcomes. It was initially planned1 that modelling would be used to inform cost-effectiveness over an extended time horizon to capture the long-term costs and health outcomes for any infants born as a result of treatment. However, the number of infants with adverse birth outcomes in this trial was too small to inform a robust analysis of any expected differences in long-term outcomes. It was felt that this analysis would be best undertaken as part of an individual patient data (IPD) meta-analysis (IPD-MA), as explained in Implications for research.

Meaning of the study

For the strategy of freeze-all to be used as routine policy, it must be better in terms of safety, efficacy and cost, as it involves a delay for couples in getting to pregnancy. There is also extra work involved for clinics. Although there is a biological plausibility that freezing all embryos, followed by frozen-embryo transfer, will be better in terms of success rates, safety and, hence, costs for all couples undergoing IVF/ICSI, this has not been proven when tested in this trial or when the data from other trials were assembled. This is not dissimilar to other interventions in this field (e.g. personalised embryo transfer,59 endometrial scratch63 and preimplantation genetic testing64 for aneuploidy) that had a biological plausibility to improve success rates, but, when put to the stringent test of RCTs, were not proven to be more effective.

The health economic findings from the E-Freeze trial of a low probability of cost-effectiveness for the freeze-all approach compared with fresh-embryo transfer, in terms of NHS treatment costs per healthy baby and per live birth and higher costs incurred by participants, do not currently support the widespread use of the freeze-all approach in the NHS. However, further research to ascertain any differences in child health outcomes and costs between the alternative approaches is required (see Implications for research). Although the data did not reach statistical significance, there is a trend towards a reduction in the healthy baby, live birth and clinical pregnancy rates with the freeze-all approach compared with fresh-embryo transfer.

The absolute number of cases with OHSS was small in our trial. Although there is no statistically significant difference in any arm in the E-Freeze trial, the data collated from all trials together show a reduction in OHSS when freezing all embryos compared with fresh-embryo transfer. Hence, from a safety viewpoint, to reduce OHSS there could be a subgroup of participants who may still benefit from freezing all embryos, especially those at high risk of OHSS.

The level of non-adherence in the freeze-all arm suggests that patients still prefer fresh-embryo transfer to the freeze-all approach.

Implications for practice and policy

Based on the fact that the strategy of freezing all embryos, followed by frozen-embryo transfer, did not lead to a higher healthy baby rate or a higher live birth rate, and given it incurred slightly higher treatment costs, its widespread adoption as part of routine NHS practice is not supported by data from the E-Freeze trial. It should be advocated only when there is an indication that the freeze-all approach would be beneficial for the individual patient. There is no indication to recommend a change to current practice in the UK, which is fresh-embryo transfer and the freezing of spare embryos.

The NICE guidance does not specifically mention freeze-all, possibly because the last guidance was published in 2004, when freeze-all was not on the horizon. An update of the guidance was conducted in 2013, but this question was not looked at.3 It is possible that the next update of the NICE guidance will specifically look at this question, as practice of freeze-all has been increasing.65 Our results clearly show that freeze all is not a cost-effective strategy from a health-care perspective, despite accounting for the extra costs associated with OHSS in fresh-embryo transfer. In fact, the freeze-all approach is more costly, with no added benefit for normal responders.

The most common reason for non-adherence was patient choice. This indicates that the strategy is not ready to be accepted by patients. This is an equally important finding, as patients’ choices must be taken into consideration, with an equal weighting as that of other clinical evidence, when recommending a policy.

Implications for regulators

The Health Fertilisation and Embryology Authority regulates and licenses all clinics providing IVF treatment in the UK. HFEA has a traffic light system for treatment add-ons.66 Treatment add-ons are optional, additional treatments that patients may be offered on top of their routine fertility treatment, usually at an additional cost. The treatment add-ons claim to improve the chances of a live birth. The traffic light rating system consists of three colours: red, green and amber. Green indicates that there is evidence, in the form of high-quality RCTs, showing that a treatment add-on can safely improve the live birth rate for someone undergoing fertility treatment. The freeze-all approach is currently amber.67 An amber symbol is given for an add-on where there is ‘conflicting evidence’ to show that an add-on can improve live birth rates or that the add-on is safe for patients to use. The amber symbol ‘means that the evidence is not conclusive and further research is required, and the add-on should not be recommended for routine use’ (© HFEA. Reproduced with permission from HFEA66). Although there is evidence that the freeze-all approach is suitable for hyper-responders, for whom the risk of OHSS is high, there is no evidence that the chance of a live birth improves for normal responders. The lack of improvement in live birth rate in the freeze-all arm compared with the fresh-embryo transfer arm is clear from this trial, but adding the data from other trials confirms it.42–46

Although the E-Freeze trial did not reach an adequate sample size on its own, the sample size is adequate when you combine the data from the other five trials42–46 that were published during the conduct of this trial. The combined sample size for these trials is much larger than the sample size that was proposed for the E-Freeze trial. The fact that adding the E-Freeze trial data to the data from other trials does not alter the precision, magnitude or direction of the effect means that the evidence is now stable for the live birth rate. There is no need for further RCTs in this area, as they would be unlikely to change the outcome for women undergoing IVF/ICSI. Any trial that has a larger sample size than this combined sample size is unlikely to be conducted in the UK given the recruitment difficulties in the current trial and the fact that it will cost a huge amount of taxpayers’ money. With the combined data from the other trials,42–46 as shown in Figure 15, the freeze-all approach could be assigned the red symbol to stop the practice of freezing all embryos (a practice which has been advocated by some clinics68) for all patients, including those who are not at a high risk of OHSS.

Implications for funders for NHS in vitro fertilisation/intracytoplasmic sperm injection treatments

Difficulties in recruiting to in vitro fertilisation studies in the UK

Most IVF/ICSI treatment across the UK is funded by patients. Where it is funded by the NHS, the full entitlement reported in the NICE guidance3 (i.e. three full cycles of IVF/ICSI) is not available for most patients. To randomise patients to a clinical IVF trial in such a set-up is always going to be difficult, especially when the intervention causes a delay and extra costs, and is the only funded treatment that patients may receive. This is, possibly, one of the reasons for the large proportion of patients who were non-compliant. RCTs are the gold standard; for successful recruitment in any future clinical trials in the UK in the field of IVF/ICSI, NHS funding must improve to the level recommended by NICE.3

Implications for research

Further analyses of the E-Freeze trial data

Cumulative live birth rates

This trial evaluated the results from the first embryo transfer after randomisation only. However, most participants had any spare embryos frozen. The field of IVF is changing, and providers’ and patients’ perceptions of the procedure’s success are drifting from focusing on the results of the first embryo transfer to the cumulative results of a single ovarian stimulation cycle. We would like to look at the CLBR after all spare embryos are used. This is especially important as it is acknowledged that the CLBR is a more relevant and important outcome for both policy and the individual patient. This has been recognised by regulatory authorities for data reporting, and the most up-to-date data reporting includes the CLBR as the headline.5,69 There have been concerns raised that the CLBR may be lower if all embryos are frozen.46,70 Participants were asked to consent to long-term follow-up, to which the majority of participants agreed.

Follow-up of children born

The primary outcome of this study was the healthy baby rate; however, there are long-term implications. We would like to follow up children born from both kinds of embryo transfer to see if there is any impact on long-term outcomes.

Further analysis of State–Trait Anxiety Inventory scores

We plan to perform in-depth analysis of the STAI self-evaluation questionnaire to:

1. investigate if there was an association between baseline demographic and clinical characteristics and anxiety at the start and end of the IVF process
2. examine whether or not anxiety prior to IVF treatment is related to non-adherence to the allocated embryo transfer method.

We will also perform qualitative analysis on the free-text responses for both those participants who complied to their allocated intervention and those participants who did not to elicit the reasons behind non-compliance.

Further research questions raised

Individual patient data meta-analysis

The question of whether or not freezing all embryos, followed by thawed frozen-embryo transfer, is better than fresh-embryo transfer has received a lot of attention. When the E-Freeze trial was planned, several other trials across the world were also planned and conducted. Despite eight RCTs in total (including the E-Freeze trial)11,12,42–46 and multiple aggregated meta-analyses on this topic, there is still a lack of clarity around the effectiveness and cost-effectiveness of a strategy of elective embryo freezing or freeze-all in IVF. What is becoming clear from the published trial data and the data from the E-Freeze trial is that freeze-all is not an effective approach in all women undergoing IVF, can reduce the chances of pregnancy in some women and carries potential risks. The current meta-analyses of aggregated data50,51 have indicated that more data are needed to resolve the current uncertainty as to the groups of patients for whom the strategy of freeze-all would be beneficial.

As the randomised trials42–46 have a degree of clinical heterogeneity that could mask the potential benefits of frozen-embryo transfer in specific groups of women, rather than investing additional time and resources into further randomised trials, we believe that a patient IPD-MA offers a more efficient and cost-effective way of identifying these subgroups and providing a definitive answer to this important clinical question. The number of published IPD-MAs has increased considerably over the past decade in other areas of medicine. Evidence synthesis involving the collection and analysis of individual patient data (IPD) are considered the best method for assessing participants’ characteristics and provide more detailed and robust meta-analysis results.71 Such an approach could also provide more power to detect any differences in the categories of adverse birth outcomes to better inform any expected differences in child health outcomes and costs.

An IPD-MA approach has both statistical and clinical advantages. Data from existing trials42–46 suggest that the freeze-all strategy is not effective for all patients but could improve the efficacy and safety of IVF in some women. Therefore, it is very important to identify subgroups of participants from the existing trials in whom a freeze-all strategy could be adopted. For example, the effectiveness of a freeze-all policy may vary by maternal age, number of available eggs and embryos, stage of development of an embryo prior to transfer and the laboratory method used to freeze embryos (i.e. slow vs. fast freezing). No individual trial is large enough to answer this question and conventional meta-analysis of aggregated data does not lend itself easily to the extraction of sufficient compatible data for meaningful subgroup analyses. By contrast, IPD will allow the effective categorisation of participants for subgroup analyses defined by single or multiple factors and, therefore, offers valuable clinical insights that are particularly relevant to our clinical question.

An IPD-MA will allow us to estimate the treatment effects adjusted for baseline factors where, previously, only unadjusted estimates were available. This has the advantage of increasing the statistical power and allowing an adjustment for potential confounding factors. Consistent inclusion and exclusion criteria could be applied across studies in a way that cannot be undertaken using published aggregated data. It is possible to verify the results of the original trials and request additional data from the triallists that are not available from the published reports.

The statistical analysis can be standardised across studies [e.g. the analysis method, how continuous variables (e.g. maternal age) are analysed] and we can combine data that have been recorded in different formats. Moreover, model assumptions can be assessed and more advanced methods can be applied when necessary.

One can estimate the incidence rate of clinically important, but less common, pregnancy and neonatal complications with greater precision in a randomised cohort of children born following either fresh-embryo transfer or frozen-embryo transfer. This will enable precise projections of any expected differences in child health outcomes and costs. An IPD approach will also help to inform future studies that develop prediction models to predict a couple’s success rate with fresh-embryo transfer compared with frozen-embryo transfer, which would not be possible with standard published data.

Conclusions

The results of this pragmatic, multicentre, two-arm, parallel-group, non-blinded RCT show no evidence of a difference in the healthy baby rate from freezing all embryos, followed by thawed frozen-embryo transfer, compared with the rate from fresh-embryo transfer. There was no statistical difference in OHSS, obstetrics or perinatal complications, or in stress and anxiety scores, between the groups.

The health economic analysis shows that freezing all embryos is not a cost-effective strategy; in fact, it is more costly in both the short and the longer terms and is, therefore, unsuitable for use in routine practice currently.

The decision to offer the freeze-all strategy should be balanced against potential benefit and harm for the mother and child. Rather than investing resources in further trials, we need to join our efforts to undertake an IPD-MA to reach definite answers and identify the subgroups who will benefit most from the freeze-all approach.