What is the most relevant standard of success in assisted reproduction?

Is there a single ‘parameter of excellence’?

Anja Pinborg1,2, Anne Loft1, Søren Ziebe1 and Anders Nyboe Andersen1

1 The Fertility Clinic, University of Copenhagen, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark

2 To whom correspondence should be addressed. e-mail: apinborg{at}rh.dk


    Abstract
 Top
 Abstract
 Is the BESST end-point...
 Outcome measures
 Three optimal parameters of...
 References
 
An optimal standard of success reflects all risk and safety aspects of the treatment-related procedures. In this debate, we question whether one single standard of success can cover the complexity of assisted reproduction technology (ART). To emphasize the use of different parameters as standards of success, we calculated data from all IVF and ICSI cycles initiated at the Fertility Clinic, Rigshospitalet, Copenhagen, Denmark between 1999 and 2001. Several already established parameters were computed and, additionally, new suggestions for end-points were calculated. Three parameters were proposed as the best standards of success in ART: (i) number of oocytes per aspiration (8.7); (ii) number of ongoing implantations per embryo transferred (20%); and (iii) number of deliveries per embryo transferred (14%). These parameters cover aspects of all steps in ART, i.e. the stimulation, laboratory and embryo transfer/outcome phase, also termed the pre-in vitro, in vitro and post-in vitro phase. We suggest that the final description of an ART programme should be a balanced choice of the three standards of success proposed herein. If these standards are implemented as national gold standards, the future goal of homogenous reporting of success rates in ART is within reach.

Key words: assisted reproductive technologies/end-point/national reporting/parameter of excellence/standard of success


    Is the BESST end-point for assisted reproduction the best?
 Top
 Abstract
 Is the BESST end-point...
 Outcome measures
 Three optimal parameters of...
 References
 
An optimal standard of success reflects all risk and safety aspects of the treatment-related procedures. In this debate, we question whether one single standard of success can cover the complexity of assisted reproduction technology (ART).

In a previous issue, Min et al. (2004Go) point to the ‘singleton term gestation live birth rate per assisted reproductive technology (ART) cycle initiated’ as the most relevant end-point of success in an ART programme. This end-point has been calculated from a single unit at Monash University in Australia and yielded an uncomfortably low success rate of 11.1%. The parameter can be estimated based on a term singleton delivery rate of 92.7% in a Danish national cohort of IVF singletons (Westergaard et al., 1999Go). In Europe, the delivery rate per transfer is typically 2.3% higher than the delivery rate per initiated cycle (Nygren and Nyboe Andersen, 2002Go). An estimated European figure of 15.2% x 0.977 x 0.927 = 13.8% is very similar to the 11.1% from the Australian data. Thus, Min et al. (2004Go) provide us with a realistic figure; however, most fertility clinics would be reluctant to report it as an end-point. As mentioned by the authors, despite universal agreement on the need for a multiple birth rate reduction, there has, until very recently, been much talk, little action and, except for Finland, no real improvements have been noticed in relation to this iatrogenic complication.

For the first time, the European IVF Monitoring (EIM) consortium in the most recent ESHRE report on European data from the year 2000 calculated the ‘singleton delivery rate per embryo transfer’ which was found to be 15.2% (Nyboe Andersen et al., 2004Go). This new standard of success differs from the Australian standard as the ‘term’ parameter was not entered and ‘number of transfers’ replaced ‘initiated cycles’.

The ESHRE EIM consortium has suggested another single ‘parameter of excellence’; the ‘number of embryos per delivery’, which can of course be reversed to the ‘number of deliveries per embryo’ (Nygren and Nyboe Andersen, 2002Go). Obviously, this parameter differs from the implantation rate. A high implantation rate is desirable; however, an ART programme will be credited with a high implantation rate if three embryos are replaced and give rise to a triplet gestation. In contrast, the ‘delivery rate per transferred embryo’ will increase the fewer embryos you transfer.

The ultimate goal of reporting ART would be to have a single ‘parameter of excellence’ comprising safety, risk and efficacy aspects. However, we believe that no single parameter describing the complexity of ART can be defined. We propose as a minimum the use of three standard parameters for reporting a successful programme, which should cover aspects of the stimulation, laboratory and embryo transfer/outcome phase, also termed the pre-in vitro, in vitro and post-in vitro phase. To emphasize the use of three such parameters, we computed data from our clinic and defined three ‘parameters of excellence’.


    Outcome measures
 Top
 Abstract
 Is the BESST end-point...
 Outcome measures
 Three optimal parameters of...
 References
 
We calculated data from all IVF and ICSI cycles initiated at the Fertility Clinic, Rigshospitalet, Copenhagen, Denmark between 1999 and 2001. Frozen embryo replacement cycles were excluded. All women were below 40 years of age, as this is the upper limit for public free of charge treatment in Denmark. As presented in Table I, several already established standards of success were computed for each year and for the entire period. Additionally, new suggestions for end-points such as ‘implantations per aspirated oocyte’ or ‘singleton live birth per transferred embryo’ were made.


View this table:
[in this window]
[in a new window]
 
Table I. Different parameters of excellence for the years 1999, 2000, 2001 calculated for the results of IVF and ICSI treatments (frozen embryo replacements were excluded) at the Fertility Clinic, Rigshospitalet, Copenhagen: the figures are crude rates (%)
 
Parameters 1–4 in the table reflect the stimulation and the laboratory phase. The mean ‘number of oocytes per aspiration’ was 8.7 for 1999–2001. This figure traditionally indicates the intensity of the hormone stimulation and it should reflect a balance between the risk of ovarian hyperstimulation syndrome (OHSS) and obtaining a sufficient number of oocytes. Parameters with oocytes as the denominator are also sensitive to the intensity of the hormone stimulation and thus to patient-related risks.

Parameters 5–7 describe the performance in the laboratory by indicating the quality of the transferred embryos. The ‘number of ongoing implantations (gestations with fetal heart beat) per transferred embryo’ (20%) takes into account both the number of implantations and the quality of the embryos, whereas the number of implantations is hidden in ‘ongoing pregnancies per transferred embryo’ (16%). The ‘number of implantations per oocyte’ (4.8%) is a new suggestion for a success parameter, but the figure is low and will rise with increasing number of transferred embryos.

Parameters 8–15 in the table are related mainly to the post-in vitro phase. Although some of these parameters such as ‘deliveries per oocyte’ (2.7%) or ‘singleton live birth per oocyte’ (2.0%) do give an overall description of an ART programme since they are related to all three phases of a programme, they have not been used in general.

Based on the IVF singleton term birth rate (92.7%) in Denmark, we estimated the ‘singleton, term live birth per initiated cycle’ rate as 16% by multiplying the ‘singleton, live birth per initiated cycle’ by a factor 0.927 (Westergaard et al., 1999Go).


    Three optimal parameters of excellence
 Top
 Abstract
 Is the BESST end-point...
 Outcome measures
 Three optimal parameters of...
 References
 
We suggest three main standards of success: (i) number of oocytes per aspiration (8.7); (ii) number of ongoing implantations per embryo transferred (20%); and (iii) number of deliveries per embryo transferred (14%). The first parameter reflects the pre-in vitro phase, i.e. the intensity of the hormone stimulation protocol and the risk of OHSS. The second parameter is an indicator of the quality of the performance in the laboratory and rises with the quality of the transferred embryos. Suboptimal laboratory performance may not only decrease the chance of pregnancy, but may also encourage the use of more aggressive stimulation protocols and thereby increase patient-related risks in order to obtain the desired number of embryos. The third parameter, the reciprocal of the ESHRE parameter, describes the transfer phase and favours single embryo transfers with high quality embryos.

There are three main concerns about the ‘term singletons delivery rate cycle’ as the optimal standard of success in ART, suggested by Min et al. (2004Go). Firstly the stimulation phase is not taken into consideration. Even though female risks related to treatment could be very high, a programme with aggressive hormone stimulation is prone to favour an increase in this end-point. Secondly, the singleton pre-term delivery rate is not influenced by the quality of our treatment and the expertise of the laboratory work. This is dependent rather on the prenatal care and the obstetric management, which could be influenced by national differences. Thirdly, it must be pointed out that a great majority of singleton babies born from gestational week 34 to 37 are completely healthy. For the last two reasons, the ‘term’ factor should not be included in an end-point with ‘singleton delivery’ as the optimal standard of success rates in ART.

Since 1997, the standard of ART in Denmark has been dual embryo transfer. Currently, there is a trend in Denmark towards elective single embryo transfer (eSET), which has to a large extent been implemented in Finland (Tiitinen et al., 2003Go). Since we carried out very few eSETs during 1999–2001 and since the data were not separated into elective and non-elective SET, an estimation of delivery rate per eSET cannot be computed. The future goal of reporting ART must also include a separate recording of the proportion of elective and non-elective SET, which will be implemented in the coming ESHRE reports.

When comparing the results of different ART programmes, a minimum of knowledge on different patient populations is necessary. Demographic data are, however, difficult to incorporate in the reporting of ART. In general, indications for treatment such as tubal factor, male factor or unexplained infertility have little impact on the results. Thus, the age of the women is the only parameter which should be stated in all reports. Our suggestion is that reporting of ART should include the mean age of the patient population and, furthermore, all results should be separated into tables for women <40 and ≥40 years of age. Another interesting and perhaps less realistic idea is to report the ‘number of oocytes retrieved per total FSH dose given to each patient’. This parameter, which we would call the ‘ovarian competence index’, is an ultimate indicator of ovarian sensitivity in the actual patient population and the intensity of the hormone stimulation programme used.

Considering a single parameter as the optimal end-point for ART, we would propose ‘singleton live birth per oocyte’. The reason for this choice is that this parameter includes most aspects of ART. The number of retrieved oocytes is dependent on the intensity of the hormone stimulation, favouring a gentle protocol, and the programme will only be credited by singleton deliveries. In cases where a few oocytes are used to achieve a singleton delivery, the programme is good, i.e. the hormone stimulation creates few follicles, the implantation rate is high and adequate embryos are transferred. However, this figure is even lower (2.0%) than the end-point proposed in the Australian study. This parameter may, in theory, be the best single end-point, but its practical application, probably, would be very limited. Furthermore, it is a parameter difficult to interpret for non-ART professionals, including our patients.

‘Initiated cycles’ was not included in any of the proposed parameters, since it may be difficult to interpret due to national and inter-clinic controversies in the definition of an initiated cycle. It could be anticipated that a programme with a high cancellation rate would be reluctant to report these data. Moreover, the reporting of IVF cycles converted to intrauterine insemination cycles is also inconsistent. Thus, we preferred ‘aspirations’ as a more valid and reliable figure. Since ‘ongoing pregnancies’ is a proxy of outcome in ART, we also avoided this figure in our final end-points. However, there is one important purpose of reporting ‘ongoing pregnancies’, which is the intermediate evaluation of changes in a programme before delivery rates can be provided.

Cryopreservation is a topic of current interest, because good cryopreservation programmes are of great importance especially concerning the implementation of eSET. The cumulative delivery rate per aspiration, which combines fresh and frozen embryo transfers, must be the optimal to report. In practical terms, the problems with long-term follow-up make this very difficult to apply. As a minimum, the number of cycles with cryopreserved embryos and related ongoing pregnancy rates should be provided.

In conclusion, we suggest that the final description of an ART programme should be a balanced choice of the three standards of success proposed herein. If these standards are implemented as national gold standards, the future goal of homogenous reporting of success rates in ART is within reach.


    References
 Top
 Abstract
 Is the BESST end-point...
 Outcome measures
 Three optimal parameters of...
 References
 
Min JK, Breheny SA, MacLachlan V and Healy DL (2004) What is the most relevant standard of success in assisted reproduction? The singleton, term gestation, live birth rate per cycle initiated: the BESST endpoint for assisted reproduction. Hum Reprod 19,3–7.[Abstract/Free Full Text]

NyboeAndersen A, Gianaroli L and Nygren KG (2004) Assisted reproductive technology in Europe, 2000. Results generated from European registers by ESHRE. Hum Reprod 19,490–503.[Abstract/Free Full Text]

Nygren KG and Nyboe Andersen A (2002) Assisted reproductive technology in Europe, 1999. Results generated from European registers by ESHRE. Hum Reprod 17,3260–3274.[Abstract/Free Full Text]

Tiitinen A, Unkila-Kallio L Halttunen M and Hyden-Granskog C (2003) Impact of elective single embryo transfer on the twin pregnancy rate. Hum Reprod 18,449–1453.

Westergaard HB, Johansen AMT, Erb K and Nyboe Andersen A (1999) Danish National In-Vitro Fertilization Registry 1994 and 1995: a controlled study of birth, malformations and cytogenetic findings. Hum Reprod 14,1896–1902.[Abstract/Free Full Text]