Human Reproduction, Vol. 14, No. 3, 689-693,
March 1999
© 1999 European Society of Human Reproduction and Embryology
How concordant are the estimated rates of natural conception and in-vitro fertilization/embryo transfer success?*
1 Department of Obstetrics and Gynaecology, St Joseph Hospital, PO Box 7777, 5500 MB Veldhoven and 2 Department of Obstetrics and Gynaecology, Academic Hospital, University of Groningen, Groningen, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Knowledge of the chance to conceive for the subfertile couple is important in the process of counselling and clinical decision making. There are no data available on the reproducibility of the clinician's ability to assess the chance to conceive, both after expectant management or treatment with in-vitro fertilization and embryo transfer (IVF–embryo transfer). We evaluated this reproducibility by means of a set of case histories presented to a panel of gynaecologists and endocrinologists. A poor reproducibility would indicate a strong need for the use of prognostic models. In 1995, 57 gynaecologists and 32 reproductive endocrinologists were asked to appraise the 1 year spontaneous conception chance as well as the cumulative success rate of three cycles for IVF–embryo transfer of four couples with different medical histories. The clinical and laboratory data of these couples were presented as case histories. The difference between the estimated spontaneous pregnancy chances and the success rate of IVF–embryo transfer was also calculated. Calculation of intra-class correlation coefficients, which can be considered as measures of the reproducibility, demonstrated a substantial reproducibility of the assessment of spontaneous conception chances, but a slight to fair reproducibility of the assessment of IVF–embryo transfer success rates. We conclude that the use of reliable prognostic models for IVF–embryo transfer in the management of subfertility is warranted.
Key words: IVF–embryo transfer success rates/prognosis/reproducibility/spontaneous conception chance
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In the clinical management of subfertility, the decision to start treatment will depend on the expected conception rate of the treatment of choice as compared to the expected conception rate without that particular treatment. Knowledge of fertility chances after treatment as well as knowledge of the probability of treatment-independent pregnancy is therefore elementary in the management of subfertility. In recent years, several authors have developed prognostic models addressing spontaneous fertility chances as well as the success rates of in-vitro fertilization (IVF)–embryo transfer (Comhaire, 1987
; Eimers et al., 1994; Collins et al., 1995; Stolwijk et al., 1996; Snick et al., 1997; Templeton et al., 1997). These models are based on characteristics of the subfertile couple that each in itself has prognostic significance for future fertility. They can be divided into general characteristics of the couple, such as female age and duration of subfertility and characteristics that are related to specific components of fertility, i.e. male factor, cervical factor, ovarian factor and tubal factor. Although the validity of these prognostic models is not well established yet, they are potentially helpful in the clinical management of subfertility (Stolwijk et al., 1996). An alternative for the use of prognostic models could be clinical experience. This clinical experience or `gut-feeling' of clinicians was the only available `tool' before the introduction of prognostic models. In fact, this `gut-feeling' is based on characteristics which are also incorporated in prognostic models in a more formal way. Although identification of prognostic factors in daily clinical practice seems easy, quantification of the relative merit of these factors and their mutual dependence is likely to be more complicated.
The aim of this study was to evaluate the reproducibility of the estimated spontaneous fertility chances and success rates of IVF–embryo transfer by clinicians with the help of a set of case histories. A poor reproducibility would indicate a strong need for prognostic models.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Fifty-seven gynaecologists and 32 reproductive endocrinologists were invited to participate in the study. The gynaecologists were all-round general specialists practising in the field of obstetrics and gynaecology, including diagnostic work-up and treatment of subfertility, with the exception of assisted reproductive technology. Almost all of them had more than 3 years' experience after registration. Reproductive endocrinologists were all senior staff members of IVF–embryo transfer centres. All the gynaecologists were Dutch, whereas 14 of the reproductive endocrinologists were Dutch and 18 were not. Four case histories with a prognostic profile of a subfertile couple were offered to each of the participating clinicians. Each clinician was asked to make an estimated assessment of the chance of spontaneous conception within 12 months and the chance of conception following three cycles of IVF–embryo transfer. For IVF–embryo transfer the baseline chance of an ongoing pregnancy was stated to be 15% per cycle. Other treatment options were not considered.
The four case histories are presented in Table I. They represent a rather wide spectrum of prognostic profiles. In case 1 the prognostic factors were chosen to be unfavourable, with a long duration of subfertility, a relatively aged female partner and a low sperm quality. In contrast, in case 2 the prognostic factors were supposed to be favourable, with a short duration of subfertility and a relatively young female and optimal sperm quality. Case 3 and 4 represented couples in the centre of the prognostic spectrum.
|
Statistical analysis
For each case, we plotted the estimated pregnancy rate after three cycles of IVF–embryo transfer against the estimated 12 months' cumulative spontaneous pregnancy rate. Subsequently, we calculated the median expected spontaneous pregnancy rate and the median expected pregnancy rate after three cycles of IVF–embryo transfer as assessed by the gynaecologists and the reproductive endocrinologists. We also calculated the difference between the estimated spontaneous pregnancy rate and the estimated pregnancy rate after IVF–embryo transfer. The estimated spontaneous pregnancy rates, the estimated success rates of IVF–embryo transfer and the difference between both estimates was tested for statistical significant differences between gynaecologists and reproductive endocrinologists using analysis of variance.
Inter-observer reproducibility was expressed by calculating intra-class correlation coefficients (ICC) and lower 95% confidence limits (95% CL) (Fleiss, 1981). The ICC expresses the degree to which the total variance can be attributed to the true variance: true differences between subjects. It not only assesses the strength of correlation between two measurements but also detects systematic errors. Thus, if a set of estimates made by one observer is systematically lower or higher than the estimates made by another observer, the ICC is correspondingly reduced. Values of the ICC are interpreted as k-statistics: slight (ICC 0–0.20), fair (ICC 0.21–0.40), moderate (ICC 0.41–0.60), substantial (ICC 0.61–0.80), or almost perfect (ICC 0.81–1.0). The ICC and 95% CL were calculated for gynaecologists and reproductive endocrinologists separately.
To counsel a subfertile couple in the decision of whether or not to proceed to IVF–embryo transfer the added value of IVF–embryo transfer as compared to the spontaneous conception chance has to be taken into account. In cases where the pregnancy rate after IVF–embryo transfer is equal to or lower than the spontaneous pregnancy rate, IVF–embryo transfer should not be offered. In cases where the pregnancy rate after IVF–embryo transfer is higher than the spontaneous pregnancy rate, IVF–embryo transfer should be considered if the spontaneous pregnancy rate is low, and if the pregnancy rate after IVF–embryo transfer is sufficiently high.
To illustrate the impact of the estimations on clinical decision-making, we arbitrarily defined the indications for IVF–embryo transfer. We defined IVF–embryo transfer to was presumed to be indicated for couples whose estimated 12 month cumulative spontaneous conception rate was <25% and in whom three cycles of IVF–embryo transfer were expected to generate a relevant improvement of the conception chances; at least 10% improvement in cases where the spontaneous chance was <10%, at least 15% improvement in cases where the spontaneous chance was 20% and at least 25% improvement in cases where the spontaneous chance was 25%. The area in which IVF–embryo transfer is supposed to be offered is shown in Figure 1.
|
Finally, we calculated expected 12 month spontaneous pregnancy rates as predicted by the prognostic models of Comhaire (1987) and Eimers et al. (1994), expected 12 month live birth rates as predicted by the models of Collins et al. (1995) and Snick et al. (1997) and the success rates after three cycles of IVF–embryo transfer as predicted by the models of Haan et al. (1991) and Templeton et al. (1997).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Table II
and Figure 2 show the estimated expected conception rates as judged by the participating clinicians. For case 1, the median expected spontaneous conception rates were 10% and 5% for the gynaecologists and the reproductive endocrinologists respectively, whereas the median expected conception rate after IVF–embryo transfer was 20% in both groups. The expected conception rates of 32 (56%) of the gynaecologists and 24 (75%) of the reproductive endocrinologists would indicate offering IVF–embryo transfer. Figure 2A shows that the expected spontaneous conception rate was almost consistently lower than the conception rate after IVF–embryo transfer.
|
|
For case 2, the expected spontaneous conception rates were higher, with a median of 80% judged by the gynaecologists, and a median of 60% judged by the reproductive endocrinologists. A majority of both the gynaecologists and the reproductive endocrinologists expected the spontaneous conception chance to be better than the expected conception chances after IVF–embryo transfer (Figure 2B
). None of the combinations of expected conception chances would have led to the offering of IVF–embryo transfer.
For case 3, a majority of the gynaecologists and all reproductive endocrinologists expected conception rates after IVF–embryo transfer to be higher than spontaneous conception rates. The median difference between the expected IVF–embryo transfer and spontaneous conception rates were 10% among gynaecologists, and 15% among reproductive endocrinologists (Figure 2C). The judgements would have indicated IVF–embryo transfer in 21 (37%) of the gynaecologists and 16 (50%) of the reproductive endocrinologists.
For case 4, the situation is comparable to case 3, with gynaecologists estimating IVF–embryo transfer to be better, whereas reproductive endocrinologists expect comparable conception after IVF–embryo transfer and expectant management (Figure 2D). The judgements would have indicated IVF–embryo transfer in 25 (44%) of the gynaecologists and 6 (19%) of the reproductive endocrinologists.
The differences in expected conception rates after IVF–embryo transfer as well as the difference between expected spontaneous and IVF–embryo transfer mediated conception rates were not statistically significant, with P values of 0.30 and 0.31 respectively. However, gynaecologists were significantly more optimistic towards spontaneous conception chances as compared with reproductive endocrinologists (P value 0.01).
As comparison, Table II also shows the conception chances as predicted by several prognostic models. For cases 1, 2 and 3 the models predicting spontaneous conception showed comparable outcomes to those made by the clinicians. For case 4, the models were predicting a lower chance, except for the model of Snick et al. (1997) using six variables. For IVF–embryo transfer, the models of Templeton et al. (1997) and Haan et al. (1991) showed comparable results.
Table III shows the ICC and lower 95% CL for the gynaecologists, as well as the reproductive endocrinologists. The ICC of the expected spontaneous conception rates were high, with values of 0.71 and 0.66 respectively. In contrast, the ICC for estimated IVF–embryo transfer success rates were low, with values of 0.24 and 0.14 respectively. The ICC for the difference between expected spontaneous conception rates and expected conception rates after IVF–embryo transfer were 0.34 for the gynaecologists and 0.46 for the reproductive endocrinologists.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This study evaluates the reproducibility of the estimated spontaneous conception chances and IVF–embryo transfer success rates in a set of case histories by gynaecologists and reproductive endocrinologists. Reproducibility of the estimation of spontaneous conception rates was substantial among both groups, but the reproducibility of assessment of success rates of IVF–embryo transfer was only slight to fair. For the expected differences between spontaneous conception rates and IVF–embryo transfer success rates, which is of crucial importance for the final treatment decision, the reproducibility of the assessment of both groups was fair to moderate.
We used four case histories, of which the first two contained rather extreme profiles. The assessments of gynaecologists and reproductive endocrinologists were consistent in the sense that the case history with the worst prognostic profile was expected to have low conception rates, whereas the case history with the best prognostic profile was expected to have high conception rates.
The study was performed at the end of 1995. At that time, some prognostic models were already available, and it might be that the answers of some of the participants, especially those of the reproductive endocrinologists, were influenced by these models. However, the case histories were completed during a general session, and none of the participants had any of the prognostic models available.
We asked the participants to appraise expected chances for spontaneous conception and IVF–embryo transfer, but we did not ask them what final counselling they would make. It might very well be possible that although participants generated different judgements of the chance of conception, their final treatment decision would not be different. However, the analysis of the differences between estimated spontaneous and IVF–embryo transfer success rates according to pre-defined rules showed that there is considerable inconsistency among gynaecologists as well as among reproductive endocrinologists in the counselling of IVF–embryo transfer. These differences are of clinical importance, especially in cases 1, 3 and 4, describing patients with the better prognostic profiles.
We found the reproducibility of the estimation of spontaneous conception rates to be substantial among both gynaecologists and reproductive endocrinologists. Since reproducibility is one of the key components of accuracy, this finding might indicate that estimated assessment of spontaneous conception chances is reliable (Kraemer, 1992). This is supported by the predictions of the prognostic models, which seem to be consistent with the estimates in three of the four case histories. However, one should also keep in mind that two of the four case histories had rather extreme profiles. In clinical practice, however, we often deal with patients in the middle of the spectrum, comparable to cases 3 and 4.
Consequently, estimates of the chance of spontaneous conception, which is important in reaching a decision on which couples should be considered for artificial reproductive treatment (i.e. those with a low chance are deemed more suitable), seem to be reliably made by both gynaecologists and reproductive endocrinologists.
In contrast, the reproducibility of assessment of success rates of IVF–embryo transfer was only slight to fair, indicating that at present unaided prediction of the results of IVF–embryo transfer is likely to be inaccurate. The prognostic models of Haan et al. (1991) and Templeton et al. (1997) also predicted success rates of IVF–embryo transfer comparable to the estimates of both gynaecologists and reproductive endocrinologists. However, the validity of these predictive models has not been established. Stolwijk et al. (1996) recently tried to validate other prognostic models for IVF–embryo transfer, but found disappointing results. Unfortunately, the present study shows that, especially for the prediction of IVF–embryo transfer, validated prognostic models are warranted, since the reproducibility of estimates of pregnancy chances after IVF–embryo transfer was found to be only slight to fair.
|
Notes |
|---|
*Presented at the 12th annual meeting of the ESHRE, 1996, Maastricht, The Netherlands.
3 To whom correspondence should be addressed 
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Collins, J.A., Burrows, E.A. and Willan, A.R. (1995) The prognosis for live birth among untreated infertile couples. Fertil. Steril., 64, 22–28.[Web of Science][Medline]
Comhaire, F.H. (1987) Simple model and empirical method for the stimulation of spontaneous pregnancies in couples consulting for infertility. Int. J. Androl., 10, 671–680.[Web of Science][Medline]
Eimers, J.M., Te Velde, E.R., Gerritse, R. et al. (1994) The prediction of the chance to conceive in subfertile couples. Fertil. Steril., 61, 44–52.[Web of Science][Medline]
Fleiss, J.L. (1981) Statistical Methods for Rates and Proportions. John Wiley, New York.
Haan, G., Bernardus, R.E., Hollanders, J.M. et al. (1991) Results of IVF from a prospective multicentre study. Hum. Reprod., 6, 805–810.
Kraemer, H.C. (1992) Evaluating Medical Tests. SAGE Publications, Newsbury Park, pp. 5–17.
Snick, H.K.A., Snick, T.S., Evers, J.L.H. and Collins, J.A. (1997) The spontaneous pregnancy prognosis in untreated subfertile couples: the Walcheren primary care study. Hum. Reprod., 12, 1582–1588.
Stolwijk, A.M., Zielhuis, G.A. Hamilton, C.J.C.M. et al. (1996) Prognostic models for the probability of achieving an ongoing pregnancy after in-vitro fertilization and the importance of testing their predictive value. Hum. Reprod., 11, 2298–2303.
Templeton, A., Morris, J.K. and Parslow, W. (1997) Factors that affect outcome of in-vitro fertilisation treatment. Lancet, 348, 1402–1406.
Submitted on June 4, 1998; accepted on November 20, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B.C.J.M. Fauser, K. Diedrich, P. Devroey, and on behalf of the Evian Annual Reproduction (EVAR) Predictors of ovarian response: progress towards individualized treatment in ovulation induction and ovarian stimulation Hum. Reprod. Update, January 1, 2008; 14(1): 1 - 14. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||