The number of oocytes associated with maximum cumulative live birth rates per aspiration depends on female age: a population study of 221 221 treatment cycles

Abstract

STUDY QUESTION

What is the number of oocytes where the maximum cumulative live birth rate per aspiration (CLBR) is observed during ART in women of different ages?

SUMMARY ANSWER

The maximum CLBR was observed when around 25 oocytes were retrieved in women between 18–35 years of age, around 9 oocytes in women more than 45 years of age and continued to increase beyond 30 oocytes in women between 36–44 years of age.

WHAT IS KNOWN ALREADY

The live birth rate per fresh or frozen/thaw embryo transfer (FET) procedure has traditionally been the main measure of ART success. However, with the introduction of highly efficient embryo cryopreservation methods, CLBR encompassing live delivery outcomes from the fresh and all subsequent FET following a single ovarian stimulation and oocyte collection is increasingly viewed as a more meaningful measure of treatment success. There is evidence suggesting that larger oocyte yields are associated with increased likelihood of cumulative live birth per aspiration. Whether this association is the same across female ages has not yet been properly investigated.

STUDY DESIGN, SIZE, DURATION

This is a large retrospective population-based cohort study using data from the Australian and New Zealand Assisted Reproduction Database (ANZARD). ANZARD contains information from all ART treatment cycles carried out in all fertility centres in Australia and New Zealand. Overall, 221 221 autologous oocyte aspiration cycles carried out between January 2009 to December 2015 were included in the analysis.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Cumulative live birth per aspiration was defined as at least one liveborn baby at ≥20 weeks gestation resulting from an ART aspiration cycle, including all fresh and FET resulting from the associated ovarian stimulation, until one live birth occurred or all embryos were used. Cycles where no oocytes were retrieved were excluded from analysis as there is no possibility of live birth. Analyses of data were performed using generalized estimating equations to account for the clustered nature of data (multiple cycles undertaken by a woman). Univariate and multivariable regression analysis was performed to identify and adjust for factors known to independently affect cumulative live birth per aspiration. An interaction term between female age and the number of oocytes retrieved was included to assess whether the age of the women was associated with a different optimal number of oocytes to achieve at least one live birth from an aspiration cycle (i.e. the effect-modifying role of female age). The likelihood of cumulative live birth per aspiration was calculated as odds ratios (ORs) with 95% CI.

MAIN RESULTS AND THE ROLE OF CHANCE

The median number of oocytes retrieved was 7 (interquartile range, 4–12) and median age of patients was 36 (interquartile range, 33–40). The overall CLBR was 32.2%. The results from the multivariable regression analysis showed

that the number of oocytes retrieved remained a significant predictor (P < 0.001) of cumulative live birth per aspiration after adjusting for female age, parity and cycle count. Compared to the reference group of 10–14 oocytes retrieved, the adjusted odds for cumulative live birth per aspiration increased with the number of oocytes retrieved: 1–3 oocytes, 0.21 (95% CI, 0.20–0.22); 4–9 oocytes, 0.56 (95% CI, 0.55–0.58); 15–19 oocytes, 1.38 (95% CI, 1.34–1.43); 20–24 oocytes, 1.75 (95% CI, 1.67–1.84); and 2.10 (95% CI, 1.96–2.25) with more than 25 oocytes. After stratifying by female age group, the rate of increase in CLBR per additional oocyte retrieved was lower in the older age groups, indicating that higher oocyte yields were more beneficial in younger women. CLBR of patients in the <30 years and 30–34 years age groups appeared to reach a plateau (with only minimal increase in CLBR per additional oocyte retrieved) after retrieval of 25 oocytes at 73% and 72%, respectively, while CLBR of patients in the 35–39 years and 40–44 years age groups continued to increase with higher oocyte yields, reaching 68% and 40%, respectively, when 30 or more oocytes were retrieved. CLBR of patients aged 45 years and above remained consistently below 5%. Findings suggest that the number of oocytes retrieved where CLBR appears to be maximized is around 25 in women between 18–35 years, more than 30 in women between 36–44 years and around 9 in women 45 years and older. However, results for women aged 45 years and older may not be as robust due to the relatively small sample size available in this age group.

LIMITATIONS, REASONS FOR CAUTION

As with all large retrospective database studies, there are potential confounders that cannot be accounted for. Despite the current study being based on complete ascertainment of ART cycles across two countries, ovarian stimulation protocols, oocyte quality parameters and a number of important patient characteristics are not collected by ANZARD. Additionally, a small number of cycles were available for women over 45 years yielding more than 15 oocytes, making these estimates unreliable.

WIDER IMPLICATIONS OF THE FINDINGS

The results from this study demonstrate that the number of oocytes retrieved where the maximum CLBR is observed during ART is dependent on female age. This provides information for clinicians and patients to understand the modifying effect of age on the number of oocytes retrieved and the likelihood of success with ART.

STUDY FUNDING/COMPETING INTEREST(S)

No external funding was used for this study. The Fertility Society of Australia funds the National Perinatal Epidemiology and Statistics Unit to manage ANZARD and conduct national reporting of ART in Australia and New Zealand. Associate Professor Georgina Chambers (G.C.) is employed by the University of New South Wales (UNSW) and is director of the National Perinatal Epidemiology and Statistics Unit at UNSW. G.C. was also a paid member of the Australian governments Medicare Benefits Scheme taskforce on assisted reproductive technologies in 2017.

cumulative live birth, oocytes number, female age, IVF treatment, assisted reproductive technology

Introduction

Controlled ovarian stimulation to generate multiple mature oocytes for aspiration has long been a central component for maximizing the chance of success from a single fresh ART cycle (Trounson et al., 1981). In fresh embryo transfer cycles, the optimal number of oocytes has been found to lie between 10–15 oocytes, above which the rates of ovarian hyperstimulation syndrome (OHSS) seem to dramatically increase (Van der Gaast et al., 2006; Sunkara et al., 2011; Ji et al., 2013). However, with improved cryopreservation techniques and the introduction of stimulation protocols as well as freeze-all strategies that greatly minimize the chance of OHSS (Devroey et al., 2011; Griesinger et al., 2011), frozen/thaw cycles have become a viable alternative to fresh embryo transfers (Loutradi et al., 2008; Noyes et al., 2009). As a result, cryopreservation has become increasingly central in clinical practice, and in some countries, the number of live deliveries following frozen/thaw embryo transfers (FET) exceeds the one following fresh transfers (Fitzgerald et al., 2018). Therefore, cumulative live birth rates per aspiration (CLBR) encompassing outcomes from fresh and all FET following a single ovarian stimulation (Zegers-Hochschild et al., 2017) has become a more meaningful measure of ART treatment success (Chambers et al., 2017).

In that context, evaluating the association between the number of oocytes retrieved with CLBR has become highly clinically relevant. A small number of studies have evaluated this association (Fatemi et al., 2012; Ji et al., 2013; Drakopoulos et al., 2016; Magnusson et al., 2017; Devesa et al., 2018; Polyzos et al., 2018), suggesting that larger oocyte yields lead to an increased likelihood of cumulative live birth per aspiration. Although results from the studies mentioned above are homogenous enough to imply a strong positive association between number of oocytes retrieved and CLBR, the potential effect-modifying role of female age (i.e. the association is not the same for all female ages) has not yet been properly investigated. It is well known that the quality of oocytes retrieved is inherently tied with female age, with older women having a significantly higher risk of developing aneuploid embryos and less ability to create high number of oocytes on aspiration (Munné et al., 1995). Hence, it is likely that retrieving the same number of oocytes in a younger woman and in an older woman would not benefit them in the same way. To date, no studies have focused on evaluating the effect-modifying role of female age on the association between the number of oocytes retrieved and CLBR in women across all age groups.

The aim of the current study was to determine the number of oocytes retrieved where the maximum CLBR is observed after a single ovarian stimulation in women of different ages.

Figure 1

Flowchart of data exclusions. Data from the Australian and New Zealand Assisted Reproduction Database (ANZARD) containing information from all treatment cycles carried out in all fertility centres in Australia and New Zealand were used in this study. Overall, 221 221 autologous oocyte aspiration cycles carried out between January 2009 to December 2015 were included in the analysis.

Materials and Methods

Study design—population

This is a large retrospective population-based cohort study with data obtained from the Australian and New Zealand Assisted Reproduction Database (ANZARD). All fertility clinics in Australia and New Zealand are required to report ART treatment information to ANZARD as part of their accreditation requirements; therefore full registration of ART cycles can be assumed. All autologous fresh and FET cycles performed between January 2009 and December 2015, with live births followed up to October 2016, were included. For this study, cycles involving gamete intra-fallopian transfer (GIFT) and PGS were excluded. In addition, natural cycle IVF/ICSI, or cycles where the purpose of treatment was long-term preservation of fertility, were also excluded. Cycles where no oocytes were retrieved (10% of initial cohort) were excluded as there is no possibility of live births from these cycles. Finally, only women who achieved a cumulative live birth per aspiration or fully utilized all cryopreserved embryos during the study period were included to ensure reliability of CLBR estimates. This study was approved by the University of New South Wales Human Research Ethics Advisory Panel Executive (HC17959).

Outcome measures

The primary outcome was cumulative live birth per aspiration, which was defined as at least one liveborn baby at ≥20 weeks gestation resulting from an ART aspiration cycle, including all fresh and FETs resulting from the associated ovarian stimulation. According to this definition, multiple deliveries from the same pregnancy were considered as one live birth. CLBRs were calculated as the proportion of cycles that achieved at least one cumulative live birth per aspiration. In any sub-analyses, CLBR was not calculated when the number of cycles was less than 20 to ensure reliability of estimates.

Statistical analysis

All continuous variables were expressed as median (interquartile range, IQR). Categorical variables were expressed as percentages.

Patients were categorized into five groups according to age upon commencement of treatment: 18–29 years, 30–34 years, 35–39 years, 40–44 years and 45 years and above. Oocyte yield was assessed in two ways: (i) as a continuous variable for graphing purposes, where all oocyte yields above 30 were considered as having 30 oocytes retrieved due to their relative rarity and (ii) categorized into six groups (1–3, 4–9, 10–14, 15–19, 20–24 and more than 25 years) for the purposes of regression analysis.

As some patients undertook more than one cycle, generalized estimating equations (GEE) with exchangeable correlation structure and robust estimation of standard errors were used to account for the clustered nature of data. Within the GEE framework, cumulative live birth per aspiration was the binary dependent variable, and oocyte groups and female age groups were the independent variables of interest. Univariate regression analysis was performed to identify factors that were associated with cumulative live birth per aspiration, and these factors (parity and complete cycle indicator) were entered as covariates in the multivariable models. The likelihood of cumulative live birth per aspiration after treatment was calculated as odds ratios (ORs) with 95% CI. Assessment of the effect-modifying role of female age was performed with an interaction term between female age groups and the number of oocytes retrieved in the GEE model. Marginal probabilities were calculated to plot the predicted probabilities for cumulative live birth per aspiration depending on the number of oocytes retrieved and female age group, while accounting for covariates.

Furthermore, multivariable fractional polynomials were used to find the most suitable functional form between cumulative live birth per aspiration and the continuous predictors, after taking into account all possible transformations. The most suitable functional form identified was then compared against the original GEE model for validation. Presence of an interaction between the number of oocytes retrieved and female age groups was confirmed using fractional polynomials (MFPIGEN routine; Royston, 2012).

All analyses were conducted using RStudio statistical software, version 3.5.0. (RStudio, Boston, MA, USA) and Stata (version 15.1, College Station, TX: StataCorp LLC).

Results

Following planned exclusions, 221 221 aspiration cycles undertaken by 116 677 women were included in the analysis (Fig. 1). Patient and cycle characteristics are summarized in Table I. The median age of women commencing ART treatment was 36 (IQR 33–40), and the median number of oocytes retrieved was 7 (IQR 4–12). The distribution of the number of oocytes retrieved differed between age groups, with older women having lower oocyte yields. This relationship is graphically depicted in Fig. 2. The overall CLBR for the entire cohort was 32.2% (95% CI, 32.0–32.4%), with the highest proportion of clinical pregnancies that did not progress to live delivery in the 1–3 oocyte yield group (27.7%). The proportion of freeze-all cycles increased as oocyte yield increased, reaching 46.0% when more than 25 oocytes were retrieved (Table II).

Table I

Patient and cycle characteristics.

Characteristics	n (cycles)	%
Country
Australia	202 583	91.6
New Zealand	18 638	8.4
Female age (group)
18–29 years	22 946	10.4
30–34 years	57 774	26.1
35–39 years	79 598	36.0
40–44 years	56 654	25.6
45 years and older	4249	1.9
Number of oocytes retrieved (group)
1–3 oocytes	42 039	19.0
4–9 oocytes	98 447	44.5
10–14 oocytes	47 462	21.5
15–19 oocytes	20 875	9.4
20–24 oocytes	7974	3.6
25 + oocytes	4424	2.0
Year of treatment cycle
2009	31 535	14.3
2010	27 789	12.6
2011	32 281	14.6
2012	33 240	15.0
2013	33 087	15.0
2014	32 889	14.9
2015	30 400	13.7
Cycle count
1	103 116	46.6
2	52 491	23.7
3	27 758	12.5
4	15 242	6.9
5	8786	4.0
≥6	13 828	6.3
Previous pregnancy(>20 weeks)
Yes	48 772	22.0
No	172 449	78.0
Insemination method1
IVF	71 995	32.5
ICSI	142 218	64.3
Mixed IVF ICSI	3968	1.8
Cause of infertility2
Endometriosis	27 399	12.4
Tubal disease	19 585	8.9
Male factor	78 044	35.3
Other	61 864	28.0
Unexplained infertility	69 373	31.4

¹In 3040 cycles fertilization was not attempted.

²Some cycles may have multiple causes of infertility.

Figure 2

Distribution of the number of oocytes retrieved stratified by female age groups. The proportion of cycles is plotted against the number of oocytes retrieved for different female ages groups (the blue curve represents 18–29 years; pink, 30–34 years; orange, 35–39 years; green, 40–44 years; and purple, 45 years and older).

Table II

Outcome characteristics according to number of oocytes retrieved.

Characteristics	Number of oocytes retrieved
	1–3	4–9	10–14	15–19	20–24	≥25	Total
% Calculated horizontally per category
Female age
18–29 years	1809	8285	6548	3613	1603	1088	22 946
% Of age group	7.9	36.1	28.5	15.7	7.0	4.7	100.0
30–34 years	5894	24 182	15 361	7589	3003	1745	57 774
% Of age group	10.2	41.9	26.6	13.1	5.2	3.0	100.0
35–39 years	14 222	37 394	17 308	6901	2553	1220	79 598
% Of age group	17.9	47.0	21.7	8.7	3.2	1.5	100.0
40–44 years	17 961	26 879	7951	2700	804	359	56 654
% Of age group	31.7	47.4	14.0	4.8	1.4	0.6	100.0
45 years and older	2153	1707	294	72	11	12	4249
% Of age group	50.7	40.2	6.9	1.7	0.3	0.3	100.0
% Calculated vertically per oocyte group
Cumulative clinical pregnancy per aspiration
No	36 225	65 381	23 506	8442	2717	1266	137 537
Yes	5814	33 066	23 956	12 433	5257	3158	83 684
% Per oocyte group	13.8	33.6	50.5	59.6	65.9	71.4	37.8
Cumulative live birth per aspiration
No	37 835	71 278	26 528	9671	3154	1495	149 961
Yes	4204	27 169	20 934	11 204	4820	2929	71 260
% Per oocyte group	10.0	27.6	44.1	53.7	60.4	66.2	32.2
Clinical pregnancies that did not progress to live delivery
Cycles	1610	5897	3022	1229	437	229	12 424
% Per oocyte group	27.7	17.8	12.6	9.9	8.3	7.3	14.8
All embryos frozen
No	41 515	95 541	44 693	18 398	6021	2388	208 556
Yes	524	2906	2769	2477	1953	2036	12 665
% Of total cycles	1.2	3.0	5.8	11.9	24.5	46.0	5.7

Bold values represent the total (calculated horizontally or vertically as indicated) per category

Association between the number of oocytes retrieved, cumulative live birth rates per aspiration and the effect-modifying role of female age

There was a strong positive association between the number of oocytes retrieved and CLBR ranging from 10% for 1–3 oocytes to 66% for ≥25 oocytes retrieved (Table II). When stratified by female age group, the rate of increase in CLBR per additional oocyte retrieved was lower in the older age groups (Fig. 3). Younger women achieved higher CLBR than older women at any given oocyte number retrieved, with patients <30 years of age achieving double the CLBR of patients 40 years and above. There was minimal difference between the CLBR of patients aged <30 years and 30–34 years, with the difference becoming more pronounced in each subsequent age group (Fig. 3). CLBR of patients in the <30 years and 30–34 years age groups appeared to reach a plateau (with only minimal increase in CLBR per additional oocyte retrieved) after retrieval of 25 oocytes at 73% and 72%, respectively (Fig. 3). CLBR of patients in the 35–39 years and 40–44 years age groups continued to increase with higher oocyte yields, reaching 68% and 40%, respectively, when 30 or more oocytes were retrieved. The rate of increase in CLBR with each additional oocyte retrieved in patients aged 45 years and above was minimal, with CLBR consistently remaining below 5% (Fig. 3).

Figure 3

Observed cumulative live birth rates according to the number of oocytes retrieved, stratified by age group. (Blue represents 18–29 years; pink, 30–34 years; orange, 35–39 years; green, 40–44 years; and purple, 45 years and older.)

Table III

Multivariable logistic regression with odds ratios for cumulative live birth per aspiration.

Variables	OR (95% CI)	P value
Oocytes retrieved
1–3	0.21 (0.20–0.22)	<0.001
4–9	0.56 (0.55–0.58)	<0.001
10–14	1.00
15–19	1.38 (1.34–1.43)	<0.001
20–24	1.75 (1.67–1.84)	<0.001
≥25	2.10 (1.96–2.25)	<0.001
Age group
<30 years	1.85 (1.79–1.91)	<0.001
30–34 years	1.62 (1.58–1.66)	<0.001
35–39 years	1.00
40–44 years	0.35 (0.33–0.36)	<0.001
≥45 years	0.05 (0.04–0.07)	<0.001
Cycle count
1	1.00
2	0.89 (0.87–0.91)	<0.001
3	0.80 (0.78–0.83)	<0.001
4	0.76 (0.73–0.79)	<0.001
5	0.76 (0.72–0.80)	<0.001
≥6	0.72 (0.69–0.76)	<0.001
Parity
No previous pregnancy	1.00
Previous pregnancy	1.24 (1.21–1.27)	<0.001

Figure 4

Predicted probabilities with 95% CI for cumulative live birth per aspiration by age group according to the number of oocytes retrieved. (The blue curve represents 18–29 years; pink, 30–34 years; orange, 35–39 years; green, 40–44 years; and purple, 45 years and older.)

The number of oocytes retrieved, female age, parity and cycle count were found to have a significant association with cumulative live birth per aspiration in the univariate logistic regression. The results from the multivariable analysis showed that the number of oocytes retrieved remained a significant predictor (P < 0.001) of cumulative live birth per aspiration after adjusting for female age, parity and cycle count (Table III). The adjusted ORs for cumulative live birth per aspiration increased with the number of oocytes retrieved, reaching 2.10 (95% CI, 1.96–2.25) when more than 25 oocytes were retrieved (Table III).

A significant interaction (P < 0.001) between the number of oocytes retrieved and female age groups was confirmed using MFPIGEN (Royston, 2012). Predicted probabilities for cumulative live birth per aspiration at each number of oocytes retrieved, stratified by female age groups were plotted using marginal probabilities that accounted for the interaction between oocyte yield and female age group, as well as parity and cycle count (Fig. 4). Results obtained through backward eliminating multivariable fractional polynomials were similar to the results from GEE modelling.

Discussion

This study suggests that a larger number of oocytes retrieved after ovarian stimulation is associated with higher CLBR. Importantly, there was a negative effect-modifying role of female age on the association between the number of oocytes retrieved with CLBR. The highest chance of achieving a cumulative live birth per aspiration in women between 18–34 years was associated with an oocyte yield of around 25 oocytes; beyond this point the rate of increase in CLBR with each additional oocyte retrieved was minimal. In women between 36–44 years of age, chances of achieving a cumulative live birth per aspiration continued to increase beyond retrieval of more than 30 oocytes; thus the more oocytes retrieved the better the treatment outcome. In women aged 45 years and older, the best chance of cumulative live birth per aspiration was associated with an oocyte yield of around 9; however even then probability was very low. It should be noted that results for women aged 45 years and older may be less robust due to the relatively smaller sample size available in this age group.

Previous studies investigating the effect of retrieval of high numbers of oocytes following ovarian stimulation have suggested that high oocyte yields may result in increased aneuploidy rates (Baart et al., 2007; Haaf et al., 2009) and hence reduced live birth rates. However, contemporary studies have found that the number of oocytes retrieved was positively associated with the absolute number of euploid embryos available for transfer (Labarta et al., 2017; Venetis et al., 2018). These findings provide an explanation for the results from the current study, as a higher availability of euploid embryos for transfer resulting from higher oocyte yields would increase the CLBR.

A small number of studies also looked at the association between the number of oocytes retrieved and cumulative live birth rates (Fatemi et al., 2012; Ji et al., 2013; Drakopoulos et al., 2016; Magnusson et al., 2017; Devesa et al., 2018; Polyzos et al., 2018). In general, all six studies found a positive association between oocyte yield and cumulative live birth rates. This is in agreement with the results from the current study, which investigated the research question in a significantly larger sample size (n = 221 221 vs. n = 102 055 combined from all 6 previous studies) that is more representative of the general sub-fertile population. Ji et al., Drakopoulos et al. and Devesa et al. were single-centre studies with a limited number of patients. Furthermore, the studies by Ji et al. and Devesa et al. were focused on a specific sub-group of patients (young women aged 34 years and below with low BMI in the study by Ji et al. and older women aged 38 years and above in the study by Devesa et al.). Patients in the studies carried out by Fatemi et al., Devesa et al. and Magnusson et al. were only followed for the duration of the study period and not until at least one live birth had occurred or all cryopreserved embryos had been utilized; thus there exists a possibility that the cumulative live birth rates may be underestimated especially at higher oocyte yields.

Two studies (Devesa et al., 2018; Polyzos et al., 2018) included an age-stratified figure of the association between CLBR and the number of oocytes retrieved. The current study focuses on investigating the effect-modifying role of female age by conducting analyses that takes into account the interaction between the number of oocytes retrieved and female age and its effect on CLBR. The current study also adds to the available evidence by provision of a large sample size from a nationwide registry that is reflective of modern changes in ART, and the inclusion of freeze-all cycles as well as cycles where no oocytes were fertilized in the analysis, which combine to provide an accurate representation of CLBR especially at the extreme lower and higher oocyte yields. Furthermore, the exclusion of cycles where no oocytes were retrieved allows for estimations of CLBR that are more clinically relevant, as it is known that if zero oocytes are retrieved then the CLBR from that cycle will be 0% (with the exception of the rare spontaneous conceptions occurring during the treatment cycle).

This study clearly shows that the association between the number of oocytes retrieved and CLBR is strongly modified by female age. This means that the optimal number of oocytes retrieved would depend upon the woman’s age, where the expected difference in the probability of achieving a cumulative live birth per aspiration with each additional oocyte retrieved is reduced with advancing age. Additionally, the expected benefit from high oocyte yields in patients between 18–34 years of age is limited as CLBR reaches a plateau after 25 oocytes were retrieved. This plateau may be due to the saturation of the possibility of achieving a cumulative live birth per aspiration after 25 oocytes. It is possible that by this point, younger women would have had an adequate number of euploid embryos to provide multiple chances of achieving a pregnancy and subsequent live birth. Younger women who still have not achieved a cumulative live birth per aspiration after exhaustion of all embryos created from retrieval of 25 or more oocytes are likely to belong to a subset of patients with extremely poor prognosis or with other severe underlying issues preventing pregnancy.

The positive association observed between number of oocytes retrieved and CLBR across all female age groups suggests that number of oocytes retrieved has a potential prognostic value in predicting cumulative live birth per aspiration. It shows that the optimal number of oocytes for women aged less than 35 years is approximately 25 and the optimal number for women aged 35–44 years is more than 30 oocytes (based on the data presented in Fig. 4). However, despite a reduction in the risk of OHSS using contemporary stimulation protocols, high oocyte yields increase the potential for patient discomfort and morbidity, embryo wastage, as well as additional laboratory and storage resources requirements. Furthermore, as shown in this study, an inherent limit exists on the number of oocytes reaching maturity following ovarian stimulation as ovarian reserve declines with age.

The ‘one and done’ approach to ovarian stimulation introduced by Vaughan et al. (2017) suggests that retrieval of more than 15 oocytes increases the chance of achieving two live births and thus completing an average sized family. According to this idea, maximizing the number of oocytes retrieved from a single ovarian stimulation may be beneficial to couples in achieving their fertility goals. The results from the current study suggest that the number of oocytes needed for this ‘one and done’ approach is likely to be different in women of different ages, as the expected benefits from maximizing oocyte yield decreases with advancing age. Therefore, future large-scale studies investigating the optimal number of oocytes retrieved to achieve more than one live birth should take female age into account.

As with all large retrospective database studies, there are potential confounders that are unmeasured and thus cannot be accounted for. In the current study, type, dose and duration of stimulation protocols were not collected by ANZARD, thus were unavailable for analysis. Although significant clinical heterogeneity exists within the cycles recorded in ANZARD, these differences are reflective of a prognostically diverse population, thus improving generalizability of the results. The current study was also limited by the relatively small sample size of patients in the 45 years and above age group when oocyte yield was greater than 15 (n = 66). However, this still represents the expected outcome following oocyte retrieval, as older women are less likely to achieve such high oocyte yields (Chuang et al., 2003). Additionally, as the current study is an observational study, a cause–effect relationship between the number of oocytes retrieved and CLBR should not be inferred since it cannot be excluded that the observed association is due to the inherent characteristics of the population examined.

In conclusion, the number of oocytes retrieved where the maximum CLBR after ART is observed is dependent on female age. Findings from the current study suggests that this number is around 25 oocytes in women between 18–35 years, more than 30 in women between 36–44 years and around 9 in women 45 years and older.

Acknowledgements

The authors thank and acknowledge the contribution of fertility clinics in Australia and New Zealand for providing ART treatment and outcome data to ANZARD.

Authors’ roles

C.V., G.C. and K.H. designed the study. Y.J.L. performed the statistical analysis and wrote the manuscript. C.V., N.Z. and K.H. provided statistical support. C.V., N.Z., G.C. and K.H. contributed to the interpretation of the results and editing of the manuscript. All authors approved the final version of the manuscript.

Funding

No external funding was used for this study.

Conflict of interest

The Fertility Society of Australia funds the National Perinatal Epidemiology and Statistics Unit to manage ANZARD and conduct national reporting of ART in Australia and New Zealand. Associate Professor Georgina Chambers (G.C.) is employed by the University of New South Wales (UNSW) and is director of the National Perinatal Epidemiology and Statistics Unit at UNSW. G.C. was also a paid member of the Australian governments Medicare Benefits Scheme taskforce on assisted reproductive technologies in 2017.