Unexplained Infertility Rates Concepts

8/13/2019 Unexplained Infertility Rates Concepts

http://slidepdf.com/reader/full/unexplained-infertility-rates-concepts 1/9

ORIGINAL ARTICLE Infertility

Unexplained infertility: overallongoing pregnancy rate and modeof conceptionM.Brandes 1,*, C.J.C.M.Hamilton 1, J.O.M.van der Steen 1, J.P.de Bruin 1,R.S.G.M. Bots 2, W.L.D.M.Nelen 3, and J.A.M. Kremer 3

1 Department of Obstetrics and Gynaecology, Jeroen Bosch Hospital, P.O. Box 90153, 5200 ME ’s-Hertogenbosch, The Netherlands2Department of Obstetrics and Gynaecology, St Elisabeth Hospital, Tilburg, The Netherlands

3Department of Obstetrics and Gynaecology,

Radboud University Medical Centre Nijmegen, Heyendaal, The Netherlands

*Corresponding address. E-mail: [email protected]

Submitted on September 17, 2010; resubmitted on October 31, 2010; accepted on November 16, 2010

background: Unexplained infertility is one of the most common diagnoses in fertility care. The aim of this study was to evaluate theoutcome of current fertility management in unexplained infertility.methods: In an observational, longitudinal, multicentre cohort study, 437 couples were diagnosed with unexplained infertility and wereavailable for analysis. They were treated according to their prognosis using standing national treatment protocols: (i) expectant management– IUI–IVF (main treatment route), (ii) IUI–IVF and (iii) directly IVF. Primary outcome measures were: ongoing pregnancy rate, patient ow over the strategies, numbers of protocol violation and drop out rates. A secondary outcome measure was the prediction of ongoing pregnancyand mode of conception.results: Of all couples 81.5% (356/437) achieved an ongoing pregnancy and 73.9% (263/356) of the pregnancies were conceived spon-taneously. There were 408 couples (93.4%) in strategy-1, 21 (5.0%) in strategy-2 and 8 (1.8%) in strategy-3. In total, 33 (7.6%) couplesentered the wrong strategy. There were 104 couples (23.8%) who discontinued fertility treatment prematurely: 26 on doctor’s advice(with 4 still becoming pregnant) and 78 on their own initiative (with 33 still achieving a pregnancy). Predictors for overall pregnancy

chance and mode of conception were duration of infertility, female age and obstetrical history.conclusions: Overall success rate in couples with unexplained infertility is high. Most pregnancies are conceived spontaneously. Werecommend that if the pregnancy prognosis is good, expectant management should be suggested. The prognosis criteria for treatment withIUI or IVF needs to be investigated in randomized controlled trials.

Key words: unexplained infertility / pregnancy / drop-out / treatment strategy / prediction model

IntroductionUnexplained infertility is one of the most common diagnoses in a fer-tility clinic (Hull et al ., 1985 ; Adamson and Baker, 2003 ; Brandes et al .,2010 ). In the absence of a specic medical cause a specic treatment

for unexplained infertility is lacking. Instead, these couples are exposedto several empirical treatments, among which medical assisted repro-ductive treatments such as, Clomiphene Citrate ( Hughes et al ., 2010 ),controlled ovarian hyperstimulation combined with intrauterine inse-mination (COH/IUI) ( Verhulst et al ., 2009 ), and/or in vitro fertilization(IVF) (Pandian et al ., 2009 ) with or without intracytoplasmic sperminjection (ICSI). Steures et al . (2006) demonstrated in a randomizedcontrolled trial (RCT) that in couples with a spontaneous pregnancyprognosis between 30 and 40%, 6 months of COH/IUI led to the

same percentage of pregnancies as 6 months of expectant manage-ment. Whether treatment in couples with a poorer prognosis issuperior to expectant management is, unfortunately, not yet suf-ciently investigated in RCTs.

Overall outcome of unexplained infertility can be studied in cohort

studies. Hull et al . (1985) published cumulative pregnancy curves for unexplained infertility. The overall pregnancy rate (including pregnancyloss) was about 70%. However, this was in the pre-IVF era, andwithout adjustment for confounders as female age and duration of infertility.

In the years thereafter, prognostic models were constructed topredict spontaneous pregnancies ( Eimers et al ., 1994 ; Collins et al .,1995 ; Snick et al ., 1997 ; Hunault et al ., 2004 ). Besides predictionmodels for spontaneous pregnancy, also prediction models for

& The Author 2010. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.For Permissions, please email: [email protected]

Human Reproduction, Vol.26, No.2 pp. 360–368, 2011Advanced Access publication on December 16, 2010 doi:10.1093/humrep/deq349



pregnancy after treatment exist, for example, prediction of pregnan-cies after IUI (Tomlinson et al ., 1996 ; Khalil et al ., 2001 ; MontanaroGauci et al ., 2001 ; Steures et al ., 2004 ) and prediction of pregnanciesafter IVF (Templeton et al ., 1996 ; Stolwijk et al ., 1996a ; Lintsen et al .,2007 ). These prognostic models predict the pregnancy chance of indi-vidual fertility treatments. In daily practice, however, the treatment of rst choice is often followed by other therapies if no pregnancy occurs,for example, an unsuccessful series of IUI is often followed by IVF. Thementioned prediction models do not take these treatment strategiesinto account. Yet, patients could be interested in their overallongoing pregnancy chance irrespective of the therapy. Is it possibleto predict this overall ongoing pregnancy chance before any treatmenthas been started? This information could be helpful for couples whendeciding if fertility treatment is worth trying.

The aim of this study is to investigate the outcome of currentevidence-based fertility management of unexplained infertility in anobservational study. This information could be useful for the counsel-ling of the infertile couples in the outpatient clinic.

Materials and Methods

Study population and settingA longitudinal multicentre cohort study was performed. The cohort ( n ¼

2476) consisted of all infertile couples referred by their general prac-titioner (GP) to the Jeroen Bosch Hospital (JBZ), ‘s-Hertogenbosch ( n ¼

1391), St Elisabeth Hospital (EZ), Tilburg ( n ¼ 944) and Radboud Univer-sity Nijmegen Medical Centre (RUNMC), Nijmegen ( n ¼ 141), in theNetherlands. Couples from the JBZ and the EZ entered the studybetween 2002 and 2006, the couples referred to the RUNMC enteredthe study in 2004. General data of the JBZ cohort were described inour previous study ( Brandes et al ., 2010).

JBZ and EZ are large, secondary care, training hospitals both offering IVFin collaboration with the IVF licensed laboratory of the EZ. RUNMC is a

tertiary care, IVF licensed and academic training hospital. In theNetherlands, GPs hardly perform any fertility work-up. They rather refer couples to a gynaecologist if they fail to conceive despite unprotectedintercourse for at least 1 year.

For this study, couples with unexplained infertility were selected for analysis (n ¼ 443). Relevant information on medical fertility history for both male and female partner as well as the results of the standard fertilitywork-up and outcome of fertility treatment was collected in a comprehen-sive electronic health record (Fertibase w ). Patients were followed untiltheir rst ongoing pregnancy. Ongoing pregnancy was dened as positivefetal cardiac activity at 12 weeks amenorrhea. Type of pregnancy (single-ton or multiple), miscarriages and ectopic pregnancies were not analysedin this study.

Because of the 5-years inclusion period, the follow-up period per

couple varied. For the JBZ and the RUNMC the database was closed inDecember 2008, for the EZ the database was closed in July 2009. Thefollow-up period of the patients was between 2 years (for the couplesentering the infertility clinic in December 2006) and 8 years (for thecouples entering the infertility clinic in January 2002). At the moment of database closure and analyses of the data, some couples ( n ¼ 14) werestill receiving treatment and six couples had discontinued fertility care,directly after fertility work-up was completed.

Fertility work-upAll couples in any of the three hospitals went through a standard fertilitywork-up. This work-up was according to the national guidelines and

consisted of: history taking of both partners; semen analysis; cycle moni-toring, including ultrasonographic ovulation detection, preovulatory post-coital test (PCT) and mid-luteal progesterone; and tubal work-up, includ-ing Chlamydia antibody test (CAT), hysterosalpingography (HSG) or laparoscopy with dye test.

Couples were diagnosed as unexplained infertility is the ndings in thestandard work-up were normal. Couples who conceived an ongoing preg-nancy before fertility work-up was completed, were also diagnosed with

unexplained infertility, if at least no abnormalities were found until thatmoment. A positive CAT was overruled by normal laparoscopic ndings.

Basal FSH was not tested routinely in all couples. Only if the womenwere older than 36 years or had a positive family history of prematureovarian failure, basal FSH level was measured ( van der Steeg et al .,2007 ). Women who started IVF were all tested for basal FSH, in order to choose the correct dosage of recombinant FSH ( Broekmans et al .,2009 ).

Fertility treatmentAfter completing the basic fertility work-up, the treatment strategy wasdetermined according to the spontaneous pregnancy chance of thecouple calculated with the models of Eimers et al. (1994) and Hunaultet al. (2004) . These models include the following variables: female age,infertility type, outcome of PCT, percentage of motile sperm and referralstatus. The cut-off point between good and poor prognosis was 30%spontaneous ongoing pregnancy chance in the following 6–12 months(Hunault et al ., 2004 , Steures et al ., 2006 ). As a result of this policy thefollowing three therapeutic strategies could be distinguished:

Strategy-1 (EXP–IUI–IVF ¼ main treatment route): In case of a goodprognosis ( ≥ 30%), expectant management (EXP) was recommendedfor up to 2 years of infertility, followed by 6 cycles COH/IUI. If no preg-nancy was achieved, three cycles of IVF were recommended.

If the chance of conceiving spontaneously was , 30%, fertility treatmentwas started and the following two strategies could be followed:

Strategy-2 (IUI–IVF): If female age was , 38 years, three to six COH/

IUI cycles were offered before IVF was started.Strategy-3 (direct IVF): If female age was ≥ 38 years, IVF was offereddirectly.

The cut-off point of 38 years female age was based on the drop in preg-nancy rate after IUI in our clinics among women older than 38 (unpub-lished data).

For all three strategies, in case of a fertilization failure in regular IVF, thetreatment was converted to ICSI in a subsequent cycle. IVF, ICSI andtransfer of cryopreserved embryos will all be indicated as ‘IVF’ in thispaper.

Drop outDecisions to stop treatment were either on the patients’ or the doctors’

initiative. In order to gain more information about the dropouts, per-mission was obtained from the hospital’s Medical Ethics Committees tocontact those patients who were documented without an ongoing preg-nancy. Couples were asked about the reason to stop fertility care,about spontaneous pregnancies and about fertility treatments at another hospital and the outcome of such treatment. The data of this inquirywere included into the analysis. General data concerning drop-out of the JBZ cohort were described in a previous study ( Brandes et al ., 2009).

Outcome measuresPrimary outcome measure of this study were cumulative overall ongoingpregnancy rate (OPR) and mode of conception, patient ow, number of

Outcome with unexplained infertility 361



protocol violations (i.e. couples that did not follow doctors’ advise) anddrop out rates. As secondary outcome measures, prediction modelswere constructed for OPR and mode of conception after a follow-upperiod of 2 years. One model was constructed for the total cohort(n ¼ 437) and one for those couples that entered the main treatmentroute and did not discontinue treatment prematurely on their owninitiative (n ¼ 331).

StatisticsIndependent t -test, x

2 test and Fishers’ Exact test were used, if appropri-ate, by comparing patient characteristics between couples who did andcouples who did not achieve an ongoing pregnancy. Realistic cumulativeOPRs were calculated for the total cohort and per prognosis group.P , 0.05 was considered as a statistically signicant difference.

In order to inform patients about their overall ongoing pregnancychance, the predicted overall pregnancy chance including mode of con-ception, was visualized. Two prediction tables were established. Therst includes all couples in the cohort ( n ¼ 437). The second includes allcouples who entered the main treatment route and did not discontinueprematurely on their own initiative. Pregnancy status was assessed 2years after their rst visit. Variables in the univariate analysis with P ,

0.15 were selected for multivariate analysis. Female FSH was excludedfrom the multivariate analysis because 42% of the data were missing. Fur-thermore, male age was excluded because of interaction with female age.Despite the P -value in univariate analysis, ‘obstetric history’ was stillincluded in the multivariate analysis as its predictive value has beenproven before.

Multinomial logistic regression analysis was performed because thedependent variable consisted of four nominal outcome measures (nopregnancy, spontaneous pregnancy, pregnancy after IUI and pregnancyafter IVF). OPRs including mode of conception were predicted for thetotal cohort and for those couples who entered the main treatmentroute without prematurely discontinuing on patient’s initiative, after a2-year follow-up period. This interval was chosen as all couples had afollow-up period of at least 2 years. Furthermore, the number of couples who conceived after that period was limited. No pregnancy wasused as reference. Only variables that showed P , 0.15 in the multivariatemodel, were maintained, others were excluded.

The following two variables were used in the rst model: female age(≤ 30, 31–35, ≥ 36 years) and duration of infertility ( , 12, 12– 23, ≥ 24months). The following three variables were used in the second model:female age (≤ 30, 31– 35, ≥ 36 years), duration of infertility ( , 12, 12– 23, ≥ 24 months) and obstetrical history (primary or secondary). Analyseswere performed using SPSS 16.0 program (Statistical Package for the SocialSciences; SPSS, Chicago, IL, USA).

Results

Study populationOf the total cohort ( n ¼ 2476), 443 (18%) were labelled as unex-plained subfertile. Six couples dropped out after completing thework-up. These couples were excluded for further analysis, resultingin a study group that consisted of 437 couples.

Table I shows the patient characteristics for this group, and for thecouples who did ( n ¼ 356) and did not conceive an ongoing pregnancy(n ¼ 81). Those who conceived had a signicantly lower mean female(P , 0.001) and male ( P , 0.001) age, a shorter mean duration of infertility (P ¼ 0.02) and a lower female basal FSH level (P ¼ 0.02).

Other patient characteristics showed no signicant difference. At thetime of database closure, 14 couples were still receiving treatment.

Patient ow Figure 1 shows, in a ow diagram, the routes patients have followed.The majority of the couples entered the main treatment route, startingwith expectant management ( n ¼ 408). More than half (228, 55.9%)

of these couples became pregnant in this rst step.In total, 184 couples started treatment (IUI and/or IVF) of whom

128 (69.6%) achieved an ongoing pregnancy. The majority, 93(50.5%), conceived as a result of treatment but 35 (19.0%) still con-ceived spontaneously despite treatment.

Ongoing pregnancy Cumulative ongoing pregnancy curves for the total unexplained infer-tility cohort are shown in Fig. 2. These curves are an exact represen-tation of our cumulative OPRs in time and are not Kaplan–Meier curves. In a 5-year follow-up period, 81.5% (356/437) of the patientsconceived an ongoing pregnancy. Of the pregnancies, 98.6% wereconceived within 3 years after rst visit. Of the pregnancies, 73.9%(263/356) were conceived spontaneously. IUI and IVF were respon-sible for 12.6% (45/356) and 13.5% (48/356) of all pregnancies,respectively.

In general, 42.3% of the couples who received IUI and 62.8%of those who received IVF achieved a pregnancy. Yet, 14.7% of couples who were treated with IUI and 11.7% of couples who weretreated with IVF conceived spontaneously. Amongst couples whowere treated with IUI then discontinued treatment, 32.2% still con-ceived spontaneously. This gure was 18.1% if they discontinued IVF.

About half of the pregnancies were conceived in the rst 6 monthsafter the rst visit at the fertility clinic, and most pregnancies were con-ceived within 1 year the after rst visit. IUI pregnancies occurred

mainly between 6 and 18 months after rst visit, while IVF pregnanciesoccurred between 12 and 30 months after rst visit.

Treatment strategies and protocol violationCouples were divided, directly after completing the work-up, intogood ( ≥ 30%) and poor ( , 30%) prognosis groups, according to theprediction model for spontaneous pregnancy of Hunault et al .(2004). Table II shows the division of the couples in the good andpoor prognosis groups, over the three different treatment strategies.Most couples (91.3%) had a good prognosis after completing the fer-tility work-up, 97.0% of whom started in the recommended strategy,namely Strategy-1.

Of the couples with a poor prognosis, 55% started, despite their indication for treatment, with expectant management. This wasmainly because couples preferred to postpone fertility treatment.

As can be seen in Fig. 1, 13 couples received IVF after a period of expectant management. The majority of these couples had already anIVF indication but wanted to wait ( n ¼ 9). Others claimed an IVFtreatment without having a proper indication ( n ¼ 3) and onecouple started IVF right after expectant management, because thehusband had to start medication for a physical illness.

There were 12 couples in the good prognosis group who enteredStrategy-2 ( n ¼ 7) or Strategy-3 ( n ¼ 5) wrongly. There were sixcouples were treated directly with IUI because they had a duration

362 Brandes et al.



of infertility . 24 months and one couple was treated directly becausethey insisted on treatment. For Strategy-3, two couples were directlytreated with IVF because of a combination of female age and durationof infertility, two couples because of the duration of infertility and onecouple because they insisted on this treatment.

In the good prognosis group, 83.7% conceived an ongoing preg-nancy and in 65.4% the mode of conception was spontaneous. In

the poor prognosis group, the OPR was 57.9 and 26.3% of the preg-nancies were conceived spontaneously ( P , 0.001 and P , 0.001).

Overall time to pregnancy was 8.1 months in the good prognosisgroup versus 14.0 months in the poor prognosis group ( P ¼ 0.005).The time to spontaneous pregnancy was 5.7 months in the goodprognosis group versus 12.1 months in the poor prognosis group(P ¼ 0.02).

.............................................................................................................................................................................................

Table I Patient characteristics for the total cohort and for those who conceived an ongoing pregnancy and for those whodid not.

Total cohort(n 5 437)

Ongoing pregnancy (n 5 356)

No ongoingpregnancy ( n 5 81)

P -value* P -value** P -value***

Female age (Mean + SD) 31.8 + 4.6 31.3 + 4.3 34.1 + 4.9 , 0.001 0.033 0.064

Male age (Mean + SD) 33.8 + 5.3 33.3 + 4.9 35.9 + 6.3 , 0.001 – –

Duration of infertility (Mean + SD) 21.0 + 15.9 20.2 + 14.7 24.6 + 19.2 0.02 0.001 0.004

Female FSH (Mean + SD) 7.3 + 3.0 7.0 + 2.6 8.0 + 3.9 0.02 – –

Female BMI (Mean+ SD) 23.7 + 4.4 23.7 + 4.2 23.5 + 5.1 0.68 – –

TMSC (Mean + SD) 96.5 + 86.3 94.7 + 83.2 103.3 + 97.8 0.43

Obstetrical history

Primary infertile 294 238 56 0.69 – 0.015

Secondary infertile 143 118 25

Missing – – –

Smoking male

Yes 122 99 23 0.99

No 292 237 55

Missing 23 20 3Smoking female

Yes 96 81 15 0.41

No 320 258 62

Missing 21 17 4

Alcohol use male

Yes 306 255 51 0.06 - -

No 95 71 24

Missing 36 30 6

Alcohol use female

Yes 216 179 37 0.37

No 204 162 42

Missing 17 15 2CAT

Positive 40 29 11 0.2

Negative 349 283 66

Missing 48 44 4

PCT

Positive 375 300 75 –

Negative – – –

Missing 62 56 6

*Univariate analysis, comparison between ‘ongoing pregnancy’ and ‘no ongoing pregnancy’.**Multivariate analysis for the total cohort.**Multivariate analysis for the total cohort, excluding the couples that did not follow the main treatment route, and couples that prematurely discontinued fertility care on their owninitiative.Female and male age in years at time of rst visit, duration of infertility in months at time of rst visit, FSH, follicle stimulating hormone (IU/l); BMI, body mass index (kg/m 2 ); TMSC,total motile sperm count ( × 10 6 ) of the rst semen analysis; CAT, Chlamydia antibody titer; PCT, post-coital test.




DiscontinuationAbout one quarter ( n ¼ 104) of the couples discontinued fertility carebefore completing the strategy and before achieving a pregnancy.During expectant management and during IUI treatment, discontinu-ation was mainly on patients’ initiative (93.0 and 89.3%, respectively).During IVF treatment, discontinuation was mainly based on doctors’initiative (60.6%).

In total, 78 of 104 couples discontinued prematurely on their owninitiative. For couples who discontinued anywhere in the strategy,female (P , 0.001) and male ( P , 0.001) age was higher, durationof infertility was longer (P , 0.001), female FSH was higher ( P ¼

0.05) and the male partner more frequently used alcohol ( P ¼ 0.01)compared with couples who continued treatment. ‘Emotional pro-blems’ (22/104) and ‘poor response (doctor’s initiative)’ (20/104)were the most common reasons to stop fertility care.

As is shown in Fig. 1, 42.3% (33/78) of the couples who discontin-ued on their own initiative, still spontaneously conceived an ongoing

pregnancy. This gure was 15.4% (4/26) for the couples that discon-tinued on doctors’ advice.

Prediction of ongoing pregnancy Figure 3a–c show in bars the predicted overall ongoing pregnancychance and the predicted mode of conception in the 2 years after rst visiting the fertility clinic.

Figure 3a is based on the total cohort ( n ¼ 437). For this group,female age and duration of infertility predict ongoing pregnancychance and mode of conception.

Figure 3b and c are based on all couples who entered the maintreatment route and did not discontinue treatment prematurely ontheir own initiative ( n ¼ 331). For this group, female age, durationof infertility and obstetrical history predict ongoing pregnancychance and mode of conception.

We used bar diagrams to express the results of the predictionmodels, as they can be useful tools in patient counselling at the end

Figure 1 Flow diagram of the total cohort.

364 Brandes et al.



of the fertility work-up, when the diagnosis of unexplained infertility ismade. They visualize the predicted ongoing pregnancy chance and themode of conception within 2 years.

For example, the female partner of a primary infertile couple is 32years old. The duration of infertility is 16 months. The overall OPR,based on the total cohort (Fig. 3a), will be about 75% (the secondtriplet of bars represents female age 30–36 years, and the middlebar of that triplet represents 12–24 months infertility). Furthermore,it is predicted that about 55% of the couples will conceive spon-taneously and that 10% will conceive by IUI and 10% by IVF.

If the same couple will not discontinue the treatment strategy prema-turely, the overall OPR will be about 80% (note, the diagram for primary

infertility is selected rst, in this example Fig. 3b) than again according toage andduration of infertility theright bar is sought. Thisbar indicates thatthe about 55% of the couples will conceive spontaneously, and that now 15% would conceive with IUI and 10% with IVF.

DiscussionIn this study, we showed that the overall success rate in an unselectedcohort of couples with unexplained infertility, newly referred bytheir GP, is quite high. Most pregnancies (75%) were conceivedspontaneously.

In couples diagnosed with unexplained infertility, the pregnancyrate, including pregnancy loss, before the era of IVF was 70% ( Hullet al ., 1985 ). Although the two cohorts cannot be compared directly,a small increase of the OPR as a result of the introduction of IVFseems likely (Brandes et al ., 2009 ). The cumulative OPRs in our study represent realistic gures and are not based on Kaplan–Meier curves in which drop outs are considered to have the same prognosisas couples who continued treatment ( Stolwijk et al ., 1996b ).

In our setting, couples were in principle treated according to thepresent national guidelines ( www.nvog.nl ), which can be deducedfrom the ow diagram. In this study, for the rst time insight is giveninto the proportion of couples for whom the initial therapy remainedunsuccessful and further steps needed to be taken. The treatmentstrategies were established by prognosis. Female age and duration of infertility indirectly inuenced treatment policy as they are part of the prediction model of Hunault et al . (2004).

Our study, also, gives insight into the proportion of couples who,for any reason, did not follow the strategy they were allocated to.As we already know from previous studies, there is a large variationin infertility care in different clinics and a large variation in thedegree national infertility guidelines are implemented ( Mourad et al .,

2008 ; Haagen et al ., 2010 ). However, as shown in our study, the fol-lowed national guidelines are likely to be effective as many pregnancies

Figure 2 Cumulative ongoing pregnancy curve for the total cohortunexplained infertility (n ¼ 437) with special attention for the contri-bution per mode of conception. Note: these curves are an exact rep-resentation of the cumulative OPRs, not Kaplan–Meier curves. IUI,intrauterine insemination; IVF, in vitro fertilization, including IVF,

ICSI, frozen embryo transfer and Oocyte donation.

.............................................................................................................................................................................................

Table II Subdivision in strategies per prognosis group according to Hunault’s prediction model for spontaneouspregnancy.

Ongoingpregnancies, n (%)

Spontaneous, n (%) IUI, n (%) IVF, n (%)

Prognosis ≥ 30%

Strategy-1 ( n ¼ 387) 324 83.7 253 65.4 35 9.0 36 9.3

Strategy-2 ( n ¼ 7) 6 85.7 2 28.6 2 28.6 2 28.6

Strategy-3 ( n ¼ 5) 4 80.0 1 20.0 – – 3 60.0

Subtotal ( n ¼ 399) 334 a 83.7 256 b 64.2 37 c 9.3 41 d 10.3Prognosis , 30%

Strategy-1 ( n ¼ 21) 12 57.1 2 9.5 4 19.0 6 28.6

Strategy-2 ( n ¼ 14) 9 64.3 7 50.0 – – 2 14.3

Strategy-3 ( n ¼ 3) 1 33.3 1 33.3 – – – –

Subtotal ( n ¼ 38) 22 a 57.9 10 b 26.3 4 c 10.5 8 d 21.1

Total ( n ¼ 437) 356 81.5 266 60.9 41 9.4 49 11.2

IUI, conception after intrauterine insemination; IVF, conception after in vitro fertilization, including ICSI, frozen embryo transfer and oocyte donation.x

2 or Fisher’s Exact test if appropriate.a P , 0.001; b P , 0.001; cP ¼ 0.77; d P ¼ 0.06.


http://www.nvog.nl/

http://www.nvog.nl/

http://www.nvog.nl/

http://www.nvog.nl/



were conceived spontaneously and average pregnancy rates wereachieved after treatment. Steures et al . (2006), also recommendedexpectant management for couples with a pregnancy prognosis of atleast 30%, because in an RCT COH/IUI did not result in higher preg-nancy rates. In case of a good prognosis, we also recommend anexpectant management instead of starting IVF as long as superiority

has not been explored sufciently yet. However, we think it is wisethat ovarian reserve for couples at risk for ovarian failure are measuredby basal FSH or anti-mullerian hormone before expectant manage-ment is started.

We showed that a relatively high number of couples discontinuetreatment before the treatment strategy has been completed. If

Figure 3 Prediction bars for ongoing pregnancy after a follow-up of 2 years. The bars indicate the overall predicted ongoing pregnancy chance. Thesubdivision in gray and black indicate how these pregnancies—according to the prediction model—will be achieved. ( a ) Total cohort ( n ¼ 437). Pre-diction bars are based on female age and duration of infertility. ( b ) All primary infertile couples from the cohort, excluding couples entering a different

treatment route than the main treatment route (exp-IUI-IVF) and excluding couples that discontinued prematurely on their own initiative ( n¼

214).Prediction bars are based on female age, duration of infertility and obstetrical history. ( c) All secondary infertile couples from the cohort, excludingcouples entering a different treatment route than the main treatment route (exp-IUI-IVF) and excluding couples that discontinued prematurely on their own initiative (n ¼ 117). Prediction bars are based on female age, duration of infertility and obstetrical history.

366 Brandes et al.



couples discontinue on their own initiative their overall ongoing preg-nancy chance could be inuenced. However, we agree that for somecouples this is the best option, as fertility treatment could lead tosevere emotional distress ( Olivius et al ., 2004a ,b; Smeenk et al .,2004 ; Verhaak et al ., 2005; Rajkowa et al ., 2006; Verberg et al .,2008 ; Brandes et al ., 2009 ).

The already existing prediction models predict pregnancy for onlyone treatment step. For example, some models ( Eimers et al ., 1994 ;Collins et al ., 1995 ; Snick et al ., 1997 ; Hunault et al ., 2004 ) onlypredict spontaneous pregnancy and others models ( Templetonet al ., 1996 ; Tomlinson et al ., 1996; Stolwijk et al ., 1996a ; Khalilet al ., 2001 ; Montanaro Gauci et al ., 2001; Steures et al ., 2004;Lintsen et al ., 2007 ) only predict pregnancy after IUI or IVF treatment.Our model, on the other hand, predicts overall pregnancy chancesand the mode of conception. The models are applicable for thetotal cohort and for couples who followed the main treatmentroute (EXP–IUI–IVF) without premature discontinuation on patients’initiative, respectively.

Therefore, the histograms which show per subgroup the predictedongoing pregnancy chance and mode of conception could be used in

the outpatient clinic in daily practice to inform couples with unex-plained infertility before treatment is started.

Pregnancy chance and mode of conception is predicted for a 2years period. We think this is a clinical relevant time frame as fromour cumulative ongoing pregnancy curves show only a fraction of the pregnancies were conceived after 2 years of follow-up. Further-more, it is realistic that couples complete their strategy within 2 years.

The variables included in our models showed in previous studies,already, their value for predicting pregnancy. Increasing female agelowers the chance of achieving both a spontaneous ( Hunault et al .,2004 ) and an induced ongoing pregnancy ( Templeton et al ., 1996 ;Steures etal .,2004 ; Lintsen etal .,2007 ). If thecouple conceived previouslythey have more chance conceiving again ( Templeton et al ., 1996 ; Hunault

et al ., 2004 ; Lintsen et al ., 2007 ). And thirdly, a long duration of infertilitynegatively inuences ongoing pregnancy chances ( Templeton et al ., 1996 ;Hunault et al ., 2004 ; Steures et al ., 2004; Lintsen et al ., 2007). However,other than the used variables there could be some other, yet unknown,variables that inuence pregnancy rate.

A strong point of our study is that a complete cohort was followed.Couples who discontinued follow-up were contacted. This was possiblesince a comprehensive electronic health record was used. Furthermore,in contrast to RCTs where strict inclusion criteria apply and manycouples are excluded or do not want to participate; our approach ismore representative for daily medical practice. In real life, thereusually is a larger variation in patient characteristics, patient preferencesand therapies than in tidily managed randomized controlled studies.

But, this study design could also be judged as a weakness, becauseRCTs are ranked higher on the level of evidence pyramid ( Aartmanand van Loveren, 2007 ; Bluhm, 2009). However, observationalstudies are valuable in the evaluation of the outcome of evidence-based management in daily practice. Conclusions from observationalstudies may point into the direction of future RCTs ( Glasziou et al .,2007 ; Vanderbroucke et al ., 2008 ).

Furthermore, our model has to be validated with an independentdataset (external validation). However, it can only be useful in dailypractice in clinics that have a similar treatment policy and the sametreatment population (GP referred). IVF, for example, could have a

larger contribution to ongoing pregnancy in clinics that treat moreaggressively. Whether this inuences time to pregnancy and overallOPR should be subject of further research.

In conclusion, overall success rate in, GP referred, couples withunexplained infertility are high. Most pregnancies were conceivedspontaneously. We, therefore, recommend that if the pregnancy prog-nosis is good, treatment should at least begin with expectant manage-ment. Future randomized controlled trials have to nd out when (i.e.below which percentage of expected pregnancy chance) other treat-ment modalities have to be initiated. Female age, duration of infertilityand obstetrical history are predictors for ongoing pregnancy and modeof conception. These variables can inuence the treatment of rstchoice.

Authors’ rolesM.B.: concept and design, acquisition of data, analysis and interpret-ation of data; drafting the article; nal approval of the version to bepublished. C.J.C.M.H., W.L.D.M.N.: concept and design, analysisand interpretation of data; revising the article for important intellectualcontent; nal approval of the version to be published. J.O.M.S.: acqui-sition of data, analysis and interpretation of data; drafting the article;nal approval of the version to be published. J.P.B., R.S.G.M.B., J.A.M.K.: concept and design; revising the article for important intellec-tual content; nal approval of the version to be published.

AcknowledgementsWe would like to thank Daisy Wopereis for her input in this study.

ReferencesAartman IHA, van Loveren C. Onderzoeksontwerpen en de ladder van

evidence. Ned Tijdschr Tandheelkd 2007;114 :161–165.Adamson GD, Baker VL. Subfertility: causes, treatment and outcome. Best

Pract Res Clin Obstet Gynaecol 2003;17 :169–185.Bluhm R. Some observations on observational research. Perspect Biol Med

2009; 52 :252–263.Brandes M, van der Steen JOM, Bokdam SB, Hamilton CJCM, de Bruin JP,

Nelen WLDM, Kremer JAM. When and why do subfertile couplesdiscontinue their fertility care? A longitudinal cohort study in asecondary care subfertility clinic. Hum Reprod 2009;24 :3127–3135.

Brandes M, Hamilton CJCM, de Bruin JP, Nelen WLDM, Kremer JAM. Therelative contribution of to the total ongoing pregnancy rate in asubfertile cohort. Hum Reprod 2010;25 :118–126.

Broekmans FJ, Soules MR, Fauser BC. Ovarian aging: mechanisms andclinical consequences. Endocr Rev 2009;30 :465–493.

Collins JA, Burrows EA, Willan AR. The prognosis for live birth amonguntreated infertile couples. Fertil Steril 1995;64 :22–28.

Eimers JM, te Velde ER, Gerritse R, Vogelszang ET, Looman CWN,Habbema JDF. The prediction of the chance to conceive in subfertilecouples. Fertil Steril 1994;61 :44–52.

Glasziou P, Chalmers I, Rawlins M, McCulloch P. When are randomisedtrials unnecessary? Picking signal from noise. Br Med J 2007;334 :349–351.

Haagen EC, Nelen WL, Grol RP, Braat DD, Hermens RP, Kremer JA.Variation in guideline adherence in intrauterine insemination care.Reprod Biomed Online 2010;20 :533–542.




Hughes E, Brown J, Collins JJ, Vanderkerchove P. Clomiphene citrate for unexplained subfertility in women. Cochrane Database Syst Rev 2010;1:CD000057.

Hull MGR, Glazener CMA, Kelly NJ, Conway DI, Foster PA, Hinton RA,Coulson C, Lambert PA, Watt EM, Desai KM. Population study of causes, treatment, and outcome in infertility. Br Med J 1985;291 :1693–1697.

Hunault CC, Habbema JDF, Eijkemans MJC, Collins JA, Evers JLH, te

Velde ER. Two new prediction rules for spontaneouspregnancy leading to live birth among subfertile couples, base donthe synthesis of three previous models. Hum Reprod 2004;19 :2019–2026.

Khalil MR, Rasmussen PE, Erb K, Laursen SB, Rex S, Westergaard PG.Homologous intrauterine insemination. An evaluation of prognosticfactors based on a review of 2473 cycles. ACTA Obstet Gynecol Scand 2001;80 :74–81.

Lintsen AME, Eijkemans MJC, Hunault CC, Bouwmans CAM, Hakkaart L,Habbema JDF, Braat DDM. Predicting ongoing pregnancy chances after IVF and ICSI: a national prospective study. Hum Reprod 2007;22 :2455–2462.

Montanaro Gauci M, Kruger TF, Coetzee K, Smith K, van der Merwe JP,Lombard CJ. Stepwise regression analysis to study male and female

factors impacting on pregnancy rate in an intrauterine inseminationprogramme. Andrologia 2001;33 :135–141.

Mourad SM, Nelen WL, Hermens RP, Bancsi LF, Braat DD, Zielhuis GA,Grol RP, Kremer JA. Variation in subfertility care measured byguideline-based performance indicators. Hum Reprod 2008;23 :2493–2500.

Olivius C, Friden B, Borg G, Bergh C. Why do couples discontinuein vitro fertilization treatment? A cohort study. Fertil Steril 2004a;81 :258–261.

Olivius C, Friden B, Borg G, Bergh C. Psychological aspects of dis-continuation of in vitro fertilization treatment. Fertil Steril 2004b; 81 :276.

Pandian Z, Bhattacharya S, Vale L, Templeton A. In vitro fertilisationfor unexplained subfertility. Cochrane Database Syst Rev 2009;1:CD003357.

Rajkhowa M, Mcconnell A, Thomas GE. Reasons for discontinuation of IVFtreatment: a questionnaire study. Hum Reprod 2006;21 :358–363.

Smeenk JMJ, Verhaak CM, Stolwijk AM, Kremer JAM, Braat DDM.Reasons for dropout in an in vitro fertilization/intracytoplasmic sperminjection program. Fertil Steril 2004;81 :262–268.

Snick HKA, Snick TS, Evers JLH, Collins JA. The spontaneous pregnancyprognosis in untreated subfertile couples: the Walcheren primary carestudy. Hum Reprod 1997;12 :1582–1588.

Steures P, van der Steeg JW; for CECERM (Collaborative Effort onthe Clinical Evaluation in Reproductive Medicine). Prediction of anongoing pregnancy after intrauterine insemination. Fertil Steril 2004;82 :45–51.

Steures P, van der Steeg JW, Hompes PG, Habbema JD, Eijkemans MJ,Broekmans FJ, Verhoeve HR, Bossuyt PM, van der Veen F, Mol BW;for CECERM (Collaborative Effort on the Clinical Evaluation inReproductive Medicine). Intrauterine insemination with controlled

ovarian hyperstimulation versus expectant management for coupleswith unexplained subfertility and an intermediate prognosis: arandomised clinical trial. Lancet 2006;368 :216–221.

Stolwijk AM, Zielhuis GA, Hamilton CJCM, Straatsma H, Hollanders JMG,Goverde HJM, van Dop PA, Verbeek ALM. Prognostic models for theprobability of achieving an ongoing pregnancy after in-vitro fertilizationand the importance of testing their predictive value. Hum Reprod 1996a; 11 :2298–2303.

Stolwijk AM, Hamilton CJCM, Bastiaans LA, Zielhuis GA. A more realisticapproach to the cumulative pregnancy rate after in vitro fertilization.Hum Reprod 1996b; 11 :660–663.

Templeton A, Morris JK, Parslow W. Factors that affect outcome of in-vitro fertilisation treatment. Lancet 1996;348 :1402–1406.

Tomlinson MJ, Amissah-Arthur JB, Thompson KA, Kasraie JL, Bentick B.

Prognostic indicators for intrauterine insemination (IUI): statisticalmodel for IUI success. Hum Reprod 1996;11 :1892–1896.

Verberg MFG, Eijkemans MJC, Heijnen EMEW, Broekmans FJ, de Klerk C,Fauser BCJM, Macklon NS. Why do couples drop-out from IVFtreatment? A prospective cohort study. Hum Reprod 2008;23 :2050–2055.

Verhaak CM, Smeenk JM, Evers AW, van Minnen A, Kremer JA,Kraaimaat FW. Predicting emotional response to unsuccessful fertilitytreatment: a prospective study. J Behav Med. 2005;28 :181–190.

Verhulst SM, Cohlen BJ, Hughes E, Heineman MJ, Te Velde E. Intrauterineinsemination for unexplained subfertility. Cochrane Database Syst Rev 2009; 4:CD:001838.

Vandenbroucke JP, Psaty BM. Benets and risks of drug treatments: how to combine the best evidence on benets with the best data about

adverse effects. J Am Med Assoc 2008;300 :2417–2419.Van der Steeg JW, Steures P, Eijkemans MJC, Habbema JDF,

Hompes PGA, Broekmans FJ, Bouckaert PXJM, Bossuyt PMM, vander Veen F, Mol BWJ; for CECERM (collaborative effort for clinicalevaluation in reproductive medicine study group). Predictive value andclinical impact of basal follicle-stimulating hormone in subfertile,ovulatory women. JCEM 2007;92 :2163–2168.

368 Brandes et al.

Documents

Unexplained Infertility Rates Concepts