perinatal outcome following suspected fetal abnormality when managed through a fetal management unit

PRENATAL DIAGNOSISPrenat Diagn 2010; 30: 149–155.Published online 8 January 2010 in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/pd.2431

Perinatal outcome following suspected fetal abnormalitywhen managed through a fetal management unit

Sharon Lewis1*, George McGillivray2, Shelley Rowlands3 and Jane Halliday1

1Public Health Genetics, Murdoch Childrens Research Institute, Melbourne, Australia2Genetic Health Services Victoria, Murdoch Childrens Research Institute, Melbourne, Australia3The Royal Women’s Hospital, Melbourne, Australia

Objective To determine the perinatal outcomes of the first 1057 cases seen at a fetal management unit (FMU).

Methods Record linkage of FMU data with the Victorian Birth data and the Victorian Births Defects Register(BDR).

Results Ninety-nine percent of cases were followed up with 811 (77%) linked to the BDR and 202 (22%)linked to the birth data. Almost two-thirds of cases with birth defects (528) were live births surviving 28 days,52 (6%) were neonatal deaths, 26 (3%) were stillbirths and 205 (25%) were terminations. The birth defects mostprevalent were of the heart/circulatory system (31%). Cases that resulted in a termination were significantlymore likely to have multiple birth defects [OR 2.43 (95% CI 1.70, 3.48)], a chromosomal birth defect [OR3.30 (95% CI 1.96, 5.57)], a lethal birth defect [OR 1.32 (95% CI 1.25, 1.38)], or a syndrome [OR 4.81 (95%CI 2.54, 9.11)]. In this setting, 61% of cases of Down syndrome resulted in a live birth.

Conclusion Over three-quarters of cases referred to the FMU were confirmed with a birth defect and notifiedto the BDR. Notably, two-thirds of the cases (with or without a birth defect) were live births surviving 28 days.Copyright 2010 John Wiley & Sons, Ltd.

KEY WORDS: fetal management unit; fetal abnormality; record linkage; perinatal outcomes; concordance;termination of pregnancy

INTRODUCTION

In Victoria, Australia, approximately 96% of preg-nant women have an ultrasound (Chew et al., 2007)and approximately 70% have maternal serum screening(MSS) during their pregnancy (Moreira et al., 2008).One of the main goals of ultrasound and MSS is thedetection of fetal abnormalities (Kaasen et al., 2006;Nikkila et al., 2006; Martinez-Zamora et al., 2007;Sonek, 2007). Improved ultrasonographic (Grandjeanet al., 1999; Levi, 2002; Martinez-Zamora et al., 2007)as well as MSS technologies (Wald et al., 2003) haveincreased the likelihood of detecting fetal abnormali-ties prenatally (Levi, 2002; Garne et al., 2005; Lee andSimpson, 2007; Spencer, 2007). It has been proposedthat the ultrasound is the most important tool for detect-ing fetal abnormalities prenatally (Bijma et al., 2008).

Often there is uncertainty about the clinical outcomefor the fetal abnormality, while consideration of false-positives is an important aspect of any screening or diag-nostic process (Grandjean et al., 1999; Martinez-Zamoraet al., 2007). Adequate interpretation of the complex sit-uation may require a multidisciplinary approach (Lukset al., 2003; Bijma et al., 2004b; Bijma et al., 2007). Insuch a model, clinicians from various specialities such asobstetrics, genetics, neonatology, paediatrics and medi-cal imaging combine expertise as required to accurately

*Correspondence to: Sharon Lewis, Public Health Genetics,Murdoch Childrens Research Institute, Melbourne, Australia.E-mail: [email protected]

diagnose any fetal abnormality and develop a prenataland postnatal management plan. To date, the few stud-ies looking at the multidisciplinary model have focussedon the perinatal management decisions made in such asetting (Luks et al. 2003; Bijma et al., 2004a, 2007).

At the Royal Women’s Hospital (RWH), a tertiary-level hospital in Melbourne, Australia, a multidisci-plinary fetal management unit (FMU) was establishedin 1999. Currently, the pregnancy outcomes for manyof the women attending this clinic are unknown. Theaims of this study were to: (1) determine the perina-tal outcome of the cases attending the FMU in thefirst 5 years (1999–2004); (2) examine the lethality ofany conditions diagnosed; and (3) determine the concor-dance between prenatal and postnatal diagnosis.

MATERIALS AND METHODS

Setting

The FMU at the Royal Women’s Hospital, Melbourne,Australia, is a multidisciplinary clinic geared to theaccurate diagnosis and comprehensive management ofpregnancies complicated by structural, chromosomal andsyndromal fetal abnormalities. Referrals to the FMUcome from within the hospital, from other metropolitanand regional hospitals, and practising obstetricians andgeneral practitioners statewide, and from other states inAustralia. The FMU is one of two such multidisciplinaryclinics in Melbourne, Victoria, and is the largest.

Copyright 2010 John Wiley & Sons, Ltd. Received: 2 October 2009Revised: 19 November 2009

Accepted: 21 November 2009Published online: 8 January 2010

150 S. LEWIS ET AL.

Study period

The study period was 1999 to 2004.

Data sources

There were three data sources used in this study. Theyincluded:

1. Hospital data(a) The FMU database is an in-house database held

at the RWH. During the study period there were1057 cases seen at the FMU.

(b) The Clinical Lookup and Acknowledgement(CLARA) database held at the RWH con-tains details of a patient’s clinical investigationsincluding the results of ultrasound scans, pathol-ogy and autopsy reports. If a pregnant womanwas seen on more than one occasion at the FMU,the final diagnosis was used.

2. Birth DataThe Perinatal Data Collection Unit (PDCU) at the

Department of Human Services in Victoria, Australia, isthe repository of data resulting from mandatory reportingof all births equal to or over 20 weeks’ gestation (or400 g in weight if gestation is unknown). Each birth isassigned a unique identifier (Riley et al., 2005). Duringthe study period there were 377 788 births in Victoria.3. Birth Defects Register

The Victorian Birth Defects Register (BDR) is asubset of the PDCU and collects information on allbirth defects for live births, stillbirths and terminationsof pregnancy (TOPs) for birth defects occurring since1 January 1982. The BDR collects information ondiagnoses made in a child up to 18 years of age andis updated as a new diagnosis of a birth defect isnotified (Riley and Halliday, 2006). Each individualcase may have up to 13 birth defects coded using theBritish Paediatric Association Classification of DiseasesSupplement to the International Statistical Classificationof Diseases and Related Health Problems (ICD—9threvision).

The ascertainment and completeness of data in theBDR has been examined in validation studies and hasbeen shown to be excellent for the majority of condi-tions with the exception of syndromes involving pro-found developmental delay (Riley et al., 2004). Duringthe study period, there were 18 492 cases notified tothe BDR.

Record linkage

Probabilistic record linkage using the computer softwareLinkageWiz (version 4.1) was undertaken to match therecords from the FMU database with the data held atthe BDR and PDCU. Initially, records from the FMUand BDR were matched using the common identifyingvariables and potential matches with low scores werethen manually checked to identify true matches. Any

cases from the FMU database that did not match tothe BDR were subsequently matched to the PDCU birthdatabase. For cases from the FMU that did not match tothe BDR or the PDCU database, their hospital recordswere examined for further information.

Grouping of birth defects/abnormalitiesinto body systems

Birth defects/abnormalities were grouped into body sys-tem groups by reassigning the birth defect/abnormalitycode into the three-digit body system groupings codesas used by BDR (Riley and Halliday, 2006). These arenot individual cases and one case may have had manybody system birth defects reported.

Lethality groupings

For this study, the lethality of the birth defect(s)/ab-normality diagnosed was coded on the basis of perinatalmortality as measured during 1983 to 1994, before pre-natal screening was utilised widely in Victoria (Davidsonet al., 2005). A birth defect was coded as ‘lethal’ whenthere was a >50% likelihood of perinatal death, ‘Maybelethal’ 15% to 50% and ‘Non-lethal’ less than 15%.

Concordance between pre- and postnataldiagnoses

Following record linkage to the BDR, the birth defectsrecorded for each matched FMU case in the BDR werecompared with the final diagnosis recorded in the hospi-tal CLARA database for each case. Cases were classifiedas ‘completely concordant’ if all defect(s) recorded post-natally were diagnosed prenatally, ‘incompletely concor-dant’ when either the primary or secondary diagnosiswas concordant but not both, or ‘non-concordant’ whenthe abnormality recorded postnatally was not diagnosedprenatally.

Cases where the final diagnosis recorded in the hospi-tal CLARA database did not match to the BDR weresubsequently matched to the PDCU birth record andalso examined for concordance. Cases were classifiedas: ‘completely concordant’ if no birth defect(s) wererecorded both pre- and postnatally; ‘likely resolved’ ifthe birth defect(s) recorded prenatally were likely tohave resolved between the time of diagnosis prenatallyand birth; and ‘non-concordant’ if the birth defect(s)diagnosed prenatally were not recorded postnatally andwere unlikely to have resolved during pregnancy.

Data analysis

All data analyses were undertaken in SPSS 15.0.1 forWindows. The statistical significance of differences inproportions for univariable analyses was assessed usinglogistic regression. Results are expressed as odds ratios(OR) with 95% confidence interval (CI) and p values.

Copyright 2010 John Wiley & Sons, Ltd. Prenat Diagn 2010; 30: 149–155.DOI: 10.1002/pd

PERINATAL OUTCOME FOLLOWING SUSPECTED FETAL ABNORMALITY 151

FMU data 1999-2004(n = 1057)

BDR data 1999-2004(n = 18,492)

Linked to

811 matches (77%) 246 non-matches (23%)Birth data 1999-2004

(n = 377,788)

Linked to

528 (65%) live births surviving 28 days52 (6%) neonatal deaths26 (3%) stillbirths (≥ 20 weeks gestation)205 (25%) terminations

44 non-matches (4%)202 matches (19%)

192 (95%) live births surviving 28 days10 (5%) stillbirths (≥ 20 weeks gestation)

15 Interstate11 terminations < 20 weeks6 no abnormalities on u/s3 FDIU < 20 weeks

9 unaccounted for (1%)

Figure 1—Linkage of FMU cases to the BDR and PDCU databases

Ethical approval

Permission for access to the PDCU and the BDR wasobtained from the Consultative Council on Obstetric andPaediatric Mortality and Morbidity, the auspicing bodyfor these databases. Ethics approval to undertake theproject was obtained from the Victorian Department ofHuman Services [ID no: 15/06] and the Royal Women’sHospital Human Research Ethics Committees [ID no:06/08].

RESULTS

Record linkage

Of the 1057 cases seen at the FMU in 1999 to 2004inclusive, 811 (77%) were linked to the BDR (Figure 1).Two-thirds of these pregnancies resulted in live birthssurviving more than 28 days. Only a small numberresulted in a neonatal death or still birth (6% and3%, respectively), while approximately 25% resulted ina TOP.

The 246 FMU cases that did not link to the BDR werelinked to the birth data for the same time period, themajority being live births surviving more than 28 days(95%) although there were a small number of stillbirths(5%) (Figure 1).

The hospital records for the 44 FMU cases that did notmatch to the BDR or the birth data were examined. Therecords revealed that 15 of the FMU cases were referredfrom interstate and therefore would not be recorded asVictorian births, 11 cases were TOPs <20 weeks, threewere fetal deaths in utero before 20 weeks, and for sixcases there were no abnormalities found on ultrasound.

As shown in Figure 1, there were nine further cases (1%)that were unaccounted for in either the linkage or thehospital records.

Study population

Maternal characteristics and obstetric history of thelinked FMU cases (n = 1013) were compared with theVictorian population in 2004 (Table 1). The maternalage distribution was statistically different (p ≤ 0.001),the women in the study population being younger thanthe Victorian population. Also, compared with the Vic-torian population in 2004, there were significantly morewomen from a Middle Eastern country (p ≤ 0.001), andsignificantly less women from Oceania (p ≤ 0.001) orEurope (p = 0.01). There were no differences betweenthe study population and the Victorian population withregards to previous pregnancies (gravidity) or births(parity).

Birth defect/abnormalities groupings

The birth defects most prevalent were those of theheart/circulatory system (31%). The next most preva-lent birth defects were of the genital/urinary system(19%), limb/musculoskeletal system (19%), nervous sys-tem (11%), cleft palate or lip or both or digestive sys-tem (6.6%), respiratory system (3.4%), chromosomalabnormality (3.3%), eye or ear, face, neck (1.8%), andother and specified (5.9%). Five-hundred-and-forty cases(67%) had an isolated birth defect, two-hundred-and-two(25%) had multiple birth defects and sixty-nine (8%) hadchromosomal abnormalities (data not shown).


152 S. LEWIS ET AL.

Table 1—Demographics of women in study compared toVictorian population in 2004

Study womenn = 1013

Victoria 2004n = 62 348

Variable n % %

Maternal age 1013<15 1 0.1 015–19 55 5.4 2.820–24 204 20.1 11.325–29 321 31.7 25.830–34 276 27.2 37.635–39 126 12.4 18.840–44 28 2.8 3.545+ 2 0.2 0.1

Country of Birth 1013Oceania 753 74.3 78.7∗Europe 39 3.8 5.7∗Middle East 80 7.9∗ 2.3Asia 88 8.7 9.8S + N America 8 0.8 1.2Africa 30 2.9 2.0Unknown 15 1.5 0.3∗

∗ p ≤ 0.001.Mean age = 28.44 years (SD 5.7). χ2 (7df) = 282.8; p ≤ 0.001.The numbers shown for each variable differs from the overall totalnumber in study population (n = 1013) because not all respondentshave completed fields for all variables.

Lethality groupings

One-hundred-and-eighty-one cases (22%) had birth de-fects that were coded as ‘lethal’, 292 (36%) of birthdefects were coded as ‘maybe lethal’, and 288 (36%)of birth defects coded as ‘non-lethal’. The remaining 50cases (6%) were unable to be coded because there wasno concrete information available on perinatal mortalityon that birth defect.

Comparing TOP versus non-TOP

FMU cases with multiple birth defects or a chromosomalbirth defect were, respectively, 2.4 and 3.3 times more

likely to result in a TOP than those cases with an isolatedbirth defect (Table 2). FMU cases with a ‘maybe lethal’birth defect or ‘lethal’ birth defect were 5 times and 11times, respectively, more likely to result in a TOP thanFMU cases with a non-lethal birth defect; and FMUcases with a syndrome were almost five times morelikely to result in a TOP than FMU cases without asyndrome. Of the FMU cases resulting in a TOP, only11% were for birth defects that were coded as non-lethal(Table 2).

FMU cases with birth defects of the nervous sys-tem, and limb or other musculoskeletal body systemwere more likely to result in a termination of the preg-nancy (χ2 = 15.82, p < 0.001; χ2 = 18.90, p < 0.001,respectively), while FMU cases with birth defects of theheart or circulatory and genital or urinary body systemwere more likely to result in a continuation of the preg-nancy (χ2 = 60.84, p < 0.001; χ2 = 9.32, p = 0.002,respectively). Of the FMU cases with isolated birthdefects that resulted in a continuation of the pregnancy,

Table 3—Referral reasons for FMU cases that did not matchto BDR but did match to PDCU database

Referral reason Frequency Percent

Ventriculomegaly 31 15.3Cyst 19 9.4Echogenic focus 16 7.9Pyelectasis 15 7.4Limb anomaly 11 5.4Kidney anomaly 11 5.4Talipes 10 5.0Exposure to infection 10 5.0Heart anomaly 10 5.0Hydramnios 9 4.5Brain anomaly 9 4.5Genetic History 8 4.0Pleural effusion 5 2.5CCAM 4 2.0Increased nuchal 4 2.0Cleft Lip 3 1.5Pericardial effusion 3 1.5Other 24 11.9Total 202 100

Table 2—Univariable analysis of the association between birth defect grouping and the outcome termination of pregnancy (TOP)

TOP n = 205 No TOP n = 606

Variable n % n % OR 95% CI p

Type of malformationIsolated 102 49.8 438 72.3 — — —Multiple 73 35.6 129 21.3 2.43 1.70, 3.48 <0.001Chromosomal 30 14.6 39 6.4 3.30 1.96, 5.57 <0.001

Lethal birth defect

No 20 11 268 47 — — —Maybe 84 45 208 36 5.41 3.22, 9.10 <0.001Yes 83 44 98 17 11.35 6.61, 19.48 <0.001

Syndrome

No 180 87.8 589 97.2 — — —Yes 25 12.2 17 2.8 4.81 2.54, 9.11 <0.001



congenital hydronephrosis was the most common (n =97) with 97% of all congenital hydronephrosis casescontinuing, followed by all talipes equinovarus cases(n = 23), all gastroschisis cases (n = 23) and 19 out of27 (70.3%) hypoplastic left heart syndrome cases. Thelargest group continuing with a chromosomal abnormal-ity was Down syndrome (n = 18), with 11 cases (61%)resulting in a live birth.

FMU cases linked to PDCU birth data, butnot birth defects data

The referral reasons for cases that linked to the PDCUbut not the BDR (n = 202) are shown in Table 3.The most common referral reasons were for cerebralventriculomegaly (n = 31, 15.3%), cysts within a fetalorgan (n = 19, 9.4%), echogenic foci within a fetalorgan (n = 16, 7.9%) or renal pylectasis (n = 15, 7.4%).

Concordance between pre- and postnataldiagnosis

As shown in Table 4a, for cases with complete data forprenatal diagnoses (n = 757), concordance between pre-and postnatal diagnosis was demonstrated in 96% ofcases, for either the primary or secondary diagnosis, ofFMU cases that matched to the BDR. Non-concordancewas demonstrated in 31 (4%) of cases where thebirth defect(s) diagnosed postnatally was not diagnosedprenatally. These birth defects were most often those ofthe heart/circulatory system with 13 (42%) of the non-concordant cases falling into this category.

For FMU cases that did not match to the BDR,but were found as a birth in the PDCU data, com-plete concordance between pre- and postnatal diagno-sis was found for over half the cases (n = 106, 55%),i.e. no birth defect was recorded pre- or postnatally(Table 4b). These cases included those referred to theclinic for a suspected abnormality which was recordedas either being ruled out during further investigations atthe clinic, or which resolved during the pregnancy. Ina quarter of the cases (n = 55, 28%), it was not clearif resolution had occurred, but they were classified as‘likely resolved’ because many of the birth defects were

Table 4—Concordance between pre- and postnatal diagnosis

n %

(a) Cases matched to the BDR with complete data sets(n = 757)Complete concordance 630 83Incomplete concordance 96 13Non-concordance 31 4

(b) Cases that did not match to the BDR but were matched tothe PDCU with complete data sets (n = 194)

Complete concordance 106 55Likely resolved 55 28Non-concordance 33 17

renal pelvic dilatation (n = 14) or borderline ventric-uolmegaly (n = 11). Non-concordance was observed in17% of cases (n = 33).

DISCUSSION

In this study, we were able to follow up the pregnancyoutcomes for 99% of cases of the first 1057 cases thatwere referred during the study period to a multidisci-plinary FMU that is geared to the accurate diagnosis andcomprehensive management of pregnancies complicatedby structural, chromosomal and syndromal fetal abnor-malities. There appeared to be appropriate referral ofcases to the FMU with over three-quarters of them con-firmed with a birth defect and notified to the VictorianBDR. Overall, two-thirds of the cases (with or withouta birth defect) were live births surviving 28 days.

Maternal demographics for the study population werecompared to the state Victorian population, rather thanthe hospital population, as the referrals for the FMUare received from health practitioners and hospitalsstatewide and not just internally from hospital clinics.The study population was shown to be younger thanthe state population. One might have predicted that thestudy population would be older as maternal age is aknown risk factor for having a baby with a birth defect(Riley and Halliday, 2006). However, this risk is forhaving a baby with a chromosomal abnormality suchas Down syndrome (trisomy 21) (Riley and Halliday,2006), which is increasingly being detected in the firsttrimester as a result of pregnancy screening, and referralsto the FMU for this and other major chromosomalabnormalities do not usually occur.

There were more women from a Middle Easterncountry and fewer women from Oceania in the studypopulation than are seen in the state birth population.It has been shown in the BDR that women from aMiddle Eastern country are more likely to have a babywith a birth defect than women from Oceania (Rileyand Halliday, 2006). Therefore, it is not surprising thatwe found more women from a Middle Eastern countryattending the FMU.

In almost two-thirds of FMU cases where a birthdefect was diagnosed, the pregnancy continued andresulted in a live birth surviving more than 28 days.In cases diagnosed with a birth defect, the proportionof terminations (25%) is at the lower end of a rangeof proportions (15–59%) reported for 17 registries inEurope (Garne et al., 2005). The proportion is similarto the 14% reported in a French study evaluating thenumber of elective TOPs on medical grounds amongpregnancies with a suspicion of a fetal abnormalityreferred to a specialist centre (Guillem et al., 2003).TOP occurred in 205 (19%) of all cases referred andof these, only 20 (9%) were in the presence of a birthdefect which was coded as non-lethal. In interpretingthis data we do need to be mindful that 50 cases wereunable to be assigned to a lethality coding. Nevertheless,a TOP for a fetal anomaly is not a common outcome andwe concur with the statement from Wyldes et al. that


154 S. LEWIS ET AL.

one is ‘chosen when a lethal or seriously handicappingcondition is diagnosed prenatally’ (Wyldes and Tonks,2007).

The high percentage of live births in cases of Downsyndrome noted in the FMU population (61%) was anunexpected finding, as we have found in state-wide pop-ulation studies that 5.3% of pregnancies with a prena-tal diagnosis of Down syndrome were not terminated(Collins et al., 2008). However, as the majority of thewomen who attend the FMU are first seen after a second-trimester ultrasound, the diagnosis of Down syndromemay be too late for consideration of a pregnancy ter-mination. During the period this study was undertaken,the Victorian law did not recognise fetal abnormalityper se as an indication for performing a TOP. For atermination to be lawful, the medical practitioner autho-rising and performing the procedure must conclude theabortion to be ‘necessary to preserve the woman froma serious danger to her life or her physical or mentalhealth’ and that the risk of the abortion is ‘not out ofproportion to the danger to be averted’ (‘The Menhennittrule’).

A limitation of this study is that although we doknow gestation at diagnosis, we did not have informationon gestation of first visit to the FMU, which could behelpful in interpreting these findings.

There was high concordance between pre- and post-natal diagnoses for the FMU cases that matched to theBDR with 90% of these cases concordant either com-pletely or with at least the primary or secondary diag-nosis. For the cases that did not match to the BDRbut did have a birth record, concordance was not ashigh with only 55% of cases being completely concor-dant. The remainder of these cases would be classifiedas false-positives as the birth defects diagnosed prena-tally were not recorded postnatally. A somewhat similarstudy investigated the false-positives related to prena-tal ultrasound screening of fetal structural anomalies ina centre with a similar set-up to the FMU (Martinez-Zamora et al., 2007). They found that 9.3% of all pre-natal diagnoses were false-positives, while the Eurofe-tus Study, a multicentre hospital-based study, observedfalse-positive diagnoses of 9.9% (Garne et al., 2005).To look at our data in the same way, we need tocombine the groups that we have classified as ‘likelyresolved’ and ‘non-concordance’ in Table 4b. When wedo this, we have a similar percentage of false-positivediagnoses at 10.8% ((55 + 33)/(55 + 33 + 630 + 96)).As reported by Martinez-Zamora’s article (Martinez-Zamora et al., 2007) and in a Strasbourg population-based registry (Stoll et al., 1995), we found that uri-nary tract anomalies, specifically renal pelvic dilatation,ventriculomegaly, cysts and short limbs were often thefalse-positive diagnoses.

CONCLUSION

In conclusion, our study shows that of the pregnantwomen who attend the FMU during the study period,the majority of pregnancies known to be affected with

a birth defect continued to live birth. Further studiesof the experience of attendance at the FMU from apatient perspective would add to our interpretation ofthis finding. This would enable us to determine whatinformation the patients have received and understoodwith respect to the health implications of the suspectedabnormality, and of the support systems in place. Thiswould facilitate our understanding of the decision-making process of the patients after diagnosis of a fetalabnormality, a decision known to involve many complexprocesses (Bijma et al., 2008).

We suggest that there are two main explanationsfor the surprisingly high proportion of pregnanciescontinuing in the presence of a birth defect. The first isthe late gestation at which many of these birth defects arediagnosed. The second includes intrinsic reasons relatedto FMU staff counselling and support that may facilitatethe choice for the continuation of an affected pregnancy.The first issue may be as a result of a problem faced inall maternity care settings, but the second highlights theimportance of a multidisciplinary clinic with its multi-faceted mix of health professionals. This could enableand empower the pregnant woman and her partner tomake informed decisions regarding the management oftheir pregnancy.

REFERENCES

Bijma HH, Schoonderwaldt EM, van der Heide A, Wildschut HI, vander Maas PJ and Wladimiroff JW. 2004a. Ultrasound diagnosisof fetal anomalies: an analysis of perinatal management of 318consecutive pregnancies in a multidisciplinary setting. Prenat Diagn24(11): 890–895.

Bijma HH, Wildschut HI, van der Heide A, van der Maas PJand Wladimiroff JW. 2004b. Obstetricians’ agreement on fetalprognosis after ultrasound diagnosis of fetal anomalies. PrenatDiagn 24(9): 713–718.

Bijma HH, Van der Heide A, Wildschut HI, Van der Maas PJand Wladimiroff JW. 2007. Impact of decision-making in amultidisciplinary perinatal team. Prenat Diagn 27(2): 97–103.

Bijma HH, van der Heide A and Wildschut HI. 2008. Decision-making after ultrasound diagnosis of fetal abnormality. ReprodHealth Matters 16(31 Suppl): 82–89.

Chew C, Halliday JL, Riley MM and Penny DJ. 2007. Population-based study of antenatal detection of congenital heart diseaseby ultrasound examination. Ultrasound Obstet Gynecol 29(6):619–624.

Collins VR, Muggli EE, Riley M, Palma S and Halliday JL. 2008. IsDown syndrome a disappearing birth defect? J Pediatr 152(1): ,20–24 24. e21.

Davidson N, Halliday J, Riley M and King J. 2005. Influence ofprenatal diagnosis and pregnancy termination of fetuses withbirth defects on the perinatal mortality rate in Victoria, Australia.Paediatr Perinat Epidemiol 19(1): 50–55.

Garne E, Loane M, Dolk H, et al. 2005. Prenatal diagnosis of severestructural congenital malformations in Europe. Ultrasound ObstetGynecol 25(1): 6–11.

Grandjean H, Larroque D and Levi S. 1999. The performance ofroutine ultrasonographic screening of pregnancies in the EurofetusStudy. Am J Obstet Gynecol 181(2): 446–454.

Guillem P, Fabre B, Cans C, Robert-Gnansia E and Jouk PS. 2003.Trends in elective terminations of pregnancy between 1989 and2000 in a French county (the Isere). Prenat Diagn 23(11): 877–883.

Kaasen A, Tuveng J, Heiberg A, Scott H and Haugen G. 2006.Correlation between prenatal ultrasound and autopsy findings: Astudy of second-trimester abortions. Ultrasound Obstet Gynecol28(7): 925–933.



Lee MJ and Simpson LL. 2007. Major fetal structural malformations:The role of new imaging modalities. Am J Med Genet Part C SeminMed Genet 145C: 33–44.

Levi S. 2002. Ultrasound in prenatal diagnosis: polemics around rou-tine ultrasound screening for second trimester fetal malformations.Prenat Diagn 22(4): 285–295.

Luks FI, Carr SR, Feit LR and Rubin LP. 2003. Experience with amultidisciplinary antenatal diagnosis and management model infetal medicine. J Matern Fetal Neonatal Med 14(5): 333–337.

Martinez-Zamora MA, Borrell A, Borobio V, et al. 2007. Falsepositives in the prenatal ultrasound screening of fetal structuralanomalies. Prenat Diagn 27(1): 18–22.

Moreira C, Muggli E and Halliday J. 2008. Prenatal DiagnosticTesting in Victoria.Murdoch Childrens Research Institute&Victorian Government Department of Human Services: Melbourne.

Nikkila A, Rydhstroem H, Kallen B and Jorgensen C. 2006.Ultrasound screening for fetal anomalies in southern Sweden:a population-based study. Acta Obstet Gynecol Scand 85(6):688–693.

Riley M and Halliday J. 2006. Birth Defects in Victoria, 2003–2004.Victorian Government Department of Human Services: Melbourne.

Riley M., Phyland S and Halliday J. 2004. Validation study of theVictorian Birth Defects Register. J Paediatr Child Health 40(9–10):544–548.

Riley M, Davey MA and King J. 2005. Births in Victoria 2003–2004.Melbourne: Perinatal Data Collection Unit, Public Health, VictorianGovernment Department of Human Services, available online:http://www.dhs.vic.gov.au/phd/perinatal/pubs.htm.

Sonek J. 2007. First trimester ultrasonography in screening anddetection of fetal anomalies. Am J Med Genet C Semin Med Genet145C(1): 45–61.

Spencer K. 2007. Aneuploidy screening in the first trimester. Am JMed Genet C Semin Med Genet 145C(1): 18–32.

Stoll C, Dott B, Alembik Y and Roth MP. 1995. Evaluation of routineprenatal diagnosis by a registry of congenital anomalies. PrenatDiagn 15(9): 791–800.

Wald N, Rodeck C, Hackshaw AK, Walters J, Chitty L and Mackin-son AM. 2003. First and second trimester antenatal screening forDown’s syndrome: the results of the Serum, Urine and UltrasoundScreening Study (SURUSS). Health Technol Assessment 7(11):1–88.

Wyldes MP, Tonks AM. 2007. Termination of pregnancy for fetalanomaly: a population-based study 1995 to 2004. Bjog 114(5):639–642.


perinatal outcome following suspected fetal abnormality when managed through a fetal management unit

Documents