Lessons from the Australian and New Zealand Assisted Reproduction Database ANZARD

1
Lessons from the Australian and New Zealand
Assisted Reproduction Database (ANZARD)

• Peter Illingworth
• Medical Director
• IVFAustralia

2
Why audit IVF results?

• Benchmark performance
• Review usage and practice
• Study outcomes

3
System prior to 2002

• Clinics calculated own results and posted single
  summative report to NPSU
• Detailed paper page (two sides) filled out for
  each pregnancy

4
Why change?

• Multiple datasets required for different bodies
• Most of past perinatal dataset of limited
  significance
• Imprecise definitions (e.g. clinical pregnancy)
• No implantation data creating pressures to
  transfer more embryos to improve crude pregnancy
  rate
• No mechanism for verification

5
ANZARD Working Group (2000-2001)

• Peter Illingworth (Chair)
• Andrew Speirs
• Doug Saunders
• Gordon Baker
• John Tyler

• Ian Johnston
• Vivienne McLachlan
• Di Craven
• Paul Lancaster
• Tara Hurst
• Jishan Dean

Director of NPSU (since 2001) Liz Sullivan
6
ANZARD New features

• Comparison of implantation rates
• Verifiable datasets
• Standardised definitions
• Inclusion of new aspects for perinatal study
• Blastocyst transfer
• Testicular/immature sperm
• PGD

7
New definitions clinical pregnancy

• A clinical pregnancy must fulfil one of the
  following criteria
• known to be ongoing at 20 weeks
• evidence by ultrasound of an intrauterine sac
  (with or without a fetal heart)
• examination of products of conception reveal
  chorionic villi
• a definite ectopic pregnancy has been diagnosed
  laparoscopically or by ultrasound

8
ANZARD Cycles recorded

• All oocyte pickup cycles. This includes natural
  or unstimulated cycles.
• All cancelled cycles where FSH has been
  administered with the intention of ART.
• All cycles where frozen embryos are thawed
  regardless of the intention or outcome of the
  thawing process.
• All cycles where artificial insemination is
  performed using donated sperm.
• Each occasion where embryos are either donated or
  moved into or out of an IVF Unit from a different
  unit.

9
ANZARD Cycles not recorded

• Any cancelled cycle where FSH is not
  administered.
• Ovulation induction with gonadotrophins
• Artificial insemination using partners sperm

10
ANZARD Data transmission
Units
NPSU
11
ANZARD the spreadsheet
12
ANZARD Fields
13
Clinical details
14
Past pregnancies
15
Fate of eggs
16
Sperm and fertilisation
17
Embryos transferred
18
Clinical pregnancy
19
Pregnancy complications
20
Maternal complications
21
Babies
22
Morbidity
23
ANZARD Use of the data
Units
ANZARD
State and National Governments
NPSU
24
ANZARD Privacy arrangements
Units
ANZARD

• Access to de-identified data (Fields 1-3 removed)
  is under the control of FSA Council subject to
  approval by Univ NSW Human Research Ethics
  Committee
• Data identifiable by Unit MRN is only available
  to
• Staff of NPSU (AIHW Confidentiality agreement)
• Staff of the original Unit
• RTAC

NPSU
25
ANZARD Reporting times

• Cycles are considered part of a time interval
  (e.g. Jan-June 2003) according to the Cycle Date

26
ANZARD Reporting times

• Cycles are considered part of a time interval
  (e.g. Jan-June 2003) according to the Cycle Date

27
ANZARD Reporting times

• Cycles are considered part of a time interval
  (e.g. Jan-June 2003) according to the Cycle Date

E.g. If a cycle is started on 29th December 2001
and OPU is on 12th January the cycle is counted
as 2001
28
Verification process

• RTAC is supplied with a list of cycles and
  pregnancies
• RTAC asks to see a sample of case records to
  verify evidence of pregnancy
• RTAC checks Unit diaries to ensure that all
  cycles (including cancelled cycles) are entered

29
Advantages for Units

• Units download from existing databases
• No paper pregnancy outcome sheets to complete
• National/State reporting requirements covered
• Rapid intelligent reports with confidence
  intervals and comparison of many parameters of
  performance with Australia/NZ average

30
PerfectIVF Pty Ltd
Our pregnancy rate is 50
31
PerfectIVF Pty Ltd
Our pregnancy rate is 50
??????
32
Progression of IVF
33
Calculating a pregnancy rate (1)

• The denominator
• Cycles started?
• Egg pickups?
• Embryo transfers?
• Embryo thaw cycles?

34
Calculating a pregnancy rate (2)

• The numerator
• Positive hCG?
• Clinical pregnancy?
• Fetal heart?
• Livebirth?
• BESST (Birth emphasising successful singleton
  term)?

35
Calculating a pregnancy rate (3)

• What do you do about frozen embryo pregnancies?
• Per thaw cycle?
• Per egg pickup

36
Interpreting a pregnancy rate

• Which population?
• Any age limits?
• Does the clinic do artificial insemination first?

37
Assisted reproduction technology in Australia
and New Zealand 2003
38
2003 Summary (1)

• During 2003, 39,720 treatment cycles were
  attempted in Australia and New Zealand. Of these,
  90.7 (36,040) took place in Australia and 9.3
  (3,680) in New Zealand
• In Australia, there were 8.4 treatment cycles
  per 1,000 women of reproductive age (15?44
  years). Correspondingly, in New Zealand, there
  were 4.2 cycles per 1,000 women of reproductive
  age (15?44 years).
• More than half (53.8) of ART procedures involved
  fresh, non- donor oocytes or embryos, almost a
  third (31.9) used frozen, non- donor embryos,
  5.7 used oocytes or embryos received from a
  donor and 7.8 of cycles were intra- uterine
  insemination using donated sperm.

39
2003 Summary (2)

• The average age of women giving birth was 34.4
  years, 4.9 years older than the average age of
  Australian mothers in 2003 (29.5 years).
• The average gestational age of all babies was
  37.2 weeks. More than a quarter (26.6) of babies
  were born preterm with a gestational age of less
  than 37 weeks. This is a lower proportion than
  that reported in 2000 (32.6), suggesting
  improved outcomes for babies following assisted
  reproduction.

40
Activity in assisted reproduction in Australia
and New Zealand
41
Activity in assisted reproduction in Australia
and New Zealand
39,720 cycles started in 2003
42
Success rates in assisted reproduction
43
Success rates in assisted reproduction
7,147 babies born
44
Quartiles Fresh cycles
45
Quartiles Thaw cycles
46
Pregnancies

• Overall, there were 8,365 pregnancies reported in
  the 2003 cohort. Of these pregnancies, 23.1 were
  less than 20 weeks gestation and 76.9 were at
  least 20 weeks gestation resulting in 7,479
  liveborn babies and 108 fetal deaths.

47
Effects of age

• The average age of women undergoing treatment in
  2003 was 35.2 years. Their partners were aged on
  average 37.8 years.
• Women aged 25?29 years achieved the greatest
  success, with 27.7 of initiated cycles achieving
  a live delivery. Women aged 40?44 years had a
  success rate of 6.8.

48
Age and live delivery rate
49
Age and pregnancy loss
50
Where to from here?Limitations of ANZARD

• Reporting delays
• Little demographic or ethnic data
• No domicile location
• Cycle-based not participant-based
• Some perinatal outcomes but not others

51
Perinatal mortality after assisted reproduction
52
(No Transcript)
53
(No Transcript)
54
Perinatal mortality after assisted reproduction
55
Morbidity of twins
56
Effect of transferring embryos
57
Number of embryos transferred
58
Effects of ART on congenital abnormalities
Selected studies
All studies
Favours controls
Favours IVF treatment
Hansen et al., Hum Reprod 2005
59
Effects of ART on congenital abnormalities
Selected studies
All studies
Favours controls
Favours IVF treatment
Hansen et al., Hum Reprod 2005
60
Epigenetic programming
61
Next steps for ANZARD

• Additional verification data points
• Domicile information
• Couple-based database
• Linkages with perinatal databases to improve
  perinatal data collection

62
Lessons from ANZARD

• Each unit can benchmark its performance
• Effects of multiple embryo transfer clear
• Should use consistent and meaningful quotes of
  pregnancy rates (livebirth/cycle started)
• Age a big factor in conception but a lesser one
  in miscarriage