Title How many abortions are there in Australia? A discussion of abortion statistics, their limitations, and options for improved statistical collection.
Database Library Publications
Date 14-02-2005
Source RESEARCH BRIEFS (SOCIAL POLICY SECTION)
Abstract Parliamentary Library Research Brief No. 9, February 2005
Author BIGGS, Amanda
BUCKMASTER, Luke
PRATT, Angela
Citation Id CF7F6
Cover date 14 February 2005
Item Online Text: 1200511
Key Item Yes
Major subject Statistical sources
Statistical techniques
Abortion
Minor subject Notification and reporting
South Australia
Medicare
Hospitals
Pages 22p., bibl.
Text online Yes
Volume no. 9 2004-05


How many abortions are there in Australia? A discussion of abortion statistics, their limitations, and options for improved statistical collection.

Parliament of Australia

Department of Parliamentary Services

Parliamentary Library RESEARCH BRIEF

Information analysis and advice for the Parliament

14 February 2005, no. 9, 2004–05, ISSN 1832-2883

How many abortions are there in Australia? A discussion of abortion statistics, their limitations, and options for improved statistical collection

Much of the recent debate on abortion in Australia has focused on the question of how many abortions take place in Australia each year. This brief discusses the existing statistics on abortion in Australia, their limitations for accurately quantifying how many abortions take place each year, and some options for improving Australian data on abortion numbers.

Angela Pratt, Amanda Biggs and Luke Buckmaster Social Policy Section

Contents

Introduction............................................................................................................................ 2

What is abortion? ................................................................................................................... 2

Data sources and limitations .................................................................................................. 3

Medicare-funded abortion procedures .............................................................................. 3

Figure 1: Medicare claims for procedures which may result in an abortive outcome, 1995–2004 ............................................................................................. 4

Limitations of the Medicare data.................................................................................. 4

Hospital data...................................................................................................................... 6

Figure 2: Separations for medical (induced) abortions in Australian hospitals, 1998–99 to 2002–03.............................................................................................. 7

Limitations of hospital data .......................................................................................... 7

Combining Medicare and hospital data?........................................................................... 9

South Australian data on abortions ................................................................................... 9

Options for improved statistical collection ..................................................................... 11

Medicare data ............................................................................................................. 11

Hospital data............................................................................................................... 12

Mandatory notification of abortions........................................................................... 13

Conclusions.......................................................................................................................... 13

Appendix One—Abortion record-keeping in Australian states and territories ................... 15 Endnotes............................................................................................................................... 17

www.aph.gov.au/library

How many abortions are there in Australia?

Introduction

Recent public discussion and debate over abortion in Australia has focused in large part on how many abortions take place in Australia each year. The issue returned to the headlines most recently when Senator Ron Boswell placed a series of questions on notice to the Health Minister, Tony Abbott, asking for detailed information on abortion numbers in Australia.1 Mr Abbott has previously said that 100 000 abortions take place in Australia each year (though he has also acknowledged that the absence of reliable statistics makes this figure difficult to quantify with accuracy).2

Most other commentators put the number of abortions at considerably less than 100 000—at somewhere between 70 000 to 80 000.3 The truth, however, is that currently, it is impossible to accurately quantify the number of abortions which take place in Australia. This is because there is no national data collection on abortion, there is no uniform method of data collection, collation or publication across the states and territories, and the data sources that are available all have several significant limitations. This Research Brief provides an explanation of these data sources and explains why, in the absence of a national data set on abortion, it is not possible to say how many abortions take place in Australia each year.

This Research Brief does not engage in discussion or debate about the ethics of abortion, or the issue of if, when or under what circumstances abortion is morally justified. The issue of abortion is by its nature emotive, sensitive, complex and controversial, and there are many different, deeply held, and often irreconcilable, views. Broader moral and ethical issues are clearly important, but they are beyond the scope of this paper.4 Rather, this brief seeks to provide a factual commentary and analysis of one of the key areas of the recent Australian debate.

What is abortion?

The medical definition of ‘abortion’ is the expulsion or removal of a fetus from the uterus.5 Abortion can be spontaneous, or it may be induced. The former is more commonly referred to as ‘miscarriage’, and it is the latter—that is, medically induced abortion—and the circumstances in which it takes place, which is the focus of public debate (and of this brief). Medically induced abortion is also referred to as ‘termination of pregnancy’. The terms ‘abortion’ and ‘termination of pregnancy’ will be used interchangeably throughout this brief.

The most common type of induced abortion is a surgical procedure known as a suction curette. This procedure—which usually takes about 15 minutes—involves the removal of the lining and the contents of the uterus (the fetus and placental tissue) by applying suction to the inside of the uterus with a small plastic tube.6 Abortions performed later in pregnancy involve different kinds of procedures, depending on the stage of gestation and the reason for which the abortion is being performed.7

2

A discussion of abortion statistics, their limitations, and options for improved statistical collection

As with all surgical procedures, pregnancy termination is not without risk. However, suction curettage is a simple and low risk procedure for women to undergo when performed between 7 and 12 weeks of pregnancy.8 The available data suggest that it is also overwhelmingly safe: in South Australia—which, as we discuss below, is the only Australian jurisdiction where comprehensive data on abortions are published—over the last decade, on average less than 1 per cent of women who had abortions experienced complications (and in fact, the proportion of reported complications has decreased steadily: from 1 per cent in 1994, to 0.4 per cent in 2002).9

Data sources and limitations

This section of the Research Brief discusses the three sources of publicly available data on abortion in Australia: Medicare data, hospital data and South Australian data. It also discusses the limitations of each data source for accurately quantifying the number of abortions which take place in Australia.

Medicare-funded abortion procedures

The Health Insurance Commission (HIC)—the Commonwealth agency responsible for processing Medicare claims—collects data on Medicare-funded procedures which may result in an abortive outcome. These procedures include both spontaneous abortions (or miscarriages), and medical or induced abortions (or terminations). There are two items on the Medicare Benefits Schedule (MBS) for these procedures for which Medicare benefits can be claimed: item 35643—‘Evacuation of the gravid uterus by curettage or suction curettage’— and item 16525—‘Management of second trimester labour, with or without induction, for intrauterine fetal death, gross fetal abnormality or life threatening maternal disease’.10 According to the HIC’s data, there just under 73 000 Medicare-funded procedures which may have resulted in an abortive outcome performed in Australia in 2004.

The total numbers of Medicare claims processed for the two MBS items which may result in an abortive outcome in each year over the last decade (1995–2004) are shown in Figure 1. The average number of Medicare-funded abortive procedures in the years 1995 to 2004 was approximately 75 700. As Figure 1 shows, in six out of the last ten years, the number of Medicare claims processed for procedures which may have resulted in an abortive outcome has decreased.

3

How many abortions are there in Australia?

Figure 1: Medicare claims for procedures which may result in an abortive outcome, 1995–2004

60000

65000

70000

75000

80000

85000

90000

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

No. of Medicare-funded procedures which may result in an abortion

Source: HIC Statistical Reports for item numbers 16525 and 35643

Limitations of the Medicare data

The number of Medicare claims processed for the two items on the MBS which may result in an abortive outcome is commonly cited in the public debate as the number of ‘Medicare-funded abortions’.11 However, MBS items which may result in abortive outcomes also apply to procedures which are not pregnancy terminations, such as those undertaken as a result of miscarriage or fetal death, or other gynaecological conditions not necessarily related to pregnancy. Therefore Medicare claims data on these item numbers includes claims for procedures which are not pregnancy terminations per se.

It is not possible to determine with any degree of precision what proportion of Medicare claims for these item numbers are for pregnancy terminations, since Medicare claims for actual abortions cannot be disaggregated from the other procedures claimed under these item numbers when the Medicare claim is lodged and processed (and therefore they are not disaggregated in the HIC’s data on the number of claims processed for these item numbers). Estimates as to what proportion of Medicare claims under these item numbers are for pregnancy terminations vary greatly:

• Dr David Molloy, President of the National Association of Specialist Obstetricians and

Gynaecologists, has been quoted as saying that up to one third of procedures processed

4

A discussion of abortion statistics, their limitations, and options for improved statistical collection

under item 35643 (the Medicare item under which the vast majority of procedures which may result in an abortive outcome are claimed) are curettes for miscarriages rather than abortions12

• Dr Geoffrey Brodie, medical director of Australian Birth Control Services (which operates

a chain of Sydney clinics), has been quoted as saying that the proportion of procedures claimed under item 35643 for ‘nonviable’ pregnancies (that is, miscarriages or cases of intrauterine fetal death13) is more likely to be around 15 per cent,14 and possibly as low as 3 to 6 per cent,15 and

• Dr Andrew Pesce, the Australian Medical Association’s (AMA’s) obstetrics and

gynaecology spokesperson, has been quoted as saying that in his practice, 90 per cent of the procedures claimed under item 35643 are for nonviable pregnancies. While there is no accurate way of determining how many services claimed under item 35643 are for abortions, Dr Pesce estimates that, nationally, half, or possibly three-quarters, of all Medicare claims for item 35643 would be for nonviable pregnancies.16

According to these estimates, the number of abortions funded by Medicare each year could range from around 20 000 to around 65 000.

Another limitation of the HIC data is that it only includes procedures undertaken on private patients in clinics or hospitals who claim a Medicare rebate. This is because women who have abortions as public patients in public hospitals do not need to claim a Medicare rebate: public hospital treatment is provided free of charge to all Australians who choose to be treated as public patients (Medicare rebates only apply to treatment provided on a fee-for-service basis).17 This also means that the Medicare data is not sensitive to differences in the principal type of facility for performance of abortions in different states and territories.18

For example, the available evidence suggests that the majority of pregnancy terminations in Australia take place in private facilities.19 However, in South Australia, more than half of all abortions (around 3000, or 55.5 per cent in 2002) take place in the Pregnancy Advisory Centre, a state-funded public hospital service co-located with the Queen Elizabeth Hospital in Adelaide.20 Because the Pregnancy Advisory Centre operates as a public hospital service (and therefore women who have abortions there will not make a claim for Medicare rebates), abortions performed at the Pregnancy Advisory Centre will not be included in the Medicare data. Both the hospital data and South Australian data on abortions are discussed in more detail below.

The HIC’s data on procedures which may result in an abortive outcome also excludes women who, for a range of reasons (such as concerns about privacy), choose not to claim the Medicare rebate. A 1995 study conducted on abortions carried out in Sydney found that 10 per cent of eligible women did not intend to claim the Medicare rebate.21 A more recent Victorian study found that between 13.1 per cent and 33.8 per cent of women who had abortions in Victoria may not claim the Medicare rebate, and thus up to 33.8 per cent of private pregnancy terminations may not be recorded in the HIC’s Medicare data.22

5

How many abortions are there in Australia?

Therefore, as well as potentially over-counting the number of Medicare-funded abortions (because some procedures claimed against the MBS item numbers used for abortions will not be for abortions per se, as explained above), the HIC data will also exclude some abortions performed in private clinics and hospitals (and on private patients in public hospitals) because not all women who have abortions in private facilities claim the Medicare rebate.

A further limitation of the Medicare data on MBS items 35643 and 16525 is that it will not include any abortions conducted after 24 weeks of pregnancy (though the available evidence suggests that the number of abortions conducted in Australia after 24 weeks is relatively small).23 This is because the procedures used under MBS items 35643 and 16525 are only practical for abortions performed in the first or second trimesters of pregnancy, and Medicare does not provide specific funding for abortions conducted after 24 weeks of pregnancy.24

In summary, the limitations of the Medicare data for enumerating abortions in Australia are as follows:

• it potentially over-counts abortion numbers, since it includes procedures which are not

pregnancy terminations

• the Medicare data does not include pregnancy terminations performed on public patients

• it also excludes women who have terminations in private settings, but do not claim a

Medicare rebate, and

• the Medicare data does not include terminations conducted after 24 weeks (though the

available evidence suggests that the number of these is relatively small).

Hospital data

The Australian Institute of Health and Welfare (AIHW) collects some data on abortions performed in Australian hospitals, in the National Hospital Morbidity Database.25 This data, on hospital ‘separations’,26 can be used as an indicator of the number of abortions which take place in Australian hospitals each year, and in particular the number of abortions which are performed on public patients, which are not included in the HIC’s Medicare statistics, as discussed above. The advantage of the hospital data, compared to the Medicare data, is that spontaneous and induced abortions—that is, miscarriages and pregnancy terminations—are recorded separately. However, like the Medicare data, the hospital data on abortions has several significant limitations for enumerating precise abortion numbers. These are discussed below.

According to the AIHW’s hospital statistics, there were around 52 000 separations for which ‘medical abortion’ (that is, induced abortion or termination of pregnancy) was reported as the principal diagnosis in Australian hospitals in 2002–03.27 Around 15 000 of these were performed in public hospitals. However, around 16 per cent of these separations (approximately 2500) were for private patients—that is, women treated in public hospitals

6

A discussion of abortion statistics, their limitations, and options for improved statistical collection

but as private patients.28 There were just under 38 000 separations reported with a principal diagnosis of ‘medical abortion’ in private hospitals in 2002–03. (A small number of these— 800, or less than 1 per cent—were public patient separations.)29 The numbers of ‘medical abortions’ performed in Australian hospitals over the five year period 1998–99 to 2002–03 are shown in Figure 2.

Figure 2: Separations for medical (induced) abortions in Australian hospitals, 1998–99 to 2002–0330

0

10000

20000

30000

40000

50000

60000

1998-99 1999-00 2000-01 2001-02 2002-03

Public hospitals Private hospitals TOTAL

Source: AIHW National Hospital Morbidity Database

Figure 2 shows a slight decline in the number of separations with a principal diagnosis of ‘medical abortion’ reported for public hospitals since 1998–99, and an increase in both the recorded number of separations for ‘medical abortion’ in private hospitals, and in the recorded number of separations for ‘medical abortion’ in Australian hospitals overall.

However, considerable caution should be exercised when drawing conclusions on the basis of this data, because of the limitations discussed below.

Limitations of hospital data

Like the Medicare data on abortions, the data contained in the National Hospital Morbidity Database has a number of shortcomings for enumerating the total number of abortions performed in Australia.31

The first set of problems relates to extrapolating conclusions about numbers of abortion procedures from data about diagnoses:

7

How many abortions are there in Australia?

• on the one hand, it is possible that the diagnosis data on medical abortions may

overestimate the number of pregnancy terminations. For example, there may be cases where a patient is recorded as having a principal diagnosis of ‘medical abortion’ when she is admitted, but for some reason the termination is not carried out (but she will still be recorded in the database under the principal diagnosis of ‘medical abortion’). There may also be cases where a patient is readmitted to hospital following a previous admission for abortion, for follow-up treatment (for example, treatment of retained products of conception). In these cases, the principal diagnosis may still be ‘medical abortion’, even though the patient did not undergo a pregnancy termination during the course of the admission, and

• on the other hand, it is possible that the hospital data on separations for which ‘medical

abortion’ was reported as the principal diagnosis also undercounts pregnancy terminations, because terminations which take place after 20 weeks gestation are recorded in the hospital database differently. As discussed above, the available evidence suggests that the number of post-20 week terminations is (relatively) small, but nonetheless this is an important limitation of the hospital data on diagnoses for use in enumerating pregnancy terminations.32

It is also important to note that the quality of coding for abortion in the National Hospital Morbidity Database—that is, how accurately or faithfully the abortion descriptors within the database are used—has not been assessed at the national level to date.33

A further set of limitations of the National Hospital Morbidity Database for quantifying abortion numbers relates to the database’s coverage of procedures performed in certain settings, as well as its coverage of certain hospitals. For instance:

• the separations data contained in the database only includes information on admitted

patients.34 Women who have abortions in public hospital outpatient clinics or in non-hospital private day facilities may not be counted as admitted patients for National Hospital Morbidity Database purposes when statistics are recorded

• a small number of (mainly private) hospitals are not included in the database, 35 and

• the database’s coverage has changed over time: statistics from some hospitals that

previously had not reported to the database are now included, and some day facilities that had not previously been designated as hospitals are now counted as hospitals for database purposes.36

Consequently, caution should be exercised in drawing conclusions about abortion numbers on the basis of the hospital data, particularly with respect to trends over time. This is because the increase in the number of separations recorded for pregnancy terminations in Australian hospitals over the years 1998–99 to 2002–03 shown in Figure 2 may not necessarily indicate an increase in the number of abortions being performed each year. Rather, it is possible that the increase reflects an increase in the number of separations for terminations which are

8

A discussion of abortion statistics, their limitations, and options for improved statistical collection

recorded in the National Hospital Morbidity Database (as a result of improved coverage), rather than an increase in the number of terminations per se.37 Accordingly, like the Medicare data discussed above, data from the AIHW’s hospital statistics collection has limited utility for estimating the total number of abortions performed in Australia each year (and even for estimating the total number of abortions performed in Australian hospitals).

A further limitation of the hospital data, at least in the context of the current debate which has focused in part on ‘late-term’ abortion,38 is that the hospital data on abortions generally does not contain any information on the gestational age at which abortions take place (or, for that matter, detailed demographic information on women who have abortions, from which social trends in the incidence of abortion could be extrapolated).39

Combining Medicare and hospital data?

As discussed above, using either the Medicare data or the hospital data in isolation will produce a misleading picture of abortion in Australia.40 At the same time, using the two data sets in combination—for example, by adding the number of Medicare-funded abortion procedures to the number of separations for pregnancy terminations performed on public patients—will not produce a reliable estimate of the total number of abortions performed in Australia, for all of the reasons outlined above. For example:

• the Medicare data does not distinguish between pregnancy terminations and other

procedures which are not abortions per se

• not all women who have abortions in private patient settings claim the Medicare rebate,

and

• it is possible that the hospital data excludes some women who have abortions in outpatient

settings in public hospitals (who may not be counted as admitted patients).

Therefore, in the absence of any other national data set on abortions, it is impossible to quantify accurately the total number of abortions which take place in Australia each year.

South Australian data on abortions

As mentioned above, South Australia is the only Australian jurisdiction which both collects and routinely publishes comprehensive data on abortions.41 Other states and territories may collect data on abortions (for example, the Northern Territory and Western Australia collect data on abortions performed within their jurisdictions) but do not publish these statistics.42 (See Appendix One for further details on abortion record-keeping in each state and territory.)

The South Australian data is sometimes used to calculate estimates of national abortion rates. For example, in 2002 there were 5417 abortions notified in South Australia, which equals approximately 17.2 pregnancy terminations for every 1000 women aged between 15 and 44 years.43 If this rate were replicated in the total Australian population of women aged 15–44

9

How many abortions are there in Australia?

years (the so-called ‘fertile age range’) for the same time period, there would have been approximately 73 300 abortions in Australia in 2002.44

Using the South Australian data to estimate the total number of abortions in Australia does not, however, take into account potential differences between abortion rates in different states and territories. For instance, using preliminary unpublished data from Western Australia, a recent AIHW report estimates that the number of abortions per 1000 women in the 15–44 year age group in WA in 2002 was 19.4 (slightly higher than the rate for South Australia in the same time period).45 Different rates in different states and territories may be the result of a range of factors. For example, women often travel interstate to get abortions because of differences in access to pregnancy termination services, and/or for privacy reasons.46 This is particularly the case for abortions which take place at later gestations.47 The phenomenon of women going interstate to seek abortions will affect the reliability of using state-based population data to estimate the total number of abortions which take place in Australia on the basis of the South Australian data.

Further, in this context, it is important to note that the South Australian data differs from the Medicare data in the way that the data is recorded: the South Australian data is a collection of statistics on abortions performed in South Australia, whereas the Medicare data records procedures by postcode of the patient. For example, a South Australian woman who has an abortion in Victoria would not be recorded in the South Australian data, though in the Medicare data (assuming she was a private patient who chose to claim the Medicare rebate) this would be recorded as a South Australian abortion.

Nevertheless, while it has limited utility for estimating the total number of abortions which take place in Australia each year, the South Australian data is extremely valuable in that it is the only comprehensive, publicly available data set on abortion in Australia. For example, the South Australian data includes:

• demographic information on women who have abortions, such as statistics on age and

marital status

• statistics on the gestational age at which pregnancies are terminated

• the grounds for abortions taking place (that is, for reason of physical or mental health of

women having abortions, or because of suspected medical condition of the fetus)

• statistics on methods used to terminate the pregnancy

• statistics on post-operative complications experienced by women who have abortions

• information on the number of abortions which take place in metropolitan and country

hospitals, and public and private facilities, and on whether the woman undergoing the abortion is a metropolitan or country resident, and

10

A discussion of abortion statistics, their limitations, and options for improved statistical collection

• statistics on the category of doctor performing the termination (that is, whether the doctor

is an obstetrician or other kind of medical practitioner).

In the light of current debate on ‘late-term’ abortions, the South Australian data on the gestational age at which pregnancies are terminated is particularly useful. It shows that in the years 1994–2002, the vast majority of abortions performed in South Australia took place before 14 weeks gestation, and that only a very small proportion (less than 2 per cent) took place at or after 20 weeks.48

Options for improved statistical collection

The recent debate over abortion numbers has highlighted the absence of a comprehensive, reliable and systematic means of quantifying the number of abortions which take place in Australia each year.

In this context, it is important to note that there are different views on the issue of abortion statistics. Some commentators in the abortion debate do not support the need for more comprehensive or accurate data on abortion numbers. For example, some pro-choice commentators have argued that abortion statistics are only ever used by pro-life activists to restrict access to pregnancy termination services.49 Other commentators—pro-choice, pro-life, and some without declared positions on abortion—argue that it is not possible to have a proper debate about whether the rate of abortions in Australia is ‘too high’ unless we know for sure how many abortions there actually are.50

Again, this paper does not take a position on this debate. It simply presents a range of options for improving data on abortions, should the issue of improved data collection be pursued.

Medicare data

One proposal for improving abortion statistics, floated during the recent public debate on abortion, is for a change to the way abortions are recorded in the Medicare statistics, so that abortions and miscarriages would be recorded separately in the Medicare data.51 This could be done through the introduction of separate MBS item numbers for pregnancy terminations and other procedures which may result in an abortive outcome but which are not abortions per se (such as curettage of the uterus following a spontaneous abortion or miscarriage).

The idea of a separate Medicare item number for abortions and miscarriages has been criticised by several groups within the health sector, largely because of concerns about privacy.52 This is because, under the current arrangements, women who have abortions in private clinics or hospitals (or as private patients in public hospitals) claim a Medicare rebate under either of the two MBS items used for abortion (items 35643 and 16525). Because medical or induced abortion is not differentiated from the use of these MBS items for other reasons (such as curettage following miscarriage), women do not need to ‘declare’ that they have had an abortion when they are making the Medicare claim. If a separate MBS item for pregnancy terminations were introduced, women would effectively have to declare that they

11

How many abortions are there in Australia?

had had an abortion when claiming the Medicare rebate. Further, the record of the abortion would remain on their Medicare record.

Pro-choice groups have expressed concern that changing the Medicare system in this way might discourage some women from claiming the Medicare rebate altogether, thereby effectively (if inadvertently) restricting access to abortion services.53 Others in the medical profession, such as Dr Lachlan de Crespigny, an Honorary Fellow of the Murdoch Children’s Research Institute, have indicated they would be in favour of differentiating between pregnancy terminations and other procedures (such as curettes following miscarriages) in the Medicare data, ‘as long as the Government’s only intention was to keep track of abortions and not use statistics gathered to reduce access for women’.54

While changing the collection of Medicare information on abortion would help to clarify abortion numbers in Australia to some extent (by clarifying the number of abortions which are subsidised by Medicare), the statistical picture would still be incomplete. This is for two reasons: first, as explained above, women who have abortions as public patients in public hospitals do not claim a Medicare rebate, and therefore these abortions do not show up in Medicare statistics; and second, not all women who have abortions as private patients choose to claim the Medicare rebate. Therefore, even if concerns about privacy were overcome, changing the collection of information about abortion in the Medicare data would still not produce a comprehensive national data set on abortion in Australia.

Hospital data

As discussed above, one of the limitations of using hospital data to enumerate abortion numbers relates to the difficulty of drawing conclusions about numbers of procedures from data on diagnoses. Another weakness of the AIHW’s hospital statistics for enumerating abortions in Australia is uncertainty about its coverage, particularly of women who have abortions as day-only procedures in private clinics. As discussed above, the hospital statistics only include information on admitted patients. The admission status of patients who have day-only procedures in private clinics and public hospital out-patient settings is unclear (and it is possible that in some cases they may be recorded as admitted patients, and therefore included in the hospital statistics collection, and in some cases they may not).

Even if these difficulties were overcome—for example, if the existing admission status of day-only patients in private clinics were clarified for the purposes of the AIHW’s hospital statistics collection—the task of establishing uniform hospital data reporting would be a complicated undertaking: the information in the AIHW’s National Hospital Morbidity Database is compiled from data supplied by state and territory health authorities,55 which manage public hospitals. However, private hospitals and private clinics are, by and large, regulated by the Federal Government. Therefore, mandating changes in the way that state and territory health authorities supply data about private hospital procedures to a federal agency such as the AIHW would be difficult and administratively complex.

12

A discussion of abortion statistics, their limitations, and options for improved statistical collection

Mandatory notification of abortions

Another alternative would be to make a South Australian-style system of data collection (under which it is compulsory for all abortions to be notified to the state’s health department) mandatory in all states and territories. This would be different to addressing problems such as those relating to coverage in the hospital statistics collection in that it would provide a dedicated national data set on abortion (as opposed to using the existing hospital data collection to decipher abortion numbers). Arguably, this would also be preferable to the proposal to change the collection of Medicare data, for the following reasons:

• a South Australian-style system of data collection would overcome the concerns about

privacy expressed in relation to proposals to change the collection of Medicare information. The data could be ‘patient de-identified’; in other words, notification of abortions could take place in such a way that individual patients would not be identified

• a uniform South Australian-style system of data collection would provide a more

comprehensive data set. As discussed above, even if Medicare data was able to distinguish between abortions and other procedures, such as those used in the event of miscarriage, the Medicare data set would not be complete, as it would not include abortions which take place in public hospitals and abortions for which no Medicare rebate is claimed, and

• further, as well as being more comprehensive in terms of coverage, a South Australian-style system of data collection would provide a more informative data set: information on the abortion itself (such as the gestational age at which it took place, and the woman’s reason for having the abortion), as well as demographic information on women who have abortions could be collected.56 This would help to provide a more accurate picture of the incidence of abortion in Australia.

Implementing a South Australian-style system of abortion notification and data collection would require a nationally coordinated approach and legislative change in each state and territory. An agreement on a uniform approach would need to be pursued through the Australian Health Ministers’ Advisory Council and the Standing Committee of Attorneys-General. This approach to improving data collection on abortion would therefore likely take time and not be without some difficulty—for example, there was a failed attempt at recommending uniform criminal laws on abortion several years ago,57 and the leaders of some states and territories have indicated that they are not keen to pursue a uniform national reporting of abortion58—but of the options available, it would appear to be the one most likely to achieve a comprehensive national data set on abortion in Australia.

Conclusions

Much of the recent public debate on abortion in Australia has focused on the issue of how many abortions take place in Australia each year.

13

How many abortions are there in Australia?

In providing an overview of the data on abortion in Australia which is currently available, this Research Brief has demonstrated how vexed this question is. Each of the three major publicly available data sources on abortion—Medicare data, hospital data and South Australian data—can be used to estimate, in fairly crude terms, the incidence of abortion. However, none of these, either singularly or in combination, can be used to quantify accurately the number of abortions which take place in Australia each year.

Accordingly, calls for accurate or ‘truthful’ information on the number of abortions in Australia will not be able to be answered, unless modification of current systems of statistical collection takes place.

14

A discussion of abortion statistics, their limitations, and options for improved statistical collection

Appendix One—Abortion record-keeping in Australian states and territories

South Australia

Regulations made under the South Australian Criminal Law Consolidation Act 1935 require medical practitioners and hospitals to provide notification to the South Australian Director-General of Medical Services of any abortions performed.59 Information from these notifications is made publicly available through the Annual Report of the Committee Appointed to Examine and Report on Abortions Notified in South Australia.

South Australia is the only Australian jurisdiction which both collects and routinely publishes comprehensive data on abortions.

Northern Territory

According to the AIHW, the Northern Territory collects population-based data on abortions performed in the Territory, but does not publish these statistics.60

Western Australia

As is the case with the Northern Territory, Western Australia collects abortion data but does not make this data publicly available.61

NSW

It appears that NSW keeps records on abortions performed in its public hospitals, according to media reports which have quoted figures on late abortions performed in NSW, obtained through freedom of information laws.62 However, as with other jurisdictions, this data is not routinely made publicly available.

NSW Premier Bob Carr recently released figures on abortions performed in NSW in 2003– 04.63 However, the figures Mr Carr released were the numbers of Medicare-funded procedures which may result in an abortive outcome (figures which are publicly available through the HIC). As discussed in this brief, there are significant shortcomings in using Medicare data to enumerate pregnancy terminations.

Victoria

In Victoria, some data is available on late abortions through the annual reports of the Consultative Council on Obstetric and Paediatric Mortality and Morbidity. The Council publishes information on terminations which take place at or after 20 weeks gestation, as these terminations are recorded as births and perinatal deaths in Victoria.64 However, in the absence of a publicly available data set on all abortions performed in Victoria, it is not

15

How many abortions are there in Australia?

possible to accurately estimate for what proportion of abortions the late procedures reported by the Consultative Council on Obstetric and Paediatric Mortality and Morbidity account.

As is the case in NSW, media reports on abortion statistics obtained through freedom of information requests indicate that Victoria does collect other data on abortion, but does not routinely make this information public.65

ACT

In the ACT, the Health Regulation (Maternal Health Information) Act 1998 used to require the managers of approved abortion facilities to report certain statistics to the Health Minister, who in turn was obliged to table these statistics before the ACT Legislative Assembly. However, that Act (and with it the requirement for abortion statistics to be tabled before the ACT’s Legislative Assembly) was repealed in 2002.

Queensland

Abortion record-keeping practices in Queensland were unknown at the time of publication.

Tasmania

Abortion record-keeping practices in Tasmania were unknown at the time of publication.

16

A discussion of abortion statistics, their limitations, and options for improved statistical collection

Endnotes

1. O. Guerrera, ‘Senator places abortion back on political agenda’, The Age, 1 February 2005, p. 3.

2. ‘No reliable national figures on abortion: Abbott admits’, AAP News Wire, 10 November 2004. Mr Abbott’s figure of 100 000 appears to be based on an estimate of abortion numbers in 1996, compiled by the Australian Bureau of Statistics (ABS) (ABS, Australian Social Trends 1998, Catalogue No. 4102.0, ABS, Canberra, p.32). However, this figure appears to have been derived by adding together the number of Medicare claims for abortion procedures, and the number of public patient hospital admissions. As we discuss later in the paper, there are significant problems associated with using this methodology for enumerating abortion numbers in Australia.

3. See, for example, ‘Termination of pregnancy—reopening the debate’, In Touch—Newsletter of the Public Health Association of Australia Inc., vol. 21, no. 11, December 2004, pp. 14–15.

4. The authors plan to publish further papers addressing other issues in the abortion debate, such as demographic information on women who have abortions, and regulatory issues, in the coming months. For information on legal issues around abortion in Australia, see the following: Natasha Cica, ‘Abortion law in Australia’, Research Paper, no. 1, Department of the Parliamentary Library, Canberra, 1998–99.

5. ‘abortion’, Concise Medical Dictionary, Oxford University Press, 2002. See Oxford Reference Online: http://www.oxfordreference.com/views/ENTRY.html?subview=Main&entry=t60.e20, accessed 14 November 2004.

6. ibid.; see also Family Planning Australia (FPA) Health, Fact sheet—About Abortion (Termination of Pregnancy), 2 April 2004. See: http://www.fpahealth.org.au/sex-matters/factsheets/62.html, accessed 10 February 2005; and ‘dilatation and curettage’, Concise Medical Dictionary, Oxford University Press, 2002. See Oxford Reference Online: http://www.oxfordreference.com/views/ENTRY.html?subview=Main&entry=t60.e2718, accessed 14 November 2004.

7. For instance, terminations of pregnancy which are beyond 20 weeks gestation take place either by dilatation of the cervix, followed by evacuation or extraction of the contents of the uterus, or by inducing labour to deliver the fetus following injection of potassium chloride into the fetus while it is in utero—Medical Practitioners Board of Victoria, Report on late term terminations of pregnancy, Department of Human Services, Victoria, April 1998. See

http://www.dhs.vic.gov.au/ahs/archive/report/, accessed 10 February 2004.

8. FPA Health, op. cit.

9. Committee Appointed to Examine and Report on Abortions Notified in South Australia (CAERANSA), Annual Report of the Committee Appointed to Examine and Report on Abortions Notified in South Australia for the Year 1994, Adelaide: CAERANSA, November 1995; CAERANSA, Addendum of the Committee Appointed to Examine and Report on Abortions Notified in South Australia for the Year 2002, Adelaide: CAERANSA, July 2004, p. 5. Note, however, that according to the CAERANSA annual reports, complications may be under-reported.

17

How many abortions are there in Australia?

10. Department of Health and Ageing, Medicare Benefits Schedule Book, 1 November 2004 edition, Commonwealth of Australia, 2004, pp. 196, 273.

11. See, for example, P. Charlton, ‘Rebirth of choice debate only some Liberals want’, Courier Mail, 5 February 2005, p. 34; M. Devine, ‘Abortion debate takes on a new life of its own’, Sydney Morning Herald, 3 February 2005, p. 17; and A. Dunn, ‘High abortion rate dismays experts’, The Age, 30 August 2004, p. 5.

12. P. Karvelas, ‘Many late-term abortions due to abnormalities, say doctors’, The Australian, 12 November 2004, p. 2.

13. Intrauterine fetal death is where the fetus dies inside the uterus. It differs to miscarriage in that miscarriage involves spontaneous expulsion of the fetus or embryo from the womb. In cases of intrauterine fetal death, the dead fetus needs to be removed from the uterus.

14. P. Karvelas, ‘Many late-term abortions due to abnormalities, say doctors’, op. cit.

15. S. Dunlevy, ‘Abortion numbers “inflated”’, The Daily Telegraph, 9 November 2004, p. 7.

16. ibid.

17. Public hospitals are funded jointly by the Commonwealth and the states and territories (through the Australian Health Care Agreements) to provide free hospital treatment to all Australians who choose to be treated as public patients.

18. J. Ford, N. Nassar, E. Sullivan, G. Chambers and P. Lancaster, Reproductive health indicators, Australia, 2002, AIHW Cat No. PER 20, AIHW National Perinatal Statistics Unit, Canberra, 2003, p. 77.

19. National Health and Medical Research Council (NHMRC), An information paper on termination of pregnancy in Australia, NHMRC, Canberra, 1996, p. 3.

20. CAERANSA, Addendum of the Committee Appointed to Examine and Report on Abortions Notified in South Australia for the Year 2002, Adelaide: CAERANSA, July 2004, p. 9.

21. P. L. Adelson, M. S. Frommer and E. Weisberg, ‘A survey of women seeking termination of pregnancy in New South Wales’, The Medical Journal of Australia, vol. 163, no. 8, 16 October 1995, p. 421.

22. In this study, of the sample of 1329 women surveyed, 13.1 per cent either did not have a Medicare card or did not intend to claim a Medicare rebate, and 20.7 per cent were unsure about whether they would submit a claim for the Medicare rebate—C. Nickson, A. M. A. Smith, and J. M. Shelley, ‘Intention to claim a Medicare rebate among women receiving private Victorian pregnancy termination services’, Australian and New Zealand Journal of Public Health, vol. 28, no. 2, April 2004, pp. 120–123.

23. As discussed later in the paper, South Australia is the only jurisdiction in which data on all abortions, and the gestational age at which they occur, is kept. Accordingly, the South Australian data is the only reliable source for determining the relative proportion of abortions which are early, mid or late term. An analysis of the South Australian data for the years 1994–2002 that we conducted shows that the vast majority of abortions performed in South Australia take place before 14 weeks gestation, and that only a very small proportion (less than 2 per cent) take place at or after 20 weeks.

18

A discussion of abortion statistics, their limitations, and options for improved statistical collection

24. Senator Kay Patterson, ‘Question without Notice: Health: Abortion’, Senate, Debates, 15 September 2003, p. 15 104. While there is no specific Medicare item number for terminations performed after 24 weeks of pregnancy, it is possible that some Medicare rebate may be claimed for these terminations where they involve procedures which require labour to be induced and the fetus to be delivered.

25. The National Hospital Morbidity Database is a collection of electronic summary records for admitted patients separated from public and private hospitals in Australia in the years 1993–94 to 2002–03. The Database is compiled by the AIHW from data supplied by the state and territory health authorities. See AIHW website: http://www.aihw.gov.au/hospitals/datacubes/datasource.html, accessed 10 February 2005.

26. ‘Separation’ is the term used to refer to an episode of hospital care, or the process by which an admitted patient completes an episode of care:

[This] can be a total hospital stay (from admission to discharge, transfer or death), or a portion of a hospital stay beginning or ending in a change of type of care (for example, from acute to rehabilitation. ‘Separation’ also means the process by which an admitted patient completes an episode of care by being discharged, dying, transferring to another hospital or changing type of care. (AIHW, Australian hospital statistics 2002–03, AIHW, Canberra, 2004, p. 337).

27. AIHW National Hospital Morbidity Database, Interactive National Hospital Morbidity Data. See: http://www.aihw.gov.au/cognos/cgi-bin/ppdscgi.exe?DC=Q&E=/AHS/principaldiagnosis0203, accessed 10 February 2004.

28. AIHW, Australian hospital statistics 2002–03, op. cit., Table S8.2: Selected separation statistics(a) for all principal diagnoses in 3-character ICD-10-AM groupings, private hospitals, Australia, 2002–03. See: http://www.aihw.gov.au/publications/hse/ahs02-03/index.html, accessed 10 February 2004.

29. ibid.; public patient separations from private hospitals may occur where private hospital providers are contracted by state governments to provide public hospital facilities.

30. The data contained in the figure are based on ‘principal diagnosis’, recorded using the official Australian version of the International Classification of Diseases (ICD) (see the AIHW’s website—http://www.aihw.gov.au/hospitals/morbidity.html—for further information about the ICD).

31. P. J. Laws and E. A. Sullivan, Australia’s mothers and babies 2002, AIHW Cat. No. PER 28, AIHW National Perinatal Statistics Unit (Perinatal Statistics Series No. 15), Sydney, 2004, p. 43.

32. Advice on the use and interpretation of hospital data was received from staff of the AIHW’s Hospitals and Mental Health Services Unit.

33. ibid.

34. Admitted patients are patients who undergo a hospital’s formal admissions process. AIHW, Australian Hospital Statistics 2002–03, op. cit., pp. 2–5.

35. ibid., p. 310.

36. Advice on the use and interpretation of hospital data was received from staff of the AIHW’s Hospitals and Mental Health Services Unit.

19

How many abortions are there in Australia?

37. ibid.

38. Note that there is some debate about how ‘late-term’ abortion should be defined: some commentators define any abortion which takes place after 20 weeks gestation as late-term; others argue that third trimester (post-24 weeks) abortion is late-term; and others argue that the definition of ‘late-term’ should relate to fetal viability—that is, an abortion is ‘late-term’ if performed beyond the point at which the fetus could survive outside the mother’s womb.

39. Though the hospital data does include information on age of patient.

40. Ford et al, Reproductive health indicators, Australia, 2002, op. cit., p. 76.

41. Regulations made under the South Australian criminal code require that the Director-General of Medical Services be notified of all abortions which take place in the state—Criminal Law Consolidation (Medical Termination of Pregnancy) Regulations 1996 (South Australia). Statistics from these notifications are made publicly available through the Annual Report of the Committee Appointed to Examine and Report on Abortions Notified in South Australia.

42. AIHW, Australia’s Health 2004, AIHW, Canberra, 2004, p. 22.

43. A. Chan, J. Scott, A. Nguyen and P. Green, Pregnancy Outcome in South Australia 2002, Pregnancy Outcome Unit, Epidemiology Unit, Department of Human Services, Government of South Australia, Adelaide, p. 39.

44. According to the ABS’s most recent population data, the estimated resident population of women aged between 15-44 years in Australia in 2002 was 4 262 904 (ABS, Australian Historical Population Statistics (cat. no. 3105.0.65.001), Canberra, 2004). If 17.2 in every 1000 of these women had an abortion, there would have been approximately 73 300 abortions in Australia in 2002.

45. Laws and Sullivan, Australia’s mothers and babies 2002, op. cit., p. 43.

46. C. Nickson, J. Shelley and A. Smith, ‘Use of interstate services for the termination of pregnancy in Australia’, Australian and New Zealand Journal of Public Health, vol. 26, no. 5, October 2002.

47. The Consultative Council on Obstetric and Paediatric Mortality and Morbidity, Annual Report for the Year 2002, incorporating the 41st Survey of Perinatal Deaths in Victoria, Melbourne, 2004, pp. 8–9.

48. See endnote 23.

49. See, for example, D. Cronin, ‘Stanhope won’t give figures on abortion’, Canberra Times, 3 February 2005, p. 3; and Leslie Cannold, ‘Put an end to abortion whispers’, Sydney Morning Herald, 4 February 2005, p. 11.

50. See, for example, R. Boswell, ‘Abortion’s elusive truths’, The Australian, 4 February 2005, p. 13; and A. Dunn, ‘How we could have a real abortion debate’, The Age, 4 February 2005, p. 15.

51. See, for example, E. Symons, ‘Howard aborts Abbott’s inquiry’, The Australian, 15 November 2004, p. 2.

20

A discussion of abortion statistics, their limitations, and options for improved statistical collection

52. L. Wright and M. Papadakis, ‘Warning on abortion—women may go underground’, Sunday Herald Sun, 14 November 2004, p. 20.

53. C. Calcutt, spokeswoman for Children by Choice, quoted in Wright and Papadakis, ‘Warning on abortion—women may go underground’, op. cit.

54. L. de Crespigny, quoted in ibid.

55. AIHW, National Hospital Morbidity Database website. See: http://www.aihw.gov.au/hospitals/morbidity.html#nhmd1, accessed 10 February 2005.

56. In addition to the demographic data currently included in the South Australian collection (including data on age and marital status), it would also be useful to have data on abortions by postcode or local area, as long as this would not breach the privacy of women who have abortions. Local area data would be useful as there are significant differences in abortion rates in different localities—see Ann Evans, ‘The outcome of teenage pregnancy: temporal and spatial trends’, People and Place, vol. 11, no. 2, 2003, pp. 39–49.

57. In 1998, a committee appointed by state, territory and federal Attorneys-General to draft a model criminal code presented a report on ‘Non-fatal Offences Against the Person’. The committee decided that it was not in a position to make recommendations about abortion law because the issue was ‘ultimately one for political decision’—Model Criminal Code Officers Committee, Model Criminal Code, Chapter 5: Non Fatal Offences Against the Person, September 1998, p.147. See also N. Dixon, ‘Abortion Law Reform: An Overview of Current Issues’, Research Brief No 2003/09, Queensland Parliamentary Library, 2003, p.6. See:

http://www.parliament.qld.gov.au/Parlib/Publications_pdfs/books/200309.pdf, accessed 10 February 2005.

58. See, for example, D. Cronin, ‘Stanhope won’t give figures on abortion’, op. cit.

59. Criminal Law Consolidation (Medical Termination of Pregnancy) Regulations 1996 (South Australia).

60. AIHW, Australia’s Health 2004, op. cit., p. 22.

61. ibid.

62. S. Dunlevy, ‘New pro-life agenda: Later-term abortion reviewed’, Daily Telegraph, 2 November 2004, p. 4.

63. Samantha Maiden, ‘Dearth of statistics drives reform bid’, The Australian, 3 February 2005, p. 2.

64. The Consultative Council on Obstetric and Paediatric Mortality and Morbidity, Annual Report for the Year 2002, incorporating the 41st Survey of Perinatal Deaths in Victoria, op. cit., pp. 8–9.

65. A. Dunn and G. Alcorn, ‘Soaring teen abortion rate revealed’, The Age, 10 November 2004, p. 1.

21

How many abortions are there in Australia?

© Copyright Commonwealth of Australia 2005

Except to the extent of the uses permitted under the Copyright Act 1968, no part of this publication may be reproduced or transmitted in any form or by any means including information storage and retrieval systems, without the prior written consent of the Department of Parliamentary Services, other than by senators and members of the Australian Parliament in the course of their official duties.

This brief has been prepared to support the work of the Australian Parliament using information available at the time of production. The views expressed do not reflect an official position of the Information and Research Service, nor do they constitute professional legal opinion.

Acknowledgements

We would like to thank the following colleagues for their assistance and advice in the preparation of this paper: Carol Kempner, Jennifer Norberry, Judy Hutchinson, Malcolm Park, Morag Donaldson, and June Verrier. We would also like to thank the following people for their extremely helpful comments on earlier versions of this paper: Professor James King, Chair of the Victorian Consultative Council on Obstetric and Paediatric Mortality and Morbidity; Dr Ann Evans, ACSPRI Centre for Social Research, ANU Research School of Social Sciences; Carolyn Nickson, Key Centre for Women's Health in Society, University of Melbourne; Jenny Hargreaves from the Australian Institute for Health and Welfare’s (AIHW) Hospitals and Mental Health Services Unit; and Narelle Grayson from the AIHW’s National Perinatal Statistics Unit.

22