Delivery of maternal health care in Indigenous primary care services: baseline data for an ongoing quality improvement initiative

Approximately four percent of women giving birth in Australia are Aboriginal and/or Torres Strait Islander (Indigenous) [1]. Disproportionately high rates of poor pregnancy outcomes including maternal and perinatal mortality, preterm birth and low birth weight have consistently been documented in Indigenous populations relative to other Australians [2]. With the exception of perinatal mortality, little improvement in these outcomes has occurred in the past two decades [1], and for some outcomes (e.g. low birth weight), the disparity appears to be widening [3].

Providing access to appropriate and quality care in the antenatal and postnatal period is a key part of closing the gap in Indigenous perinatal outcomes. It also offers an important opportunity to ameliorate disease in adult life, given the recognised link between early life influences and chronic disease [4].

Antenatal care aims to provide appropriate screening, preventive or treatment interventions and health information to maximise the health of women and their infants [5]. Almost all pregnant Australian women have some antenatal care, which can be delivered by a diverse range of providers (e.g. midwives, obstetricians, Aboriginal health workers, general practitioners etc.) working across hospital and community settings. However, Indigenous women access care differently to other Australian women. Although there is no complete national data for Indigenous women, the available data suggests that antenatal care is underutilised, as Indigenous women are more likely to present for care later in pregnancy and have fewer antenatal visits [6–8].

The importance of achieving equity in access to antenatal care is recognised in several local and international development initiatives [9, 10]. However, improving the quality of care is also important, as poor quality care can contribute to poor outcomes [11] and be a barrier to service utilisation [12]. Previous research assessing care for Aboriginal women birthing in the Northern Territory (NT) identified the need to focus on accurate assessment and follow up of problems identified in pregnancy, and other aspects of quality of care, rather than focusing solely on the timing and frequency of antenatal visits [13]. Furthermore, there is evidence of sustained improvements in perinatal outcomes amongst women attending a large Aboriginal Medical Service in Queensland that focussed on quality improvement in maternal and child health [14, 15].

Recognising the potential benefits of quality improvement principles in primary health care, the Australian Government launched the Healthy for Life program in 2005 [16]. The program aimed to improve the capacity of and care offered by Aboriginal primary health care services for chronic disease and maternal and child health, and drew extensively from a quality improvement action research project that has been operating since 2002 - the Audit and Best practice for Chronic Disease (ABCD) Project [17]. Here we describe patterns of delivery of maternal health services and gaps in these services drawing on data from this project.

The data presented here arise from baseline clinical audits undertaken in the ABCD Project. Of 56 participating primary health care centres in early 2009, 34 elected to audit their maternal health records. These centres were located in: a) remote communities in the Top End (13) and Central Australia (2) of the NT and in North Queensland (QLD) (4); and b) small regional towns in Far West New South Wales (NSW) (6) and Western Australia (WA) (9). Participating services were widely distributed across Australia and many of the remote health centres were located in sparsely populated areas in the northern part of Australia.

Delivery of maternal health care was assessed by auditing a sample of clinical records from participating services. Records of women with an infant aged between 2-14 months and who were resident in the community for at least six months of that infant's gestation were eligible. From eligible records in each service, a random sample of up to 30 records was drawn using computer generated random numbers. Where there were fewer than 30 eligible records all were included for audit. In the ACBD project these sampling methods have been demonstrated to be sufficient to show meaningful differences between communities and trends over time.

The audit was based on previous research [18], published antenatal guidelines [19], the Women's Business Manual (WBM) [20] and key literature [21–24]. The WBM is the standard management manual for practitioners caring for women in remote areas.

The audits were conducted by trained members of the project team familiar with the auditing process and the maternal audit tool, in collaboration with local health staff. The audit form and protocol are available at: <http://www.abcdproject.org.au >. A clinical service related to antenatal/postnatal care was assessed as "delivered" if there was a record of the service being delivered at specific periods in line with antenatal/postnatal care guidelines.

We have previously published a short report on selected data from this study [25]. Here we present a more complete analysis of the data.

The study was approved by the human research ethics committees in the Top End of the NT, Central Australia, Far West NSW, WA and Cairns and Townsville, and their Indigenous sub-committees where required.

Across services the median age of women was 25.3 years (range 14-48) (Table 1). Compared with services in the NT, WA and North QLD, Far West NSW had a higher proportion of non-Indigenous women presenting for antenatal or postnatal care (P < 0.05), however, overall 89% of all women were Indigenous.

Across services, the proportion of women presenting for their first antenatal visit in the first trimester ranged from 34% to 49% and the mean gestational age at first visit was 16 weeks (range 3-40). The first antenatal assessment was undertaken by a variety of providers; with the exception of WA and North QLD services, the most common provider was a midwife (Table 1).

Documentation of prescription of folic acid was poor. Across all services only 27% of women were prescribed folic acid prior to 20 weeks gestation and even fewer (8%) prior to conception. In contrast, more than half of women in most services were prescribed iron during pregnancy. Over 80% of women had an antenatal care plan recorded in their file (Table 1).

Consistent with the late gestational age at first presentation observed across services, documentation of routine antenatal checks including maternal height/weight, blood pressure and urinalysis before 13 weeks gestation was poor (range 10-45% for all women), but improved for checks undertaken between 13-24 weeks gestation (range 61-71% for all women), and after 24 weeks gestation (range 72-81% for all women) (Table 3). There was significant variation between regions in the frequency of routine antenatal checks.

Documentation of routine ultrasound varied between regions. Overall only 42% of women had a recorded ultrasound between 16-20 weeks gestation for fetal morphology (range 31-69% across regions) (Table 4). Similarly documentation of screening or diagnostic tests for gestational diabetes mellitus (GDM) varied (51% for all women, range 33-78%) as did investigation for Group B Streptococcus (GBS) infection (35% for all women, range 10-62%).

The clinical audit data presented here provide insights into documented delivery of maternity care which should be useful for informing local, regional and national efforts to promote the quality of primary maternal health care for Indigenous women. Participating services varied in their relative strengths and weaknesses, both geographically and between service items. Nevertheless, adherence to recommended screening investigations and brief interventions/advice about health behaviours were consistently identified as areas for improvement across services.

This study extends previous work demonstrating improvements in antenatal screening at a local level [14, 15] and indicates the potential for improvements to be replicated on a broad scale. The study also demonstrates the potential for quality improvement methods to be extended to systematic assessment of other aspects of care including follow up of abnormal clinical findings. In addition, this study adds to the limited international literature on approaches to assessing the quality of antenatal care in primary care settings.

Overall, there was good documentation around routine checks in pregnancy such as blood pressure, and in the follow up of abnormal clinical findings related to hypertension and GDM, where rates of specialist examination were high. Wide variations were present in documentation of routine investigations such as morphology ultrasound and screening tests for GDM and GBS. On average documentation of these procedures was only around 40-50%, despite evidence for treatment of carbohydrate intolerance in pregnancy [26] and GBS screening [27].

Across all services, prescription of peri-conceptional folic acid supplements was low. These findings may reflect later presentation for care in pregnancy and indicate that fortified food is a potentially important source of folate to protect against neural tube defects (NTD) in those communities. Importantly, Australia recently introduced a policy of mandatory fortification of certain foods with folic acid for additional protection against NTD.

Uptake of fetal anomaly screening was low, which supports previous calls for the development of culturally safe resources around these tests [13]. There is considerable room for improvement in other aspects of antenatal counselling and education. Good evidence exists for the effectiveness of smoking cessation advice in pregnancy [28], yet documentation of this information was present for only 46% of smokers, despite the high proportion of women with recorded use of cigarettes. There was virtually no change in the proportion of women who smoked in the first compared to the third trimester overall, although in two regions the proportion declined (Top End 32%-21%; WA 32%-26%) (Table 3). Cessation advice that is inappropriate to the cultural context may be a barrier to addressing smoking in pregnancy [29] and we await the results of a controlled trial of a culturally-specific smoking intervention for pregnant Indigenous women [30]. It should be noted that many mainstream maternity units in Australia do not have relevant smoking cessation protocols [31].

In the postnatal period, only 55% of women received breastfeeding advice, and there was poor documentation of advice around smoking, hygiene, injury prevention and SIDS prevention. This is a significant area for improvement given the high prevalence of SIDS risk factors such as smoking, preterm birth and low birth-weight in these communities. Overall, this indicates there is great potential to improve the continuity of care throughout the antenatal and postnatal period. It should be noted that centres in this study do not provide birthing services; local policies currently require women in remote areas to be transferred into regional hospitals for birthing from 36 weeks gestation. Therefore, attempts to improve continuity in this setting will require better integration of services or a redesign of care to support community-based services across the spectrum of maternity care. This was acknowledged in the recent report of the Australian Maternity Services Review [32], which also recognised the importance of provision of services to Indigenous women that accommodate cultural beliefs about childbirth, including in some communities, a preference for 'birthing on country'.

The findings of this study are subject to some limitations. Participating services have a history of involvement in quality improvement activities, expanded from a focus on chronic disease management to maternal, child and other aspects of health, which may limit the generalisability of findings. The maternal audit was undertaken to provide baseline data for individual services with an interest in improving delivery of maternal health care, as part of an on-going quality improvement study (the ABCD project). Not all services participating in this larger study elected to audit maternal records, which may also limit the generalisability of findings. Our audit data provide similar estimates of mean maternal age and smoking prevalence, but lower estimates of preterm birth and low birth weight than estimates reported in routine NT, WA, QLD and NSW Indigenous perinatal information [2]. As such, our data should not be regarded as representative of all services in the regions. Furthermore, small numbers of abnormal findings in this sample means estimates of rates of follow up of abnormal findings lack precision. The data presented here are baseline service activity only; more data will be available for subsequent years of the quality improvement intervention, which includes annual cycles of organisational assessment, clinical audits, feedback to and interpretation of data with participating health centre staff, action planning, goal setting and implementing of strategies for change [18].

The measures of quality of care reported in this study provide important data for feedback to services at a local level. When aggregated, they provide a picture of regional patterns of maternity care which will be progressively more informative for broad scale policy and program purposes as increasing numbers of services become involved. Importantly, these measures provide information that is not currently available from routinely collected perinatal data [2].

There is increasing activity in Australia and internationally around assessment of maternity care, with the completion several projects designed to facilitate benchmarking of services [33–35]. The two recent Australian projects recommended a core set of indicators focussing on intrapartum and postpartum clinical outcomes. In developing our study we had difficulty choosing measures of quality of maternal health care relevant for the community health centre setting. Reporting on the proposed national maternity indicators would require information from other systems such as hospital administrative data. It is also unrealistic to expect to show improvements in outcomes such as low birth weight over annual cycles of quality improvement activities. We chose to focus on processes of antenatal and postnatal care, which reflect delivery of routine screening investigations and treatment of abnormal findings, consistent with local guidelines for the community health setting. We recognise that like the proposed national indicators, this presents an incomplete picture of the scope of maternity care. A coordinated and consistent national approach to reporting of maternity information across all services (primary, secondary and tertiary) is needed, building on the substantial work of the National Perinatal Statistics Unit within the Australian Institute of Health and Welfare in national perinatal data development. Achieving this will require consensus on standard components of antenatal care, and so it is critical for Governments to continue to support the development of national guidelines defining best practice in antenatal care in Australia, such as those in the United Kingdom [36].

Supporting services to improve maternity care for Indigenous women is a recognised priority [16] given the persistence of unacceptably high rates of poor Indigenous perinatal outcomes. This is acknowledged by Indigenous women, who consistently identify optimising the health of mothers and babies as a high priority [21, 37, 38]. This study demonstrates that although Indigenous women presented at a later gestational age they regularly attended services for maternity care, with the average total number of antenatal visits within suggested minimum numbers [5, 20]. Importantly, this indicates that there are opportunities for appropriate investigation and risk factor intervention amongst Indigenous women. The findings highlight the inadequacy of existing systems in community health centres to provide access to routine evidence-based screening and health information in the antenatal and postnatal period. More broadly, the findings reflect the lack of consistency in national standards defining antenatal care that all Australian women should expect to receive, which unless addressed, will see the persistence of inequalities in maternity care and outcomes.