About one in three Australian women give birth outside metropolitan areas. Historically, most of these women have been cared for by general practitioner obstetricians and midwives in local hospitals. However, in recent years it has been difficult to recruit and retain midwives and doctors with the necessary skills to adequately staff many rural maternity units providing traditional models of care. Whenever any essential component of obstetric, anaesthetic, paediatric or midwife infrastructure has become unavailable, maternity units have been closed, and women in the area are required to travel to the nearest maternity centre instead. It is estimated that more than 130 Australian rural maternity units have closed since 1995.1

In this issue of the Journal, a description by Scherman and colleagues (→ The first year of a midwifery-led model of care in Far North Queensland) of the first year’s experience at Mareeba District Hospital’s midwifery-led maternity unit provides some interesting and valuable insights into the role of low-intervention birth units for low-risk women in rural areas.2 Women planning to give birth at Mareeba were screened for risk factors, and low-risk women were selected for maternity care provided by a midwife. Medical back-up was provided through regular case conferences with an obstetrician at the nearby Cairns Base Hospital, who also visited Mareeba once a month, and on-site back-up from local GPs, who were able to perform caesarean section delivery if necessary. Intrapartum transfer was rarely used and largely confined to primiparous women. The authors reported no significant preventable morbidity or mortality in mothers or their babies, and their findings would appear to support the feasibility of this model of care, especially for multiparous women — although they acknowledge that the number of births so far is too low to support any conclusions regarding safety. Intervention rates were significantly below state averages.

In 2004, 207 women gave birth at Mareeba District Hospital under the previous model of care,3 and in the 12 months of this study (2005–2006), 147 low-risk women (plus 11 high-risk women) did so under the care of a midwife. Had the maternity unit been closed, these women would have had to travel over 60 km to Cairns Base Hospital for delivery, as did the 45 women who required transfer during pregnancy for pre-existing or emergent risk factors. Interestingly, the powerful sway of maternity care politics is evident. What other clinical discipline would employ three full-time and nine part-time staff in a dedicated unit with medical back-up to treat 203 patients a year, with an alternative facility 60 km away that ultimately treated almost a quarter of them anyway?

The study by Scherman and colleagues raises several other significant issues. The low use of epidural anaesthesia (1%) — which required transfer to Cairns when requested — surely reflects the lack of access at Mareeba, rather than true patient preference. Although facilities for some caesarean deliveries are in place at Mareeba, it is not always possible to perform an emergency caesarean section on site. Recorded transfer times for unplanned deliveries at Cairns Base Hospital were 80–145 minutes, with transfer required either for epidural request or first-stage failure to progress. There is little evidence upon which to base recommendations for the optimum time interval between decision to transfer and emergency delivery, and perhaps transfer would be completed more urgently in a situation where delay could compromise the mother or baby, but most clinicians would agree that a shorter interval than those reported is desirable.

A related issue is whether or not staff (both medical and midwifery) providing obstetric care at such units should also spend some time working in a higher-level unit, to maintain their skills in management of relatively rare complications. For example, the treatment of catastrophic life-threatening obstetric haemorrhage is rarely required — but will outcomes be optimal in units where this complication might only be seen by staff once every 5 or 6 years? A Cochrane review of outcomes at birth centres and medical-led units found a statistically significant higher perinatal mortality rate (relative risk, 2.38) in birth units staffed by midwives who did not also work in medically supervised hospital units.4

The search for models of maternity care best suited to women in rural areas will continue to challenge health service planners. A paradox haunts them: the more remote a maternity service is from a referral centre, the higher the perceived community value in keeping it open, but also the higher the risk in operating a low-intervention unit that can provide care for most low-risk women but which ultimately relies on emergency transfer of women with unexpected complications. It is noteworthy that 70% of women who develop pregnancy complications are classified as low risk on initial assessment.5

It seems likely that continuing provision of maternity services in rural areas will depend on optimum use of the local workforce and health facility infrastructure. Individual regional areas will need to come up with arrangements based on consultation with the local community and health workforce. This will require a flexible approach and should recognise that transfer of women may sometimes be required, because of temporary unavailability of a core of necessary staff. It is unlikely that a “one size fits all” approach will deliver solutions to all areas at all times. The key will be cooperation and consultation with the local medical and midwifery workforce, open disclosure to women of what services can and can’t be provided locally, and facilitation of transfer of pregnant women to referral centres if they so choose.