Given that accurate information is essential for collaborative decision making by parents and medical staff around the treatment for periviable babies this study demonstrates that there is greater pessimism about the outcomes of the most premature babies by all HCP groups than is indicated by the actual outcome figures. Information is also important for enabling parents of babies at older gestations who will still require tertiary level intensive care for their babies to understand the risks to their offspring, and at older gestations, HCP are more accurate in their knowledge. HCP who have the most contact with parents prior to delivery, are the least accurate in terms of both mortality and the risks of a poor outcome at the lowest gestations. This discrepancy is concerning, as proactive antenatal care improves neonatal outcome, and where the antenatal team disagrees with the neonatal team in the provision of care, the outcomes for the baby are seen to be worse . Where active care is proposed, antenatal steroids, magnesium sulphate, and monitoring of the foetus may optimise the condition of the baby and reduce later morbidity, hence decisions often need to be made well before delivery where possible [6, 18].
It is possible that the information as understood by HCP’s is merely out of date, however, whilst survival data has improved with time, there have been only modest improvements in the rates of severe disability seen in some studies . Even in previous decades, the perceptions found for survival and disability would have been unduly negative, reflecting survival rates found in the late 1990’s [23, 24]. Studies done in the mid 2000’s reflect improved survival rates for babies offered intensive care . In the Australian context with both inborn and retrieved babies improved survival rates are seen from the early 2000’s . Undue negativity may reflect a reluctance of some HCP to provide care for these babies. Previous studies have shown that pessimistic clinicians are less likely to intervene to provide intensive care for periviable babies . Hospitals with more optimistic obstetric and neonatal trainees are known to have received training from hospitals who have higher rates of providing care at the lowest gestations, and are found to be more accurate in their outlook . Higher rates of offering care leads to improved outcomes [27, 28] and in some studies this appears to be regardless of numbers of small babies being cared for . Whilst the tertiary unit described is a smaller tertiary centre in Australia, it has a high rate of offering care to babies under 25 weeks gestation  with comparative survival rates, but with more positivity it is likely that the survival and long term outlook for these babies would improve.
Extremely preterm babies will remain in the neonatal intensive care for months before going home. Parents who have experienced neonatal intensive care have been shown to have high rates of anxiety, depression, stress and trauma [29, 30] which may result in poorer long term developmental outcomes for the child . Parents tell us that they need hope and honesty to help sustain them through their neonatal stay . Whilst the potential for an adverse outcome needs to be understood by parents depending on the evolution of events during the babys care, if parents have been given a very negative outlook for their baby, the realistic hope that the baby may be healthy is removed, and the parent will need to endure the invasive painful treatment of the baby without recognising that the suffering baby has a potentially good outcome.
Staff based at smaller centres were found to be more negative about survival below 26 weeks than the tertiary HCP, but there was no difference in their perceptions of rates of disability. The origins for this are unclear. This has been noted in the Australian context in previous studies . Most HCP at all centres were negative about long term outcomes. The non-tertiary centres will deliver fewer babies at extreme prematurity as an attempt to transfer antenatally to tertiary centres is standard care. Where parents presenting to these centres discuss the prognosis for their extremely premature babies, a more negative impression for potential survival will already have been conveyed to parents prior to transfer, and may have led to less optimisation of the fetus for postnatal survival, such as the administration of steroids [21, 33] at the referring hospital. The parents, in turn will have a more negative outlook for the baby and this may influence their decision making. Work to improve the knowledge at referral centres may improve the wellbeing of the delivered baby as shown in the work by Morse .
Clinicians who are involved in the care of women prior to delivery are significantly more negative than those who care for the baby in the short and long term. This confirms previous work done and has previously been shown to adversely affect the antenatal care of the extremely preterm fetus [10, 18]. Clinicians caring for the woman presenting with complications will have earlier counselling encounters with families and their more negative knowledge may affect parental decision making. Further research may reveal the origins of the more negative opinions.
All clinicians would offer care for patients at significantly lower gestations than they would wish for themselves, which is not unexpected given their negative perceptions of outcome. This has been described previously in trainee doctors  and may reflect a respect for patient autonomy and acceptance that patients may make different choices to the clinician. Furthermore, HCP recognised that there were specific factors about each pregnancy which would alter their risk assessment for the baby, and hence influence whether they thought that intensive care should be provided. Both surgical congenital anomalies and trisomy 21 were recognised as negative factors for survival and neurodevelopment, however, emotional factors such as previous pregnancy loss or the presence of no live born children in the family would positively encourage resuscitation despite no evidence that the difficult previous history will improve the outlook for the pregnancy at risk.
The difficulty in predicting an outcome for an individual pregnancy from large epidemiological studies was reflected in several free text comments. Whilst statistics may be important to clinicians, these reflections of uncertainty may be important factors for parents to understand. In a pilot study of 15 clinicians giving antenatal counselling, Prentice et al.  showed that most interactions involved the imparting of statistics and information only (60%) and eliciting parental preferences or engaging in deliberation was less frequent (20%). The nuance of the statistics and uncertainty with their application is unlikely to form part of this type of counselling. Previous studies have demonstrated that parents of extremely premature babies perceive the risk of death as more important than the risk of disability for a baby when a decision is made to resuscitate occurs [35,36,37,38]. HCP in these studies felt that the risk of severe disability was more important. Where death usually occurs in extreme prematurity, it is usually in the first days following delivery, so the uncertainty primarily affects the prognosis for disability, and this should be a part of counselling for decision making. Our study suggests that parents in North Queensland will receive a negative message about survival at gestations below 28 weeks, and rates of severe disability at the earliest gestations. At the earliest gestations, intact survival is similarly underestimated.
Most HCP recognise a need to support autonomy in parental decision making. However, this attitude was not consistently reflected in the answers to the range of questions asked. Where parents wanted intensive care provision for their baby but the clinician did not feel that it was in the best interests of the baby, 69.1% of respondents said that this care should be provided whilst 21% said that it should not. However, where parents did not want intensive care for the baby, but the clinician did, 37.5% would follow the parental request, but 40.6% would provide resuscitation despite this preference. Below 25 weeks, over half respondent felt that parents could be the sole decision makers, but 64.6% also said that clinicians could disregard parental choice at this gestation. It seems recognised that risks of death and disability decrease with increasing gestational age, but specific gestational cut offs are relatively artificial. Ethical dilemmas in the relative roles of parents and clinicians are reflected in these findings, with a range of opinions from complete parental autonomy to decline intensive care, even at gestations over 25 weeks, and those which deem that parents should not always be the final decision makers, even if intensive care then occurs for babies whose parents did not want this for their child. The data suggests that the trend is towards clinicians as the final arbiters of decisions. Further research could clarify the underpinnings of HCP beliefs.
Parental involvement in decision making can only be based on accurate information. Most guidelines currently in use in Australia, include parental discretion around the resuscitation of babies below 24 or 25 weeks gestation [3, 4, 39]. Despite the negativity of clinicians and guidelines discouraging the resuscitation of babies under 24 weeks, many of these babies are receiving intensive care in Australia and a recent review of the use of the consensus guidelines in New South Wales and Australian Capital Territory reflect that resuscitation at 23 and even 22 weeks regularly occurs . In North Queensland, nearly all babies at 24 weeks gestation and nearly half of babies delivered at 23 weeks gestation receive tertiary intensive care, regardless of place of delivery . With increasing parental autonomy, and parental requests for active care at gestations below 23 weeks, the perinatal community as a whole in Australia needs to be aware of improving outcomes and consider if the guidelines need modification to include clarity around resuscitation and provision of care at lower gestations.
There are some limitations to this study. It is a relatively small study based on a self-designed cross sectional survey from one area of Australia only which may limit the generalisability. However, local examination is important and the findings are consistent with those found in previous studies both historically and more recent. Another potential limitation is the use of the long term follow up data from the ANZNN. Follow up rates at the tertiary unit are relatively poor, and highest rates of follow up occurs for local babies within the immediate tertiary unit area and only one other regional centre where standardised tests are available.
The strength of this study is that it there was a good response rate, and that participation was invited from regional and remote centres where many patients initially presented with complications in their pregnancy. There are few studies which examine referring HCP knowledge. The survey also included staff whose contribution towards parental knowledge might previously have been ignored such as midwifery and neonatal nursing staff, as well as more junior obstetric staff. Midwifery and neonatal nursing staff will contribute to the parents’ perception of the long term with much closer daily contact whilst providing care both antenatally and postnatally and can influence the hope that parents need to cope with their neonatal experience. The inclusion of paediatricians who see these babies long term is also uncommon, but important as they will often have a long-lasting relationship with the children. A further strength of the study is that it has been done in an area with a high Indigenous population where Indigenous babies are over-represented on the neonatal unit. A strength of the study not reflected in most studies is the content analysis of the qualitative data. Qualitative data adds to the richness of the quantitative data in studies of knowledge and attitudes.