How safe is GP obstetrics? An assessment of antenatal risk factors and perinatal outcomes in one rural practice

full article:

Introduction

Western Australia had 29 622 obstetric deliveries in 2007, of which 7213 or 24.4% live in rural or remote regions¹. Of these, 5651 or 18.8% delivered in rural and remote regions¹; therefore, about 80% of women resident in country areas deliver in or near their region of residence². The medical workforce servicing these patients is made up predominantly of general practitioner (GP) obstetricians, 137 of whom were identified working across the state in 2007. There were just six specialist obstetricians of which four were in one town, Bunbury³. Thus the burden of obstetric care in rural Western Australia falls to non-specialist doctors or GPs with an interest in obstetrics.

As a GP obstetrician, the challenge is to refer high-risk patients to an appropriate centre in a timely fashion and develop and maintain the skills to adequately deal with whatever else occurs. In addition there are well-recognized aspects of rural populations that make them a higher risk as a group. These factors include higher rate of Aboriginality, remote locations and poor access to health services and, in some areas, poor socioeconomic indicators. Given these challenges, it is worthwhile to look at what is actually happening in a rural GP obstetric practice.

This article reports the finding of an audit of 2 years of one obstetric practice in Kalgoorlie Western Australia. Kalgoorlie is a regional centre in Western Australia 600km east of the state capital Perth. This article explores how safe is GP obstetrics. In particular it asked the following questions:

1. What are the antenatal risk factors?
   
2. What are the obstetric outcomes for the mother?
   
3. What are the obstetric outcomes for the baby?

The audit population comprised all deliveries conducted by the author over a two-year period from July 2007 to June 2009.

Methods

All patients delivering in Kalgoorlie hospital and in whose obstetric care and delivery the author was involved were reviewed for the two-year period from July 2007 to June 2009. There were 195 patients identified, which is about 15% of the total number of deliveries at Kalgoorlie Hospital for this period. Not included in this audit were patients seen antenatally and referred elsewhere. This included those patients electively choosing to be seen privately by an obstetrician in Perth, patients with high antenatal risks such as insulin-dependent diabetes or morbid obesity who were referred to King Edward Memorial Hospital, the sole tertiary centre for Western Australia or those patients transferred out in pre-term labour prior to 35 weeks gestation, which was the cut off for managing these pregnancies in Kalgoorlie.

A list of antenatal complications was developed from reviewing the cases. Intrapartum and immediate postnatal complications were also documented. General maternal characteristics were recorded including age, gravida, parity, body mass index (BMI), Aboriginal and Torres Strait Islander background and existing medical conditions. The babies' gender, gestation at delivery, birth weight and Apgar scores at 1 and 5 minutes were also recorded.

Results

Results are presented in Table 1. For each measurement there is a comparison with figures published elsewhere. Comparisons were made with the most recent Western Australian data where possible. This was obtained from the Midwife Notification Report data in the Perinatal Morbidity Report¹.

Table 1: Study results^1,2,4,5



Discussion

The patient cohort in this study was a relatively low-risk obstetric group, given the low rate of Aboriginality and referral of the more complicated patients to specialist services. Despite this, a substantial number of antenatal complications occurried including particularly hypertension, obesity, pre-eclampsia and a significantly high level of gestational diabetes. There were a number of intrapartum and post-partum complications, although mostly these were in line with expected levels. In particular, maternal sepsis, antepartum haemorrhage, shoulder dystocia, failure to progress and foetal distress occurred at rates similar to the averages for Western Australia. Post-partum haemorrhage at 10.3% seems high but is also on a par with the Western Australia state averages and only 2% required transfusion.

There were 195 live births from 195 singleton deliveries. These were better than expected results in this audit for high vaginal birth rate, low caesarean section rate and a relatively high vaginal birth after caesarian rate. These figures make the case that despite a significant number of antenatal and peripartum complications it is possible to have a safe outcome for mother and baby in a selected obstetric population. It supports the hypothesis that GP obstetrics can be practiced safely in rural Western Australia.

Of note was a higher than expected rate of gestational diabetes. Gestational diabetes was the most common maternal antenatal complication in Western Australia in 2007 at a rate of 4.6%¹. A prevalence of 8.2% gestational diabetes recorded in this study is markedly higher than the average for Western Australia. It would be interesting to investigate this further to determine the true rate of gestational diabetes in the rural obstetric population.

About a fifth of Western Australian women deliver their babies in rural and remote regions of the state and 77.5% of those living in rural areas choose to deliver in rural areas². In absolute numbers that was 5587 births in 2007². Perinatal morbidity data from Western Australia tells us that certain rural regions fare worse than others⁶. Data from the 12th Report on Perinatal and Infant Mortality⁶ show that the perinatal infant mortality rates were significantly higher in the Kimberley and the Great Southern regions than in metropolitan areas. The report from the previous triennium showed higher mortality rates in the Goldfields and the Pilbara⁶. Perinatal mortality is the tip of the iceberg of antenatal and perinatal morbidity and one of the more obvious differences is the higher rate of Aboriginal women giving birth in rural and remote regions and the disproportionately high rate of Aboriginal babies represented in the perinatal morbidity statistics⁶. Other contributing factors that apply to all people in rural areas regardless of ethnicity include reduced access to all levels of health care due to physical isolation or distance from facilities and fewer numbers of health personnel per head of population, particularly obstetrically qualified doctors and midwives. In Western Australia, the medical workforce caring for these women is predominantly non-specialist, that is GP obstetricians³. Data from the BEACH study⁷ reflects a similar trend Australia-wide, showing that rural GPs see a significantly higher number of obstetric consultations than their urban colleagues. In summary, a significant proportion of the total births for Western Australia occur in rural regions. We know from perinatal mortality data that some of the highest risk groups are in the remotest regions. Access to a tertiary centre is problematic in emergencies and difficult even for planned consultations. And the majority of doctors providing obstetric care in rural Western Australia are GPs with an interest in obstetrics.

The safety of remote and rural centres has been looked at elsewhere. For example, Grzybowski et al. conducted a five-year prospective review of obstetric outcomes in a remote rural community in Canada⁸. In that study all deliveries in one small remote hospital over a four-year period were reviewed. The outcome measures included perinatal mortality, low birth weight, neonatal transfer and low Apgar at 5 minutes. A total of 286 women were included and there were 6 perinatal deaths of which 5 were very premature births. Predictably there were no significant differences between these extreme endpoints and expected results. It was calculated that the study would have to continue for 85 years to collect enough cases to test for significant differences in perinatal mortality⁸.

Morbidity outcomes rather than mortality allow for a more meaningful comparison when there are small numbers of patients. Mortality figures are effectively the tip of the iceberg when compared with morbidity data. Because mortality events are so infrequent, it is difficult to get meaningful sample sizes, particularly in smaller communities. Morbidity data reports more frequent events and a comparison in smaller sample sizes is possible The results reported in this study looked at similar numbers of patients to the Canadian study but reviewed antenatal and perinatal complications and outcomes in greater detail. The absence of any perinatal or maternal mortality in this study is reassuring but of no statistical consequence due to the small numbers. The results for factors such as hypertension, pre-eclampsia and diabetes can be meaningfully compared with expected averages.

The isolated and scattered environment of rural health often means that there are fewer resources per head of population in rural and remote areas than urban areas. However rural people still need adequate and appropriate health care. This study supports the argument that appropriately trained generalist doctors can safely care for rural obstetric patients in areas where specialist services are less accessible.

Conclusion

The results of this audit show that in a rural setting where options for referral may be less than in an urban setting it is still possible to practice obstetrics safely. These results support the hypothesis that it is safe to practice GP obstetrics particularly in a rural setting. An unexpected result in this audit was the high rate of gestational diabetes, which is significantly greater than the average published for Western Australia. Further investigation of rates of gestational diabetes in other rural populations is warranted to explore this further.

References

1. Nguyen N, Gee V, Le M. Perinatal statistics in Western Australia, 2007: Twenty-fifth Annual Report of the Western Australian Midwives' Notification System. Department of Health, Western Australia. In: Maternal and Child Health Unit Information Collection and Management. Perth, WA: Department of Health Western Australia, 2008.

2. Gee V, Hu QM, Ernstzen AN. Perinatal statistics in Western Australia, 2005: Twenty-third Annual Report of the Western Australian Midwives' Notification System. Department of Health, Western Australia. In: Maternal and Child Health Unit Information Collection and Management. Perth, WA: Department of Health Western Australia. 2007.

3. Snowball K. Maintaining an effective procedural workforce in rural Western Australia. Commissioned Report by Australian General Practice Training. Canberra, ACT: AGPT, 2007.

4. Australian Bureau of Statistics. National Health Survey: summary of results, 2007-2008 (Re-issue) 4364.0. (Online) 2009. Available: http://www.abs.gov.au/ausstats/abs@.nsf/Latestproducts/4364.0Main%20Features52007-2008%20%28Reissue%29?opendocument&tabname=Summary&prodno=4364.0&issue=2007-2008%20%28Reissue%29&num=&view= (Accessed 25 August 2009).

5. Cunningham GE, Leveno KJ, Bloom SL, Hauth JC, Gilstrap L, Wenstrom KD (Eds). Williams obstetrics, 22nd edn. New York: McGraw-Hill, 2005.

6. Perinatal and Infant Mortality Committee of Western Australia. 12th Report of the Perinatal and Infant Mortality Committee of Western Australia, Deaths 2002-2004. Perth, WA: Department of Health, Western Australia, 2007.

7. Knox S, Britt H, Pan Y, Miller GC, Bayram C, Valenti L et al. Bettering the evaluation and care of health. Locality matters: the influence of geography on general practice activity in Australia 1998-2004. In: General Practice Series, A joint report by the University of Sydney and the Australian Institute of Health and Welfare. Canberra, ACT: Australian Institute of Health and Welfare, 2005.

8. Grzybowski SCW, Cadesky AS Hogg WE, Rural obstetrics: a 5-year prospective study of the outcomes of all pregnancies in a remote northern community. Canadian Medical Association Journal 1991; 144(8): 987-994.