Letter published in BMJ - Birthplace study flaws

My letter to the BMJ outlining concerns about flaws in the Birthplace study was published today.

Titled, "Concerned that flawed analysis may be used to restrict birth choices", it reads:

This analysis of the 2011 Birthplace study(1) suffers from some of the same inherent flaws in the study’s collection and analysis of comparative data. But while all research has its limitations and weaknesses, the current economic climate and imminent restructuring of the NHS mean that the promise of “cost-effectiveness” here may well be used for political and ideological influence. As such, these are my concerns:

Incomplete cost analysis
As others have pointed out, the analysis does not include the downstream costs of adverse birth outcomes such as litigation, treatment of infant and maternal injuries, or counseling for psychological trauma. NICE, for example, in its cost model comparison of a planned caesarean versus planned vaginal delivery,(2) found that by introducing just ONE adverse downstream outcome (i.e. urinary incontinence), the headline difference of £720 was reduced to just £84. Further analysis like this is necessary before cost-effectiveness of any birth plan/place can be adequately demonstrated.

Other costs, women’s choice and midwifery ratios are not accounted for

A number of FMUs have (or are being) closed in the UK due to unsustainable running costs, undependable service(3) and not enough women choosing to give birth there.(4,5) In Kent, one unit managed just 300 births in 2010 and a third of women required transfer.(6) In Hull, births at a remote FMU cost £2,000 each, with a financial loss of £120,000 per year.(7) So it is crucial that we read this cost analysis of the Birthplace study (which itself cost “in the region of £12m”) in the context of real world events. That is, unit closures, and the fact that despite years of encouragement, the vast majority of women simply DO NOT want to give birth away from immediate obstetric care.

Study is weighted against OU births

The discrepancy between 20% of women with “complicating conditions” at the start of labour in OUs versus less than 7% in all others should have been adjusted throughout, and not consigned to the Appendix and occasional “sensitivity analysis”. For instance, the latter found a “significant increase in the odds of the primary outcome in planned FMU births” for nulliparous women. There were also more women at OUs with an intermediate level of risk due to “other” conditions, which “would have tended to make outcomes appear worse”. And furthermore, not only was there some general loss of paperwork/data during and after labour transfers, the authors state it’s possible that specifically, cases of intrapartum stillbirths or early neonatal deaths may not have been included. Finally, the ratio of midwives to women/babies at some OUs is very poor in comparison to midwife-led units, which can affect health outcomes and costs (e.g. 1 to 40 versus 1 to 9 in Kent(8) and 1 to 53 versus 1 to 4 in Portsmouth(9)).

Focus is on place, not mode of birth

The Birthplace study design excluded planned caesareans of women with low risk pregnancies; therefore conclusions could not be made about the comparative (short- and long-term) cost-effectiveness of consultant-led surgical care for women who choose this. In light of NICE’s conclusion that a caesarean should not be refused “on cost-effectiveness grounds", and given that a) more women request a caesarean than a home birth(10), and b) research has shown greater maternal satisfaction with maternal request CS than a planned homebirth,(11,12) future studies need to compare mode (as well as place) of birth – particularly when informing nulliparous women of birth risks (for whom, in all planned vaginal birth settings, intervention and adverse perinatal and maternal outcomes were more common).

Excludes stillbirths prior to onset of labour

The majority of stillbirths occur prior to labour, and indeed many full-term stillbirths occur in low risk pregnancies, yet the Birthplace study excluded all stillbirths prior to the onset of labour - it shouldn’t have. U.S. research into 10 years of births(13) found a significant increase in the stillbirth rate after 39 weeks’ gestation, and women deserve to know if there is a greater risk at term with a specific birth or birthplace plan. Other birth injury risks for babies were excluded too, all of which have physical, psychological and financial costs that should not be ignored.

Potential for self-serving interests

The data used to inform this study was compiled by midwives, whose own reputation, livelihood and job satisfaction are intrinsically linked with the study’s outcomes. For instance, “If the numbers of home births increases, the experience of midwives providing care at home will increase”. Also, the study design “only included women who received labour care from a midwife in their planned place of birth”, which excludes comparative health outcomes and costs with consultant-led care.

Assumption that midwife-led is the safest and most cost-effective care for all ‘low risk’ births

Firstly, incremental cost savings with “no significant effect on adverse perinatal outcomes” in home births for low risk multiparous women is not the same as finding ‘no effect’, especially given the potential loss of data described above. Secondly, since a “variability in experience and training is likely to affect midwives’ ability to provide safe and effective care”, we should be cautious about affecting any fundamental nationwide changes to maternity care policy (e.g. midwife-led care for ALL low risk women), especially at a time of well documented staff shortages. Thirdly, recent Dutch research found “a higher risk of delivery related perinatal death and the same risk of admission to the NICU” for babies under the (initial) supervision of a midwife compared to the supervision of an obstetrician.(14) Therefore, to reiterate an earlier BMJ letter,(15) it’s not proven that midwife-led care is most cost-effective, identifying “low risk” women can be very unreliable, and women should have the choice of consultant-led care too.

Summary

Somewhat ironically, while the UK seems to be trying to move towards a model of care closer to that of The Netherlands, the homebirth rate there is falling, and a spotlight remains on its poor perinatal mortality rate. Certainly homebirth and midwifery-led care deserve their place in the spectrum of birth choices, but I find it very concerning that the Birthplace study and its cost analysis might actually be used to restrict ‘low risk’ women’s choices. Surely if national maternity policy is to be informed by any study’s conclusions, then all modes of birth (incl. planned caesarean), all types of professional care (incl. consultant-led), and the most common postpartum consequences need to be examined?
References
(1) Perinatal and maternal outcomes by planned place of birth for healthy women with low risk pregnancies: the Birthplace in England national prospective cohort study. BMJ2011;343:d7400
(2) NHS National institute for Clinical Excellence, “Caesarean Section: Full Guideline” November 2011.
(3) “Ascot Birthing Centre should not reopen, says board,” BBC News Online, February 24, 2012.
(4) “Canterbury and Dover birthing centres 'should close',” BBC News Online, September 9, 2011.
(5) “NHS recommends closure of Derbyshire birth centres,” BBC News Online, December 5, 2011.
(6) “Kent and Canterbury birth unit remains shut for review,” BBC News Online, June 7, 2011.
(7) “Jubilee Birth Centre to shut after watchdog blow,” BBC News Online, July 27, 2011.
(8) “Health board expected to close birth centre,” thisiskent.co.uk, April 26, 2012.
(9) “Portsmouth hospital criticised over maternity care,” BBC News Online, August 4, 2011.
(10) National Collaborating Centre for Women’s and Children’s health, “Caesarean Section: Clinical Guideline 13,” NHS National institute for Clinical Excellence, April 2004.
(11) Perinatal factors related to negative or positive recall of birth experience in women 3 years postpartum in the Netherlands (Rijnders et al). Birth. 2008 Jun;35(2):107-16.
(12) Mothers' Satisfaction with Planned Vaginal and Planned Cesarean Birth (Blomquist et al). Am J Perinatol. 2011 Mar 4.
(13) Cesarean Section on request at 39 Weeks: impact on Shoulder Dystocia, Fetal trauma, Neonatal Encephalopathy, and intrauterine Fetal Demise (Hankins, Clark, Munn). Seminars in Perinatology 30, no. 5 (October 2006): 276–87.
(14) Perinatal Mortality and Severe Morbidity in low and high risk term Pregnancies in the Netherlands: Prospective Cohort Study (Evers et al). BMJ2010;341:c5639
(15) Midwife led care may not be appropriate or cost effective (PM Hull). BMJ2011;342:d2298

Competing interests:Co-author of "Choosing Cesarean: A Natural Birth Plan" (Prometheus Books 2012)