Rapid Responses to:

RESEARCH: Imelda Balchin, John C Whittaker, Roshni R Patel, Ronald F Lamont, and Philip J Steer Racial variation in the association between gestational age and perinatal mortality: prospective study BMJ 2007; 0: bmj.39132.482025.80v1 [Abstract]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Need to relook into the gestational age at spontaneous labor and cut-off gestational age for induction in South-Asian women

Kishan Kumar Jani, Neeru Gupta   (20 April 2007)

There is only one, Human, race

Woody Caan   (20 April 2007)

Use of wrong denominator potentially undermines conclusion

Gordon C S Smith, Ian R White, Senior Scientist, MRC Biostatistics Unit, Cambridge, UK.   (28 April 2007)

Need to relook into the gestational age at spontaneous labor and cut-off gestational age for induction in South-Asian women 20 April 2007
Kishan Kumar Jani, Chief Medical Officer Civil Hospital, Shahadra, Delhi-110032., Neeru Gupta Send response to journal: Re: Need to relook into the gestational age at spontaneous labor and cut-off gestational age for induction in South-Asian women	Editor-Perinatal mortality is the yardstick of obstetric and pediatric care before and around the time of birth. It is also affected by the birth weight and the gestational age of the baby and various maternal disorders. This is an interesting article by Balchin et al with very important finding that racial differences in perinatal mortality exist despite the fact that they all were given care in maternity hospitals in London, which is supposedly same for pregnant women belonging to the three races, viz. whites, blacks and South Asians1. The pregnancy is safest if it is between 37-42 weeks. The paper by Dr. Balchin et al nicely brings out the fact that induction should be contemplated at 40 weeks in south-Asian women as risk for perinatal mortality rapidly increases beyond this gestational age irrespective of the reasons (being socially or biologically disadvantaged)1. A study in Singapore (2002-03) revealed that among the 1,094 Singaporean women, who went into spontaneous labour, the mean gestational age was 272.1 9 days (80% delivered spontaneously between 261-280 days) and perinatal mortality and predictors of perinatal morbidity (e.g. meconium staining of liquor) increased significantly beyond 280 days (40 weeks) among them2. Studies support that induction of labour per se is not a risk factor for ceasarean section3,4. Increased risk of Ceasarean sections is associated with low bishop score or unfavorable cervix4. This too can be remedied by using cervical ripeners e.g. Prostaglandin E2 etc. Generally the benefits to mother and child incurred due to induction outweigh the potential risk to them and there are established practice guidelines for safe clinical use of methods of induction5. More studies are required to know the proportion of women delivering spontaneously beyond 40 weeks of gestation in South Asian populations. If the onset of spontaneous labor or length of pregnancy is 259- 280 days (37 -40) weeks in majority of women in these populations, then a small proportion of pregnancies going beyond 40 weeks of gestation, requiring induction of labor, will not impose much of extra burden on health care system. Moreover, till 40 weeks we already have a back-up of three extra weeks beyond 37 weeks (safeguarding against prematurity due to irregularity in cycles or wrong dates or lack of precision in ultrasound checkups), therefore there is no need to allow the pregnancy to continue beyond 40 weeks in the vulnerable South Asian populations. The intrapartum deaths in these populations should also be studied in terms of gestational age and if the mortality is found higher beyond 40 weeks of gestation, it will further substantiate the findings of this study and all the more necessitate the induction of labor at 40 weeks of pregnancy. A randomized controlled trial (taking a large sample size to give sufficient power to results) involving close foetal surveillance beyond 40 weeks rather than 41 weeks, as suggested by authors, is the perfect answer to determine the cut-off gestational age at which the induction should be done in these populations. It is especially required to know whether induction at 40 weeks would also rescue the growth retarded babies which is a common occurrence in South Asian popultions. Referrences: 1. Imelda Balchin, John C Whittaker, Roshni R Patel, Ronald F Lamont, Philip J Steer Racial variation in the association between gestational age and perinatal mortality: prospective study. BMJ, doi:10.1136/bmj.39132.482025.80 (published 2 March 2007). 2. Bhat RA, Kushtagi P. A re-look at the duration of human pregnancy. : Singapore Med J. 2006;47(12):1044-8. 3. Vrouenraets FP, Roumen FJ, Dehing CJ, van den Akker ES,Aarts MJ, Scheve EJ. Bishop score and risk of cesarean delivery after induction of labor in nulliparous women. Obstet Gynecol. 2005;105(4):690-7. 4. Alexander JM, MCIntire DD, Leveno KJ. Prolonged pregnancy: induction of labor and cesarean births Obstet Gynecol. 2001;97(6):911-5. 5. [No authors listed] Induction of labor. Int J Gynaecol Obstet. 2000;69(3):283-92. Competing interests: None declared
There is only one, Human, race 20 April 2007
Woody Caan, Professor of public health Anglia Ruskin University, Cambridge CB1 1PT, UK. Send response to journal: Re: There is only one, Human, race	The important scientific findings of Balchin et al [1] contain one offensive flaw, the use of the concept of Race as if it were a scientifically validated description in the way that they apply ‘racial classification’, ‘racial groups’ and ‘racial patterns’ in their paper. First in the Imperial powers of the late 19th century, and then in some 20th century Totalitarian states, many attempts at a scientific racism were made, most notoriously by Himmler’s medical anthropologists during the 1930s. The Nazis assumed that what these authors label ‘racial variation’ really did correspond with what they discuss as ‘genetic variation’ [1] – but then Himmler and Hitler believed in an ideal Aryan Race descended from the lost civilization of Atlantis and genetically diluted with inferior stock along a unidimensional spectrum of mischlings and untermenchen. In states where Race Laws existed within living memory (especially the USA) skin colour is often equated with these classifications, and reading the Abstract I thought this was the mistake made by the authors, but in the Methods they say women ‘self reported’ belonging to groups such as Caribbean or Bangladeshi. The Greek term Ethnos (a people) describes such belonging one attributes oneself, and throughout the paper it would have been better to employ correct terms like Ethnic Groups. The authors assume genetic patterns account for their birth data, which is possible but unproven. Londoners who identify with descriptions like Caribbean or Bangladeshi may also share common experiences of poor housing, poor education and limited employment opportunities, that could also influence children’s growth…. [1] Balchin I, Whittacker JC, Patel RR, Lamont RF, Steer PJ. Racial variation in the association between gestational age and perinatal mortality: prospective study. BMJ 2007; 334: 833 . Competing interests: None declared
Use of wrong denominator potentially undermines conclusion 28 April 2007
Gordon C S Smith, Professor of Obstetrics & Gynaecology Cambridge University, CB2 2SW, UK, Ian R White, Senior Scientist, MRC Biostatistics Unit, Cambridge, UK. Send response to journal: Re: Use of wrong denominator potentially undermines conclusion	A key methodological weakness in the paper of Balchin et al[1] undermines their main conclusions. Balchin et al describe a logistic regression model where the outcome is any perinatal death, and the predictors are gestational age, ethnic group and their interaction. Using this approach, the gestation-specific perinatal mortality rate is expressed with all births at the given week of gestation as the denominator. However, Yudkin et al proposed, 20 years ago, that the correct denominator for antepartum stillbirth at a given week of gestation is the number of on -going pregnancies.[2] Since over 60% of the losses are due to antepartum stillbirth, the majority of perinatal deaths are being related to an inappropriate denominator. This inappropriate analytic approach is likely to explain the lower risk of perinatal death before 32 weeks among black women. Black women in their population are at increased risk of spontaneous preterm birth.[3] Hence, at a given preterm gestation, more black women will give birth. This increases the denominator and leads to an apparently lower risk of stillbirth compared with white women. However, this is not a true lower risk of stillbirth, it merely reflects the fact that the distribution of this incorrect denominator in relation to gestational age varies with race. Consistent with this, we have previously shown an increased risk of antepartum stillbirth among black women in London between 24 and 32 weeks when analysed using the appropriate denominator.[4] We have developed Yudkin’s approach by using time to event methods where gestational age is the time scale, antepartum stillbirth is the event and all other births are censored,[5] an approach which has now been used by other authors.[6] The issues of appropriate denominators for different types of perinatal death have been discussed in detail elsewhere.[7] We demonstrate the importance of using correct denominators by re- analysing some previous work. Time-to-event methods show that women with a previous preterm birth are at increased risk of both explained and unexplained antepartum stillbirth in the second pregnancy.[8] Using a Cox model and a test of the proportional hazards assumption, we demonstrated that the increased risk of unexplained stillbirth was stronger prior to 33 weeks.[8] However, if we re-analyse the data using births before 33 weeks as the denominator, we find that the risk of unexplained stillbirth prior to 33 weeks appears to be reduced within this group (odds ratio 0.43, 95% confidence interval 0.25 to 0.75). This is not a true finding and merely reflects the fact that any factor associated with an increased risk of preterm birth will appear to be associated with a reduced risk of stillbirth at preterm gestations when the denominator is wrongly restricted to births at preterm gestations. We urge these authors to re-analyse their data by separating antepartum stillbirth from other perinatal deaths and using a time to event method for the antepartum stillbirths: we predict that they will find an increased risk prior to 33 weeks among black women. Such statistical issues might seem arcane. However, incorrect analytic methods have the potential to cause harm by failure to identify women at high risk and failure to identify the gestational age where the risk of stillbirth is increased. In the case of statistical analysis of perinatal mortality and gestational age, the devil is in the detail. Statistical modelling of antepartum stillbirth risk in relation to gestational age is best performed using time to event methods, as described above, or logistic regression, where the event is death within a narrow gestational window and the denominator is all births at that or a later gestational age.[8] References 1. Balchin I, Whittaker JC, Patel RR, Lamont RF, Steer PJ. Racial variation in the association between gestational age and perinatal mortality: prospective study. BMJ 2007;334:833. 2. Yudkin PL, Wood L, Redman CW. Risk of unexplained stillbirth at different gestational ages. Lancet 1987;1:1192-4. 3. Steer P. The epidemiology of preterm labour. BJOG 2005;112 Suppl 1:1-3. 4. Smith GCS, Yu CK, Papageorghiou AT, Cacho AM, Nicolaides KH. Maternal uterine artery Doppler flow velocimetry and the risk of stillbirth. Obstet Gynecol 2007;109:144-51. 5. Smith GCS, Pell JP, Dobbie R. Caesarean section and risk of unexplained stillbirth in subsequent pregnancy. Lancet 2003;362:1779-84. 6. Smith GCS. Re: "A proportional hazards model with time-dependent covariates and time-varying effects for analysis of fetal and infant death". Am J Epidemiol 2005;161:100-1. 7. Smith GCS. Estimating risks of perinatal death. Am J Obstet Gynecol 2005;192:17-22. 8. Smith GCS, Shah I, White IR, Pell JP, Dobbie R. Previous pre-eclampsia, preterm delivery and delivery of a small for gestational age infant and the risk of unexplained stillbirth in the second pregnancy: a retrospective cohort study, Scotland 1992-2001. Am J Epidemiol 2007;165:194-202. Competing interests: None declared