To give you my nonspecific thoughts on "this kind of stuff," here is an excerpt from the manuscript for the new edition of Obstetric Myths Versus Research Realities by Henci Goer and Amy Romano to be published by Michigan University Press. It comes from a section entitled "How to Avoid Being Misled by the Obstetric Research."
Do investigators display medical model assumptions and biases? Distrust studies where investigators display medical model assumptions and biases. These can distort research on every level from study design to data interpretation to conclusions drawn. . . . Watch for:
- . . .
- loaded language: Classic examples of these are calling estimated 500 mL blood loss at vaginal birth—the same amount as taken when donating blood—“hemorrhage,” terming subnormal blood pH in the fetus or newborn “asphyxia,” and deeming elective cesarean surgery “maternal request” cesarean.
- calls for intervening more often to improve outcomes rather than refraining from intervention, such as, for example, assigning healthy women to undergo prophylactic intervention, as is happening with calls for trials of elective cesarean surgery and “preventive” induction of labor.
- calls for using a different intervention instead of using supportive techniques, as for example, replacing the now discredited episiotomy with planned cesarean surgery as a means of protecting the pelvic floor.
The Nicholson trial engages in all of the above.
In addition, the trial was carried out in an institution that invented and is promoting Active Management of Risk in Pregnancy at Term (AMOR-IPAT), and participating clinicians knew to which group women had been assigned. The authors acknowledge this point but dismiss the potential for bias saying it is "unlikely" (p. e12) that knowledge of group assignment would affect the decision to perform cesarean. Their opinion is not backed up by the facts. A large body of research attests to the effect of provider beliefs on cesarean rates.
Finally, Nicholson and his team are arguing that preventive induction is the key to a low cesarean rate and superior newborn outcomes. If a study can be found showing that a low cesarean rate and good outcomes can be achieved without preventive induction, their theory is disproved. Here is one such study:
Leeman L, Leeman R. Do all hospitals need cesarean delivery capability? An outcomes study of maternity care in a rural hospital without on-site cesarean capability. J Fam Pract 2002;51(2):129-34.
The investigators studied 1132 women in the catchment area of a rural hospital without on-site capability to perform cesareans. The population was at substantially higher risk than the one in the Nicholson 2008 trial in that the study reported outcomes for all women, including those not eligible for care at the rural hospital and those with medical indications for induction--women who would have been excluded from the Nicholson trial--as well as those transferred during labor. The cesarean rate in women with no prior cesarean was 5% (Nicholson rates: 10% preventive induction group and 15% usual care group), and the induction rate was 14% (Nicholson rates: 58% preventive induction group and 22% usual care). Neonatal outcomes in the Leeman study were similar to or compared favorably with national statistics.
In other words, whatever was responsible for the improved outcomes in the Nicholson trial, it wasn’t preventive induction.
-- Henci