Death by confounding

If there is one thing that natural childbirth and homebirth advocates are absolutely, positively, 100% certain about, it is that interventions increase bad outcomes. They are particularly adamant on the issue of C-sections, declaiming at every possible opportunity that C-sections increase the risk of neonatal and maternal mortality. In fact, they’ve turned their belief into a virtual cult, the ‘cult of the unnecessarean’.

As with so many things, they are flat out wrong, and as with so many issues, their perverse certainty in the face of a mountain of evidence to the contrary stems from a lack of understanding of statistics.

Statistics is more than just number crunching. Statistics also teaches us how to choose the particular numbers to crunch. When NCB and homebirth advocates talk about interventions and death rates, their math is right, but they’ve used the wrong numbers to start with. Specifically, their claims about interventions and death rates are wrong because of confounding bias.

According to Ghaemi and Thommi in Death by confounding: bias and mortality:

The confounding factor is associated with … what we think is the cause … and leads to the result. The real cause is the confounding factor; the apparent cause, which we observe, is a bystander. An example of this relationship is the statement: coffee causes cancer. Even though large epidemiological studies show that those who drink coffee are more likely to have cancer (2), this is resulting from the fact that the coffee drinkers are more likely to be cigarette smokers (3), which is the cause of cancer in those persons. Coffee is the apparent cause, whereas cigarette smoking, the real cause, is the confounding factor.

In the case of C-sections and mortality rates, what NCB advocates think is the cause (C-sections) is a bystander. The real cause is the confounding factor. What is the confounding factor in this case? It is the underlying problem that lead to the C-section in the first place.

A detailed explanation can be found in the chapter on Maternal Mortality and Severe Maternal Morbidity in the book Reproductive and Perinatal Epidemiology:

In 2000-2002, the MMR among women in the United Kingdom who delivered by the vaginal roule was 4.8 per 100,000, compared with … 17.2 per 100,000 among women who delivered by the cesarean route… [I]nference regarding the safety of the mode of delivery based on the above rules would be seriously compromised because the association is confounded by the indication for cesarean delivery. As the CEMACH [Confidential Enquiry into Maternal and Child Health] report stated “For the large majority of deaths that followed caesarean section. however, there were serious prenatal complications or illness that, in many cases, precipitated the caesarean section…”

Confounding bias occurs in other medical and obstetrical situations:

It is the underlying phenomenon that explains why death rates among patients in an intensive care unit are higher than among patients in the non- intensive care setting, … why perinatal mortality rates are higher among hospital births compared with home births, and why perinatal mortality rates among women cared for by obstetricians arc typically higher than those cared (or by nonobstetricians.)

Ghaemi and Thommi offer a humorous real life example to illustrate how confounding bias leads to “solutions” that are utterly ineffective.

… The physicist Richard Feynman described it well in analogy to the curious behaviour of natives of the South Sea Islands. During the Second World War, the natives had grown accustomed to receiving goods dropped by cargo planes. After the war, the soldiers left, and the cargo droppings stopped. The natives then created a ‘cargo cult’ whereby they built runways, put fires along them and constructed huts for ‘controllers’ to summon the planes. Everything was set up the way it looked when the planes used to come; but the planes did not come.

Sadly, NCB and homebirth advocates are just like the South Sea Island natives. They compare mortality rates of vaginal births and C-sections and are anxious to produce the mortality rates associated with vaginal births. They have created the ‘cult of the unnecessarean’ whereby they demand lower C-section rates, insist that most C-section are unnecessary and actually refuse C-section in the belief that this will reduce mortality rates. Everything is set up to imitate the lower mortality rates associated with vaginal delivery, but the lower mortality rates do not occur.

The South Sea Island natives can be excused for their futile actions because the source of the confounding bias (the reason for the original cargo drops) was literally beyond their understanding. They had no idea that the there was a war going on, that the cargo was intended to fight it, and that there was no further reason to drop cargo when the war ended.

NCB and homebirth advocates have no such excuse. It should be obvious, or at the very least understandable, that direct comparisons between interventions and mortality rates are futile. Without understanding the reasons for the C-sections and taking those reasons into account, the data is worse than useless.