In this story, we hear about a long-term comparison of three kinds of birth in Scotland: planned C-section, unplanned C-section and vaginal delivery. This enormous observational study included more than 320,000 children, and followed their health for many years. There was a slight increase in risk of Type 1 diabetes between the scheduled and unscheduled C-sections, which the story uses as a window for interviews with experts about speculation on how vaginal birth benefits babies. But the story misses key points about baseline differences in the women giving birth in three ways, which may have impacted the findings. Additionally, the story only discusses the increased risk (from the hazard ratios given) in relative percent increase, not absolute increase. This results in a misleading description of the findings and the size of the risk. Moreover, the study itself carefully warns that the results don’t prove any causation between the style of birth and the risk. Some readers of the story may mistakenly believe that the limited results prove causation.
As the story points out, almost one-third of babies born in the United States are born by C-section. This rate is higher than many other countries. Some of those deliveries are determined by medical conditions that require a C-section, and some are more related to the preference of physicians or patients. Women who may be offered an elective c-section may want to participate in the decision more actively and learn more about the benefits and risks, including those raised by this study.
The article does not include any discussion of costs. We believe that information could have added a further dimension to the topic.
Cesarean delivery is estimated to cost 50% more than vaginal delivery, in an analysis from 2013 of US births. [http://www.medscape.com/viewarticle/803426_2]
At a time when health-care costs are important drivers of policy, this is not a small factor. Health insurers, including the government’s Medicaid program, could adjust policies to discourage elective C-sections.
The story could have added this dimension to the discussion of impact. Unwarranted variation in procedures of any kind that carry risk and may not always improve care should raise questions.
The article doesn’t go far enough in explaining the difference in risk between the planned and unplanned C-sections, and the vaginal deliveries.
The story presented only the relative increased risk of 35% (from a hazard ratio of 1.35) in type 1 diabetes, despite the fact that the absolute risk was 0.22 (less than 1/4 of 1 percent!). This is a VERY modest increase and was only statistically significant because of the huge sample size.
Readers need this kind of context to be able to interpret the findings.
The story is a comparison of harms and it did a reasonable job of describing relatively slight differences between the study groups: planned C-sections, unplanned C-sections and vaginal deliveries.
The headline overstates what the story describes carefully. In the excerpt below, italics by editor.
“Over all, the differences between a scheduled C-section and an emergency C-section were slight. However, the data do begin to shed light on why babies born through vaginal birth may have fewer health risks than babies born by C-section.
The biggest difference between babies born by scheduled and unscheduled C-section appeared in risk for Type 1 diabetes. The results showed that babies born by planned C-section had a 35 percent higher risk of Type 1 diabetes compared with babies born by emergency C-section, after adjusting for differences among the mothers.”
As noted above, the risk sounds much higher when given in relative terms, but since we’ve already docked points for that concern above,we’ll award a Satisfactory rating here.
The story explains that the study was observational and of more than 250,000 Scottish births to first-time mothers. A key part of the study’s importance is the long-term follow up of 15 years on the children.
We would have liked the story to emphasize that the study does not prove causation between the style of delivery and the subsequent health risks, it only shows a correlation. In addition, the story did not explain that in this observational study, there were substantial baseline differences in the women who had the different types of births. For example the women who had scheduled c-sections were older, more obese, and had higher rates of type I diabetes themselves. Although multivariable regression adjusts for baseline factors, there is strong possibility of residual confounding (meaning that differences among the moms may be responsible for the outcomes observed even after statistical adjustment). This is particularly true for the outcome of diabetes in the children.
There was no disease mongering.
Several independent experts were quoted.
The story did a careful job of comparing the different delivery methods. There could have been perhaps a little exploration about recovery time and adverse events in the moms who have c-section.
All three of the birth methods described are widely available.
As the story points out, the existence of additional health risks for babies born by C-section are already well known. That is not a novel finding. What was new was the comparison between planned and unplanned C-sections.
The story did not rely on a news release. There is evidence of independent reporting.