Reader Comments

Post a new comment on this article

Conclusions are not supported by data

Posted by morriand on 21 Aug 2015 at 13:29 GMT

Interesting paper. With the small sample size and the ongoing preceding downward trend in mortality, it is strident to claim that mortality has increased. In fact, the reductions over time are dramatically noteworthy. A statistician as author/reviewer would have strengthened this paper considerably.

No competing interests declared.

RE: Conclusions are not supported by data

maartjemvdberg replied to morriand on 25 Aug 2015 at 10:35 GMT

The comment claims that because of the small sample size, it is 'strident' to conclude that there has been an increase. The sample size is in fact quite large; it is the number of 'events' - deaths in newborns - that is relatively small. Statistical testing allows for the sample size. For a difference to be statistically significant, the point estimates must show large differences. This was the case for the data in the paper.

No competing interests declared.