I was saying something that happened in the hospital– a treatment, intervention, or procedure– could have caused the death, making it not the “fault” of being a planned out-of-hospital birth. In cases like that, it would be misleading to label the case an out-of-hospital fatality, even though that was the intended place of birth.

But designing a study that takes that into account is tricky. Some deaths that take place in a hospital (after transferring) will be unpreventable, regardless of location of birth; some will have been preventable, but the result of decisions made prior to the transfer (ie out of the hospital); and others will have been preventable, but are the result of something that happened after the transfer (ie in the hospital). Researchers could try to ascertain which was the cause in each death, and assign the cases to a study group based on that, but that would require access to more than death certificates and birth certificates, and would introduce more human error.