How Can a Combination of Historical Demography and Prosopographical Methods Aid the Understanding of Causes of Death? An Illustration Using Maternal Mortality as an Example

Abstract

The growing availability of individual-level historic cause of death data is allowing increased insight into the construction of official mortality statistics, the role of changing medical provision and knowledge, and the practices of individual medical practitioners. Even the most detailed demographic data can shed little light on the particular choices that doctors made, however. We argue that a mixed methods approach, combining demographic and prosopographical approaches, can help to resolve such questions. We illustrate this using a particular conundrum relating to cause of death recording: why doctors "hid" deaths in childbirth by allocating them to causes which cannot be assumed to be maternal mortality. Triangulating different types of evidence from different sources for a particular, but fairly typical, Scottish doctor in the mid-19th century, we argue that doctors were unlikely to have deliberately obscured the maternal nature of deaths. The evidence suggests that they were more likely to have failed to realise that although they knew a woman had recently delivered, this fact was not indicated in the death register and thus the causes of death they offered could often not be identified as maternal mortality.