Power in numbers

Abstract

Collecting useful data on a sufficiently large cohort of pregnancies in women with rheumatic disease is a challenge. The original manuscripts that demonstrated the dangers of pregnancy in women with lupus were relatively small case series. As larger prospective cohorts were collected by university-based experts, however, greater safety was demonstrated and the current norms of treatment were determined. In recent years, larger administrative databases have been tapped to study pregnancies not managed within university clinics and to study the long-term impact of maternal rheumatic disease on the offspring. Each of these methods of study has both strengths and weaknesses, adding a unique piece of data to our overall knowledge. We will discuss a range of approaches to the study of rheumatic disease in pregnancy, covering the potential benefits that each brings as well as the biases that can impact study results. When the results of studies are viewed through these lenses, each can contribute to our larger understanding of the rheumatic diseases in pregnancy.

Fig. 1

Example of immortal-time bias in study of association of NSAIDs with fetal loss Reprinted from American Journal of Obstetrics & Gynecology, 212/3, Daniel S et al. Immortal time bias in drug safety cohort studies: spontaneous abortion following nonsteroidal antiinflammatory drug exposure, pp. 307.e1–6, Copyright (2015), with permission from Elsevier.