Validation of a tool for determining the clinical utility of stillbirth investigations

Abstract

Background: Up to 20% of all stillbirths and 45% of term stillbirths are currently classified as unexplained. Many of these stillbirths do not undergo currently recommended investigations. This may leave questions unanswered and not identify stillbirths with a recurrence risk in subsequent pregnancies.

Aims: To validate a new tool (Stillbirth Investigation Utility Tool) to identify the clinical utility of investigations in stillbirth and the inter-rater agreement on cause of stillbirth using the Perinatal Society of Australia and New Zealand-Perinatal Death Classification (PSANZ-PDC).

Materials and methods: Thirty-four stillbirths were randomly selected for inclusion, each assessed independently by five blinded assessors. The investigations were grouped into three categories: clinical and laboratory; placental pathology; and autopsy examination. The cause of death was assigned at the end of each group. Outcome measures were clinical utility of investigations measured by assessor rated usefulness and inter-rater agreement on the assigned cause of death.

Results: Comprehensive maternal history, maternal full blood count, maternal blood group and screen and placenta histopathology were useful in all cases. Clinical photographs were not performed and should have been performed in 50% of cases. The inter-rater agreement on cause of death assigned after all investigation results was 0.93 (95% CI 0.87-1.0).

Conclusions: The new Stillbirth Investigation Utility Tool showed very good agreement in assigning the cause of death using PSANZ-PDC. Four investigations were useful in all cases. Minor refinements will be made based on feedback to enhance usability for wider implementation in research studies to assess the yield of investigations in stillbirths.

Keywords: autopsy; clinical utility; death classification; investigations; placenta; stillbirth; yield.