Diagnostic value of radiography in cases of perinatal death: a population based study

Objective: To examine the yield of radiographic abnormalities in a population based set of perinatal deaths, the diagnostic value of whole body postmortem radiographs in the same set, and previous factors that may increase the proportion of useful examinations. Design: Retrospective population based study. Setting: A region of Norway. Patients: All infants from a well defined geographical area who were stillborn or had died soon after birth over an 11 year period (n=542), who had routinely undergone whole body radiography and autopsy. Main outcome measures: (a) Proportion of cases with abnormal radiographic findings. (b) Proportion of abnormal radiographs providing new information that was useful for postmortem diagnosis. Results: Radiographs were abnormal in 162/542 cases (30%). These provided new information about, but did not help to confirm, the pathological process leading to death in 14/162 (8.6%), may have helped to confirm, but not establish, the cause(s) of death in 1/162 (0.6%), and were of vital importance for establishing the cause(s) of death in 5/162 (3.1%). Among infants with external malformations, the proportion of useful radiographs was 12/100 (12%), and among the remainder it was 8/436 (1.8%), a difference of 10.2% (95% confidence interval 3.7% to 16.7%; data missing for six cases). Conclusions: The diagnostic value of postmortem radiography in this population based set was low. However, radiographic findings were of vital importance for establishing the cause(s) of death in 5/542 cases (0.9%).