Diagnostic performance of minimally invasive autopsy in a respiratory ICU: A retrospective analysis

Post-mortem tissue sampling can help determine the cause of death, particularly in resource-limited settings. We conducted a retrospective audit of minimally invasive autopsies (MIA) performed in a respiratory intensive care unit.

Lung biopsies were obtained using either a core biopsy needle (CBN) or through a small thoracic incision at the fourth intercostal space (open incision method [OIM]). We performed a kidney or liver biopsy using a CBN with ultrasound guidance. The final cause of death, ascertained by a multidisciplinary team, served as the reference standard. The primary objective was to assess biopsy yield and the frequency of diagnostic revision based on histopathological findings.

One hundred and sixteen patients underwent MIA. Lung biopsies were done in all cases (66 [56.9%] by CBN and 50 [43.1%] by OIM). Kidney and liver were biopsied in less than 10% of cases. Twelve biopsies (10.3%) were non-representative and excluded from further analysis. The overall biopsy yield was 74%. Pathological findings led to a change in the final diagnosis in three cases (2.9%). Infections aetiologies were most common, including bacterial pneumonia (24, [23.1%]), fungal pneumonia (8, [7.7%]), and tuberculosis (2, [1.9%]). Malignancy was identified in 8 (7.7%) cases. The diagnostic yield was significantly higher with OIM than with CBN (87.5% vs. 62.5%, P = 0.004). Agreement between clinical and pathological diagnoses was fair (kappa = 0.41, P < 0.001).

MIA achieved a 74% diagnostic yield with fair agreement with the MDT-determined cause of death. OIM provided a superior yield compared to CBN and may be preferable where feasible.