Non-invasive or minimally invasive autopsy compared to conventional autopsy of suspected natural deaths in adults: a systematic review

Britt M. Blokker, Ivo M. Wagensveld, Annick C. Weustink, J. Wolter Oosterhuis, M. G. Myriam Hunink

Autopsies are used for healthcare quality control and improving medical knowledge. Because autopsy rates are declining worldwide, various non-invasive or minimally invasive autopsy methods are now being developed. To investigate whether these might replace the invasive autopsies conventionally performed in naturally deceased adults, we systematically reviewed original prospective validation studies. We searched six databases. Two reviewers independently selected articles and extracted data. Methods and patient groups were too heterogeneous for meaningful meta-analysis of outcomes. Sixteen of 1538 articles met our inclusion criteria. Eight studies used a blinded comparison; ten included less than 30 appropriate cases. Thirteen studies used radiological imaging (seven dealt solely with non-invasive procedures), two thoracoscopy and laparoscopy, and one sampling without imaging. Combining CT and MR was the best non-invasive method (agreement for cause of death: 70 %, 95%CI: 62.6; 76.4), but minimally invasive methods surpassed non-invasive methods. The highest sensitivity for cause of death (90.9 %, 95%CI: 74.5; 97.6, suspected duplicates excluded) was achieved in recent studies combining CT, CT-angiography and biopsies. Minimally invasive autopsies including biopsies performed best. To establish a feasible alternative to conventional autopsy and to increase consent to post-mortem investigations, further research in larger study groups is needed. • Health care quality control benefits from clinical feedback provided by (alternative) autopsies. • So far, sixteen studies investigated alternative autopsy methods for naturally deceased adults. • Thirteen studies used radiological imaging modalities, eight tissue biopsies, and three CT-angiography. • Combined CT, CT-angiography and biopsies were most sensitive diagnosing cause of death.