Background and aims In a previous study on 135 cases, the rate of autopsies in infants dying after 28 days of life appeared low. The aim of the present study was to analyse in a larger sample whether age at death influences autopsy rates.

Methods Retrospective cohort study. Descriptive statistics was used for continuous variables. To study the relationship between rates of autopsy and days of age at death, Chi² test and Kruskal Wallis rank test were performed. The variable “days of age at death” was stratified in four groups: A: 0–2 days, B: 3–7 days, C: 8–28 days, D: >28 days.

Results 402 infants died during the period 1/1998–12/2012 and 239 had an autopsy (59.4%). The mean value of days-of age-at-death in infants with autopsies was 9.8 ± 23 days, CI 95% (6.8–12.9) vs. 18.3 ± 28 days CI 95% (13.9–22.6) for cases without autopsy, p < 0.01. Group A had a high percentage of cases with autopsy: 115/ 171 (67%) In group D a low percentage of cases had an autopsy: 13/46 (28%). Using Kruskal Wallis rank test differences between Group A and group D were significant (p < 0.001).

Conclusions Autopsy rates in infants dying during the first 2 days of life were significantly higher than rates after 28 days of life. Reasons for these differences need to be evaluated in a prospective manner that should include social, educational, religious and behavioural aspects of parents and caregivers.