Pulmonary embolism - temporal changes.

BACKGROUND

The clinical diagnosis of pulmonary thromboembolism (PE) is notoriously inaccurate, with many cases either wrongly diagnosed (overdiagnosed) or missed (underdiagnosed), and autopsy is still regarded as the diagnostic "gold standard." Unfortunately, statistics from recent, larger autopsy studies are scarce and, moreover, the results of such studies are often controversial.

METHODS

The autopsy incidence of PE from 1960 to 2005 was retrospectively reviewed.

RESULTS

In the six reference years 1960, 1970, 1980, 1990, 2000, and 2005, a total of 3877 adult autopsies were performed. PE was found in 628 patients; the overall incidence of PE in the six reference years was 16.2%. The incidence showed a steady decline, from 19.2% in 1960 to 10.0% in 2005. It was higher in women (19.4%) than in men (13.7%). The average age at the time of death for all 628 patients with PE was 65.9 years (64.3 in men; 67.3 in women). It showed a tendency to increase over the period studied, from 63.7 years in 1960 to 69.6 years in 2005. We also assessed variation of PE deaths throughout the year. The incidence was highest in November (21.4%) and lowest in May (11.1%) and in March (12.1%).

CONCLUSIONS

Our results confirm those of most pathological studies, in that the autopsy incidence of PE is declining. In patients dying with PE, the underlying causes have been more or less the same over the 45 years covered by this study. Despite the progress in clinical diagnostic approaches, PE remains an ill-diagnosed condition.