Abstract
Background: In recent years, Greece has been in the spotlight as one of the countries hit the hardest by the European financial crisis. Yet evidence on the impact of the crisis on total and cause-specific mortality remains unclear. The paper explores whether the economic crisis has interrupted the level and/or the trend of mortality rates. Methods: We use regional panel data for the period 2001-2013 and adopt an interrupted time-series approach in order to test for any changes in a series of gender, age and cause-specific mortality trends. Findings: Our results show that overall mortality has continued to decline during the years of the financial crisis but at a slower pace (-0∙13 vs. -0∙065) and the trend difference is statistically significant (0∙062, p<0∙001). The trend difference is more evident for women (0∙087, p<0∙001) than for men (0∙040, p<0∙01). Older age groups have experienced more negative effects than the younger population, who have even seen some improvements in mortality trends. Deaths by diseases of the circulatory system have declined more slowly, while deaths from vehicular accidents have declined faster during the crisis, most prominently among young men. On the other hand, deaths from suicides, diseases of the nervous system and mental disorders have followed a positive upward trend during the crisis. We also find evidence that deaths due to medical errors during treatment have increased considerably during the crisis, affecting mainly older adults. Comparing the expected values of the during-crisis versus extrapolated before-crisis fitted, one estimates an extra 241 deaths per month during the crisis period with the largest contributor being an excess of 153 deaths per month from circulatory causes. Interpretation: The findings suggest that mortality trends have been interrupted during the years of the crisis, but that the changes vary by age, gender and cause of death. The increase of deaths due to medical errors may reflect the effects of quality deterioration during economic recessions.