The use of linked registries to assess long-term mortality of ICU patients.

Clinical registries are frequently used to monitor and analyze the quality of health care by assessing the in-hospital mortality. However, long-term mortality is often ignored as it is rarely recorded in such clinical registries. In this study linkage of a clinical registry and administrative database is used to assess the longterm mortality of a large ICU sample. Information about long-term mortality may be used to inform patients about their prognosis, to get insight in factors that influence long-term mortality, and to adjust admission policy to the ICU. This study showed that the observed mortality in the total ICU population at 3, 6, and 12 months after ICU admission was 20.3%, 22.9%, and 26.6% respectively. Medical and urgent surgery patients showed a higher long-term mortality risk and planned surgery patients showed a lower long-term mortality risk compared to the other ICU patients. In this study we have focused on the general ICU population, though linkage of clinical and administrative databases can also be used to perform analyses in specific diagnostic ICU populations or for non-ICU patients. In this study 71.4% of the clinical records could be linked with the administrative database. Future studies should focus on improving linkage of different registries.