UK national recommendations include 24 h non-resident availability of consu
ltants with a sessional commitment to intensive care (intensivists). We tes
ted whether continual availability of such specialists improved standardise
d mortality ratios compared with non-specialist cover by anaesthetists who
also cover other hospital departments. The case-mix-adjusted hospital morta
lity of intensive-care patients improved significantly in the intensivist g
roup compared with the nonspecialist group (standardised mortality ratios 0
.81 vs 1.11 ratio 0.73 [95% CI 0.55-0.97]). Introduction of 24 h intensivis
t cover, therefore, seems to improve outcomes in intensive-care units.