Is the volume of mechanically ventilated admissions to UK critical care units associated with improved outcomes?

It is unknown whether a volume-outcome relationship exists for mechanically ventilated admissions to UK critical care units. This study was conducted to evaluate the volume-outcome relationship for mechanically ventilated admissions to adult, general critical care units in the UK with a view to informing policy, service delivery and organisation of specialist, advanced respiratory care. A retrospective cohort study using data from the Case Mix Programme Database was conducted. The primary exposure of interest was annual volume (absolute number) of mechanically ventilated admissions per critical care unit per year. The primary outcome was ultimate acute hospital mortality. A multivariable analysis was performed to assess the relationship between annual volume and outcome while adjusting for a priori selected confounders. Two interaction tests were performed. The first interaction test was between annual volume and admission type and the second between annual volume and initial acute severity of respiratory failure. Sensitivity analysis excluding volume outlier units and using restricted cubic splines to model volume was also performed. After adjusting for confounding, there was a significant relationship between annual volume and ultimate acute hospital mortality (p < 0.02). The first interaction test revealed a strong interaction between annual volume and admission type, with a more pronounced volume-outcome relationship for non-surgical admissions (p < 0.001). The second interaction test between annual volume and initial acute severity of respiratory failure was not statistically significant (p = 0.12). The analysis using restricted cubic splines demonstrated a similar graphical relationship but the results were not statistically significant (p = 0.87). A volume-outcome relationship was demonstrated for mechanically ventilated admissions to adult, general critical care units in the UK. The relationship is sensitive to the modelling approach used.