Variation in ventilation time after coronary artery bypass grafting: an analysis from the society of thoracic surgeons adult cardiac surgery database.

BACKGROUND: Short postoperative ventilation times are accepted as a marker of quality. This analysis assesses center level variation in postoperative ventilation time in a subset of patients undergoing isolated coronary artery bypass grafting (CABG). METHODS: In 2009 and 2010, 325,129 patients in the STS Adult Cardiac Surgery Database underwent isolated CABG. Patients were excluded if they were intubated before entering the operating room, required ventilation for greater than 24 hours, or had missing data on key covariates. The final study cohort was 274,231 isolated CABG patients from 1,008 centers. Bayesian hierarchical models were used to assess between-center variation in ventilation time and to explore the effect of center-level covariates. Analyses were performed with and without adjusting for case mix. RESULTS: After adjusting for case mix, the ratio of median ventilator time at the 90th percentile of the center-level distribution compared with the tenth percentile was 9.0:5.0=1.8 (95% credible interval: 1.79 to 1.85). This ratio illustrates the scale of between-center differences: centers above the 90th percentile have a ventilation time of at least 1.8 times that of centers below the tenth percentile. Smaller hospital volume, presence of a residency program, and some census regions were associated with longer ventilation times. CONCLUSIONS: After adjustment for severity of illness, substantial inter-center variation exists in postoperative ventilation time in this subset of patients undergoing isolated CABG. This finding represents an opportunity for multi-institutional quality improvement initiatives designed to limit variations in ventilator management and achieve the shortest possible ventilation times for all patients, thus benefiting both clinical outcomes and resource utilization.

Duke Scholars

Author Sean Michael O'Brien Biostatistics & Bioinformatics, Division of Biostatistics