| Context: Existing research has associated higher provider volume with a lower rate of adverse patient outcomes after coronary artery bypass graft (CABG). However, the relationship between surgical procedural volume and specialization and clinical outcomes has been understudied. Research Objectives: This research examined the effect of surgical procedural volume and specialization on patient outcomes for 119,559 patients undergoing CABG in Florida short-term acute hospitals from 2000-2006. Methods: Florida Hospital Discharge Data were linked with Practitioner Profile Database available from Florida Department of Health by using unique surgeon identifiers. Surgeon on-pump and off-pump CABG volume was assessed in quartiles. In-hospital complications were measured by using Patient Safety Indicators developed by the Agency for Healthcare Research and Quality (AHRQ). Analyses included chi-square, t-test, logistic regression and multilevel regression to adjust for nested surgeon and hospital effects. Results: In adjusted analyses stratified by on- and off-pump CABG, patients operated by surgeons with lower volume of a specific CABG type were more likely to have in-hospital mortality: for off-pump CABG quartile 1 OR=3.05, 95% CI: 1.68-5.53, quartile 2 OR=1.57, 95% CI: 1.10-2.26 and quartile 3 OR=1.35, 95% CI: 1.01-1.81, and for on-pump CABG quartile 2 OR=1.82, 95% CI: 1.34-2.47 and quartile 3 OR=1.51, 95% CI: 1.21-1.90. Surgeries performed by physicians in lower on-pump CABG quartiles were also significantly associated with increased odds of complications (quartile 1 OR=1.97, 95% CI: 1.19-3.26, quartile 2 OR=1.43, 95% CI: 1.14-1.80 and quartile 3 OR=1.33, 95% CI: 1.14-1.57). For off-pump CABG only quartile 2 physicians retained significance (OR=1.80, 95% CI: 1.29-2.51) for complications. Discussion: The volume/outcome relationship for CABG surgery is specific to the type of procedure, but not total (all procedures) volume. This finding may suggest the need of specialized and focused training of cardiac surgeons as well as development of specific CABG outcome reporting protocols to enable sufficient differentiation in outcomes of two different types of CABG. |
