Impact of Hospital Practice & Staffing Differences on Transesophageal Echocardiography Use During Cardiac Valve or Coronary Artery Bypass Graft Surgery
Key Points
Among 261,860 patients, 123,702 underwent valve and 138,158 underwent isolated CABG surgery. TEE was more frequently used in valve (76%) vs isolated CABG (50%).
The hospital in which cardiac surgery was the strongest predictor for TEE use – more predictive than any patient-level or hospital-level factor. The median odds of two identical patients undergoing surgery at two randomly-selected hospitals varied more than 2.5-fold in valve surgery and more than 4-fold in isolated CABG surgery.
A percentage of this TEE variability across hospitals was explained by differences in TEE staffing models. This finding was particularly pronounced in isolated CABG surgery where the addition of TEE staffing resulted in a 21% decrease in the median odds ratio for TEE receipt (i.e. the inclusion of TEE staffing explained 21% of the across hospital variability in TEE practice patterns).
Anesthesiologist TEE staffing associated with a higher odds of TEE receipt among patients undergoing either valve or CABG surgery and was the strongest, observed, fixed effect, predictor for TEE use among those undergoing isolated CABG surgery.
Figure 1.a: Probability of Intraoperative TEE Use in Cardiac Valve Surgery by TEE Staffing Provider
After identifying the specialty of the provider performing the TEE in valve surgery, we graphed the percentage of TEE performed (y-axis) by each of the three TEE provider types (e.g. anesthesiologist, cardiologist, and other provider) against the probability of TEE use by hospital (x-axis) Among the valve surgery cohort, the median probability of an anesthesiologist-performed TEE was: 0.70 (interquartile range (IQR): 0.40, 0.83) and the median probability of a cardiologist-performed TEE was: 0.23 (IQR: 0.14, 0.46).
Figure 1.b Probability of Intraoperative TEE Use in Isolated Coronary Artery Bypass Graft Surgery by TEE Staffing Provider
After identifying the specialty of the provider performing the TEE in isolated CABG surgery, we graphed the percentage of TEE performed (y-axis) by each of the three TEE provider types (e.g. anesthesiologist, cardiologist, and other provider) against the probability of TEE use by hospital (x-axis) Among the isolated CABG surgery cohort, the median probability of an anesthesiologist-performed TEE was: 0.71 (IQR: 0.40, 0.84) and the probability of a cardiologist-performed TEE was:0.20 (IQR: 0.12, 0.42).
Figure 2.a: Anesthesiologist vs Cardiologist TEE Staffing: Comparison of the Probability of TEE by Hospital in Valve Surgery
Once the TEE Staffing models were categorized, we plotted the density of the probability of TEE according to anesthesiologist and cardiologist TEE staffing models. The y-axis is now the probability density, which facilitates the direct comparison of the two primary TEE staffing patterns: (e.g. anesthesiologist vs cardiologist) on the overall probability of TEE by hospital (x-axis). Now, among the valve surgery cohort, both the anesthesiologist and cardiologist TEE staffing models demonstrated similar median probabilities for TEE: 0.83 [IQR 0.75, 0.88] for anesthesiologist and 0.77 [IQR 0.62, 0.85] for cardiologist TEE staffing.
Figure 2.b: Anesthesiologist vs Cardiologist TEE Staffing: Comparison of the Probability of TEE by Hospital in Isolated Coronary Artery Bypass Graft Surgery
Once the TEE Staffing models were categorized, we plotted the density of the probability of TEE according to anesthesiologist and cardiologist TEE staffing models. The y-axis is now the probability density, which facilitates the direct comparison of the two primary TEE staffing patterns: (e.g. anesthesiologist vs cardiologist) on the overall probability of TEE by hospital (x-axis). In contrast to the valve surgery cohort (Figure 2.a), the probability for TEE use by hospital according to anesthesiologist vs cardiologist is bimodal (i.e. higher probability for TEE among hospitals with anesthesiologist TEE staffing vs lower probability for TEE among hospitals with cardiologist TEE staffing). Specifically, among the isolated CABG surgery cohort, the median probability for TEE among hospitals was: 0.73 [IQR 0.54, 0.81] for hospitals with anesthesiologist TEE staffing vs. 0.24 [IQR 0.11, 0.52] for hospitals with cardiologist TEE staffing.