///2017 Abstract Details
2017 Abstract Details2019-08-02T15:54:53-06:00

Anesthesiologist/Obstetrician Combinations Influence Choice of Anesthesia for Cesarean Delivery at an Urban New York Hospital

Abstract Number: F-33
Abstract Type: Original Research

Ivan Velickovic M.D.1 ; Joanna Serafin Ph.D.2; Jeremy Weedon Ph.D, MA, BS3; Antonella Stanko M.D.4; Marko Velickovic B.S.5; Ming Zhang M.D.6

Introduction: General anesthesia (GA) is used in less than 5% of elective Cesarean deliveries (CD) nationwide.1 There is evidence that the ratio of GA to regional anesthesia (RA) for CD varies as a function of practice environment2, but it is unclear if physician preference or patient risk factors are related to the ratio of GA to RA. XXXXXXXX Medical Center, which mostly serves socioeconomically disadvantaged populations, delivers 1,400 babies every year, of which about 40% are CD. This study examines factors that influence the probability of GA for CD.

Methods: All CD at XXXXXXXXXX 2011-2015 were included in this study. A generalized linear mixed model (GLMM) was constructed, dependent variable was whether GA was used (modeled as Bernoulli-distributed). Preoperative diagnosis (repeat/elective CD; arrest of labor; category 3 tracing; category 2 tracing; other) was a fixed factor. Surgeon-anesthesiologist dyad was a random effect. The technique of best linear unbiased estimators (BLUEs) was used to generate lists of estimated prevalence of general anesthesia for each anesthesiologist-surgeon dyad, controlling for diagnostic category. It is important to note that because 93% of the patients were African-American, only African-American patients (N=3409) were used for GLMM.

Results: Controlling for inter-operator effects, there were highly significant differences in prevalence of GA among diagnostic categories. Category 3 differed (p<0.001) from all other diagnostic groups. Controlling for differences in diagnosis type, there were highly significant differences among anesthesiologist-surgeon dyads (p<0.001): Adjusted for diagnostic category, dyad-specific prob(GA) varied from 0.165 (95% CI 0.064, 0.365) (based on 16 cases), to 0.012 (95% CI 0.004, 0.034) (based on 183 cases). Variability in the percentage of GA was similar for 17 anesthesiologists who performed 95% of CD, max = 8.3% (168 CD) and min = 1.5% (1121 CD) and 15 surgeons who performed 96% of CD, max = 8.5% (106 CD) and min = 2.3% (43CD). Although the overall rate of GA for CD was within national standards (4.9%) in the highest five dyads it was 3-6 times greater (16%-33%).

Conclusions: The Interaction between the obstetrician and anesthesiologist is a significant factor that affects the type of anesthesia for CD, even when controlling for differences in diagnosis type. Obstetricians and anesthesiologists who belong to dyads with the highest percentage of GA should be advised about the risks vs. benefits of GA and RA, and be potentially reassigned to work with colleagues who have a lower rate of GA. We were not able to evaluate the effect of the race on GA/RA ratio, however our rate of GA for CD for African-American patients is much less than what’s reported in literature3.

References:1.Bucklin B et al. Anesthesiology. 2005;103:645-653. 2.Johnson D et al. CJA. 2002;49:954-957. 3.Butwick A et al. Anesth Analg. 2016;122:472-479.

SOAP 2017