AUTHORS: Markley, John C., MD, PhD et al

Anesthesia & Analgesia: February 23, 2018 – Volume Publish Ahead of Print – Issue – p

BACKGROUND: General anesthesia (GA) is often selected for cesarean deliveries (CD) with placenta previa and suspected morbidly adherent placenta (MAP) due to increased risk of hemorrhage and hysterectomy. We reviewed maternal outcomes and risk factors for conversion to GA in a cohort of patients undergoing CD and hysterectomy under neuraxial anesthesia (NA).

METHODS: We performed a single-center, retrospective cohort study of parturients undergoing nonemergent CD for placenta previa with suspected MAP from 1997 to 2015. Patients were classified according to whether they received GA, NA, or intraoperative conversion from NA to GA. The primary outcome measure was postoperative acuity, defined as the need for intensive care unit admission, arterial embolization, reoperation, or ongoing transfusion with ≥3 units packed red blood cells. We additionally identified variables positively associated with intraoperative conversion from NA to GA during hysterectomy. Confounding was controlled with logistic regression models.

RESULTS: Of 129 patients undergoing nonemergent CD for placenta previa with suspected MAP, 122 (95%) received NA as the primary anesthetic. NA was selected in the majority of patients with a body mass index ≥40 kg/m2 (9 of 10, 90%), a history of ≥3 prior CDs (18 of 20, 90%), suspected placenta increta or percreta (29 of 35, 83%), and Mallampati classification ≥3 (19 of 21, 90%). Of 72 patients with NA at the time of delivery who required hysterectomy, 15 (21%) required conversion to GA intraoperatively. Converted patients had a higher rate of major packed red blood cell transfusion (60% vs 25%; P = .01), with similar rates of massive transfusion (9% vs 7%; P = 1.0). Converted patients also had a higher incidence of postoperative acuity (47% vs 4%; P < .0001), including 5 intensive care unit admissions for airway management after large-volume resuscitation. After adjusting for multiple confounders, the only independent predictors of conversion among hysterectomy patients were longer surgical duration (adjusted odds ratio 1.54, 95% CI, 1.01.2.42) and a history of ≥3 prior CDs (adjusted odds ratio, 6.45; 95% CI, 1.12.45.03).

CONCLUSIONS: NA was applied to and successfully used in the majority of patients with suspected MAP. Our findings support selective conversion to GA during hysterectomy in these patients, focusing on those with the highest levels of surgical complexity.