Materials and Methods
The Mayo Clinic Department of Obstetrics electronic database was reviewed to identify all patients undergoing a surgical procedure within 12 months of a documented pregnancy and/or delivery from January 1992 through March 2014. For study purposes, a GA of ≥23 weeks’ gestation was considered to represent the threshold of fetal viability in our institution. Each electronic medical record identified was reviewed to confirm an antepartum surgical procedure was performed, and then abstracted for maternal demographic variables, GA at time of surgery, and procedure and anesthetic types, with GA at delivery selected as the primary outcome.
Anesthetic records were reviewed in detail for all patients in whom intraoperative fetal monitoring was performed. As modifications in anesthetic technique to address potentially compromised fetal status are difficult to objectively qualify and printed fetal heart rate tracings were not retained for review, we considered the administration of vasoactive medications to represent a surrogate marker for fetal compromise unless recorded exclusively for maternal indication. Entry into the abdominal cavity was empirically considered to impart higher fetal risk due to potential for direct uterine manipulation or injury. Student t test was employed for comparison of population demographics and Fisher exact test utilized for comparison of outcomes between groups, with a P value of < .05 considered to represent statistical significance. This study was approved by the Mayo Clinic Institutional Review Board under protocol no. 06-002729.
Results
During the 23-year study interval, approximately 36,100 patients were delivered at the Mayo Clinic in Rochester, MN. A total of 121 surgical procedures were performed in 111 patients at ≥23 weeks’ gestation, for an incidence of approximately 1 procedure per 325 pregnancies. Singleton gestations comprised 108 of the 111 pregnancies, with 3 twin pregnancies included. Multiple procedures were required in 8 patients during a single pregnancy, 5 of which were performed for recurrent urologic indications. Demographics are shown in Table 1 . Surgical procedures were subdivided based on peritoneal cavity entry: all procedures requiring an abdominal incision with peritoneal entry (eg, laparoscopy or laparotomy) were classified as ‘abdominal entry,’ while those procedures not involving an abdominal incision (comprised chiefly of endoscopic gastrointestinal or genitourinary procedures) were categorized as ‘nonabdominal entry’; a single transvaginal cerclage performed at 24 0/7 weeks was included in this group. Mean GA at time of operation was 29.2 weeks (range, 23 3/7 to 36 5/7 weeks), and the majority of procedures (88/121, 73%) were completed under a general anesthetic ( Table 2 ). Consistent with standardization of our obstetric and anesthesiology practices over the last decade ( Figure ), intraoperative electronic fetal monitoring was performed in only 14 (12%) cases during the study interval: 7 abdominal, 3 urologic, 2 thoracic, and 1 each orthopedic and parathyroid procedures. Reviewing these operative notes and anesthesia records in detail, 1 patient was administered a single dose of ephedrine for hypotension following induction of general anesthesia, but there were no recorded instances during which vasoactive medications were administered specifically for fetal heart rate abnormalities. In the 5 cases of laparoscopy or laparotomy during which fetal monitoring was performed, monitoring technique was described in only 1 operative report. A single episode of fetal loss was recognized postoperatively following cardioplegia for aortic valve replacement; in this situation a patient was referred to our institution in heart failure due to valvular regurgitation and cardiomyopathy, with the pregnancy not disclosed preoperatively (a pregnancy test was not performed) and only discovered at the conclusion of surgery. Immediate postoperative ultrasound revealed an intrauterine demise at approximately 25 weeks’ GA, but as fetal viabilty was not verifed prior to surgery the timing of loss was not able to be ascertained with certainty. Subsequent induction of labor and transvaginal delivery was accomplished without incident.
Demographic variable | All patients (n = 111) | Patients undergoing abdominal-entry procedures (n = 40) | Patients undergoing nonabdominal-entry procedures (n = 71) | P value |
---|---|---|---|---|
Age, y | 27.6 | 27.3 | 27.8 | .64 (NS) |
Gravidity | 2.5 | 2.4 | 2.6 | .47 (NS) |
Parity | 1.0 | 0.9 | 1.0 | .21 (NS) |
Previous preterm deliveries, total | 13 (11.7%) | 7 (18.0%) | 6 (8.0%) | .22 (NS) |
• 1 previous | 11 | 6 | 5 | |
• ≥2 previous | 2 | 1 | 1 |
Procedure type | Mean EGA procedure, wk | Mean EGA delivery, wk | Mean procedure-to-delivery interval, wk | Incidence of preterm delivery |
---|---|---|---|---|
GI, abdominal entry, n = 32 | 28.0 (23–36) | 37.2 (24–41) | 9.0 | 12/32 (38%) |
GI, nonabdominal entry, n = 10 | 30.9 (26–38) | 35.1 (28–40) | 4.4 | 6/10 (60%) |
Gynecologic, n = 4 | 25.5 (23–29) | 36.2 (33–39) | 10.7 | 2/4 (50%) |
Orthopedic, n = 7 | 33.4 (26–37) | 39.9 (38–41) | 6.2 | 0/7 (0%) |
Renal, n = 17 | 31.0 (23–36) | 37.5 (32–41) | 6.7 | 7/17 (41%) |
Neurosurgical, n = 4 | 27.8 (25–33) | 37.1 (35–40) | 9.3 | 2/4 (50%) |
Other, n = 12 | 27.7 (24–33) | 35.1 (25–40) | 7.5 | 7/12 (58%) |
As patients were often transferred for management of an acute surgical condition and afterward returned to their primary providers elsewhere for prenatal care and delivery, perinatal outcome data were available for only 86/111 patients (78%). The mean GA at delivery for the group was 36.9 weeks, with a cesarean delivery rate (combined primary and repeat) of 37% (32/86). Thirty-five patients (35/86, 41%) ultimately delivered preterm at an average GA of 33.1 weeks, and only 9 (9/86, 10%) within 1 week of the surgical procedure.
We attempted to identify any unique predictive characteristics of the 9 patients who delivered within 1 week of their surgical procedure. Operative reports were reviewed in detail; 3 cases were complicated by peritonitis or intestinal necrosis and 1 by malignancy, but no specific intraoperative complications or hemodynamic instability were described. Reflective of the remainder of the cohort, general anesthesia was used for the majority of procedures (7/9, 78%) and intraoperative monitoring was recorded in 3 cases; none of the subsequent deliveries were performed due to nonreassuring intraoperative fetal status. Analysis of delivery within ≤7 days vs >7 days from procedure based on abdominal entry also did not show a significant association ( P = .29). Specific characteristics of these cases are shown in Table 3 .
Patient no. | EGA, wk | Surgical procedure | Indication | Anesthetic | Interval to delivery, d | Indication for delivery | Mode of delivery |
---|---|---|---|---|---|---|---|
1 | 28 1/7 | Subtotal colectomy and ileostomy | Refractory inflammatory bowel disease | General | 2 | Preterm labor | Vaginal |
2 | 32 | Intestinal resection | Ischemic bowel | General | 2 | Placental abruption, preterm labor | Vaginal |
3 | 30 1/7 | Appendectomy (open) | Acute appendicitis | General | 2 | Preterm labor | Vaginal |
4 | 31 4/7 | Appendectomy | Acute appendicitis | General | 6 | Preterm labor | Cesarean |
5 | 24 3/7 | Appendectomy | Appendiceal abscess | General | 0 | Postoperative nonreassuring fetal status | Cesarean |
6 | 27 | Colonoscopy | Suspected bowel obstruction | IV sedation | 7 | Preterm labor | Cesarean |
7 | 36 5/7 | Cystoscopy with stent placement | Obstructive nephrolithiasis | General | 7 | Elective | Cesarean |
8 | 27 5/7 | EGD | Gastric malignancy | IV sedation | 7 | Deteriorating maternal clinical status | Cesarean |
9 | 25 1/7 | Aortic valve replacement | Heart failure, cardiomyopathy | General | 2 | Intrauterine demise | Vaginal |