Abortion is one of the most frequently performed surgical procedures in the United States, with approximately 1.06 million operations performed in 2011.1 Surgical abortion is an extremely safe procedure with a low risk of morbidity and a mortality rate of 0.7 per 100,000 procedures.2 Although mortality is rare, national epidemiologic data from 1988 to 1997 and 1998 to 2010 have demonstrated that anesthesia complications account for a significant proportion of abortion-related mortality.2,3 The risk factors for and underlying etiologies of anesthesia-related deaths, however, are poorly characterized in the literature. Furthermore, a paucity of data has been published on anesthesia-related morbidity or risk factors for anesthesia-related complications among women undergoing surgical abortion. Therefore, data informing an evidence-based approach to anesthesia care for women undergoing surgical abortion are limited.
The overwhelming majority of surgical abortions in the United States are performed in outpatient clinics, with only 4% occurring in hospitals in 2008.4 Empirically, there is a discrepancy in the approach to anesthesia between hospital settings, where either neuraxial anesthesia or general anesthesia with tracheal intubation often is performed routinely, and outpatient abortion facilities, where local anesthesia or IV sedation is the predominant approach and tracheal intubation is rarely performed.5,6 Underlying this discrepancy is the paucity of published data on the risks of anesthesia for pregnant women undergoing surgical abortion. Contrary to historical teaching, recent data have suggested that the risk of aspiration in pregnant women may not differ from the nonpregnant general population. Studies by Carp et al.7 and Wong et al.8,9 have demonstrated that gastric emptying does not differ in obese and non-obese term gravidas compared with nonpregnant controls. The Society for Obstetric Anesthesia and Perinatology Serious Complication Repository (SCORE) Project showed zero aspiration events in >5000 general anesthetics for cesarean delivery, a rate similar to the background risk in the population undergoing general anesthesia.10 The majority of the data on pregnant women, however, occurs at the time of delivery and therefore may not apply to surgical procedures in the first 2 trimesters of pregnancy.
Although limited, available evidence does not support an increased risk of aspiration with the use of IV sedation without tracheal intubation for surgical abortion in the first and second trimesters. Wilson et al.11 reported a complication rate of 0.3% among 1433 women receiving IV fentanyl and midazolam for surgical abortion up to 18 weeks’ gestation. None of the documented complications were anesthesia-related. The largest and only other published study in the literature is a retrospective series of >62,000 outpatient abortions performed in patient under deep sedation without intubation up to 24 0/7 weeks’ gestation in which no aspiration events occurred.12 The generalizability of this study, however, is limited by the use of a single anesthetic regimen (propofol with or without fentanyl) and the exclusion of women with class III obesity (body mass index [BMI], > 40 kg/m2). Notably, obese women have an increased risk of airway complications and have been demonstrated to have longer operative times, and therefore anesthesia exposure, during second-trimester pregnancy termination.13,14 Furthermore, obese patients undergoing sedation have changes in their respiratory physiology, including reduced chest wall compliance, elevated pulmonary artery systolic pressure, and increased work of breathing, all of which may impact their tolerance of sedation.15 In the background of the US obesity epidemic, an evidence-based approach to anesthesia for surgical abortion must include obese women of reproductive age. Therefore, the primary objective of this study was to assess the risk of perioperative anesthesia-related complications in a large cohort of obese and nonobese women undergoing outpatient surgical abortion under IV sedation without tracheal intubation.
We performed a retrospective cohort study of outpatient surgical abortions at a freestanding abortion clinic in Cleveland, Ohio, from January 1, 2012, through December 31, 2013. The licensed center, which serves all of Ohio and the surrounding regions, performs approximately 5000 procedures annually. Surgical abortions were performed through 22 6/7 weeks’ gestation. First-trimester procedures were performed by suction dilation and curettage, and second-trimester procedures were performed by dilation and evacuation. Anesthesia options offered at the facility include local anesthesia (paracervical block), oral analgesia and sedation (ibuprofen, hydrocodone, diazepam), and IV analgesia and sedation (fentanyl, midazolam, propofol) without tracheal intubation. The choice of anesthesia (local, oral sedation, IV sedation) was made by individual patients; however, by institutional protocol, all women undergoing dilation and evacuation at or beyond 17 weeks of gestation received IV sedation. The primary cohort for analysis in this study included all women undergoing surgical abortion who received IV sedation without intubation. This study was approved by the Case Western Reserve University IRB (IRB no. 2014-759), and the requirement for patient consent was waived.
Patients were screened routinely for medical and obstetric comorbidities that would preclude outpatient surgical abortion (such as severe hypertension, pharmacologic anticoagulation, and placenta accreta) and, if present, were referred to a tertiary center for care. All abortions were performed by board-certified obstetrician/gynecologists or by resident physicians under the supervision of the attending obstetrician/gynecologist. IV sedation was administered by registered nurses and certified registered nurse anesthetists (CRNAs). Propofol was only administered by CRNAs. Given that Ohio does not allow independent CRNA practice, all medical supervision for IV sedation was performed by the attending obstetrician/gynecologist.
The regimen for IV sedation included either IV fentanyl and midazolam or IV propofol with or without fentanyl or midazolam. In rare circumstances in which drugs were temporarily unavailable or in short supply, an alternative agent was administered (such as methohexital or meperidine). All women receiving IV sedation also received a paracervical block consisting of 20 mL 1% lidocaine with or without vasopressin (2 units of vasopressin for gestational age ≤12 weeks and 4 units of vasopressin for gestational age >12 weeks). The dose of lidocaine and use of vasopressin varied by the attending obstetrician/gynecologist. Exclusion criteria for IV sedation included alcohol or illicit drug use <24 hours before the procedure. Patients receiving IV sedation without propofol were instructed to avoid intake of solid food 8 hours before the procedure and remain nil per os (NPO) for 4 hours before the procedure. Patients receiving propofol were advised to be NPO starting at midnight the night before the procedure.
For all patients receiving IV sedation, intraoperative anesthesia administration and monitoring followed the American Society of Anesthesiologist’s Practice Guidelines for Sedation and Analgesia by Non-Anesthesiologists and is summarized as follows.16 A complete vital sign set (arterial blood pressure, heart rate, oxygen saturation, respiratory rate, and temperature) was recorded before the start of all procedures. Intraoperatively, all patients received continuous supplemental oxygen by nasal cannula, continuous pulse oximetry monitoring, electrocardiography, serial arterial blood pressure evaluation (every 5 minutes), and serial assessment of level of consciousness and response to commands/stimuli. Capnography was used at the discretion of the CRNA. Serial recording of vital sign data was performed by the dedicated nurse or CRNA every 5 minutes if the procedure had not concluded already. The initial dosing of agents for IV sedation varied by nurse and CRNA and was based on patient BMI and anticipated procedure duration. Intraoperative dose was titrated to physiologic effect with the target depth of sedation being purposeful response after repeated or painful stimuli (“deep sedation/analgesia”).16 Jaw thrust was used at the discretion of the nurse or CRNA; however, it generally was applied routinely throughout procedures. The threshold for naloxone use varied by provider; however, generally it was administered for persistent respiratory depression as evidenced by bradypnea, hypoxemia, or diminished respiratory effort. Ohio law requires outpatient abortion clinics to be licensed ambulatory surgical centers, and therefore, standard emergency equipment and resources (a stocked and maintained resuscitation cart) are readily available and providers trained in advanced cardiac life support are always present at the facility.
Postoperatively, patients were monitored in the postanesthesia care unit (PACU) by a team of providers that included medical assistants, registered nurses, and a certified nurse midwife with a staff to patient ratio of approximately 1:3. Vital signs (arterial blood pressure, heart rate, respiratory rate, and spot pulse oximetry) were monitored and recorded every 15 minutes until discharge, which occurred approximately 60 minutes after the procedure when patients met all discharge criteria. Prespecified discharge criteria included the ability to ambulate to and use the bathroom unassisted, tolerate oral intake, have appropriate pain control and postoperative bleeding (pad checks performed by PACU staff), and demonstrate stable vital signs. The PACU is on the same floor directly adjacent to the procedure rooms; therefore, all patient monitors, medications, and the stocked resuscitation cart are available immediately if needed. Physicians in the clinic are available for bedside assessment of patients in the PACU if clinically indicated but do not routinely assess all patients before discharge. All patients receiving sedation are required to have transportation and in-person support for the first 24 hours postoperatively.
The clinic maintains a database containing pertinent demographic, clinical, obstetric, and procedure variable fields, which served as the primary data source for this study. Procedure-related complications are distinctly coded in the database (such as hemorrhage, infection, or perforation). However, anesthesia-related complications are not coded as discrete fields in the database. Therefore, data confirmation and additional data abstraction, including all primary outcome data, were performed by one member of the research team. In addition, the clinic Quality Assurance database, which contains detailed data on all adverse events, including surgical and anesthesia-related complications and hospital transfers, was queried to ensure that all potential patients with complications were identified. Lastly, all complications were verified against the internal Quality Assurance database.
The primary outcome for this study was the rate of perioperative anesthesia complications, defined as tracheal intubation, pulmonary aspiration, or hospital transfer for an anesthesia-related indication. Anesthesia-related adverse events included hypoxemia (persistent oxygen saturation < 95% on room air) or allergic reaction and also were included in the primary outcome measure.
The study by Dean et al.12 demonstrated an aspiration risk of at most 1 in 20,708 anesthetics for surgical abortion. Given the broader definition of anesthesia-related complications, a more diverse range of drug regimens and a greater-risk patient population compared with the Dean et al. study, we chose a 2-year time frame with an anticipated 10,000 total abortions, 0 to 5 primary outcome events, and 10 to 30 secondary outcome events. In addition, clinical practice during the chosen study epoch was highly consistent (stable group of providers) and complete electronic data were available for analysis.
The use of opioid reversal (naloxone) was assessed as a secondary outcome measure. The association of obesity on primary and secondary outcome measures was also assessed. BMI calculations were based on recorded height and weight values obtained on the day of the procedure. Obesity was defined according to World Health Organization criteria as follows: class 1 obesity (BMI, 30.0–34.9 kg/m2), class 2 obesity (BMI, 35.0–39.9 kg/m2), and class 3 obesity (BMI ≥ 40.0 kg/m2). In addition, the group of women with “super obesity,” defined as BMI ≥ 50 kg/m2, was assessed separately. Univariate analysis was performed with the χ2 and Fisher exact tests when appropriate. Multivariate logistic regression was performed for the secondary outcome measure, naloxone use, with adjustment for confounding factors, which were selected based on biologic plausibility (obesity) and the results of univariate analysis (included in regression if P < 0.10). Testing for interactions among covariates was performed to inform the final multivariable model. Statistical significance was determined based on a 2-tailed P value of <0.05. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated, and analysis was performed with Stata 13.1 (College Station, TX).
During the study period, 9348 abortions were provided. Our primary study cohort consisted of the 5579 (59.7%) procedures that were performed under IV sedation without tracheal intubation. Demographic, obstetric, and procedure characteristics of the study population are depicted in Table 1. As demonstrated in Table 2, the majority of women (94.7%) received a combination of fentanyl with either midazolam or propofol, or both. The median and range of dosing of fentanyl, midazolam, and propofol in the study cohort were 200 μg (100–400 μg), 4 mg (1–10 mg), and 100 mg (20–470 mg), respectively. The distributions of fentanyl, midazolam, and propofol by gestational age and BMI are depicted as box-and-whisker plots (Supplemental Digital Content, http://links.lww.com/AA/B420).
No patient in the study cohort experienced the primary outcome measures of perioperative pulmonary complication or anesthesia-related adverse event. Therefore, there were no cases of intubation, pulmonary aspiration, or hospital transfer for an anesthesia-related indication. Based on the upper 95% CI for the sample size, the maximal risk of an anesthesia-related complication is 1 in 1860 procedures.17 Data on the secondary outcome measure, the use of naloxone, are presented in Table 3. Naloxone was administered to 13 patients in the cohort (0.2%). The most common indications for naloxone administration were hypoxemia (median oxygen saturation 69%, range 42%–100%) and diminished respiratory effort. The median dose of naloxone was 0.4 mg (range, 0.2–1.0 mg). Naloxone administration was not more frequent among obese compared with nonobese patients (0.14% vs 0.27%; 95% CI of OR, 0.12–2.36; P = 0.54), procedures performed in the second (≥13 weeks) compared with first trimester (0.41% vs 0.15%; 95% CI of OR, 0.89–7.91; P = 0.08), or procedures performed ≥17 weeks compared with <17 weeks (0.47% vs 0.19%; 95% CI of OR, 0.76–8.06; P = 0.12). In addition, the use of naloxone was not associated with patient age, history of asthma, history of opioid use, procedure time, or increasing severity/class of obesity. The use of naloxone was assessed for doses of anesthetic agents that exceeded the median for the study population (200 μg for fentanyl and 4 mg for midazolam). Naloxone administration was associated with the use of fentanyl at doses >200 μg (0.82% vs 0.13%; 95% CI of OR, 2.07–18.40; P = 0.002) but was not associated with the use of midazolam at doses >4 mg (0.80% vs 0.22%; 95% CI of OR, 0.99–13.13; P = 0.07). Importantly, any negative association between clinical, procedural, and anesthesia characteristics and naloxone use should be interpreted with caution given the study sample size.
Multivariate regression was performed to further characterize the relationship between naloxone use and the use of fentanyl at doses >200 μg, controlling for confounding factors. Obesity (BMI ≥ 30 kg/m2), second-trimester procedure (≥13 weeks’ gestation), and use of midazolam at a dose >4 mg were explored as potential covariates based on biologic plausibility and the results of the univariate analysis (P < 0.1). An evaluation was performed to test for interactions between fentanyl dose >200 μg and other covariates. An interaction was present between fentanyl dose and procedure trimester but not the other covariates (Table 4).
The interaction between fentanyl dose and procedure trimester was explored further in univariate analysis. Tables 5 and 6 demonstrate the use of naloxone by procedure trimester and fentanyl dose with univariate relative risks and 95% CIs. Expressed as incident rate ratios (IRRs), the risk of naloxone use was associated with fentanyl dose >200 μg in the first trimester (IRR, 9.02; 95% CI, 3.73–21.80) but not in the second trimester (IRR, 0.92; 95% CI, 0.23–3.70).
During the 24-month study period, 4 patients were transferred to the hospital for suction curettage-related or dilation and evacuation–related adverse events. The indication for all the hospital transfers was hemorrhage. No patients were transferred for anesthesia-related adverse events.
In this large cohort study, we found that IV sedation without intubation was not associated with a significant risk of perioperative anesthesia complications, including the need for tracheal intubation, pulmonary aspiration, anesthesia-related adverse events, or hospital transfers. In addition, the use of opioid reversal, a surrogate marker of “near-miss” morbidity, was not greater among traditional high-risk subgroups such as obese women, second-trimester procedures, or procedures at a gestational age of 17 weeks or greater. Given the low rate of complications, however, our cohort is too small to make definitive conclusions regarding the overall safety of IV sedation without tracheal intubation or the relative safety of IV sedation with and without tracheal intubation. Our findings support the consideration of IV sedation without tracheal intubation for obese and non-obese women undergoing first- and second-trimester surgical abortion. This study also supports current abortion practice, in which the overwhelming majority of procedures are performed in the outpatient setting without tracheal intubation, including during the midtrimester.5,6
This study has many notable strengths, including the large sample size and the verification of detailed clinical data from patient medical records and a Quality Assurance database, which limits the possibility of ascertainment bias. In addition, the diversity in patient characteristics and anesthesia regimens represents one of the primary contributions of our study to the literature. The largest abortion-specific publication on the anesthesia risks of surgical abortion reported no perioperative pulmonary aspiration events in >62,000 procedures12; however, the generalizability of that study is limited by the use of a single anesthetic regimen (propofol with or without fentanyl) and the exclusion of women with class III obesity. Our study included women receiving various anesthetic agents (opioids, benzodiazepines, and propofol) and anesthetic doses, as well as women with a BMI >30, 40, or even 50 kg/m2. Furthermore, no increased risk of anesthesia-related adverse events was evident in our cohort of women undergoing procedures at or beyond 17 weeks’ gestation, which historically has been considered a greater-risk gestational age range. Therefore, our study should be generalizable to a broader population of US women undergoing surgical abortion.
Our study is not without limitations, including the retrospective nature of the study design and data. Although the clinic has a robust system for reporting and tracking adverse events, the possibility of an anesthesia-related adverse event occurring after the patient was discharged (and therefore remaining unreported) cannot be excluded; however, given that all anesthetic medications administered for procedures have short half-lives and that patients were monitored routinely in recovery for approximately 60 minutes and not discharged unless vital signs remained stable over serial measurements, the risk of an unrecognized anesthesia-related complication is unlikely. The primary limitation of our study stems from the rarity of anesthesia-related adverse events, which significantly limits the power of our study to detect clinically meaningful differences in subgroups. Given that the upper limits of the 95% CIs for naloxone use indicate as much as a 2-fold increased risk with obesity, an 8-fold increased risk with procedures at or beyond 17 weeks, and a 13-fold increased risk with the use of larger than median doses of midazolam, larger studies are necessary to better characterize the anesthesia risks for surgical abortion under IV sedation. Furthermore, the association of fentanyl dose >200 μg with naloxone use among first-trimester procedures requires exploration in larger studies. Our sample size limited the ability to generate an appropriate multivariable model that controlled for confounding variables and the interaction between fentanyl dose and trimester. Physiologically, given that only a small proportion of blood volume expansion in pregnancy has occurred by the end of the first trimester,18 we speculate that equivalent doses of fentanyl may have greater potential for toxicity in the first compared with the second trimester.
Our study, along with data by Dean et al.,12 demonstrates rates of anesthesia-related complications with surgical abortion similar to that of the general obstetric population during childbirth.10 Ideally, the risk of anesthesia-related complications for surgical abortion should be similar to other equivalent-risk procedures in a population of reproductive age women. Given the political nature of abortion, however, any complication may be treated with a significantly elevated degree of scrutiny. Therefore, supposing a minimal acceptable risk for anesthesia-related complications of 1 in 30,000 (which translates into approximately 37 complications in 1.1 million procedures annually), a study of >90,000 patients would be necessary to satisfy the upper limit of the 95% CI.1,17 With 96% of abortions occurring at clinics outside the hospital setting, and given that detailed data apart from mortality surveillance by the Centers for Disease Control and Prevention are not captured centrally, obtaining these data, particularly for second-trimester procedures, would be exceptionally challenging.
Last, our study may not be generalizable to patients with different clinical characteristics than our study population, including medical comorbidities, known underlying airway problems, patients who are not NPO, patients at a gestational age beyond 22 completed weeks, or patients undergoing emergency surgical procedures. These characteristics may increase the risk of aspiration or anesthesia-related complications.19 In addition, given the individualized dosing of anesthetic agents in our study, which were titrated to both operative need and the ability to maintain appropriate oxygenation and ventilation by facemask, our results should be constrained to similar clinical environments with comparable staffing and resources.
In conclusion, anesthesia complications among women undergoing surgical abortion with IV sedation are rare and may not be greater among obese women, second-trimester procedures, or procedures at or beyond 17 weeks’ gestation. Our findings support the consideration of IV sedation without intubation for surgical abortion across a diverse group of patients. Although very rare, anesthesia-related complications remain a leading cause of morbidity and mortality after surgical abortion.2 Therefore, we recommend larger, national studies to better characterize risk factors for anesthesia-related adverse events and inform evidence-based anesthesia practices for surgical abortion.
Name: Priyanka Gokhale, BS.
Contribution: This author helped design the study, conduct the study, and write the manuscript.
Conflicts: Priyanka Gokhale reported no conflicts of interest.
Attestation: Priyanka Gokhale has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Justin R. Lappen, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Conflicts: Justin R. Lappen reported no conflicts of interest.
Attestation: Justin R. Lappen has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Jonathan H. Waters, MD.
Contribution: This author helped analyze the data and write the manuscript.
Conflicts: Jonathan H. Waters reported no conflicts of interest.
Attestation: Jonathan H. Waters has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Lisa K. Perriera, MD, MPH.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Conflicts: Lisa K. Perriera consulted for Merck.
Attestation: Lisa K. Perriera has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
The authors thank Preterm Cleveland for supporting this research collaboration.
1. Jones RK, Jerman J. Abortion incidence and service availability in the United States, 2011. Perspect Sex Reprod Health 2014;46:3–14.
2. Zane S, Creanga AA, Berg CJ, Pazol K, Suchdev DB, Jamieson DJ, Callaghan WM. Abortion-related mortality in the United States: 1998-2010. Obstet Gynecol 2015;126:258–65.
3. Bartlett LA, Berg CJ, Shulman HB, Zane SB, Green CA, Whitehead S, Atrash HK. Risk factors for legal induced abortion-related mortality in the United States. Obstet Gynecol 2004;103:729–37.
4. Jones RK, Kooistra K. Abortion incidence and access to services in the United States, 2008. Perspect Sex Reprod Health 2011;43:41–50.
5. Lichtenberg ES, Paul M, Jones H. First trimester surgical abortion practices: a survey of National Abortion Federation members. Contraception 2001;64:345–52.
6. O’Connell K, Jones HE, Simon M, Saporta V, Paul M, Lichtenberg ESNational Abortion Federation Members. National Abortion Federation MembersFirst-trimester surgical abortion practices: a survey of National Abortion Federation members. Contraception 2009;79:385–92.
7. Carp H, Jayaram A, Stoll M. Ultrasound examination of the stomach contents of parturients. Anesth Analg 1992;74:683–7.
8. Wong CA, Loffredi M, Ganchiff JN, Zhao J, Wang Z, Avram MJ. Gastric emptying of water in term pregnancy. Anesthesiology 2002;96:1395–400.
9. Wong CA, McCarthy RJ, Fitzgerald PC, Raikoff K, Avram MJ. Gastric emptying of water in obese pregnant women at term. Anesth Analg 2007;105:751–5.
10. D’Angelo R, Smiley RM, Riley ET, Segal S. Serious complications related to obstetric anesthesia: the serious complication repository project of the Society for Obstetric Anesthesia and Perinatology. Anesthesiology 2014;120:1505–12.
11. Wilson LC, Chen BA, Creinin MD. Low-dose fentanyl and midazolam in outpatient surgical abortion up to 18 weeks of gestation. Contraception 2009;79:122–8.
12. Dean G, Jacobs AR, Goldstein RC, Gevirtz CM, Paul ME. The safety of deep sedation without intubation for abortion in the outpatient setting. J Clin Anesth 2011;23:437–42.
13. Adams JP, Murphy PG. Obesity in anaesthesia and intensive care. Br J Anaesth 2000;85:91–108.
14. Murphy LA, Thornburg LL, Glantz JC, Wasserman EC, Stanwood NL, Betstadt SJ. Complications of surgical termination of second-trimester pregnancy in obese versus nonobese women. Contraception 2012;86:402–6.
15. Porhomayon J, Papadakos P, Singh A, Nader ND. Alteration in respiratory physiology in obesity for anesthesia-critical care physician. HSR Proc Intensive Care Cardiovasc Anesth 2011;3:109–18.
16. American Society of Anesthesiologists Task Force on S, Analgesia by Non- Anesthesiologists. Practice guidelines for sedation and analgesia by non-anesthesiologists. Anesthesiology 2002;96:1004–17.
17. Hanley JA, Lippman-Hand A. If nothing goes wrong, is everything all right? Interpreting zero numerators. JAMA 1983;249:1743–5.
18. Bernstein IM, Ziegler W, Badger GJ. Plasma volume expansion in earlypregnancy. Obstet Gynecol 2001;97:669–72.
19. Warner MA, Warner ME, Weber JG. Clinical significance of pulmonary aspiration during the perioperative period. Anesthesiology 1993;78:56–62.
Supplemental Digital Content
© 2016 International Anesthesia Research Society