Silvestri, Mark T. MD; Pettker, Christian M. MD; Brousseau, E. Christine MD; Dick, Madeline A. MD; Ciarleglio, Maria M. PhD; Erekson, Elisabeth A. MD, MPH
Nonobstetric surgery is performed during the antepartum period on one to two out of every 1,000 pregnant women.1–3 Appendectomy and cholecystectomy are the two most common nonobstetric surgeries performed in pregnant patients.4
Evidence on outcomes of nonobstetric surgery in pregnancy is limited primarily to fetal and pregnancy-related complications. A systematic review accumulating 54 articles and more than 12,000 nonobstetric antenatal surgeries highlights the incidence of fetal loss, prematurity, and major birth defects reported in the literature.5 Maternal surgical outcomes after these operations are rarely reported, however, and available literature is primarily from small patient cohorts. Even fewer studies have compared postoperative outcomes in pregnant patients with those in their nonpregnant counterparts. A recent American College of Obstetricians and Gynecologists Committee Opinion acknowledges that nonobstetric surgery on the pregnant patient is an important concern for physicians who care for women and that “there are no data to allow for specific recommendations.”6 This highlights the need for further research into maternal complications of nonobstetric surgery. Improved data regarding surgical outcomes in pregnant patients would allow for better patient counseling and improved quality of care.
The objective of this study is to use the data from the American College of Surgeons (ACS) National Surgical Quality Improvement Program to estimate major postoperative morbidity after 1) appendectomy in pregnant compared with nonpregnant women; and 2) cholecystectomy in pregnant compared with nonpregnant women.
MATERIALS AND METHODS
We analyzed data from the ACS National Surgical Quality Improvement Program participant use data sets (2005–2009) to perform a retrospective cohort study. The ACS National Surgical Quality Improvement Program is a national program that collects data on preoperative patient characteristics, intraoperative variables, and postoperative complications with the primary goal of enhancing surgical quality and improving patient outcomes. Participating institutions gather data on more than 130 variables on a sample of individual surgeries. Every 8 days, the first 40 patients undergoing surgical cases at each participating institution are sampled.7 The variables collected from these patients' charts include patient demographic and preoperative characteristics, operative procedure information, and 30-day major postoperative complications. The ACS National Surgical Quality Improvement Program data are collected by chart review by surgical clinical reviewers, who are trained nurses at each hospital. In 2009, more than 400 community and academic hospitals voluntarily participated in the ACS National Surgical Quality Improvement Program. The interrater reliability of variables in this data set has been demonstrated to be 97%–98%, and data audits are regularly conducted to ensure the quality of the data.8 Additional information regarding the ACS National Surgical Quality Improvement Program database is available online (www.acsnsqip.org). The timeframe of 2005–2009 was selected because these were the only years for which data sets were available. Because this study was a secondary analysis of a deidentified data set, we received exemption status in writing from the institutional review board for the Yale University School of Medicine.
Our first target population included all patients who underwent appendectomy, according to the Physicians' Current Procedural Terminology Coding System, 4th edition (CPT-4) codes listed for their primary procedure. We included patients with the following CPT-4 codes: 44900, 44950, 44960, 44970, and 44979. Definitions of these codes are listed in Table 1. Patients were excluded for the following reasons: 1) male sex; 2) age 51 years or older; or 3) previous surgical procedure within 30 days of appendectomy; or 4) unknown pregnancy status. No cases of concomitant appendectomy with either cesarean or vaginal delivery were identified in the data set. The ACS National Surgical Quality Improvement Program data set does not collect information on the procedure type or the CPT-4 code of the surgical procedures performed within 30 days before the index procedure.
Our second target population included all patients who underwent cholecystectomy according to the following CPT-4 codes: 47480, 47562, 47563, 47564, 47570, 47600, 47605, 47610, 47612, and 47620. Definitions of these codes are listed in Table 1. Patients were excluded for 1) male sex; 2) age 51 years or older; 3) previous surgical procedure within 30 days of cholecystectomy; or 4) unknown pregnancy status. No cases of concomitant and cholecystectomy with either cesarean or vaginal delivery were identified in the data set.
Because our primary comparison was between women who were listed as pregnant at the time of surgery and their reproductive aged nonpregnant counterparts, all women aged 51 years or older were excluded from this study. The ACS National Surgical Quality Improvement Program does not include patients younger than 16 years.
We examined preoperative characteristics including age, race, and ethnicity. We also explored several preoperative risk factors, including pre-existing medical diseases and functional status. These preoperative risk factors and pre-existing medical diseases that were analyzed are listed in Table 2 for appendectomy and Table 3 for cholecystectomy. Functional status was reported as the fraction of women who were partially or totally dependent on another person for activities of daily living (bathing, feeding, dressing, toileting, and mobility). Intraoperative variables such as operative time, type of anesthesia, and surgical approach (open or laparoscopic) were also analyzed. Unfortunately, gestational age, trimester of pregnancy, and pregnancy outcomes are not recorded in the data set.
Our primary outcome was composite 30-day major morbidity. This was a composite (not exclusive) outcome of all women experiencing postoperative mortality, cardiac arrest, acute myocardial infarction, postoperative coma greater than 24 hours, cerebrovascular accident with neurologic deficit, acute renal failure, progressive renal insufficiency, deep wound surgical site infection, organ-space surgical site infection, wound dehiscence, pulmonary embolism, deep vein thrombosis, prolonged mechanical ventilation greater than 48 hours, unplanned reintubation, pneumonia, sepsis, septic shock, postoperative blood transfusion, or return to the surgical operating room. Urinary tract infections and superficial surgical site infections were not considered to be major morbidities and therefore were not included in the composite outcome. A 30-day postoperative timeframe was used for our composite outcome because the ACS National Surgical Quality Improvement Program collects data on complications up to 30 days postoperatively.
Continuous variables were compared between cohorts using the Student's t test, and categorical variables were analyzed with the chi-square test and Fisher's exact test as appropriate. To adjust for differences in preoperative characteristics between the pregnant and nonpregnant women, a logistic regression model was constructed. Variables were selected for inclusion in the model based on univariable analysis (P<.1). Variables were added to the model in a stepwise fashion using forward and backward selection (P<.05). Goodness of fit of the model was verified using the Hosmer-Lemeshow test (P<.05). Variables included in the final logistic regression model were assessed for potential interactions. Statistical analysis was performed using STATA 11.0, PASS 2008, and SAS 9.2.
We estimated the prevalence of complications after open appendectomy to be 6.1% after excluding superficial surgical site infection and urinary tract infection.9 We assumed a clinically meaningful difference in the occurrence of postoperative complications to be a difference of 3.0% or an increase in the occurrence of postoperative complications to 9.1%. We were limited by a finite sample size of 857 pregnant women and 19,712 nonpregnant reproductive-aged women available for analysis. Our analysis achieved 91% power to detect a difference of major postoperative complications of 3.0% with a significance level (α) of .05 (two-sided).
We estimated the prevalence of complications after cholecystectomy to be 3.1%.10 We assumed a clinically meaningful difference in the occurrence of postoperative complications to be a difference of 3.0% or an increase in the occurrence of postoperative complications to 6.1%. We were limited by a finite sample size of 436 pregnant women and 32,479 nonpregnant reproductive-aged women available for analysis. Our analysis achieved 88% power to detect a difference of major postoperative complications of 3.0% with a significance level (α) of .05 (two-sided).
A total of 971,455 surgical cases were available for review in the combined 2005–2009 ACS National Surgical Quality Improvement Program data set. Of these, 1,969 (0.2%) were nonobstetric procedures performed on pregnant women.
A total of 61,909 patients were identified as having undergone appendectomy based on CPT-4 coding, of whom 31,978 patients were excluded for male sex. Women were excluded for the following reasons: 1) age 51 years or older (n=8,123); 2) prior surgical procedure within 30 days (n=116); and 3) unknown pregnancy status (n=1,663). This left a total of 20,029 women included in our final analysis of appendectomy in women (Fig. 1). Of these cases, 857 (4.3%) involved pregnant women.
Results of the univariable analysis of preoperative characteristics in pregnant and nonpregnant women are presented in Table 2. Compared with nonpregnant women, pregnant women undergoing appendectomy were younger (27.3±6.1 years compared with 32.0±9.8 years, P<.001), less frequently white (66.6% compared with 71.8%, P=.001), and more frequently Hispanic (19.3% compared with 12.3%, P=.001). The two groups were similar with respect to most preoperative medical conditions; however, pregnant women had a higher incidence of preoperative systemic infection (39.7% compared with 33.6%, P<.001) and a lower incidence of diabetes mellitus (1.1% compared with 2.6%, P=.006) and hypertension (1.2% compared with 6.4%, P<.001) than nonpregnant women. Pregnant women underwent more emergency procedures (82.4% compared with 73.4%, P<.001), were more likely to have types of anesthesia other than general anesthesia (5.8% compared with 0.02%, P<.001), and more frequently had an open surgical approach compared with a laparoscopic approach (36.4% compared with 16.6%, P<.001).
Postoperative complications in pregnant and nonpregnant women after appendectomy are presented in Table 4. Pregnant and nonpregnant women had similar composite 30-day major morbidity (3.9% compared with 3.1%, P=.212). Additionally, the two groups had similar occurrences of all specific complications except pneumonia, which occurred more frequently in pregnant women (0.7% compared with 0.2%, P=.004). All cases of postoperative pneumonia were observed in women who underwent general endotracheal anesthesia.
Superficial surgical site infections (cellulitis) and urinary tract infections were not considered as a part of composite 30-day major morbidity and not significantly different between pregnant and nonpregnant women. Superficial surgical site infections occurred in 1.52% of pregnant women and 1.36% of nonpregnant women undergoing appendectomy (P=.70). Urinary tract infections occurred in 0.7% of pregnant women and 0.71% of nonpregnant women undergoing appendectomy (P=.98).
A logistic regression model was created incorporating preoperative characteristics as predictors of composite 30-day major morbidity. Pregnancy status was not associated with increased postoperative morbidity (adjusted odds ratio [OR] 1.26, 95% confidence interval [CI] 0.87–1.82) after adjusting for age per 5-year increase (adjusted OR 1.17, 95% CI 1.12–1.22), preoperative ascites (adjusted OR 2.31, 95% CI 1.42–3.76), preoperative systemic infection (adjusted OR 2.21, 95% CI 1.88–2.60), diabetes mellitus (adjusted OR1.65, 95% CI 1.13–2.41), and open procedure (adjusted OR 1.81, 95% CI 1.51–2.16). Results of the logistic regression model are listed in Table 5. We assessed our logistic regression model for interactions between variables. Interactions among age, ascites, and open procedures were not observed. An interaction between diabetes mellitus and preoperative systemic infection was observed with women who had diabetes mellitus more likely to have preoperative systemic infection.
A total of 90,696 patients were identified as undergoing cholecystectomy based on CPT-4 coding, of whom 26,779 patients were excluded for male sex. Women were excluded for the following reasons: 1) age 51 years or older (n=22,235); 2) prior surgical procedure within 30 days (n=219); and 3) unknown pregnancy status (n=8,548). This left a total of 32,915 women included in our final analysis of cholecystectomy in women (Fig. 1). Of these cases, 436 (1.3%) involved pregnant women.
Results of the univariable analysis of preoperative characteristics in pregnant and nonpregnant women are presented in Table 3. Compared with nonpregnant women, pregnant women undergoing cholecystectomy were younger (27.8±6.5 years compared with 35.8±9.6 years, P<.001) and more frequently Hispanic (22.5% compared with 14.9%, P<.001). The two groups were similar with respect to most preoperative medical conditions; however, pregnant women had a higher incidence of preoperative systemic infection (11.9% compared with 5.2%, P<.001) and a lower incidence of both diabetes mellitus (1.8% compared with 4.8%, P=.004) and hypertension (2.8% compared with 13.8%, P<.001). Pregnant women underwent more emergency procedures (21.1% compared with 8.2%, P<.001) and more frequently had an open surgical approach (9.6% compared with 4.5%, P<.001).
Postoperative complications in pregnant and nonpregnant women after cholecystectomy are presented in Table 6. Pregnant and nonpregnant women had similar composite 30-day major morbidity (1.8% compared with 1.8%, P=.954). Additionally, the two groups had similar occurrences of all specific complications. Minor complications of superficial surgical site infections and urinary tract infections were not considered as a part of the composite 30-day major morbidity and were not significantly different between pregnant and nonpregnant women. Superficial surgical site infections occurred in 1.4% of pregnant women and 1.0% of nonpregnant women undergoing cholecystectomy (P=.32). Urinary tract infections occurred in 0.9% of pregnant women and 0.6% of nonpregnant women undergoing cholecystectomy (P=.39).
We were unable to conduct logistic regression exploring the effect of preoperative characteristics in pregnant and nonpregnant women on major postoperative complications after cholecystectomy resulting from the low frequency of pregnant women with postoperative complications.11
Major postoperative complications are rare in pregnant women after both appendectomy and cholecystectomy. Rates of postoperative complications among the women in this study were similar to previously reported postoperative complication rates for appendectomy12–14 and cholecystectomy4,10,15,16 in general populations. Additionally, our analysis indicates that the occurrence of major postoperative morbidity is similar among pregnant and nonpregnant reproductive-aged women. This suggests that pregnancy alone does not significantly increase the risk of major surgical morbidity after these two procedures.
Pregnant women were significantly more likely than their nonpregnant counterparts to be diagnosed with systemic infection before both appendectomy and cholecystectomy. This may simply reflect that a pregnant woman's physiologic leukocytosis, increased heart rate, and respiratory alkalosis make her more likely to meet clinical criteria for systemic inflammatory response syndrome than a nonpregnant woman. However, it is also possible that the increase in preoperative systemic infection is the result of a delayed diagnosis of acute appendicitis or acute cholecystitis in pregnant women resulting from the challenges presented by the anatomic and physiologic changes of pregnancy17,18 combined with a reluctance to operate on pregnant women until the diagnosis is certain. Emergency procedures were significantly more common among pregnant women undergoing both appendectomy and cholecystectomy, suggesting that some degree of delayed diagnosis was present in the pregnant women studied. The higher rate of emergent cholecystectomies in the pregnant population also may be the result of attempts to manage cases expectantly without surgery and defer elective procedures until after pregnancy.
We also found that pregnant women underwent more open procedures compared with nonpregnant patients. This increase in open procedures may have been the result of a delayed diagnosis or increased disease severity. Alternatively, the open approach may have been the result of surgeons' reluctance to perform laparoscopic procedures in pregnant women. Because neither gestational age nor specific hospital and surgeon data are available from the ACS National Surgical Quality Improvement Program participant use data set, we were unable to determine if the open approach for appendectomy and cholecystectomy was selected more often for advanced gestations or by specific providers. The increased rate of open procedures in the pregnant population did not appear to increase the risk of postoperative complications.
Kuy et al also compared maternal postoperative outcomes with those in nonpregnant women. This case–control analysis of patients undergoing cholecystectomy used a population-based data set built from medical coding data, the Healthcare Cost and Utilization Project–Nationwide Inpatient Sample, to compare immediate postsurgical outcomes between 8,933 pregnant and 53,598 nonpregnant women. In multivariate analysis, pregnancy was not found to be an independent predictor for surgical complications after cholecystectomy.4 In our analysis of the ACS National Surgical Quality Improvement Program data, we similarly found no difference in major morbidity after cholecystectomy in pregnant compared with nonpregnant patients. Our study builds on that by Kuy et al in several important ways. First, our report adds data on appendectomy in pregnancy, aggregating information for the two most commonly performed nonobstetric antenatal surgeries. Second, by focusing on maternal morbidity rather than an array of outcomes, we were able to present an extensive list of individual complication rates for each procedure in addition to our primary outcome of composite morbidity. Finally, the database we used for our study was amassed from a comprehensive, standardized, and validated chart review process rather than from medical coding data. A systematic chart review process, by which the data used in this report were gathered, has been shown to be the most sensitive and accurate method for accruing perioperative complication data19,20 and avoids some of the problems encountered in analyses of administrative data sets like that by Kuy et al.21 As demonstrated by Heisler et al, postoperative complications are much more likely to be identified through a formal chart review process, like the ACS National Surgical Quality Improvement Program process, than by collecting coding data from medical billing, which substantially underestimates postoperative morbidity.19 The comprehensive nature of the ACS National Surgical Quality Improvement Program database allowed us to incorporate 19 postoperative complications involving multiple organ systems into our composite outcome measurement.
There was a nonstatistically significant increase in return to the operating room in within 30 days in pregnant compared with nonpregnant women undergoing appendectomy (1.6% compared with 1.0%, P=.07) and pregnant compared with nonpregnant women undergoing cholecystectomy (1.4% compared with 0.8%, P=.24). Unfortunately, we do not have information on the CPT-4 codes of the procedures performed during the return to the operating room; therefore, we cannot determine whether these follow-up procedures were for cesarean delivery.
The primary limitation of this study is that it is a secondary review of an existing data set and as such, it is constrained by the breadth of information available. A few relevant pieces of data—specifically gestational age, pregnancy outcome, fetal wellbeing, and hospital and surgeon data—are not available in the ACS National Surgical Quality Improvement Program database for analysis. Therefore, we were not able to stratify by gestational age or report pregnancy and neonatal outcomes. We also were unable to account for the effect of clustering of observations within centers, although others have previously demonstrated that the clustering effect from the ACS National Surgical Quality Improvement Program is minimal and does not change overall adjusted outcomes.22,23 Additionally, as a result of the fact that the ACS National Surgical Quality Improvement Program exclusively uses CPT-4 codes to identify patients, there is a possibility that some cases with incorrect coding were missed, and the sampling of only the first 40 patients in each 8-day cycle implies that not every case at each hospital is included in the data set. Participation in the ACS National Surgical Quality Improvement Program is voluntary, and many hospitals choose not to participate. Therefore, data from ACS National Surgical Quality Improvement Program cannot be considered a representative sample of the entire United States. However, current ACS National Surgical Quality Improvement Program participation includes more than 400 hospitals with a wide range of both community-based care facilities and tertiary care centers. This represents a steady increase of participation from 37 hospitals in 2005 to 230 hospitals in 2008 as the ACS National Surgical Quality Improvement Program continues to grow.22 By analyzing hundreds of pregnant women undergoing appendectomy and cholecystectomy at participating centers, we were able to make a meaningful comparison of complication rates between pregnant and nonpregnant women. Our composite outcome by its nature includes postoperative complications with a wide range of severity and long-term health implications, which limits its direct clinical relevance. Given the paucity of pregnant women undergoing appendectomy and cholecystectomy, however, and given the rarity of complications after these procedures, using a composite outcome allowed an important comparison to be made with nonpregnant women. Additionally, by presenting a composite outcome that included all of the major postoperative complications included in the ACS National Surgical Quality Improvement Program, we were able to present an assessment of the overall surgical morbidity of these procedures in the antenatal period.
In conclusion, we found that maternal postoperative complications are similar in pregnant and nonpregnant women after appendectomy and cholecystectomy. This study provides important information on the maternal safety of both appendectomy and cholecystectomy during pregnancy for both obstetricians and consulting surgeons caring for these women.
1. Mazze RI, Kallen B. Reproductive outcome after anesthesia and operation during pregnancy: a registry study of 5405 cases. Am J Obstet Gynecol 1989;161:1178–85.
2. Jenkins TM, Mackey SF, Benzoni EM, Tolosa JE, Sciscione AC. Non-obstetric surgery during gestation: risk factors for lower birthweight. Aust N Z J Obstet Gynaecol 2003;43:27–31.
3. Guidelines for diagnosis, treatment, and use of laparoscopy for surgical problems during pregnancy. 2011. Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) web site. Available at: www.sages.org/publication/id/23/
. Retrieved February 8, 2011.
4. Kuy S, Roman SA, Desai R, Sosa JA. Outcomes following cholecystectomy in pregnant and nonpregnant women. Surgery 2009;146:358–66.
5. Cohen-Kerem R, Railton C, Oren D, Lishner M, Koren G. Pregnancy outcome following non-obstetric surgical intervention. Am J Surg 2005;190:467–73.
6. Nonobstetric surgery during pregnancy. ACOG Committee Opinion No. 474. American College of Obstetricians and Gynecologists. Obstet Gynecol 2011;117:420–1.
7. Dhungel B, Diggs BS, Hunter JG, Sheppard BC, Vetto JT, Dolan JP. Patient and peri-operative predictors of morbidity and mortality after esophagectomy: American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP), 2005–2008. J Gastrointest Surg 2010;14:1492–501.
8. Shiloach M, Frencher SK Jr, Steeger JE, Rowell KS, Bartzokis K, Tomeh MG, et al.. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16.
9. Page AG, Pollock JD, Perez S, Davis SS, Lin E, Sweeney JF. Laparoscopic versus open appendectomy: an analysis of outcomes in 17,199 patients using ACS/NSQIP. J Gastrointest Surg 2010;14:1955–62.
10. Ingraham AM, Cohen ME, Ko CY, Hall BL. A current profile and assessment of north american cholecystectomy: results from the american college of surgeons national surgical quality improvement program. J Am Coll Surg 2010;211:176–86.
11. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 1996;49:1373–9.
12. Guller U, Hervey S, Purves H, Muhlbaier LH, Peterson ED, Eubanks S, et al.. Laparoscopic versus open appendectomy: outcomes comparison based on a large administrative database. Ann Surg 2004;239:43–52.
13. Ingraham AM, Cohen ME, Bilimoria KY, Pritts TA, Ko CY, Esposito TJ. Comparison of outcomes after laparoscopic versus open appendectomy for acute appendicitis at 222 ACS NSQIP hospitals. Surgery 2010;148:625–35.
14. Nguyen NT, Zainabadi K, Mavandadi S, Paya M, Stevens CM, Root J, et al.. Trends in utilization and outcomes of laparoscopic versus open appendectomy. Am J Surg 2004;188:813–20.
15. Carbonell AM, Lincourt AE, Kercher KW, Matthews BD, Cobb WS, Sing RF, et al.. Do patient or hospital demographics predict cholecystectomy outcomes? A nationwide study of 93,578 patients. Surg Endosc 2005;19:767–73.
16. Shea JA, Healey MJ, Berlin JA, Clarke JR, Malet PF, Staroscik RN, et al.. Mortality and complications associated with laparoscopic cholecystectomy. A meta-analysis. Ann Surg 1996;224:609–20.
17. Tracey M, Fletcher HS. Appendicitis in pregnancy. Am Surg 2000;66:555–9.
18. Augustin G, Majerovic M. Non-obstetrical acute abdomen during pregnancy. Eur J Obstet Gynecol Reprod Biol 2007;131:4–12.
19. Heisler CA, Melton LJ 3rd, Weaver AL, Gebhart JB. Determining perioperative complications associated with vaginal hysterectomy: code classification versus chart review. J Am Coll Surg 2009;209:119–22.
20. Campbell PG, Malone J, Yadla S, Chitale R, Nasser R, Maltenfort MG, et al.. Comparison of ICD-9-based, retrospective, and prospective assessments of perioperative complications: assessment of accuracy in reporting. J Neurosurg Spine 2011;14:16–22.
21. Grimes DA. Epidemiologic research using administrative databases: garbage in, garbage out. Obstet Gynecol 2010;116:1018–9.
22. Henderson WG, Daley J. Design and statistical methodology of the National Surgical Quality Improvement Program: why is it what it is? Am J Surg 2009;198:S19–27.
23. Cohen M, Dimick J, Bilimoria K, Ko C, Richards K, Hall B. Risk adjustment in the American College of Surgeons National Surgical Quality Improvement Program: a comparison of logistic versus hierarchical modeling. J Am Coll Surg 2009;209:687–93.