Management of Complicated Appendicitis During Pregnancy in the US

IMPORTANCE Data are sparse regarding the optimal treatment for complicated appendicitis during pregnancy. OBJECTIVE To compare nonoperative and operative management in complicated appendicitis during pregnancy. DESIGN, SETTING, AND PARTICIPANTS This cohort study was conducted using National Inpatient SampledatafrombetweenJanuary2003andSeptember2015.Thisdatabaseapproximatesa20% stratified sample of US inpatient hospital discharges. Included individuals were pregnant women discharged with the diagnosis of complicated appendicitis. Data were analyzed from February 2020 through February 2022.


Introduction
Acute appendicitis is one of the most common nonobstetrical emergencies in pregnant women.
2][3] By the same token, more than 1 in 20 women of childbearing age who present with acute appendicitis are pregnant. 410][11] Unlike uncomplicated appendicitis, complicated appendicitis (eg, perforation with peritonitis or abscess) among pregnant women does not have a clear optimal management strategy. 12,13For complicated appendicitis within the general population, immediate operation and nonoperative management, including antibiotics and percutaneous drainage for appendiceal abscess, have been found to be associated with effective outcomes. 14,15However, when nonoperative management failed, the need for open surgery and bowel resection portending to increased morbidity was increased. 16Current guidelines for the management of complicated appendicitis in the general population recommend laparoscopic appendectomy given that data suggest the feasibility and safety of laparoscopic appendectomy, with shorter hospital stays and decreased morbidity and mortality compared with nonoperative intervention. 6,17,18Given the scarcity of literature comparing management strategies for complicated appendicitis during pregnancy, the aim of this study was to compare clinical outcomes between nonoperative and operative management of complicated appendicitis in pregnant women.We hypothesized that immediate operation among pregnant women who presented with complicated appendicitis would be associated with decreased morbidity and hospital length of stay (LOS).

Methods
This cohort study was approved and a waiver of informed consent granted owing to the use of deidentified data by the institutional review board at the University of Southern California.The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was followed to summarize the performance of the observational study.

Data Source and Patient Eligibility
This is a retrospective cohort study using Healthcare Utilization Project National Inpatient Sample (NIS) data from between January 2003 and September 2015.The NIS database is deidentified and approximates a 20% stratified sample of US inpatient hospital discharges. 19We queried the database for pregnant women with complicated appendicitis during the study period using diagnostic codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).
Complicated appendicitis included ICD-9-CM codes 540.0 (acute appendicitis with generalized peritonitis) and 540.1 (acute appendicitis with peritoneal abscess) (Table 1). 20,21The study end date of September 2015 was chosen because ICD-9-CM was retired at this time in favor of the International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM).Pregnant patients were identified using the neonatal and/or maternal (NEOMAT) variable built into the NIS database, which identifies patients who are discharged with neonatal and maternal diagnoses. 22A NEOMAT code of 1 includes only patients discharged with maternal codes and excludes neonatal codes.Patients who had pregnancies with complications, such as ectopic pregnancy and hydatidiform mole, were excluded using associated ICD-9-CM diagnosis codes (Table 1).

Baseline Demographic and Outcome Variables
Baseline patient and hospital characteristics were abstracted from NIS.These included age, hospital location, hospital region, race and ethnicity, payer, income quartile, and discharge year.Race and ethnicity categories reflect options provided for Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project (HCUP) coding of data elements, which include Asian or Pacific Islander, Black, Hispanic, Native American, White, and other.The source of race and ethnicity classification was the HCUP State Inpatient Databases disparities analysis file, which is collected by self-report on admission to the hospital using fixed categories.Race and ethnicity were assessed in this study to investigate the association between race and ethnicity and management of complicated appendicitis in pregnant patients.Using coding algorithms validated for defining comorbidities in ICD-9-CM administrative data, 23 a Charlson Comorbidity Index (CCI) score was identified for each patient.Clinical outcome variables included maternal death, preterm delivery, preterm labor, abortion, antepartum hemorrhage, premature rupture of membranes, amniotic infection, systemic inflammatory response syndrome, sepsis, severe sepsis, and pneumonia and reflect outcomes that occurred at any point during hospitalization (Table 1).Owing to the inability of ICD-9-CM codes to stratify patients by trimester, a composite perinatal outcome of preterm labor or delivery and abortion was used given that these perinatal outcomes are likely to be distributed differently across trimesters.Other outcome variables included hospital LOS and total hospital charges.The latter outcome was adjusted for inflation using the Consumer Price Index measured by the US Bureau of Labor Statistics. 24

Statistical Analysis
Using ICD-9-CM procedure codes, we divided study patients into 3 groups: those with successful nonoperative management, failed nonoperative management, and immediate operation for complicated appendicitis.Successful nonoperative management was defined as no appendectomy during the hospital stay.Failed nonoperative management was defined as a trial of at least 1 day of nonoperative management followed by operative intervention (ie, laparoscopic or open appendectomy).Patients with unknown operation timing were excluded from the study.Patient baseline characteristics were compared across groups using univariate analysis.The χ 2 test of independence was used to examine associations between treatment strategy and dichotomous outcome variables.The Kruskal-Wallis test was used to examine associations between treatment strategy and continuous outcome variables.
Multivariate regression analysis was used to compare clinical and economic outcomes across groups after adjusting for baseline characteristics, including patient characteristics (ie, age, CCI score, and race and ethnicity), economic characteristics (ie, payer and income quartile), and hospital characteristics (ie, region, location, and teaching status).Multivariate logistic regression analysis was used to examine the association between treatment strategy and dichotomous outcome variables.
Effect size was expressed with an adjusted odds ratio (OR) and corresponding 95% CI.Multivariate linear regression analysis was used for continuous outcome variables.To satisfy statistical assumptions necessary for linear regression, base 10 logarithmic transformations were required for hospital LOS and total charges.
Subgroup analysis was then performed on patients who underwent operative intervention.
Multivariate regression analyses were performed to examine the association of delay in performing operation with maternal and perinatal outcomes after adjusting for the baseline characteristics previously listed.Effect size was expressed with an adjusted OR and corresponding 95% CI.Delay was defined as the number of days between hospital admission and operative intervention.
Weighted values for national estimates provided by NIS were used per the program's recommendation.The complete case analysis (CCA) method was used for handling missing data; this has been shown to be an effective method for controlling relative bias when performing analysis of the NIS database. 25   In a subgroup analysis of patients undergoing an operation and examining the association of delay in operative intervention with maternal and perinatal outcomes, multivariate logistic regression analysis found that delays to surgery were uniformly associated with higher odds of complications (Table 4).Each day in delay to surgery was associated with an increase in odds of preterm delivery, preterm labor, or abortion by 23% (OR, 1.23; 95% CI, 1.18-1.29;P < .001).Antepartum hemorrhage (OR, 1.29; 95% CI, 1.16-1.42;P < .001)and premature rupture of membranes (OR, 1.45; 95% CI, 1.26-1.67;P < .001)also had higher odds per day in delay.Each day in delay to operation was also associated with higher odds of maternal infectious complications, including amniotic infection (OR, 1.38; 95% CI, 1.24-1.53;P < .001),sepsis (OR, 1.13; 95% CI, 1.06-1.22;P = .001),and pneumonia (OR, 1.15; 95% CI, 1.04-1.27;P = .005)(Table 4).

Discussion
This cohort study found that 56% of pregnant women with complicated appendicitis underwent an immediate operative intervention.This is lower than the recently published 85% appendectomy rate for all pregnant women with appendicitis, combining uncomplicated and complicated appendicitis from 28 hospitals in the US. 26 This is likely associated with well-defined recommendations supporting immediate surgery for uncomplicated appendicitis; however, no definitive guidelines have been established on how best to manage complicated appendicitis among pregnant patients.
Therefore, it is critical to understand the association of operative and nonoperative management of acute appendicitis in this special population with maternal and perinatal outcomes.To our knowledge, this is the largest study using a nationwide database to find significant benefits in outcomes associated with immediate operation during pregnancy in patients with complicated appendicitis.
We found that immediate operation was associated with lower odds of maternal infectious complications, including amniotic infection, pneumonia, and sepsis, compared with successful and unsuccessful nonoperative management, with no association with odds of preterm delivery, preterm labor, or abortion or with antepartum hemorrhage.It is also important to note that a trial of nonoperative management is not always successful.In this study, 74% of pregnant women who trialed nonoperative management of complicated appendicitis failed and subsequently underwent operative intervention during the same hospital stay.In this group of women, maternal and perinatal outcomes were uniformly worse compared with the immediate operation group.Furthermore, each day in delay to surgery was associated with higher odds of complication for every maternal and perinatal outcome assessed in this study.Additionally, as hospital cost containment and LOS play a greater role in decision-making, it is important to note that immediate operation was associated with shorter hospital stays and decreased hospital costs.
Existing literature that evaluates appendectomy during pregnancy tends to focus on uncomplicated appendicitis or does not define severity of presentation. 2,4,8,9,27National guidelines like those published by the Society of American Gastrointestinal and Endoscopic Surgeons strongly support appendectomy in acute uncomplicated appendicitis in pregnant women, going so far as to state, "There is no role for nonoperative management of uncomplicated acute appendicitis." 11In Pregnant women with complicated appendicitis are unlike pregnant women with uncomplicated appendicitis given that complications like perforation are clearly associated with increased maternal and fetal morbidity; however, it is unclear how these elevated stakes should affect treatment decisions. 12Studies 15,28 in the general population have found that complicated appendicitis is also associated with increased morbidity compared with uncomplicated appendicitis, with management historically focused on antibiotics and percutaneous drainage, with or without interval appendectomy.However, updated consensus guidelines for the management of complicated appendicitis in the general population have suggested the feasibility and safety of laparoscopic appendectomy, with an association with shorter hospital stays and decreased morbidity and mortality compared with nonoperative intervention. 6,15,17,18In their retrospective cohort study in a general population presenting with complicated appendicitis, excluding pregnant patients, Nimmagadda et al 29 found a similar difference in outcomes between immediate operation and successful nonoperative intervention as in our study, with immediate operation associated with a shortened hospital LOS.However, in their population, the failure rate for nonoperative management was 13.9% compared with the greater than 70% failure rate observed in our study, with both studies finding increased morbidity and LOS in those patients who failed a trial of nonoperative management of acute appendicitis.These results suggest that our study findings may help define the preferred management strategy in complicated appendicitis during pregnancy to be immediate operation.

Limitations
There are several limitations to this study.The NIS is a retrospective administrative database that includes patients based on discharge diagnoses, leading to possible misclassification bias as the diagnoses evaluated in this study would be subject to facility-level definitions.Similarly, the NIS does not maintain longitudinal follow-up data for its patients; therefore, it fails to capture complications manifesting on readmission encounters, which can yield unmeasured confounding.Given that successful nonoperative management in this study was defined as not having undergone an operation during the hospital stay being examined, the number of readmissions for these patients was not quantified.Previous literature suggests that the recurrence rate after nonoperative management of appendicitis and subsequent operation can be as high as 30%. 30Our study was unable to include clinical and economic outcomes associated with these recurrences.Additionally, this study was unable to capture clinical data to delineate why patients may have failed nonoperative treatment.Prior published data suggests reasons for nonoperative failure may include persistent pain and signs of systemic infection, including tachycardia and fever. 29Additional considerations must be made for the fetus in pregnant patients, and therefore a prospective study to understand nonoperative failure is warranted.Interestingly, among patients in our study who failed nonoperative treatment and required operative intervention, an increased proportion underwent open appendectomy compared with laparoscopy.There may be many reasons for a surgeon to choose an open surgical approach over laparoscopy that we were unable to define in this study; this warrants further investigation.Another limitation of this study is the inability of ICD-9-CM codes to reflect trimesters in pregnancy.Operative and nonoperative risks are likely to differ across trimesters, although current literature suggests that operative intervention is safe in any trimester. 7,27,31ditionally, our study does not capture the outcomes associated with negative appendectomy during pregnancy, which has been associated with increased maternal morbidity, preterm labor, and fetal loss. 10,32However, our data used discharge diagnosis codes and so was unlikely to include patients with a negative appendectomy given that the diagnosis was eventually confirmed prior to discharge.

Table 2 .
Patient and Institutional Characteristics b Suppressed per the Healthcare Cost and Utilization Project requirement.c Race and ethnicity categories reflect options provided for Healthcare Cost and Utilization Project coding of data elements.