Clinical Gastroenterology and Hepatology 2015;-:-–-
Pregnancy Is a Risk Factor for Pancreatitis After Endoscopic Retrograde Cholangiopancreatography in a National Cohort Study Sumant Inamdar,* Tyler M. Berzin,‡ Divyesh V. Sejpal,* Douglas K. Pleskow,‡ Ram Chuttani,‡ Mandeep S. Sawhney,‡ and Arvind J. Trindade* *Department of Medicine, Division of Gastroenterology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore Long Island Jewish Health System, New Hyde Park, New York; ‡The Center for Advanced Endoscopy, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts BACKGROUND & AIMS:
There are limited data on the safety of endoscopic retrograde cholangiopancreatography (ERCP) during pregnancy, only data from observational series (approximately 350 cases) have been published. We aimed to evaluate the safety of ERCP in pregnant women by evaluating a large nationwide database.
METHODS:
We performed a retrospective matched-cohort study, collecting data from the National Inpatient Sample from 2008 through 2009. We compared data from pregnant women who underwent ERCP (n [ 907) with those from nonpregnant women who underwent ERCP (controls, n [ 2721). Complications related to ERCP were measured against the matched controls. Obstetric and fetal complications were measured against the national average of all obstetric admissions.
RESULTS:
ERCP-associated complications of perforation, infection, and bleeding were infrequent in both cohorts (P > .05). Post-ERCP pancreatitis (PEP) occurred in 12% of pregnant women and in 5% of controls (P < .001). There was a significantly lower rate of PEP in teaching hospitals (9.6%) than in nonteaching hospitals (14.6%; P < .001). The adjusted odds ratio for developing PEP among pregnant women vs controls was 2.8 (95% confidence interval [CI], 2.1–3.8). This risk of PEP was higher among nonteaching hospitals (adjusted odds ratio, 3.5; 95% CI, 2.3–5.2) than teaching hospitals (adjusted odds ratio, 2.5; 95% CI, 1.6–3.9).
CONCLUSIONS:
Pregnancy is an independent risk factor for PEP; PEP and PEP among pregnant women are each more prevalent in community hospitals than teaching centers. Proper precautions therefore should be considered for pregnant women undergoing ERCP, including transfer to a tertiary care center if appropriate.
Keywords: Pancreas; Pancreaticobiliary Disease; Morbidity; Choledocholithiasis.
ancreaticobiliary disease is the second-leading gastrointestinal disorder requiring surgery during pregnancy and can lead to significant morbidity and mortality.1 Physiological changes, such as an increase in weight and hormonal changes, can predispose pregnant women to gallstone and biliary sludge formation. Estrogen increases bile lithogenicity whereas progesterone impairs gallbladder emptying.2,3 These hormonal changes contribute to cholelithiasis and choledocholithiasis. The estimated prevalence of cholelithiasis has been estimated to be as high as 12% in pregnancy. Fortunately, only 0.1% of pregnancies are complicated by choledocholithiasis.3,4 Endoscopic retrograde cholangiopancreatography (ERCP) is widely considered to be the safest approach for the management of choledocholithiasis; avoiding more invasive surgeries.4
P
Choledocholithiasis is the most common indication for ERCP in pregnancy. Other less common indications include bile duct leaks, bile duct strictures, and therapy for the pancreatic duct.4 To date, the literature for safety in ERCP has consisted mainly of small retrospective studies. The medical
Abbreviations used in this paper: aOR, adjusted odds ratio; CI, confidence interval; dx1, primary diagnosis; dx2, secondary diagnosis; ERCP, endoscopic retrograde cholangiopancreatography; GI, gastrointestinal; HCUP, Healthcare Cost and Utilization Project; ICD-9-CM, International Classification of Diseases, 9th revision, Clinical Modification; NIS, Nationwide Inpatient Sample; PEP, post–endoscopic retrograde cholangiopancreatography pancreatitis. © 2015 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2015.04.175
2
Inamdar et al
literature consists of approximately 350 total cases of ERCP during pregnancy.5 The individual studies were small and were limited for various reasons; sample size being the most problematic. Nevertheless, the consensus is that ERCP should be performed in pregnant patients for strong indications because the risks have been understood to be generally comparable with those in nonpregnant patients.5 Given the limited data on ERCP performed in pregnant patients, our study aim was to examine ERCP complications in a large nationwide database of pregnant patients who underwent ERCP and to compare it with age-matched nonpregnant controls. This was performed in an effort to provide stronger evidence for the use of ERCP in the pregnancy population.
Methods Data Source We used the Nationwide Inpatient Sample (NIS) database to collect data for the years 2008 and 2009. The NIS is the largest all-payer inpatient database in the United States. It is publicly available and is part of the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality. This database contains records from approximately 8 million hospital admissions per year from approximately 1000 hospitals to achieve a 20% stratified sample of hospitals in the United States. Patient records include clinical information extracted from inpatient and discharge data. The NIS data set includes a unique identifier, demographic variables, discharge disposition, primary and secondary diagnoses (up to 25), primary and secondary procedures (up to 15), hospital characteristics, and information on nearly 100 patient- and hospital-related variables. The NIS database is representative of the national population and it concurs with the National Hospital Discharge Survey, supporting data reliability.6 Quality control procedures performed by HCUP have shown reliability and accuracy, specifically when it pertains to the principal diagnosis and dates of hospitalization.7 Thus, results extracted from this database can be used reliably to represent the population of patients undergoing ERCP in the United States.
Study Population Our study population consisted of all patients, older than the age of 18 years in the NIS database (2008 and 2009), who were hospitalized and had inpatient ERCP. The patient population and the controls used in this study were selected based on procedural coding in accordance with the International Classification of Diseases, 9th revision, Clinical Modification (ICD-9 CM).
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
These codes were validated previously and have been shown to have good sensitivity and positive predictive value. All women who underwent inpatient ERCP in 2008 and 2009 were identified using ICD-9 codes. The procedure codes used to identify patients for inclusion in the data set are shown in Table 1. The control population was selected among nonpregnant women who underwent ERCP by using stratified sampling. The controls were selected randomly using SAS 9.2 (SAS Institute, Inc, Cary, NC) using proc survey select, in which 3 nonpregnant control women were selected randomly for each pregnant woman drawn from the same 5-year age group. The sampling design used for the study is shown in Figure 1.
Outcome and Predictor Variables The goal of this study was to evaluate the safety of ERCP in pregnancy to both the patient and the fetus. The primary outcomes evaluated were ERCP-related complications (pancreatitis, bleeding, infection, perforation, and mortality) and obstetric complications (premature delivery, fetal complications/distress, and fetal mortality). The secondary outcomes evaluated included length of stay and procedural failure rates. Procedural failure was defined by the need to perform percutaneous
Table 1. ERCP ICD-9 Procedure Codes Used for Inclusion of Patients in the Study Procedure codes Diagnostic ERCP codes 51.10 51.11 51.14 52.13 52.14 Therapeutic ERCP codes 51.84 51.85 51.86 51.87 51.88 52.93 52.94 52.97 52.98
Procedure description
Endoscopic retrograde cholangiopancreatography [ERCP] Endoscopic retrograde cholangiography [ERC] Other closed [endoscopic] biopsy of biliary duct or sphincter of Oddi Endoscopic retrograde pancreatography [ERP] Closed [endoscopic] biopsy of pancreatic duct
Endoscopic dilation of ampulla and biliary duct Endoscopic sphincterotomy and papillotomy Endoscopic insertion of nasobiliary drainage tube Endoscopic insertion of stent (tube) into bile duct Endoscopic removal of stone(s) from biliary tract Endoscopic insertion of stent (tube) into pancreatic duct Endoscopic removal of stone(s) from pancreatic duct Endoscopic insertion of nasopancreatic drainage tube Endoscopic dilation of pancreatic duct
-
2015
Pregnancy and Pancreatitis After ERCP
3
Figure 1. Diagram of study population showing sampling scheme for the nonpregnant control group who had ERCP. For each pregnant woman, 3 nonpregnant women were selected randomly from the source population of all nonpregnant women from the same 5-year age strata.
transhepatic biliary drainage or open common bile duct exploration after ERCP. We were able to differentiate between acute pancreatitis at admission from post-ERCP pancreatitis (PEP). The methodology in deriving PEP from the database has been used before.8 In the NIS database the primary and secondary diagnoses are the medical conditions requiring admission to the hospital. Patients with a primary diagnosis (dx1) or secondary diagnosis (dx2) as pancreatitis were classified as patients with “acute pancreatitis” not related to ERCP. Patients who had a diagnosis of acute pancreatitis that was not a primary diagnosis (dx1) or a secondary diagnosis (dx2) and who underwent ERCP were classified as patients with PEP. In addition, post-ERCP pancreatitis requires a billing code of ICD-9 code 997.4 (complication of gastrointestinal [GI] procedure) along with the pancreatitis code. This is how post-ERCP pancreatitis is coded by medical billers and can be captured the same way in the NIS database.
Given that the methodology of deriving PEP was essential to the conclusion of the study, we aimed to validate this methodology by comparing the rates of PEP among all patients who underwent ERCP classified by age (from the NIS database) with rates of PEP from previous studies in the literature. The rates of PEP among the groups from the NIS database were similar to the rates observed in previous studies. Hence, we concluded that this variable was a good indicator of PEP. The rates of PEP were 5.92% for patients aged younger than 70 years, 1.52% for patients between 70 and 79 years, 1.20% for patients between 80 and 89 years, and 0.25% for patients older than 90 years of age; these rates are similar to those in the previous literature.9–12 Bleeding after ERCP was identified by specific ICD-9 codes used to define post-ERCP hemorrhage (998.11, 909.3, and V58.89). We also included patients who did not have a dx1 or a dx2 of GI bleeding, who underwent ERCP and were coded to have a GI bleed as one of their
4
Inamdar et al
diagnoses. Cholecystitis after ERCP was identified by using ICD-9 codes 575.0 and 575.1 among patients who did not have a dx1 or a dx2 of cholecystitis, who underwent ERCP and were coded to have cholecystitis as one of their other diagnoses. Perforation after ERCP was identified by using ICD-9 code 569.83. The Elixhauser comorbidity index13,14 was used to identify and adjust for comorbidities in the study. The Elixhauser comorbidity index is well established and validated in predicting in-patient mortality in previous studies.15 The NIS database contains 29 Agency for Healthcare Research and Quality comorbidity measures used by Elixhauser et al.13 We used the Comorbidity Software created by HCUP for the creation and analysis of these comorbidities.
Statistical Analyses Data analysis was performed using SAS software version 9.2. During the analyses, we took into account the stratified 2-stage cluster design by using the SAS proc survey commands and incorporating individual discharge-level weights. We applied the appropriate hospital and discharge level weights published by the HCUP for weighted analysis to reflect nationwide data during the study period. Statistical analysis using chi-square tests for categoric data and the Student t test for continuous data were performed. A P value of less than .05 was considered statistically significant. Univariate and bivariate analyses were performed to assess the indications and complications of ERCP in pregnancy. We also used logistic regression to assess the association between pregnancy and PEP while adjusting for age, race/ethnicity, health insurance, comorbidity, teaching hospital status, and the presence of sepsis.
Ethical Considerations The NIS database consists of completely unidentified data with no risk of loss of confidentiality. The present study was exempt from Institutional Review Board review. We completed a data user agreement with the Agency for Healthcare Research and Quality before using the NIS database. According to the data user agreement, any individual table cell counts of 10 or fewer cannot be presented to preserve patient confidentiality. In such instances, data are designated as information suppressed.
Results Patient Characteristics and Demographics There were 907 pregnant women who underwent ERCP identified in the NIS database, yielding a national estimate of 4474 ERCPs performed among pregnant women in the United States in 2008 and 2009. The computer software randomly selected 2721 nonpregnant
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Table 2. Patient Characteristics of Pregnant Women and Nonpregnant Controls Who Underwent ERCP Pregnant (n ¼ 907)
Not pregnant P (n ¼ 2721) valuea
Age, mean, y 26.06 26.24 .4197 (95% CI) (25.66–26.45) (26.01–26.47) .05). There was no difference among pregnant (1.2%) and nonpregnant women (1.6%) in the failure rate of ERCP and requirement of percutaneous drainage. PEP occurred in 12% of pregnant women compared with 5% of nonpregnant controls (P < .001). Of note, there was a significantly lower rate of PEP in teaching hospitals (9.6%) compared with nonteaching hospitals (14.6%; P < .001). In addition, pregnant women (4.1%) were less likely (P ¼ .01) to have a pancreatic stent placed compared with nonpregnant women (6.4%). The factors associated with risk for PEP in our study included age, hospital size, presence of comorbid conditions, and sepsis. After adjustment for age, race, hospital size, comorbidities, teaching hospital status, pancreatic stent placement, and presence of sepsis, the adjusted odds ratio (aOR) for developing PEP among pregnant women vs nonpregnant women was 2.8 (95% confidence interval [CI], 2.1–3.8). This risk of PEP was higher among nonteaching hospitals (aOR, 3.5; 95% CI, 2.3–5.2) compared with teaching hospitals (aOR, 2.5; 95% CI, 1.6–3.9).
Table 4. Complications Among Pregnant Women Undergoing ERCP
Table 3 shows the indications for ERCP in our study, which were based on admission diagnosis codes. The most common indication for ERCP among pregnant women was common bile duct stones with or without gallstones (93%). Approximately 4% of the patients in both groups had cholangitis. Pregnant women were significantly less likely to have ERCP for acute pancreatitis (20% vs 29% in controls) and biliary stricture (3% vs 7% in controls). Compared with nonpregnant
Maternal–fetal complications Maternal mortality Fetal loss Fetal distress/ complication Preterm labor
Pregnant (n ¼ 907), %
National estimates, %
0 IS IS
0.028 0.65 0.2
1.87
11.5
Pregnant (n ¼ 907), %
Not pregnant (n ¼ 2721), %
P value
93.38
79.53
Pregnancy Is a Risk Factor for Pancreatitis After Endoscopic Retrograde Cholangiopancreatography in a National Cohort Study Sumant Inamdar,* Tyler M. Berzin,‡ Divyesh V. Sejpal,* Douglas K. Pleskow,‡ Ram Chuttani,‡ Mandeep S. Sawhney,‡ and Arvind J. Trindade* *Department of Medicine, Division of Gastroenterology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore Long Island Jewish Health System, New Hyde Park, New York; ‡The Center for Advanced Endoscopy, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts BACKGROUND & AIMS:
There are limited data on the safety of endoscopic retrograde cholangiopancreatography (ERCP) during pregnancy, only data from observational series (approximately 350 cases) have been published. We aimed to evaluate the safety of ERCP in pregnant women by evaluating a large nationwide database.
METHODS:
We performed a retrospective matched-cohort study, collecting data from the National Inpatient Sample from 2008 through 2009. We compared data from pregnant women who underwent ERCP (n [ 907) with those from nonpregnant women who underwent ERCP (controls, n [ 2721). Complications related to ERCP were measured against the matched controls. Obstetric and fetal complications were measured against the national average of all obstetric admissions.
RESULTS:
ERCP-associated complications of perforation, infection, and bleeding were infrequent in both cohorts (P > .05). Post-ERCP pancreatitis (PEP) occurred in 12% of pregnant women and in 5% of controls (P < .001). There was a significantly lower rate of PEP in teaching hospitals (9.6%) than in nonteaching hospitals (14.6%; P < .001). The adjusted odds ratio for developing PEP among pregnant women vs controls was 2.8 (95% confidence interval [CI], 2.1–3.8). This risk of PEP was higher among nonteaching hospitals (adjusted odds ratio, 3.5; 95% CI, 2.3–5.2) than teaching hospitals (adjusted odds ratio, 2.5; 95% CI, 1.6–3.9).
CONCLUSIONS:
Pregnancy is an independent risk factor for PEP; PEP and PEP among pregnant women are each more prevalent in community hospitals than teaching centers. Proper precautions therefore should be considered for pregnant women undergoing ERCP, including transfer to a tertiary care center if appropriate.
Keywords: Pancreas; Pancreaticobiliary Disease; Morbidity; Choledocholithiasis.
ancreaticobiliary disease is the second-leading gastrointestinal disorder requiring surgery during pregnancy and can lead to significant morbidity and mortality.1 Physiological changes, such as an increase in weight and hormonal changes, can predispose pregnant women to gallstone and biliary sludge formation. Estrogen increases bile lithogenicity whereas progesterone impairs gallbladder emptying.2,3 These hormonal changes contribute to cholelithiasis and choledocholithiasis. The estimated prevalence of cholelithiasis has been estimated to be as high as 12% in pregnancy. Fortunately, only 0.1% of pregnancies are complicated by choledocholithiasis.3,4 Endoscopic retrograde cholangiopancreatography (ERCP) is widely considered to be the safest approach for the management of choledocholithiasis; avoiding more invasive surgeries.4
P
Choledocholithiasis is the most common indication for ERCP in pregnancy. Other less common indications include bile duct leaks, bile duct strictures, and therapy for the pancreatic duct.4 To date, the literature for safety in ERCP has consisted mainly of small retrospective studies. The medical
Abbreviations used in this paper: aOR, adjusted odds ratio; CI, confidence interval; dx1, primary diagnosis; dx2, secondary diagnosis; ERCP, endoscopic retrograde cholangiopancreatography; GI, gastrointestinal; HCUP, Healthcare Cost and Utilization Project; ICD-9-CM, International Classification of Diseases, 9th revision, Clinical Modification; NIS, Nationwide Inpatient Sample; PEP, post–endoscopic retrograde cholangiopancreatography pancreatitis. © 2015 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2015.04.175
2
Inamdar et al
literature consists of approximately 350 total cases of ERCP during pregnancy.5 The individual studies were small and were limited for various reasons; sample size being the most problematic. Nevertheless, the consensus is that ERCP should be performed in pregnant patients for strong indications because the risks have been understood to be generally comparable with those in nonpregnant patients.5 Given the limited data on ERCP performed in pregnant patients, our study aim was to examine ERCP complications in a large nationwide database of pregnant patients who underwent ERCP and to compare it with age-matched nonpregnant controls. This was performed in an effort to provide stronger evidence for the use of ERCP in the pregnancy population.
Methods Data Source We used the Nationwide Inpatient Sample (NIS) database to collect data for the years 2008 and 2009. The NIS is the largest all-payer inpatient database in the United States. It is publicly available and is part of the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality. This database contains records from approximately 8 million hospital admissions per year from approximately 1000 hospitals to achieve a 20% stratified sample of hospitals in the United States. Patient records include clinical information extracted from inpatient and discharge data. The NIS data set includes a unique identifier, demographic variables, discharge disposition, primary and secondary diagnoses (up to 25), primary and secondary procedures (up to 15), hospital characteristics, and information on nearly 100 patient- and hospital-related variables. The NIS database is representative of the national population and it concurs with the National Hospital Discharge Survey, supporting data reliability.6 Quality control procedures performed by HCUP have shown reliability and accuracy, specifically when it pertains to the principal diagnosis and dates of hospitalization.7 Thus, results extracted from this database can be used reliably to represent the population of patients undergoing ERCP in the United States.
Study Population Our study population consisted of all patients, older than the age of 18 years in the NIS database (2008 and 2009), who were hospitalized and had inpatient ERCP. The patient population and the controls used in this study were selected based on procedural coding in accordance with the International Classification of Diseases, 9th revision, Clinical Modification (ICD-9 CM).
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
These codes were validated previously and have been shown to have good sensitivity and positive predictive value. All women who underwent inpatient ERCP in 2008 and 2009 were identified using ICD-9 codes. The procedure codes used to identify patients for inclusion in the data set are shown in Table 1. The control population was selected among nonpregnant women who underwent ERCP by using stratified sampling. The controls were selected randomly using SAS 9.2 (SAS Institute, Inc, Cary, NC) using proc survey select, in which 3 nonpregnant control women were selected randomly for each pregnant woman drawn from the same 5-year age group. The sampling design used for the study is shown in Figure 1.
Outcome and Predictor Variables The goal of this study was to evaluate the safety of ERCP in pregnancy to both the patient and the fetus. The primary outcomes evaluated were ERCP-related complications (pancreatitis, bleeding, infection, perforation, and mortality) and obstetric complications (premature delivery, fetal complications/distress, and fetal mortality). The secondary outcomes evaluated included length of stay and procedural failure rates. Procedural failure was defined by the need to perform percutaneous
Table 1. ERCP ICD-9 Procedure Codes Used for Inclusion of Patients in the Study Procedure codes Diagnostic ERCP codes 51.10 51.11 51.14 52.13 52.14 Therapeutic ERCP codes 51.84 51.85 51.86 51.87 51.88 52.93 52.94 52.97 52.98
Procedure description
Endoscopic retrograde cholangiopancreatography [ERCP] Endoscopic retrograde cholangiography [ERC] Other closed [endoscopic] biopsy of biliary duct or sphincter of Oddi Endoscopic retrograde pancreatography [ERP] Closed [endoscopic] biopsy of pancreatic duct
Endoscopic dilation of ampulla and biliary duct Endoscopic sphincterotomy and papillotomy Endoscopic insertion of nasobiliary drainage tube Endoscopic insertion of stent (tube) into bile duct Endoscopic removal of stone(s) from biliary tract Endoscopic insertion of stent (tube) into pancreatic duct Endoscopic removal of stone(s) from pancreatic duct Endoscopic insertion of nasopancreatic drainage tube Endoscopic dilation of pancreatic duct
-
2015
Pregnancy and Pancreatitis After ERCP
3
Figure 1. Diagram of study population showing sampling scheme for the nonpregnant control group who had ERCP. For each pregnant woman, 3 nonpregnant women were selected randomly from the source population of all nonpregnant women from the same 5-year age strata.
transhepatic biliary drainage or open common bile duct exploration after ERCP. We were able to differentiate between acute pancreatitis at admission from post-ERCP pancreatitis (PEP). The methodology in deriving PEP from the database has been used before.8 In the NIS database the primary and secondary diagnoses are the medical conditions requiring admission to the hospital. Patients with a primary diagnosis (dx1) or secondary diagnosis (dx2) as pancreatitis were classified as patients with “acute pancreatitis” not related to ERCP. Patients who had a diagnosis of acute pancreatitis that was not a primary diagnosis (dx1) or a secondary diagnosis (dx2) and who underwent ERCP were classified as patients with PEP. In addition, post-ERCP pancreatitis requires a billing code of ICD-9 code 997.4 (complication of gastrointestinal [GI] procedure) along with the pancreatitis code. This is how post-ERCP pancreatitis is coded by medical billers and can be captured the same way in the NIS database.
Given that the methodology of deriving PEP was essential to the conclusion of the study, we aimed to validate this methodology by comparing the rates of PEP among all patients who underwent ERCP classified by age (from the NIS database) with rates of PEP from previous studies in the literature. The rates of PEP among the groups from the NIS database were similar to the rates observed in previous studies. Hence, we concluded that this variable was a good indicator of PEP. The rates of PEP were 5.92% for patients aged younger than 70 years, 1.52% for patients between 70 and 79 years, 1.20% for patients between 80 and 89 years, and 0.25% for patients older than 90 years of age; these rates are similar to those in the previous literature.9–12 Bleeding after ERCP was identified by specific ICD-9 codes used to define post-ERCP hemorrhage (998.11, 909.3, and V58.89). We also included patients who did not have a dx1 or a dx2 of GI bleeding, who underwent ERCP and were coded to have a GI bleed as one of their
4
Inamdar et al
diagnoses. Cholecystitis after ERCP was identified by using ICD-9 codes 575.0 and 575.1 among patients who did not have a dx1 or a dx2 of cholecystitis, who underwent ERCP and were coded to have cholecystitis as one of their other diagnoses. Perforation after ERCP was identified by using ICD-9 code 569.83. The Elixhauser comorbidity index13,14 was used to identify and adjust for comorbidities in the study. The Elixhauser comorbidity index is well established and validated in predicting in-patient mortality in previous studies.15 The NIS database contains 29 Agency for Healthcare Research and Quality comorbidity measures used by Elixhauser et al.13 We used the Comorbidity Software created by HCUP for the creation and analysis of these comorbidities.
Statistical Analyses Data analysis was performed using SAS software version 9.2. During the analyses, we took into account the stratified 2-stage cluster design by using the SAS proc survey commands and incorporating individual discharge-level weights. We applied the appropriate hospital and discharge level weights published by the HCUP for weighted analysis to reflect nationwide data during the study period. Statistical analysis using chi-square tests for categoric data and the Student t test for continuous data were performed. A P value of less than .05 was considered statistically significant. Univariate and bivariate analyses were performed to assess the indications and complications of ERCP in pregnancy. We also used logistic regression to assess the association between pregnancy and PEP while adjusting for age, race/ethnicity, health insurance, comorbidity, teaching hospital status, and the presence of sepsis.
Ethical Considerations The NIS database consists of completely unidentified data with no risk of loss of confidentiality. The present study was exempt from Institutional Review Board review. We completed a data user agreement with the Agency for Healthcare Research and Quality before using the NIS database. According to the data user agreement, any individual table cell counts of 10 or fewer cannot be presented to preserve patient confidentiality. In such instances, data are designated as information suppressed.
Results Patient Characteristics and Demographics There were 907 pregnant women who underwent ERCP identified in the NIS database, yielding a national estimate of 4474 ERCPs performed among pregnant women in the United States in 2008 and 2009. The computer software randomly selected 2721 nonpregnant
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Table 2. Patient Characteristics of Pregnant Women and Nonpregnant Controls Who Underwent ERCP Pregnant (n ¼ 907)
Not pregnant P (n ¼ 2721) valuea
Age, mean, y 26.06 26.24 .4197 (95% CI) (25.66–26.45) (26.01–26.47) .05). There was no difference among pregnant (1.2%) and nonpregnant women (1.6%) in the failure rate of ERCP and requirement of percutaneous drainage. PEP occurred in 12% of pregnant women compared with 5% of nonpregnant controls (P < .001). Of note, there was a significantly lower rate of PEP in teaching hospitals (9.6%) compared with nonteaching hospitals (14.6%; P < .001). In addition, pregnant women (4.1%) were less likely (P ¼ .01) to have a pancreatic stent placed compared with nonpregnant women (6.4%). The factors associated with risk for PEP in our study included age, hospital size, presence of comorbid conditions, and sepsis. After adjustment for age, race, hospital size, comorbidities, teaching hospital status, pancreatic stent placement, and presence of sepsis, the adjusted odds ratio (aOR) for developing PEP among pregnant women vs nonpregnant women was 2.8 (95% confidence interval [CI], 2.1–3.8). This risk of PEP was higher among nonteaching hospitals (aOR, 3.5; 95% CI, 2.3–5.2) compared with teaching hospitals (aOR, 2.5; 95% CI, 1.6–3.9).
Table 4. Complications Among Pregnant Women Undergoing ERCP
Table 3 shows the indications for ERCP in our study, which were based on admission diagnosis codes. The most common indication for ERCP among pregnant women was common bile duct stones with or without gallstones (93%). Approximately 4% of the patients in both groups had cholangitis. Pregnant women were significantly less likely to have ERCP for acute pancreatitis (20% vs 29% in controls) and biliary stricture (3% vs 7% in controls). Compared with nonpregnant
Maternal–fetal complications Maternal mortality Fetal loss Fetal distress/ complication Preterm labor
Pregnant (n ¼ 907), %
National estimates, %
0 IS IS
0.028 0.65 0.2
1.87
11.5
Pregnant (n ¼ 907), %
Not pregnant (n ¼ 2721), %
P value
93.38
79.53
