j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Postoperative pneumoperitoneum: is it normal or pathologic? Brandon C. Chapman, MD,a Kelsey E. McIntosh, BSN,b Edward L. Jones, MD,a Daniel Wells, MD,c Greg V. Stiegmann, MD,a and Thomas N. Robinson, MDa,* a
Department of Surgery, University of Colorado at Denver, Aurora, Colorado University of Colorado Hospital, Aurora, Colorado c Department of Radiology, University of Colorado at Denver, Aurora, Colorado b
article info
abstract
Article history:
Background: Pneumoperitoneum on computed tomography (CT) after abdominal surgery is
Received 28 October 2014
common, but its incidence, duration, and clinical significance is widely debated.
Received in revised form
Materials and methods: A retrospective, cohort study of patients who underwent abdominal
5 February 2015
CT within 30 days of abdominal surgery.
Accepted 27 March 2015
Results: Among 344 patients, pneumoperitoneum was found in 39% (135/344) of patients on
Available online 1 April 2015
postoperative days 0e6 in 53%, 7e13 in 41%, 14e20 in 23%, 21e27 in 13%, and 28e30 in 0%. Pneumoperitoneum was associated with the presence of a drain (P ¼ 0.014) but not with
Keywords:
age, gender, body mass index, smoking history, lung disease, or open versus laparoscopic
Free air
surgery (P > 0.05 for all variables). Eight patients required intervention (6%), most
Pneumoperitoneum
commonly for anastomotic leak (4 patients, 50%).
Postoperative
Conclusions: Postoperative pneumoperitoneum on abdominal CT can be seen in up to 23% of patients 3-weeks postoperatively; however, only 6% of the patients required intervention emphasizing the typically benign consequences of postoperative free air. Published by Elsevier Inc.
1.
Introduction
Plain radiographs and computed tomography (CT) are frequently performed for suspected intra-abdominal complications after abdominal operations. A common finding on imaging is pneumoperitoneum. The prevalence and duration of postoperative abdominal free air is dependent on which imaging modality is used. Most free air on plain radiographs resolves within 48 h and rarely occurs more than 5 d after surgery [1]. CT has been shown to be more sensitive in identifying free air. Pneumoperitoneum can be seen on 87% of CT scans and 53% of plain radiographs 3 d after uncomplicated abdominal surgery and on 50% of CT scans and 8% of plain
radiographs obtained 6 d after surgery [2]. The quandary that the clinician faces with the finding of pneumoperitoneum on a postoperative CT scan is whether this is a pathologic finding, which requires surgical intervention, or simply a normal finding after abdominal operation. The objective of this study was to characterize the incidence and duration of postoperative pneumoperitoneum in patients undergoing CT of the abdomen and pelvis and its impact on the need for additional surgical interventions in a large patient population. We hypothesize that postoperative pneumoperitoneum will be a common finding that will decrease in a linear fashion from time of surgery and will rarely result in surgical intervention.
* Corresponding author. Department of Surgery, University of Colorado at Denver, Mail Stop C313, 12631 East 17th Avenue, Room 6001, Aurora, CO 80045. Tel.: þ1 303 724 2728; fax: þ1 303 724 2733. E-mail address: [email protected] (T.N. Robinson). 0022-4804/$ e see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jss.2015.03.083
108
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
2.
Materials and methods
2.1.
Patients
Using key search words “free air”, “pneumoperitoneum”, and/ or “intraperitoneal gas,” a retrospective review identified all patients who had an abdominal CT examination of the abdomen and pelvis at the University of Colorado Hospital from September 1, 2011eAugust 27, 2013.
2.1.1.
Patient demographics
Patient demographics recorded included age, gender, body mass index, current or history of smoking, and history of pulmonary disease including obstructive sleep apnea, asthma, chronic obstructive pulmonary disease, and lung cancer. This study was approved by COMIRB protocol 13-2975.
2.1.2.
Surgical specialties
We identified patients of all surgical specialties operating in the abdomen, including general surgery, transplant, neurosurgery, gynecology, and urology patients, who had undergone open or laparoscopic intraperitoneal operations and had abdominal CT examination within 30 postoperative days (PODs) of the operation.
2.2.
Radiology reports
Radiology reports dictated by board-certified radiologists were reviewed individually for the presence or absence of “free air,” “pneumoperitoneum,” and/or “intraperitoneal gas,” presence of an intra-abdominal drain, and the indication for imaging. If present, the magnitude of pneumoperitoneum quantified by the radiologist ordinally as tiny and/or scant, small, moderate, or large was recorded.
2.3.
Operative notes
Operative notes were reviewed for surgical specialty, surgical operation, open versus laparoscopic, and indication for surgery. To determine how many patients required an operation after imaging findings of pneumoperitoneum, operative notes within 30 d of initial surgery were reviewed. Among these patients, intraoperative findings were recorded.
2.4.
and PODs 28e30. The presence of pneumoperitoneum was correlated with the sex, age, body mass index, smoking history, previous lung disease, presence of intra-abdominal drain, and laparoscopic or open abdominal surgery. The size of the pneumoperitoneum on radiology reports was correlated with need for additional surgical intervention.
3.
Results
Three hundred forty-four patients underwent abdominal CT scans up to 30 days after abdominal surgery. The most common rationale for ordering a CT scan were as follows: evaluate abscess (86), abdominal pain (68), fever (26), evaluate obstruction (24), evaluate for anastomotic leak (21), nausea/ vomiting (12), evaluate free air (9), leukocytosis (9), tachycardia (6), abdominal distention (5), evaluate bleeding (5), and indication not recorded (73). Overall, postoperative pneumoperitoneum was found in 39% of patients (135 of 344). Pneumoperitoneum was demonstrated on PODs 0e6 in 53% (64 of 120), 7e13 in 41% (56/139), 14e20 in 23% (12 of 53), 21e27 in 13% (3/22), and 28e30 in 0% (0 of 10; P ¼ 0.0454 PODs 0e6 versus 7e13; P ¼ 0.0168 for PODs 7e13 versus 14e20; P ¼ 0.5302 for PODs 14e20 versus 21e27; P ¼ 0.534 for PODs 21e27 versus 28e30; Figure). No free air (0 of 22 patients) was found on POD 24 or after. Comparing patients with and without pneumoperitoneum, there was no difference in age (P ¼ 0.208), gender (P ¼ 0.052), body mass index (P ¼ 0.068), smoking history (P ¼ 0.218), lung disease (P ¼ 0.456), or open versus laparoscopic surgery (P ¼ 0.309). However, the incidence of pneumoperitoneum was associated with the presence of a drain (P ¼ 0.014; Table 1). Only 6% (8 of 136) of patients required surgery after the diagnosis of pneumoperitoneum. Most of these patients had a combination of free fluid and/or fluid collections and pneumoperitoneum on CT scan. Six of the eight patients requiring an operation had an absolute indication for surgical intervention as follows: 1 large volume hemorrhage from intraperitoneal drain, 2 feculent material from intraperitoneal drain, 1 feculent material from surgical incision, 1 tracheoesophageal fistula, and 1 bowel obstruction failing medical management. The remaining two patients were taken to the operating room based on pneumoperitoneum. Intraoperative findings were the following: 4 anastomotic leaks (PODs 7, 8, 10,
Statistical analysis
Statistical analysis was used to determine the natural history of postoperative pneumoperitoneum and to compare patients with and without postoperative pneumoperitoneum. Nominal variables are expressed as proportions and percentages; continuous variables are expressed as means and standard deviations. Between groups differences for proportions were tested using two-sided chi-square or Fisher exact tests where appropriate; between groups differences for continuous variables were tested using two-sided t-tests not assuming equal variance. Outcomes analyzed included the number of days postoperatively on which pneumoperitoneum was identified defined as PODs 0e6, PODs 7e13, PODs 14e20, PODs 21e27,
Figure e Incidence of postoperative pneumoperitoneum. (Color version of the figure is available online.)
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
Table 1 e Patient characteristics. Characteristics
No free air (n ¼ 209)
Free air (n ¼ 135)
P value
53 16 27.3 7.2
55 17 25.8 7.2
0.208 0.068
77 (54.6) 132 (65.0)
64 (45.4) 71 (35.0)
0.052
168 (64.6) 40 (47.6)
92 (35.4) 44 (52.4)
0.007
172 (62.5) 46 (43.8)
103 (37.5) 59 (56.2)
0.075
109 (56.8) 98 (65.3)
83 (43.2) 52 (34.7)
0.119
61 (67.0) 143 (60.3)
30 (33.0) 94 (39.7)
0.309
Age (y) Body mass index (kg/m2) Gender, n (%) Male Female Drain, n (%) No Yes Lung disease, n (%) No Yes Smoking history, n (%) No Yes Surgical approach, n (%) Laparoscopic Open
Values stated are mean standard deviations unless otherwise noted.
and 11), 2 colon perforations (PODs 4 and 7), 1 no evidence of perforation or leak (POD 9), and 1 jejunal serosal tear (POD 10; Table 2).
4.
Discussion
Postoperative pneumoperitoneum is a common finding on CT scans after abdominal surgery. In our cohort of patients, we found that up to 39% of patients who undergo abdominal CT postoperatively have pneumoperitoneum. The incidence of pneumoperitoneum decreases in a dose-response fashion from time of surgery. However, up to 23% of patients have evidence of pneumoperitoneum 3 weeks postoperatively. Pneumoperitoneum was found to persist up to 23 days. Only 6% of patients with pneumoperitoneum required further surgical intervention. The patient’s complete clinical presentation must be taken into consideration in addition to imaging to minimize the morbidity that can be seen with a negative exploratory laparotomy. Postoperative patients frequently have abdominal pain, nausea, vomiting, abdominal distention, fevers, or leukocytosis that often leads to imaging of the abdomen. Pneumoperitoneum can be seen on plain radiographs and/or CT and may signify a benign finding of residual air that will eventually be resorbed by the body or it may represent a surgical complication including anastomotic leak or perforation of the bowel that requires urgent surgical intervention. Postoperative pneumoperitoneum is reported in more than one-third of patients with postoperative imaging. In a prospective study by Gayer et al. [3], the presence of pneumoperitoneum was observed in 31.5% (29 of 92) of patients undergoing CT after abdominal surgery. McAllister et al. [4] reported on 27 patients who were evaluated by computerized axial tomography of the abdomen within 24 h after laparoscopic cholecystectomy and found 19 (70.4%) demonstrated residual pneumoperitoneum.
109
The prevalence of pneumoperitoneum has been shown to decrease with the length of time between surgery and CT examination as follows: 48% of patients on PODs 0e3 and 31% on PODs 4e7 [3]. In our patient population, pneumoperitoneum was seen in up to 53% at 1 wk, 41% at 2 wk, 23% at 3 wk, and 13% at 4 wk. Similarly, Earls et al. [2] reported pneumoperitoneum in 87% of patients on POD 3 and 50% on POD 6. However, pneumoperitoneum can persist for several days and has been reported to occur 48 d after a total laparoscopic hysterectomy. In this particular patient, an exploratory laparotomy was performed, and there was no evidence of perforation or organic trauma [5]. A single patient in our cohort was found to have pneumoperitoneum on POD 23 and did not require additional surgical intervention. Previous studies have evaluated several variables that may be associated with the presence of postoperative pneumoperitoneum, and findings are quite variable. In agreement with Earls et al. [2], we found that pneumoperitoneum was not associated with sex, age, or body habitus; however, Gayer et al. [3] found a correlation with sex and body habitus, but not age. Millitz et al. [6] observed an inverse correlation between the patient’s body weight and the duration of residual postoperative pneumoperitoneum. Other studies have shown no association with patient age [7,8]. Similar to Gayer et al. [3], laparoscopic surgery did not have an effect on the incidence of postoperative pneumoperitoneum in our patient population. However, a study comparing open to laparoscopic cholecystectomy found that the incidence of subdiaphragmatic air after laparoscopic cholecystectomy was 24% compared with 60% in the open cholecystectomy group [9]. Likewise, in study of five pigs that underwent laparoscopic (4) and open (1) cholecystectomy, pneumoperitoneum resolved on upright radiographs in all pigs by POD 1 and serial CT scans demonstrated that the laparoscopic group had resolution or minimal persistence of free intraperitoneal gas by POD 2 compared with that by POD 6 in the open group [10]. In our patient population, the presence of an intraperitoneal drain was significantly associated with an increased incidence of postoperative pneumoperitoneum. Likewise Gayer et al. [3] found that patients with one or more drains had significantly more often free air than patients without a drain. Intraperitoneal drains may provide a point of entry for air to pass into the peritoneal cavity from the atmosphere. Gayer et al. [3] demonstrated that air bubbles can be seen adjacent to the drain supporting this theory. Furthermore, Bryant et al. [7] reported that one patient had a gradually increasing pneumoperitoneum as a result of air passing into the peritoneal cavity along the drainage tract, which resolved after removal of the drain. However, Bryant et al. and others have found that the presence of a drain is not associated with the incidence of postoperative pneumoperitoneum [7,11]. Gayer et al. sought to determine how many patients were found to have a disrupted gastrointestinal tract on reexploration and had the presence of pneumoperitoneum. Among the 7 of 89 patients that had a disrupted gastrointestinal tract, only 4 had pneumoperitoneum [3]. In our patient population, the vast majority of patients had a benign abdominal examination and only 6% of these patients required an operation. Although there were no consistent
110
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
Table 2 e Patients with pneumoperitoneum who required additional surgical intervention. Type of surgery
Preoperative diagnosis
POD
Significant CT findings High grade mechanical bowel obstruction Free fluid paracolic gutters/pelvis Pneumoperitoneum in pelvis and abdominal wall Multiple fluid collections with gas Splenic vein thrombosis Increased free air from prior CT Thickened stomach Tiny focus of air adjacent to pull through Fluid collection Small free air in operative bed No fluid collections Pelvic abscess Diffuse bowel wall thickening Small pneumoperitoneum Focal gas and contrast material near anastomosis Very large pneumoperitoneum outlining liver and spleen Extensive retroperitoneal air in pelvis, right paracolic gutters, pararenal space, and psoas margins Fluid collection with gas Large pelvic fluid collection Moderate pneumoperitoneum Obstruction near anastomosis
Total colectomy and end ileostomy
Ulcerative colitis
9
Exploratory laparotomy, tumor debulking
Pseudomyxoma peritonei
7
Esophagectomy
Esophageal cancer
11
Whipple
Pancreatic cyst
10
Exploratory laparotomy
Spontaneous perforation
Cystoprostatectomy
Bladder cancer
10
Total colectomy with ileal-sigmoid anastomosis
Colonic inertia
8
Sigmoidectomy
Diverticulitis
7
4
findings among these patients on CT scan, most patients had a combination of free fluid and/or fluid collections and pneumoperitoneum. However, 6 of the 8 patients had an absolute indication for surgical intervention as follows: 1 large volume hemorrhage from intraperitoneal drain, 2 feculent material from intraperitoneal drain, 1 feculent material from surgical incision, 1 tracheoesophageal fistula, and 1 bowel obstruction failing medical management. These patients would have been operated on regardless of the presence or absence of pneumoperitoneum on CT scan. The remaining two patients were taken to the operating room based on pneumoperitoneum and only one was found to have an anastomotic leak. Previous studies are limited by a small sample size. To our knowledge, this study is the largest reported to date and
Operative Indication
Intraoperative findings
Leukocytosis Free air
No perforation Interloop abscess
Feculent material from drain
Colonic perforation
Tracheoesophageal fistula
Ischemic gastric conduit, anastomotic leak
Bleeding from drain
Arterial bleeding, jejunal serosal tear
Feculent material from drain
Perforated viscus
Feculent material from incision
Anastomotic leak
Extensive pneumoperitoneum
Anastomotic leak
Obstruction
Anastomotic leak
includes a diverse range of surgical operations across all specialties making the findings generalizable to a large patient population. However, our study does have important limitations. As with all retrospective studies, our study is prone to selection bias. However, we reviewed consecutive patients that underwent postoperative CT scans to limit this bias. In addition, sampling bias may have been present in our patient population. Some patients that had surgery at our institutions may have followed up with outside institutions with postoperative complications. Furthermore, because our inclusion criteria only included patients who had an abdominal operation and an abdominal CT examination within 30 PODs of the operation, we cannot determine the frequency in which patients required a reoperation and did not have a CT scan. Similarly, it is unknown if all patients who had an unusual
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
postoperative course underwent a CT scan; however, most patients who fail to progress postoperatively and show signs of fever, tachycardia, or leukocytosis undergo CT scan at our institution.
5.
Conclusions
In conclusion, postoperative pneumoperitoneum is a common finding on CT scans after abdominal surgery, and up to 23% of patients 3 weeks postoperatively continue to have pneumoperitoneum. However, most patients with postoperative pneumoperitoneum do not require an intervention. Better understanding the natural history of postoperative pneumoperitoneum will aid surgeons in interpreting their postoperative CT scans.
Acknowledgment Authors’ contributions: B.C.C., E.L.J., D.W., G.V.S., and T.N.R. contributed to the conception and design of study. B.C.C., K.E.M., and D.W. did the acquisition of data. B.C.C., E.L.J., G.V.S., and T.N.R. did the analysis and interpretation of data. B.C.C., K.E.M., and D.W. did the drafting of the article. B.C.C., K.E.M., E.L.J., D.W., G.V.S., and T.N.R. did the final approval.
Disclosure The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
111
references
[1] Nielsen KT, Lund L, Larsen LP, Knudsen P. Duration of postoperative pneumoperitoneum. Eur J Surg 1997;163:501. [2] Earls JP, Dachman AH, Colon E, Garrett MG, Molloy M. Prevalence and duration of postoperative pneumoperitoneum: sensitivity of CT vs left lateral decubitus radiography. AJR Am J Roentgenol 1993;161:781. [3] Gayer G, Jonas T, Apter S, Amitai M, Shabtai M, Hertz M. Postoperative pneumoperitoneum as detected by CT: prevalence, duration, and relevant factors affecting its possible significance. Abdom Imaging 2000;25:301. [4] McAllister JD, D’Altorio RA, Snyder A. CT findings after uncomplicated percutaneous laparoscopic cholecystectomy. J Comput Assist Tomogr 1991;15:770. [5] Smith KS, Wilson TC, Luces L, Stevenson AA, Hajhosseini B, Siram SM. Pneumoperitoneum 48 days after laparoscopic hysterectomy. JSLS 2013;17:661. [6] Millitz K, Moote DJ, Sparrow RK, Girotti MJ, Holliday RL, McLarty TD. Pneumoperitoneum after laparoscopic cholecystectomy: frequency and duration as seen on upright chest radiographs. AJR Am J Roentgenol 1994;164:837. [7] Bryant LR, Wiot JF, Kloecker RJ. A study of the factors affecting the incidence and duration of postoperative pneumoperitoneum. Surg Gynecol Obstet 1963;117:145. [8] Draper K, Jefson R, Jongeward R Jr, McLeod M. Duration of postlaparoscopic pneumoperitoneum. Surg Endosc 1997;11: 809. [9] Schauer PR, Page CP, Ghiatas AA, Miller JE, Schwesinger WH, Sirinek KR. Incidence and significance of subdiaphragmatic air following laparoscopic cholecystectomy. Am Surg 1997; 63:132. [10] Feingold DL, Widmann WD, Calhoun SK, et al. Persistent post-laparoscopy pneumoperitoneum. Surg Endosc 2003;17: 296. [11] Rice RP, Thompson WM, Gedgaudas RK. The diagnosis and significance of extraluminal gas in the abdomen. Radiol Clin North Am 1982;20:819.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Postoperative pneumoperitoneum: is it normal or pathologic? Brandon C. Chapman, MD,a Kelsey E. McIntosh, BSN,b Edward L. Jones, MD,a Daniel Wells, MD,c Greg V. Stiegmann, MD,a and Thomas N. Robinson, MDa,* a
Department of Surgery, University of Colorado at Denver, Aurora, Colorado University of Colorado Hospital, Aurora, Colorado c Department of Radiology, University of Colorado at Denver, Aurora, Colorado b
article info
abstract
Article history:
Background: Pneumoperitoneum on computed tomography (CT) after abdominal surgery is
Received 28 October 2014
common, but its incidence, duration, and clinical significance is widely debated.
Received in revised form
Materials and methods: A retrospective, cohort study of patients who underwent abdominal
5 February 2015
CT within 30 days of abdominal surgery.
Accepted 27 March 2015
Results: Among 344 patients, pneumoperitoneum was found in 39% (135/344) of patients on
Available online 1 April 2015
postoperative days 0e6 in 53%, 7e13 in 41%, 14e20 in 23%, 21e27 in 13%, and 28e30 in 0%. Pneumoperitoneum was associated with the presence of a drain (P ¼ 0.014) but not with
Keywords:
age, gender, body mass index, smoking history, lung disease, or open versus laparoscopic
Free air
surgery (P > 0.05 for all variables). Eight patients required intervention (6%), most
Pneumoperitoneum
commonly for anastomotic leak (4 patients, 50%).
Postoperative
Conclusions: Postoperative pneumoperitoneum on abdominal CT can be seen in up to 23% of patients 3-weeks postoperatively; however, only 6% of the patients required intervention emphasizing the typically benign consequences of postoperative free air. Published by Elsevier Inc.
1.
Introduction
Plain radiographs and computed tomography (CT) are frequently performed for suspected intra-abdominal complications after abdominal operations. A common finding on imaging is pneumoperitoneum. The prevalence and duration of postoperative abdominal free air is dependent on which imaging modality is used. Most free air on plain radiographs resolves within 48 h and rarely occurs more than 5 d after surgery [1]. CT has been shown to be more sensitive in identifying free air. Pneumoperitoneum can be seen on 87% of CT scans and 53% of plain radiographs 3 d after uncomplicated abdominal surgery and on 50% of CT scans and 8% of plain
radiographs obtained 6 d after surgery [2]. The quandary that the clinician faces with the finding of pneumoperitoneum on a postoperative CT scan is whether this is a pathologic finding, which requires surgical intervention, or simply a normal finding after abdominal operation. The objective of this study was to characterize the incidence and duration of postoperative pneumoperitoneum in patients undergoing CT of the abdomen and pelvis and its impact on the need for additional surgical interventions in a large patient population. We hypothesize that postoperative pneumoperitoneum will be a common finding that will decrease in a linear fashion from time of surgery and will rarely result in surgical intervention.
* Corresponding author. Department of Surgery, University of Colorado at Denver, Mail Stop C313, 12631 East 17th Avenue, Room 6001, Aurora, CO 80045. Tel.: þ1 303 724 2728; fax: þ1 303 724 2733. E-mail address: [email protected] (T.N. Robinson). 0022-4804/$ e see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jss.2015.03.083
108
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
2.
Materials and methods
2.1.
Patients
Using key search words “free air”, “pneumoperitoneum”, and/ or “intraperitoneal gas,” a retrospective review identified all patients who had an abdominal CT examination of the abdomen and pelvis at the University of Colorado Hospital from September 1, 2011eAugust 27, 2013.
2.1.1.
Patient demographics
Patient demographics recorded included age, gender, body mass index, current or history of smoking, and history of pulmonary disease including obstructive sleep apnea, asthma, chronic obstructive pulmonary disease, and lung cancer. This study was approved by COMIRB protocol 13-2975.
2.1.2.
Surgical specialties
We identified patients of all surgical specialties operating in the abdomen, including general surgery, transplant, neurosurgery, gynecology, and urology patients, who had undergone open or laparoscopic intraperitoneal operations and had abdominal CT examination within 30 postoperative days (PODs) of the operation.
2.2.
Radiology reports
Radiology reports dictated by board-certified radiologists were reviewed individually for the presence or absence of “free air,” “pneumoperitoneum,” and/or “intraperitoneal gas,” presence of an intra-abdominal drain, and the indication for imaging. If present, the magnitude of pneumoperitoneum quantified by the radiologist ordinally as tiny and/or scant, small, moderate, or large was recorded.
2.3.
Operative notes
Operative notes were reviewed for surgical specialty, surgical operation, open versus laparoscopic, and indication for surgery. To determine how many patients required an operation after imaging findings of pneumoperitoneum, operative notes within 30 d of initial surgery were reviewed. Among these patients, intraoperative findings were recorded.
2.4.
and PODs 28e30. The presence of pneumoperitoneum was correlated with the sex, age, body mass index, smoking history, previous lung disease, presence of intra-abdominal drain, and laparoscopic or open abdominal surgery. The size of the pneumoperitoneum on radiology reports was correlated with need for additional surgical intervention.
3.
Results
Three hundred forty-four patients underwent abdominal CT scans up to 30 days after abdominal surgery. The most common rationale for ordering a CT scan were as follows: evaluate abscess (86), abdominal pain (68), fever (26), evaluate obstruction (24), evaluate for anastomotic leak (21), nausea/ vomiting (12), evaluate free air (9), leukocytosis (9), tachycardia (6), abdominal distention (5), evaluate bleeding (5), and indication not recorded (73). Overall, postoperative pneumoperitoneum was found in 39% of patients (135 of 344). Pneumoperitoneum was demonstrated on PODs 0e6 in 53% (64 of 120), 7e13 in 41% (56/139), 14e20 in 23% (12 of 53), 21e27 in 13% (3/22), and 28e30 in 0% (0 of 10; P ¼ 0.0454 PODs 0e6 versus 7e13; P ¼ 0.0168 for PODs 7e13 versus 14e20; P ¼ 0.5302 for PODs 14e20 versus 21e27; P ¼ 0.534 for PODs 21e27 versus 28e30; Figure). No free air (0 of 22 patients) was found on POD 24 or after. Comparing patients with and without pneumoperitoneum, there was no difference in age (P ¼ 0.208), gender (P ¼ 0.052), body mass index (P ¼ 0.068), smoking history (P ¼ 0.218), lung disease (P ¼ 0.456), or open versus laparoscopic surgery (P ¼ 0.309). However, the incidence of pneumoperitoneum was associated with the presence of a drain (P ¼ 0.014; Table 1). Only 6% (8 of 136) of patients required surgery after the diagnosis of pneumoperitoneum. Most of these patients had a combination of free fluid and/or fluid collections and pneumoperitoneum on CT scan. Six of the eight patients requiring an operation had an absolute indication for surgical intervention as follows: 1 large volume hemorrhage from intraperitoneal drain, 2 feculent material from intraperitoneal drain, 1 feculent material from surgical incision, 1 tracheoesophageal fistula, and 1 bowel obstruction failing medical management. The remaining two patients were taken to the operating room based on pneumoperitoneum. Intraoperative findings were the following: 4 anastomotic leaks (PODs 7, 8, 10,
Statistical analysis
Statistical analysis was used to determine the natural history of postoperative pneumoperitoneum and to compare patients with and without postoperative pneumoperitoneum. Nominal variables are expressed as proportions and percentages; continuous variables are expressed as means and standard deviations. Between groups differences for proportions were tested using two-sided chi-square or Fisher exact tests where appropriate; between groups differences for continuous variables were tested using two-sided t-tests not assuming equal variance. Outcomes analyzed included the number of days postoperatively on which pneumoperitoneum was identified defined as PODs 0e6, PODs 7e13, PODs 14e20, PODs 21e27,
Figure e Incidence of postoperative pneumoperitoneum. (Color version of the figure is available online.)
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
Table 1 e Patient characteristics. Characteristics
No free air (n ¼ 209)
Free air (n ¼ 135)
P value
53 16 27.3 7.2
55 17 25.8 7.2
0.208 0.068
77 (54.6) 132 (65.0)
64 (45.4) 71 (35.0)
0.052
168 (64.6) 40 (47.6)
92 (35.4) 44 (52.4)
0.007
172 (62.5) 46 (43.8)
103 (37.5) 59 (56.2)
0.075
109 (56.8) 98 (65.3)
83 (43.2) 52 (34.7)
0.119
61 (67.0) 143 (60.3)
30 (33.0) 94 (39.7)
0.309
Age (y) Body mass index (kg/m2) Gender, n (%) Male Female Drain, n (%) No Yes Lung disease, n (%) No Yes Smoking history, n (%) No Yes Surgical approach, n (%) Laparoscopic Open
Values stated are mean standard deviations unless otherwise noted.
and 11), 2 colon perforations (PODs 4 and 7), 1 no evidence of perforation or leak (POD 9), and 1 jejunal serosal tear (POD 10; Table 2).
4.
Discussion
Postoperative pneumoperitoneum is a common finding on CT scans after abdominal surgery. In our cohort of patients, we found that up to 39% of patients who undergo abdominal CT postoperatively have pneumoperitoneum. The incidence of pneumoperitoneum decreases in a dose-response fashion from time of surgery. However, up to 23% of patients have evidence of pneumoperitoneum 3 weeks postoperatively. Pneumoperitoneum was found to persist up to 23 days. Only 6% of patients with pneumoperitoneum required further surgical intervention. The patient’s complete clinical presentation must be taken into consideration in addition to imaging to minimize the morbidity that can be seen with a negative exploratory laparotomy. Postoperative patients frequently have abdominal pain, nausea, vomiting, abdominal distention, fevers, or leukocytosis that often leads to imaging of the abdomen. Pneumoperitoneum can be seen on plain radiographs and/or CT and may signify a benign finding of residual air that will eventually be resorbed by the body or it may represent a surgical complication including anastomotic leak or perforation of the bowel that requires urgent surgical intervention. Postoperative pneumoperitoneum is reported in more than one-third of patients with postoperative imaging. In a prospective study by Gayer et al. [3], the presence of pneumoperitoneum was observed in 31.5% (29 of 92) of patients undergoing CT after abdominal surgery. McAllister et al. [4] reported on 27 patients who were evaluated by computerized axial tomography of the abdomen within 24 h after laparoscopic cholecystectomy and found 19 (70.4%) demonstrated residual pneumoperitoneum.
109
The prevalence of pneumoperitoneum has been shown to decrease with the length of time between surgery and CT examination as follows: 48% of patients on PODs 0e3 and 31% on PODs 4e7 [3]. In our patient population, pneumoperitoneum was seen in up to 53% at 1 wk, 41% at 2 wk, 23% at 3 wk, and 13% at 4 wk. Similarly, Earls et al. [2] reported pneumoperitoneum in 87% of patients on POD 3 and 50% on POD 6. However, pneumoperitoneum can persist for several days and has been reported to occur 48 d after a total laparoscopic hysterectomy. In this particular patient, an exploratory laparotomy was performed, and there was no evidence of perforation or organic trauma [5]. A single patient in our cohort was found to have pneumoperitoneum on POD 23 and did not require additional surgical intervention. Previous studies have evaluated several variables that may be associated with the presence of postoperative pneumoperitoneum, and findings are quite variable. In agreement with Earls et al. [2], we found that pneumoperitoneum was not associated with sex, age, or body habitus; however, Gayer et al. [3] found a correlation with sex and body habitus, but not age. Millitz et al. [6] observed an inverse correlation between the patient’s body weight and the duration of residual postoperative pneumoperitoneum. Other studies have shown no association with patient age [7,8]. Similar to Gayer et al. [3], laparoscopic surgery did not have an effect on the incidence of postoperative pneumoperitoneum in our patient population. However, a study comparing open to laparoscopic cholecystectomy found that the incidence of subdiaphragmatic air after laparoscopic cholecystectomy was 24% compared with 60% in the open cholecystectomy group [9]. Likewise, in study of five pigs that underwent laparoscopic (4) and open (1) cholecystectomy, pneumoperitoneum resolved on upright radiographs in all pigs by POD 1 and serial CT scans demonstrated that the laparoscopic group had resolution or minimal persistence of free intraperitoneal gas by POD 2 compared with that by POD 6 in the open group [10]. In our patient population, the presence of an intraperitoneal drain was significantly associated with an increased incidence of postoperative pneumoperitoneum. Likewise Gayer et al. [3] found that patients with one or more drains had significantly more often free air than patients without a drain. Intraperitoneal drains may provide a point of entry for air to pass into the peritoneal cavity from the atmosphere. Gayer et al. [3] demonstrated that air bubbles can be seen adjacent to the drain supporting this theory. Furthermore, Bryant et al. [7] reported that one patient had a gradually increasing pneumoperitoneum as a result of air passing into the peritoneal cavity along the drainage tract, which resolved after removal of the drain. However, Bryant et al. and others have found that the presence of a drain is not associated with the incidence of postoperative pneumoperitoneum [7,11]. Gayer et al. sought to determine how many patients were found to have a disrupted gastrointestinal tract on reexploration and had the presence of pneumoperitoneum. Among the 7 of 89 patients that had a disrupted gastrointestinal tract, only 4 had pneumoperitoneum [3]. In our patient population, the vast majority of patients had a benign abdominal examination and only 6% of these patients required an operation. Although there were no consistent
110
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
Table 2 e Patients with pneumoperitoneum who required additional surgical intervention. Type of surgery
Preoperative diagnosis
POD
Significant CT findings High grade mechanical bowel obstruction Free fluid paracolic gutters/pelvis Pneumoperitoneum in pelvis and abdominal wall Multiple fluid collections with gas Splenic vein thrombosis Increased free air from prior CT Thickened stomach Tiny focus of air adjacent to pull through Fluid collection Small free air in operative bed No fluid collections Pelvic abscess Diffuse bowel wall thickening Small pneumoperitoneum Focal gas and contrast material near anastomosis Very large pneumoperitoneum outlining liver and spleen Extensive retroperitoneal air in pelvis, right paracolic gutters, pararenal space, and psoas margins Fluid collection with gas Large pelvic fluid collection Moderate pneumoperitoneum Obstruction near anastomosis
Total colectomy and end ileostomy
Ulcerative colitis
9
Exploratory laparotomy, tumor debulking
Pseudomyxoma peritonei
7
Esophagectomy
Esophageal cancer
11
Whipple
Pancreatic cyst
10
Exploratory laparotomy
Spontaneous perforation
Cystoprostatectomy
Bladder cancer
10
Total colectomy with ileal-sigmoid anastomosis
Colonic inertia
8
Sigmoidectomy
Diverticulitis
7
4
findings among these patients on CT scan, most patients had a combination of free fluid and/or fluid collections and pneumoperitoneum. However, 6 of the 8 patients had an absolute indication for surgical intervention as follows: 1 large volume hemorrhage from intraperitoneal drain, 2 feculent material from intraperitoneal drain, 1 feculent material from surgical incision, 1 tracheoesophageal fistula, and 1 bowel obstruction failing medical management. These patients would have been operated on regardless of the presence or absence of pneumoperitoneum on CT scan. The remaining two patients were taken to the operating room based on pneumoperitoneum and only one was found to have an anastomotic leak. Previous studies are limited by a small sample size. To our knowledge, this study is the largest reported to date and
Operative Indication
Intraoperative findings
Leukocytosis Free air
No perforation Interloop abscess
Feculent material from drain
Colonic perforation
Tracheoesophageal fistula
Ischemic gastric conduit, anastomotic leak
Bleeding from drain
Arterial bleeding, jejunal serosal tear
Feculent material from drain
Perforated viscus
Feculent material from incision
Anastomotic leak
Extensive pneumoperitoneum
Anastomotic leak
Obstruction
Anastomotic leak
includes a diverse range of surgical operations across all specialties making the findings generalizable to a large patient population. However, our study does have important limitations. As with all retrospective studies, our study is prone to selection bias. However, we reviewed consecutive patients that underwent postoperative CT scans to limit this bias. In addition, sampling bias may have been present in our patient population. Some patients that had surgery at our institutions may have followed up with outside institutions with postoperative complications. Furthermore, because our inclusion criteria only included patients who had an abdominal operation and an abdominal CT examination within 30 PODs of the operation, we cannot determine the frequency in which patients required a reoperation and did not have a CT scan. Similarly, it is unknown if all patients who had an unusual
j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 7 ( 2 0 1 5 ) 1 0 7 e1 1 1
postoperative course underwent a CT scan; however, most patients who fail to progress postoperatively and show signs of fever, tachycardia, or leukocytosis undergo CT scan at our institution.
5.
Conclusions
In conclusion, postoperative pneumoperitoneum is a common finding on CT scans after abdominal surgery, and up to 23% of patients 3 weeks postoperatively continue to have pneumoperitoneum. However, most patients with postoperative pneumoperitoneum do not require an intervention. Better understanding the natural history of postoperative pneumoperitoneum will aid surgeons in interpreting their postoperative CT scans.
Acknowledgment Authors’ contributions: B.C.C., E.L.J., D.W., G.V.S., and T.N.R. contributed to the conception and design of study. B.C.C., K.E.M., and D.W. did the acquisition of data. B.C.C., E.L.J., G.V.S., and T.N.R. did the analysis and interpretation of data. B.C.C., K.E.M., and D.W. did the drafting of the article. B.C.C., K.E.M., E.L.J., D.W., G.V.S., and T.N.R. did the final approval.
Disclosure The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
111
references
[1] Nielsen KT, Lund L, Larsen LP, Knudsen P. Duration of postoperative pneumoperitoneum. Eur J Surg 1997;163:501. [2] Earls JP, Dachman AH, Colon E, Garrett MG, Molloy M. Prevalence and duration of postoperative pneumoperitoneum: sensitivity of CT vs left lateral decubitus radiography. AJR Am J Roentgenol 1993;161:781. [3] Gayer G, Jonas T, Apter S, Amitai M, Shabtai M, Hertz M. Postoperative pneumoperitoneum as detected by CT: prevalence, duration, and relevant factors affecting its possible significance. Abdom Imaging 2000;25:301. [4] McAllister JD, D’Altorio RA, Snyder A. CT findings after uncomplicated percutaneous laparoscopic cholecystectomy. J Comput Assist Tomogr 1991;15:770. [5] Smith KS, Wilson TC, Luces L, Stevenson AA, Hajhosseini B, Siram SM. Pneumoperitoneum 48 days after laparoscopic hysterectomy. JSLS 2013;17:661. [6] Millitz K, Moote DJ, Sparrow RK, Girotti MJ, Holliday RL, McLarty TD. Pneumoperitoneum after laparoscopic cholecystectomy: frequency and duration as seen on upright chest radiographs. AJR Am J Roentgenol 1994;164:837. [7] Bryant LR, Wiot JF, Kloecker RJ. A study of the factors affecting the incidence and duration of postoperative pneumoperitoneum. Surg Gynecol Obstet 1963;117:145. [8] Draper K, Jefson R, Jongeward R Jr, McLeod M. Duration of postlaparoscopic pneumoperitoneum. Surg Endosc 1997;11: 809. [9] Schauer PR, Page CP, Ghiatas AA, Miller JE, Schwesinger WH, Sirinek KR. Incidence and significance of subdiaphragmatic air following laparoscopic cholecystectomy. Am Surg 1997; 63:132. [10] Feingold DL, Widmann WD, Calhoun SK, et al. Persistent post-laparoscopy pneumoperitoneum. Surg Endosc 2003;17: 296. [11] Rice RP, Thompson WM, Gedgaudas RK. The diagnosis and significance of extraluminal gas in the abdomen. Radiol Clin North Am 1982;20:819.
