Predictive Factors of Splenic Injury in Colorectal Surgery: Data From the Nationwide Inpatient Sample, 2006-2008

Objectives To determine frequency of splenic injury and to evaluate predictive risk factors of splenic injury during colorectal surgery.

Design Retrospective database analysis.

Setting The National Inpatient Sample database.

Patients Patients who underwent a colorectal resection during the period from 2006 to 2008 in the United States.

Main Outcome Measures Patient characteristics, patient comorbidities, type of pathology, type of resection, surgical technique used, type of admission, and teaching hospital status were evaluated for splenic injury during colorectal surgery.

Results A total of 975 825 patients underwent colorectal resection during this period. Overall, the rate of splenic injury was 0.96%, of which 84.75% were treated with complete splenectomy (splenorrhaphy, 13.55%; partial splenectomy, 1.70%). The most common procedure associated with splenic injury was transverse colectomy (3.40%). Using multivariate regression analysis, we found that transverse colectomy (adjusted odds ratio [AOR], 5.30), left colectomy (AOR, 5.08), total colectomy (AOR, 2.85), open operation (AOR, 2.68), malignant tumor (AOR, 2.11), diverticulitis (AOR, 1.93), teaching hospital (AOR, 1.73), male sex (AOR 1.20), peripheral vascular disease (AOR, 1.14), and emergent admission (AOR, 1.06) were associated with a higher risk of splenic injury. There was no association between age, race, hypertension, diabetes, chronic lung disease, congestive heart failure, renal failure, liver disease, obesity, sigmoidectomy, proctectomy, ulcerative colitis, or Crohn disease and splenic injury.

Conclusions Type of resection (transverse, total, or left colectomy), type of pathology (malignancy or diverticulitis), open operation, and teaching hospital are potent independent predictors of splenic injury. Male sex, peripheral vascular disease, and emergent admission are less effective predictors. Surgeons should be aware of these risk factors and inform patients accordingly. In higher-risk circumstances, it may be appropriate to consider prophylactic vaccination.

Splenic injury is a major intraoperative complication of abdominal surgery and places patients at a higher risk of morbidity and mortality, of longer operating time, and of longer hospital stay.^1,2 Splenic injuries may be managed nonoperatively or by splenorrhaphy, partial splenectomy, or complete splenectomy.³ Because splenectomy is reported to have higher mortality rates than splenorrhaphy,⁴ great effort is taken to preserve the spleen; however, excessive blood loss can mandate splenectomy.⁵Quiz Ref IDStudies have documented capsular tears to be the most common type of splenic injury. Other frequently occurring splenic injuries include lacerations, avulsions, and subcapsular hematomas.^6-11 The highest rate of iatrogenic splenic injury is attributed to colonic surgery; this has been attributed to the close vicinity of the spleen and the colonic splenic flexure.^4,12

Few studies have been conducted that examine splenic injury during colorectal surgery.^13,14 Understanding the risk factors associated with splenic injury in colorectal surgery may help the surgical team in instituting risk-reduction strategies and properly informing patients of the risks prior to surgery. Because the rate of splenic injury after colorectal surgery is relatively low,^8,13-15 an analysis of the factors that predict splenic injury will require a large series of patients that is not frequently available from most single institutional case series. In our study, we used the Nationwide Inpatient Sample (NIS) database to (1) determine the prevalence of splenic injury in colorectal surgery and (2) evaluate the effect of patient characteristics, patient comorbidities, type of pathology, type of resection, surgical technique used, type of admission, and teaching hospital status on splenic injury in colorectal surgery.

The NIS is the largest inpatient care database in the United States in which approximately 1000 hospitals participate; it contains information from nearly 8 million hospital stays each year across the country. The NIS consists of a nationally representative sample of approximately 20% of US community hospitals, resulting in a sampling frame that comprises approximately 90% of all hospital discharges in the United States. Data elements within the NIS are drawn from hospital discharge abstracts that allow determination of all procedures performed during a given hospital admission. It also contains discharge information on inpatient hospital stay, including patient characteristics, length of stay, specific postoperative morbidity, and observed in-hospital mortality. The NIS database has no information available on complications occurring after hospital discharge. Not all data elements in the NIS are provided by each state's data source. These data elements are provided in the NIS because they can be valuable for research purposes, but they should be used cautiously. For example, data on race are missing for a number of states; therefore, race-specific estimates may be biased. Approval for use of the NIS patient-level data in our study was obtained from the institutional review board of the University of California, Irvine Medical Center and the NIS.

Using the NIS database, we analyzed discharge data on patients who underwent elective/emergent laparoscopic or open colorectal resection from 2006 to 2008. However, patients who underwent colorectal surgery without colon or rectal resection (eg, ostomy closure) were excluded form our study. We used International Classification of Diseases, Ninth Revision (ICD-9) procedure codes and principal diagnosis codes to identify the type of colorectal resection performed and the pathologic reason for surgery (Table 1). To identify splenic injuries during colorectal surgery, we used ICD-9 procedure codes (not as a principal procedure) for complete splenectomy (code 41.5), partial splenectomy (code 41.43), and splenorrhaphy (codes 41.95 and 41.99). Also, we identified procedures that were performed laparoscopically by specific laparoscopic procedure codes or by identifying additional ICD-9 procedure codes for diagnostic laparoscopy (code 54.21) or laparoscopic lysis of adhesion (code 54.51) in combination with open procedure codes.

The preoperative factors that were analyzed included patient characteristics (age, sex, and race), patient comorbidities, type of admission (emergent or elective), pathologic conditions (malignancy, diverticulitis, ulcerative colitis, Crohn disease, or benign disease), type of surgery (right colon, transverse colon, left colon, sigmoid, or rectum), surgical technique used (laparoscopic vs open), and teaching status of hospital. The overall rate of splenic injury following laparoscopic and open colorectal resection and the rate of splenic injury by surgical site, type of pathology, and type of admission were examined. Because the NIS database only contains procedure codes, without any information regarding indications, we assumed that all splenic surgeries were performed for incidental splenic injuries.

Univariate and multivariate logistic regression analyses were performed, and the adjusted odds ratio (AOR) with a 95% CI was calculated to determine the combined effect of various preoperative factors on splenic injury. All statistical analyses for the NIS database were conducted using SAS version 9.2 (SAS institute), incorporating recommended discharge and hospital weights. Discharge weight was used to create national estimates for all analysis. Statistical significance was set at P < .05 (with ORs and 95% CIs that excluded 1).

Among the 118 877 284 reported admissions during this 3-year period in the United States, a total of 975 825 patients underwent colon or rectal resection. The majority of patients were female (53.0%) and white (79.01%). The mean age was 62.46 years, with 49.80% of patients older than 65 years (Table 2). The most prevalent comorbidities were hypertension (45.40%), chronic lung disease (15.77%), and diabetes mellitus (14.24%). The most common indication for colorectal resection was colorectal malignancy (34.73%). The majority of procedures were performed in nonteaching hospitals (52.0%), and 58.0% of these patients were admitted electively. Overall, 7.37% of all procedures were performed laparoscopically.

Quiz Ref IDThe overall rate of splenic injury was 0.96% (0.72% elective surgery vs 1.28% emergent surgery; P < .01); 84.75% of these splenic injuries were treated with complete splenectomy (splenorrhaphy, 13.55%; partial splenectomy, 1.70%).Table 3 shows splenic injury rates for different types of surgery and for different pathologies. Quiz Ref IDConsidering procedure type, the highest rate of splenic injury was observed in transverse colectomy (3.40%). Considering pathologic conditions, the highest rate of splenic injury was observed in patients with a malignant tumor (0.55%). The rate of splenic injury was lower following laparoscopic procedures than following open procedures (0.30% vs 1.01%; P < .01). Also, univariate regression analysis (Table 4) showed that patients who underwent an open procedure had a 3.41 times higher chance of having a splenic injury compared with patients who underwent a laparoscopic procedure (OR, 3.41 [95% CI, 2.98-3.91]; P < .01).

Tables 4 and 5 present the data on the univariate and multivariate logistic regression analyses for in-hospital splenic injury in colorectal surgery. Multivariate regression analysis (risk adjusted) showed that, for patients' characteristics, the factor associated with a higher rate of splenic injury was male sex (AOR, 1.20). For comorbidities, only peripheral vascular disease (AOR, 1.14) was an independent factor associated with an increased rate of splenic injury. For procedure type, transverse colectomy (AOR, 5.30), left colectomy (AOR, 5.08), and total colectomy (AOR, 2.85) had independently increased rates of splenic injury compared with right colectomy. Considering pathologic conditions, malignant tumor (AOR, 2.11) and diverticulitis (AOR, 1.93) were independently associated with a higher rate of splenic injury. Quiz Ref IDAlso, open procedure (AOR, 2.68), emergent admission (AOR, 1.06), and teaching hospitals (AOR, 1.73) were associated with higher rates of splenic injury in this patient population. There was no association between age, race, hypertension, diabetes, chronic lung disease, congestive heart failure, renal failure, liver disease, obesity, sigmoidectomy, proctectomy, ulcerative colitis, or Crohn disease and splenic injury.

To our knowledge, this study is the largest retrospective analysis to evaluate the incidence of splenic injuries during colorectal surgery using a large administrative database. We found that the incidence of splenic injuries was 0.96%. Previous studies have documented the incidence of splenic injury to occur in 0.5% to 8% of all colonic procedures^8,13-15 and were based on retrospective analyses of single hospital experiences. Although studies have shown the advantages of splenorrhaphy compared with complete splenectomy,⁵ our study shows that the majority of splenic injuries were treated with complete splenectomy (85%). Using multivariate regression analysis, we found several factors that were identified as being significantly associated with a higher rate of splenic injury in colorectal resection. Resection type (transverse, total, or left colectomy), open operation, type of pathology (malignancy or diverticulitis), teaching hospital, male sex, peripheral vascular disease, and emergent admission were factors associated with a higher rate of splenic injury. Knowledge of these risk factors will help surgeons in their decision-making process and in properly informing patients regarding their risks.

According to our findings, the location of the procedure also influenced the frequency of iatrogenic splenic injury. Although isolated transverse colectomy is performed so infrequently (3.77% of colorectal operations), it had the highest rate of iatrogenic splenic injury (AOR, 5.30). Also, the second most performed operation associated with a higher risk of splenic injury was left colectomy (AOR, 5.08). This may be related to the proximity and mobilization of the splenic flexure. Similarly, McGory et al,² in evaluating 41 999 (non-T4) colorectal cancers using the California Cancer Registry, found that tumor locations from the transverse colon to the rectosigmoid significantly increased the odds of inadvertent splenectomy. In a single-institution case-matched study of 118 patients who underwent colon surgery (59 patients with or without splenic injury), Wang et al¹⁴ used multivariate regression analysis and found that only splenic flexure mobilization was independently associated with an increased risk of splenic injury. We have observed that other procedures that involve splenic flexure mobilization also are associated with an increased rate of splenic injury; transverse colectomy, left colectomy, and total colectomy were associated with significantly high rates of splenic injury. Also, we found that the type of pathology significantly influenced the rate of splenic injury. Malignant tumor and diverticulitis were independent risk factors of splenic injury. An extensive resection because of a malignant tumor might have increased the risk of injury to the spleen. Wang et al¹⁴ found that cancer and diverticulitis were the main diagnoses associated with splenic injury; however, the difference was not significant when compared with the control group with no splenic injury.

Quiz Ref IDWhen examining the type of procedure, we found that laparoscopic colectomy was associated with fewer splenic injuries than was open colectomy; risk-adjusted analysis indicated that open cases had almost a 3-fold higher chance of splenic injury than did laparoscopic cases (OR, 2.68). Laparoscopic procedures might have a lower splenic injury rate owing to better visualization, precise movements, and the natural tendency to place less tension on the colon with laparoscopic retraction. Our finding is in agreement with previous studies^13,14 that have also reported fewer splenic injuries with laparoscopic colectomy. In fact, Malek et al¹⁶ even reported that there were no injuries for 1911 cases of laparoscopic colectomy.

In general, patients' comorbidities and characteristics did not correlate with the rate of splenic injury; however, our study showed that male sex (AOR, 1.20) and peripheral vascular disease (AOR, 1.14) were associated with a slightly increased risk for splenic injury. Although univariate regression analyses showed that congestive heart failure, chronic renal failure, and liver disease were associated with a higher risk of splenic injury (Table 4), multivariate analysis did not show these comorbidities to be independent risk factors for splenic injury (Table 5). Unlike our findings, previous studies^2,17-19 have also found age to be a risk factor for splenic injury. This difference might be due to the increased fragility of the spleen and to the strong retraction of the left costal margin caused by decreased rib elasticity.¹⁹ Interestingly, we found that obesity was not an important factor in determining splenic injury during colorectal surgery. This contrasts with previous findings by Liberman and Welch,²⁰ who found that more than half of the 176 patients with iatrogenic splenic injury in their study were obese. The change in findings might be due to improved surgical equipment and navigation in obese patients.

Teaching hospital (AOR, 1.73) was associated with a higher risk of splenic injury, although this finding is more likely related to the complexity of the disease than to the complexity of the procedure, to the advanced stage of the disease, or to the inexperience of the trainees involved with the operations found in teaching hospitals. The NIS is limited in determining the difference in disease severity and surgeon's experience between teaching and nonteaching hospitals. Interestingly, emergent surgery had a lesser effect on splenic injury than we expected (AOR, 1.06). Conversely, other studies^13,14 have shown emergent indications associated with increased splenic injury and even increased mortality. In a 1992 to 2007 retrospective review of 13 897 colectomies, Holubar et al¹³ reported 59 splenic injuries (0.42%); they found that emergent surgery was independently related to lower long-term survival (hazard ratio, 2.75; P = .009) because 22 cases resulted from emergent surgery and because 50% these emergent cases resulted in death within 30 days or in major complications.

There are limitations to our study that are similar to the limitations of other studies making use of a large administrative database. The NIS database has no information about indication for procedures; therefore, patients who underwent therapeutic splenic surgery were included as having an incidental splenic injury. For example, if a patient was scheduled for a multivisceral resection that included splenectomy, this would appear in our study as a “splenic injury.” Additionally, the NIS database has no information about previous history of abdominal surgery, which could be an important factor in iatrogenic injury. Lastly, dedicated laparoscopic colectomy ICD-9 codes were not available for all types of procedures, which may have led to inaccuracies in the technique of procedure. Additionally, selection bias may have accounted for the difference between open and laparoscopic operation outcomes, and we were also unable to determine the converted cases. Despite these limitations, to our knowledge, this study is the largest to date, with a large number of patients to evaluate factors predictive of splenic injury in colorectal surgery.

In conclusion, our study, like other studies, has shown that the rate of splenic injury during a colorectal resection is relatively low (<1%). We identified multiple risk factors for splenic injury during colorectal surgery. For example, we found that the rate of splenic injury is lower in laparoscopic as opposed to open operations. Our better understanding of these high-risk groups can help in (1) facilitating a more comprehensive preoperative discussion with patients concerning their risks, (2) preparing for possible splenic injury in high-risk patients, and (3) considering prophylactic vaccination in high-risk circumstances such as open transverse colectomy for a malignant tumor.

Correspondence: Michael J. Stamos, MD, Department of Surgery, University of California, Irvine Medical Center, 333 City Blvd W, Ste 700, Orange, CA 92868 (mstamos@uci.edu).

Accepted for Publication: September 16, 2011.

Published Online: December 19, 2011. doi:10.1001/archsurg.2011.1010

Author Contributions:Study concept and design: Masoomi, Carmichael, and Stamos. Acquisition of data: Masoomi. Analysis and interpretation of data: Masoomi, Carmichael, Mills, Ketana, Dolich, and Stamos. Drafting of the manuscript: Masoomi and Ketana. Critical revision of the manuscript for important intellectual content: Masoomi, Carmichael, Mills, Dolich, and Stamos. Statistical analysis: Masoomi. Administrative, technical, and material support: Masoomi and Ketana. Study supervision: Carmichael, Mills, Dolich, and Stamos.