BACKGROUND Selective nonoperative management (SNOM) of blunt kidney injuries has been the standard of care for decades. However, the role of SNOM after renal gunshot wounds (GSWs) remains unclear. The purpose of this study was to assess the safety and outcomes of SNOM of renal GSWs at a national level.
METHODS The National Trauma Data Bank was queried for patients who sustained a GSW to the kidney (January 2007 to December 2014). Patients with emergency department death, transfer, nonsurvivable (Abbreviated Injury Scale score = 6) injuries, absent vitals on arrival, associated hollow viscus or major abdominal vascular injury, or missing procedure/discharge data were excluded. Study groups were defined according to management strategy, with operative management (OM) defined as laparotomy 4 hours or less from admission and SNOM defined by the lack of laparotomy 4 hours or less from admission. Demographics, injury data, and outcomes (mortality, complications, need for nephrectomy, failure of SNOM [defined as laparotomy 5–72 hours after admission], and hospital length of stay [LOS]) were collected. Logistic regression compared outcomes between SNOM and OM.
RESULTS Over the study period, 1,329 patients met inclusion and exclusion criteria. Of these, 459 (34.5%) underwent SNOM (38.6% of American Association for the Surgery of Trauma kidney injury grades I–II, 39.4% of grade III, 24.0% of grade IV, and 5.2% of grade V). SNOM was associated with significantly shorter hospital LOS (6 days vs. 9 days, p < 0.001). Failure of SNOM occurred in 10.2%. Logistic regression showed no association between SNOM and mortality (odds ratio [OR], 0.614, p = 0.244). However, SNOM was independently associated with fewer complications (OR, 0.711; p = 0.008) and a reduction in need for nephrectomy (OR, 0.056; p < 0.001).
CONCLUSION SNOM of GSWs to the kidney has become a common practice in the United States. It is safe and has a high success rate, even in severe kidney injuries. It decreases hospital LOS and is independently associated with fewer complications and reduced need for nephrectomy.
LEVEL OF EVIDENCE Therapeutic/Care Management, level IV.
Selective nonoperative management (SNOM) is a well-established treatment option for blunt abdominal trauma that has also gained support for patients with penetrating abdominal trauma. With SNOM, patients with hemodynamic instability, peritonitis, or an unevaluable abdomen are brought for immediate laparotomy while the absence of these criteria qualify a patient for CT scan and potential SNOM with a period of observation. However, the role of nonoperative management after gunshot wounds (GSWs) to abdominal solid organs has remained a controversial issue. Many trauma surgeons have been reluctant to adopt SNOM in these patients, despite encouraging results from trauma centers who have adopted this approach.1–4
SNOM of renal GSWs was first introduced in 1998.5 Since that time, the practice gained some acceptance, although many surgeons remain reluctant.6–9 The acceptance of SNOM for these injuries on a national level and the effect on outcomes are currently unknown. The purpose of this study was to define the national use of SNOM for renal GSWs over time and to compare outcomes after SNOM versus operative management (OM) of GSWs to the kidney. Our hypothesis was that SNOM of renal GSWs would gain acceptance over time and be associated with improved outcomes.
METHODS
After obtaining approval from the institutional review board of the University of Southern California, the American College of Surgeons National Trauma Data Bank (NTDB) was used to identify all patients from January 2007 to December 2014 who sustained a GSW to the kidney, based on mechanism of injury and Abbreviated Injury Scale [AIS] pre-dot codes. The NTDB is the largest American trauma registry in existence and is based on clinical data collected from participating trauma centers in the United States by the American College of Surgeons Committee on Trauma.10 Patients with associated hollow viscus or major abdominal vascular injuries, those who died in the emergency department (ED) or arrived without vital signs, those with AIS score of 6 in any body region, those who were transferred to or from an outside hospital, and patients with missing procedure codes, procedure times, or discharge disposition were excluded.
Patient demographics (age, sex), admission year, clinical data (systolic blood pressure [SBP], heart rate [HR], and Glasgow Coma Scale [GCS] score upon arrival to the ED), injury data (American Association for the Surgery of Trauma [AAST] grade of renal injury; AIS scores for the head, chest, and abdomen; associated intra-abdominal solid organ injuries), operative procedures (type and timing from admission), the use of angioembolization (defined by procedure codes 88.45, 39.77, and 39.79), and outcomes were collected. The primary outcome was in-hospital mortality. Secondary outcomes included 24-hour mortality, in-hospital complications captured by NTDB coding (herein referred to as complications), failure of SNOM, need for nephrectomy, and hospital length of stay (LOS).
The study groups were defined based on management strategy. Based on the existing literature,11 patients who underwent laparotomy 4 hours or less from hospital admission were defined as the OM group. The SNOM group was defined by patients who did not undergo laparotomy 4 hours or less from admission. Failure of SNOM was considered as the need for laparotomy between 5 hours and 72 hours after hospital admission.
The trend in use of SNOM versus OM was plotted over time, with significance determined using linear-by-linear association. Univariate analyses of demographics, clinical/injury data, and outcomes were performed using Pearson's χ2 test, Fisher's exact test, or Mann-Whitney U test, as appropriate. Categorical variables are expressed as number (%) and continuous variables as median (interquartile range [IQR]).
Independent risk factors for mortality, complications, failure of SNOM, and nephrectomy were identified using logistic regression, adjusting for age, sex, hypotension (SBP <90 mm Hg), tachycardia (HR >120 bpm), GCS score less than 9, year of injury, AAST grade of kidney injury, AIS score greater than 3 in the head or chest, level of trauma center (Level I vs. Levels II–IV), management strategy (SNOM or OM), and associated solid organ injury (liver or spleen). Covariates for the logistic regression models were selected a priori on the basis of the existing literature and expert consensus. Covariates with p < 0.2 were retained in the models to identify potential independent predictive variables. Multicollinearity testing was performed by examination of the variance inflation factor to ensure a lack of correlation between variables. A cutoff of variance inflation factor greater than 2.5 was utilized to conservatively define collinearity, although no collinear variables were identified in this study. Accuracy was examined with area under receiver operating characteristic (AUROC) curve with 95% confidence interval (CI). A p-value of <0.05 was considered statistically significant. Analysis was performed using SPSS Statistics version 20 (IBM Corporation, Chicago, IL).
RESULTS
A total of 1,329 patients from 301 centers were included in the study (Fig. 1). The maximum patient contribution from a single center was 3% (n = 42). Overall, 459 (34.5%) underwent SNOM and 870 (65.5%) underwent OM for a renal GSW (Table 1). The rate of SNOM increased significantly over the study period (p = 0.033), from 34.5% in 2007 to 40.0% in 2014 (Fig. 2). According to AAST grade of kidney injury, SNOM was used in 38.6% of grades I to II injuries, 39.4% of grade III injuries, 24.0% of grade IV injuries, and 5.2% of grade V injuries.
The patient demographics and injury characteristics between patients undergoing SNOM versus OM are detailed in Table 1. Patients undergoing SNOM were significantly less likely to be hypotensive (SBP, <90 mm Hg) on arrival (5.5% vs. 11.2%, p < 0.001). Additionally, grade V injuries were significantly less likely to be managed with SNOM than OM (5.2% vs. 94.8%, p < 0.001). Patients undergoing OM had a higher abdominal AIS score than those undergoing SNOM (3 [IQR, 2–4] vs. 2 [IQR, 2–3], p < 0.001), with a greater proportion of patients with abdominal AIS score greater than 3 (36.6% vs. 20.7%, p < 0.001). Rates of angioembolization did not differ between OM and SNOM groups (2.4% vs. 3.7%, p = 0.180). On univariate analysis of outcomes, patients managed with SNOM had significantly fewer complications (40.3% vs. 53.0%, p < 0.001), lower rates of in-hospital and 24-hour mortality (3.3% vs. 5.9%, p = 0.038 and 1.7% vs. 4.6%, p = 0.008), and shorter hospital LOS (6 days vs. 9 days, p < 0.001) (Table 2).
Multivariate analyses were performed, assessing independent predictors of mortality (Table 3), complications (Table 4), need for nephrectomy (Table 5), and failure of SNOM (Table 6) after GSW to the kidney. Independent risk factors for mortality after GSW to the kidney were older age (odds ratio [OR], 1.046, p < 0.001), hypotension (OR, 3.105, p = 0.003), tachycardia (OR, 4.506, p < 0.001), GCS score less than 9 (OR, 8.347, p < 0.001), and chest AIS score greater than 3 (OR, 2.586, p = 0.013). SNOM was not associated with mortality (OR, 0.614, p = 0.244) (Table 3).
Independent risk factors for complications after renal GSWs were older age (OR, 1.011, p = 0.042), hypotension (OR, 2.077, p = 0.001), tachycardia (OR, 1.468, p = 0.035), GCS score less than 9 (OR, 2.188, p = 0.005), and earlier year of injury (OR, 0.854, p < 0.001). Selective nonoperative management was independently associated with fewer complications (OR, 0.711, p = 0.008) (Table 4). Independent predictors of nephrectomy were hypotension (OR, 2.923, p < 0.001), tachycardia (OR, 1.979, p = 0.014), grades IV–V renal injuries (OR, 6.828, p < 0.001; and OR, 22.491, p < 0.001; respectively), and head AIS score greater than 3 (OR, 4.394, p = 0.040). Selective nonoperative management was independently associated with reduced need for nephrectomy (OR, 0.056, p < 0.001) (Table 5).
Among the 459 patients who underwent SNOM, 47 patients (10.2%) failed. Of these, 7 (14.9%) presented with unstable vital signs (defined by SBP <90 mm Hg or HR >120 bpm). If these patients had been managed operatively from the outset, the failure rate of SNOM would have been reduced to 8.7%. On multivariate analysis, failure of SNOM was independently predicted by GCS score less than 9 (OR, 6.356, p = 0.003), earlier study year (OR, 1.299, p = 0.002), and associated liver (OR, 2.266, p = 0.042) or spleen (OR, 10.399, p < 0.001) injuries (Table 6). Patients who failed SNOM were compared against patients managed with OM from the outset to establish if any adverse sequelae were associated with the delay in operation. The median time to laparotomy among patients who failed SNOM was 13 hours (IQR, 6–36), compared with 1 hour (IQR, 1–2) in the OM group (p < 0.001). There was no difference in the rate of angioembolization between patients who failed SNOM and those who were managed with OM (n = 2, 4.3% vs. n = 21, 2.4%; p = 0.332). There were no differences between patients who failed SNOM and those who underwent OM in terms of overall complications (53.2% vs. 53.0%, p = 0.978); mortality, either in-hospital (4.3% vs. 5.9%, p = 1.000) or within 24 hours (0% vs. 4.6%, p = 0.259); or hospital LOS (9 days vs. 9 days, p = 0.717).
There were two patients who failed SNOM and died, both with grade II kidney injuries. A 48-year-old man died on hospital day 8 with an ISS of 29 after arriving hemodynamically normal with a GCS score of 15. He underwent laparotomy 12 hours after admission. In addition to the grade II kidney injury, his injuries included a pulmonary laceration, hemopneumothorax, grade II splenic injury, and a grade V thoracic spine fracture and cord injury. A 51-year-old man died on hospital day 5 with an ISS of 34. He arrived hypotensive, tachycardic, and with a GCS score of 3. He underwent laparotomy 11 hours after admission. His injuries included a grade V liver injury and hemopneumothorax in addition to the grade II renal injury. In summary, both of the patients who died following failed SNOM of renal GSWs had significant associated injuries in the presence of a low grade renal injury.
To examine the effect of failing SNOM on the need for nephrectomy, nephrectomy rates after SNOM versus OM were examined. The nephrectomy rate after failure of SNOM was lower than after OM, although this difference was not significant (12.8% vs. 20.7%, p = 0.188).
Subgroup analysis by AAST grade of renal injury was performed in order to examine the effect of injury grade on the association between SNOM and outcomes. Multivariate analysis with logistic regression revealed that the association between SNOM and outcomes did not vary by AAST grade of renal injury (p > 0.05), although analyses of moderate and high grade injuries (AAST grades III–V) were limited by small study size.
DISCUSSION
The decision to manage a patient nonoperatively after a renal GSW should be made on the basis of the clinical examination and CT scan findings. Only patients with hemodynamic stability and no signs of peritonitis should be considered for CT scan evaluation. A CT scan with intravenous contrast can provide valuable information about the presence or absence of other associated injuries, the extent of the renal injury, and the presence of active bleeding, false aneurysms, or arteriovenous fistulae which could benefit from angioembolization and increase the chance of kidney salvage.
Renal injuries are more amenable to nonoperative management than other intra-abdominal solid organ injuries, in part because of the tamponade effect on bleeding imparted by Gerota's fascia and the retroperitoneum. Routine exploration of the injured kidney may result in unnecessary nephrectomy, because of the loss of this tamponade effect.5,12 In addition, the rich renal blood supply may promote healing after even severe parenchymal injuries.
In the present study, after controlling for potential confounders, SNOM was found to be independently associated with reduced need for nephrectomy and fewer complications. Selective nonoperative management was also associated with shorter hospital LOS as compared to OM. The lower incidence of complications in patients managed nonoperatively is not surprising, considering that even with a negative laparotomy for trauma, the incidence of abdominal complications is high.13,14 In addition, the risk of complications increases if the kidney exploration results in nephrectomy. A significant concern in the nonoperative management of a renal GSW is the potential adverse effect of delaying a needed operation. This study showed that the small group of patients who failed SNOM had comparable outcomes to those who were managed operatively from the outset, and therefore there was no apparent consequence to the delay in operation.
There are limitations to this study which must be acknowledged. First, it is a registry-based study with the associated inherent limitations of this type of study design. Second, NTDB angioembolization coding is relatively nonspecific and it is possible that a patient underwent angioembolization of an associated injury and not the renal injury itself. Therefore, the use of angioembolization after renal GSWs is not well delineated by this study. The subspecialization of the operating surgeons is not detailed by the NTDB, and thus, the involvement of urologists, which is likely to vary between centers, is not defined by this study. It is also possible that the laterality of renal injuries impacts outcomes as a result of associated injuries, but the NTDB is insufficiently granular to capture these details. The NTDB also does not capture specific complications which can be important in kidney injuries, such as urinoma or pseudoaneurysm. Because these complications may be more frequent after SNOM than OM, this study may underestimate in-hospital complications following SNOM for GSWs to the kidney. Additionally, the NTDB does not have long term follow up available, and therefore delayed complications after either management strategy, such as hypertension, false aneurysms, renal failure, adhesive small bowel obstruction, and incisional hernia, are not captured. Finally, the possibility of selection bias between the two study groups must be considered. Although attempts were made to control for potential confounders, it is possible that there are inherent differences between patients managed with SNOM versus OM that have not been accounted for. This may have an effect on outcomes.
It must be emphasized that SNOM for GSWs to the kidney should be considered only in carefully selected patients who are hemodynamically stable, have no signs of peritonitis, and in whom the CT scan does not show any suspicion of other significant intraabdominal injuries requiring operative intervention. To perform SNOM, the hospital should have in-house surgical coverage for continuous and close clinical observation. Ideally, the center should also have the capacity for angioembolization. Early recognition of failed SNOM and timely surgical intervention for these patients are essential in preventing adverse effects.
In summary, this study provides a large, contemporary analysis of SNOM of GSWs to the kidney at the national level. Selective nonoperative management is safe and effective across all grades of renal injury. It is associated with significantly lower need for nephrectomy, reduced rates of complications, and shorter hospital LOS. Finally, when patients fail SNOM, renal salvage rates remain higher and outcomes are no worse than if the patients had undergone OM from the outset. These data suggest that, in the right clinical setting, SNOM of renal GSWs is safe and may improve outcomes and renal salvage rates.
AUTHORSHIP
M.S., E.B., and D.D. provided the study concept. M.S., E.B., and A.P. performed the data collection. M.S., E.B., and A.P. performed the data analysis. M.S., E.B., K.I., and D.D. performed the data interpretation. All authors participated in writing and critically reviewing the final article.
DISCLOSURE
The authors have no conflicts of interest or disclosures of funding to declare.
REFERENCES
1. Berg RJ, Inaba K, Okoye O, Pasley J, Teixeira PG, Esparza M, Demetriades D. The contemporary management of penetrating splenic injury. Injury. 2014;45:1394–1400.
2. DuBose J, Inaba K, Teixeira PG, Pepe A, Dunham MB, McKenney M. Selective non-operative management of solid organ injury following abdominal gunshot wounds. Injury. 2007;38:1084–1090.
3. Demetriades D, Hadjizacharia P, Constantinou C, Brown C, Inaba K, Rhee P, Salim A. Selective nonoperative management of penetrating abdominal solid organ injuries. Ann Surg. 2006;244:620–628.
4. Omoshoro-Jones JA, Nicol AJ, Navsaria PH, Zellweger R, Krige JE, Kahn DH. Selective non-operative management of liver gunshot injuries. Br J Surg. 2005;92:890–895.
5. Velmahos GC, Demetriades D, Cornwell EE III, Belzberg H, Murray J, Asensio J, Berne TV. Selective management of renal gunshot wounds. Br J Surg. 1998;58:523–525.
6. Serafetinides E, Kitrey ND, Djakovic N, Kuehhas FE, Lumen N, Sharma DM, Summerton DJ. Review of the current management of upper urinary tract injuries by the EAU Trauma Guidelines Panel. Eur Urol. 2015;67(5):930–936.
7. Shariat SF, Jenkins A, Roehrborn CG, Karam JA, Stage KH, Karakiewicz PI. Features and outcomes of patients with grade IV renal injury. BJU Int. 2008;102:728–733.
8. Buckley JC, McAninch JW. Selective management of isolated and nonisolated grade IV renal injuries. J Urol. 2006;176:2498–2502.
9. Nicol AJ, Theunissen D. Renal salvage in penetrating kidney injuries: a prospective analysis. J Trauma. 2002;53:351–353.
10. American College of Surgeons. National Trauma Data Bank. 2019. Available at https://www.ntdbdatacenter.com/ [accessed 7 Jan 2019].
11. Zafar SN, Rushing A, Haut ER, Kisat MT, Villegas CV, Chi A, Stevens K, Efron DT, Zafar H, Haider AH. Outcome of selective non-operative management of penetrating abdominal injuries from the North American National Trauma Database. Br J Surg. 2012;(99 Suppl 1):155–164.
12. Velmahos GC, Degiannis E. The management of urinary tract injuries after gunshot wounds of the anterior and posterior abdomen. Injury. 1997;28(8):535–538.
13. Morrison JJ, Poon H, Garner J, Midwinter MJ, Jansen JO. Nontherapeutic laparotomy in combat casualties. J Trauma Acute Care Surg. 2012;73(6 Suppl 5):S479–S482.
14. Renz BM, Feliciano DV. Unnecessary laparotomies for trauma: a prospective study of morbidity. J Trauma. 1995;38:350–356.
Keywords:
Selective nonoperative management; penetrating abdominal trauma; gunshot wounds; renal trauma; kidney injuries
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου