- To detail the 9-year experience of renal trauma at a modern Level 1 trauma centre and report on patterns of injury, management and complications.
Renal injury occurs in 1–5% of all traumas presenting to the Emergency Department [1, 2]. Renal trauma is usually classified according to mechanism of injury, although the relative preponderance of penetrating and blunt injuries varies greatly depending upon the study population. Most cases of blunt renal trauma are associated with injury of other abdominal organs, but they are commonly low-grade injuries .
Most haemodynamically stable patients with blunt renal injuries can be successfully managed conservatively , and expectant non-operative management is increasingly accepted for some penetrating injuries .
Non-operative management has been facilitated by improved staging of injury severity with advancing radiographic techniques and the development of a validated renal injury scoring system, and improved management of patients with multiple injuries within specialised trauma units. Furthermore, selective treatment with interventional radiology procedures has reduced the need for surgical intervention in patients with blunt and penetrating renal injuries [2, 6, 7].
The Alfred Hospital is one of two adult level 1 trauma centres in Victoria, Australia. In our institution, all patients who are haemodynamically stable after resuscitation are considered for expectant non-operative management regardless of injury grade or mechanism (blunt or penetrating). Patients with ongoing blood loss who do not require laparotomy for concomitant injuries are referred for angio-embolisation. Patients undergoing laparotomy without preoperative CT are explored according to established international guidelines.
We detail our 9-year experience of renal trauma at a single institution and report on patterns of injury, management and complications at a modern Level 1 trauma centre.
All patients admitted with renal trauma at the Alfred Hospital between May 2001 and January 2010 were identified using a prospectively recorded and maintained trauma database (TraumaNet) and cross referenced with data from the hospital's Health Information Service. Data on the incidence of renal injuries in Victoria was provided by the Victorian State Trauma Registry (VSTR) . This registry was commenced in 2001 and is voluntary with an opt out consent process, with the estimated case capture of >90% in December 2001, increasing over the years until reaching complete population-based coverage and capture of major trauma patients in Victoria since July 2005 . VSTR data are checked for completeness and accuracy through manual and automated processes, and checks are performed to ensure major trauma patients are captured with capture-recapture methods to cross-reference different data sources .
Data on the age, gender, type and mechanism of injury, length of stay (LOS), clinical findings on presentation (blood pressure, heart rate, Glasgow Coma Scale score, Focused Assessment with Sonography for Trauma [FAST]), injury severity score (ISS), management (laparotomy, embolisation, nephrectomy, conservative), outcome and complications were collected. Being an adult trauma centre, the population studied was primarily adults aged 16 years and above.
The imaging for all patients was retrospectively reviewed by a radiologist experienced in renal trauma who was ‘blinded’ to the original reports before reviewing the films, and graded using the American Association for the Surgery of Trauma (AAST) organ injury severity scale . Patients with haematuria who did not have signs of renal injury on imaging were excluded.
Numerical variables were compared using the Student's t-test or Kruskal–Wallis test as appropriate, and categorical variables were compared using chi-square or Fisher's exact test. All tests are two-sided with significance assumed at P < 0.05.
This study was approved by the Alfred Hospital Ethics Committee.
There were 338 patients admitted with renal injury over a 9-year period, from May 2001 until January 2010. On average 38 patients with renal injuries presented per year at our institution, ranging from 26 in 2004 to 50 in 2008. There was no consistent change in number of admissions per year over the 9-year period.
The median (IQR) LOS for patients admitted with renal injury over the 9-year period was 10.7 (5.7–19) days, with a cumulative total of 5210 hospital days for the total cohort of patients with renal trauma. There was no consistent change in median LOS over the 9-year period. Furthermore there was no correlation between LOS and the grade of renal injury.
Over the 9-year period, the median (IQR) LOS for patients treated conservatively was 10.7 (6–18.2) days), embolisation was 12.7 (10.5–23.4) days and nephrectomy was 20 (13–34) days. There was no significant change in the LOS and the type of renal management over the 9-year period.
Males comprised 74.9% (n = 253) of the total cohort. The highest incidence of renal injury was amongst those aged 20–24 years (17.8%, n = 60), after which the incidence of renal trauma generally decreased with increasing age. The median (IQR) age on admission was 32 (22–50) years.
Between 2001 and 2010, a mean (range) of 42 (29.7–65)% of renal injuries occurring in Victoria each year were treated at the Alfred . However, there was a greater preponderance of high-grade injuries, with an mean (range) of 71 (51.1 to >95)% of grade 3–5 and only 22 (10.5–30.5)% of grade 2 Victorian injuries being treated at the Alfred. There was no consistent change in the proportion of renal injuries seen at the Alfred, nor of the proportion of low- and high-grade injuries relative to the total incidence in Victoria, over the 9-year period.
The median (IQR) age was similar between males and females; males 30 (22–47) years, females 33 (25–50.5) years. However, there was a different pattern of age distribution between males and females. Males had a clear increase in incidence from 15–29 years, after which the incidence generally decreased with increasing age. A similar increase was not seen amongst females, where incidence was more consistent across the ages; however, there were too few patients in this group to comment on an overall pattern of incidence (Fig. 1).
Blunt injuries comprised 96.2% (n = 325) of all the renal injuries over the 9-year period. Of the blunt renal injuries, the predominant mechanism of injury was motor vehicle accidents, accounting for 60.9% (n = 198) of injuries, followed by; falls 11.4%, pedestrians 8.9%, motorbike accidents 4.9% and sporting injuries 7.1%. Of the sporting injuries, 19 (5.8% of blunt injuries) were incurred playing Australian Rules football. Less common mechanisms include injuries from animals 2.5%, cycling accidents 2.5% and assault 1.5% (Fig. 2).
There were 13 penetrating injuries (3.9%), comprising one gunshot injury and the remainder were stabbing injuries. One female incurred a penetrating injury (stabbing), the rest were male.
There was no consistent change in incidence of injury from blunt or penetrating mechanisms over the 9-year period. There were comparatively fewer females who incurred injuries from motorbike accidents (3.5% total injuries vs 5.1% in males) and stabbing (1.2% vs 4%), and no feamles incurred injuries from sport, assaults or gunshots.
The distribution of grade of renal injury over the 9-year period was; 21.6% grade 1 (n = 73), 24.3% grade 2 (n = 82), 24.9% grade 3 (n = 84), 16.6% grade 4 (n = 56), and 12.7% grade 5 (n = 43).
Of the 13 penetrating injuries, the gunshot trauma was a grade 1 injury, and the stabbing injuries comprised four grade 2, four grade 3, two grade 4 and two grade 5 injuries.
There was no statistically significant correlation between injury grade and the year of admission over the 9-year period (Fig. 3), nor was there a correlation in the incidence of renal injuries across Victoria over the same period.
There was a correlation between injury grade and patient age (P = 0.001), with a decreased incidence of high-grade injury with increasing age.
Females had a comparatively higher presentation of grade 5 injuries at 17.6% (15 of 85 injuries) vs 11.1% (28 of 253) for males (Fig. 4).
There was a statistical correlation (P < 0.05) between the grade of injury and the mean systolic blood pressure, haemoglobin level, FAST result, ISS and new ISS (nISS).
Of the 338 patients with renal injuries that were identified, 287 (84.9%) were treated conservatively, 25 underwent emergency nephrectomy and eight angioembolisation, of which one patient had a delayed nephrectomy after embolisation. In all, 17 patients died during the first 24 h due to significant multisystem trauma.
Of those patients who survived the first 24-h period, all patients with grade 1–2 blunt injuries were successfully managed conservatively. One patient from this group was transferred to another institution. Three of 80 patients with grade 3 blunt injuries underwent angioembolisation and one underwent nephrectomy. Of the 54 grade 4 blunt injuries, four were embolised, one underwent nephrectomy and the remaining 47 (87%) were treated conservatively. In all, 12 (29.3%) of the 41 grade 5 blunt injuries were successfully managed conservatively, one was embolised and 22 underwent nephrectomy (Table 1).
|Conservative, n||Laparotomy, n||Embolisation, n||Nephrectomy, n||Deceased <24 h, n||Total, n|
There were 158 (46.7%) patients requiring urgent surgery for non-urological trauma, involving orthopaedic, neurosurgical or general surgical procedures. Of the 25 patients who underwent nephrectomy, six (all with grade 5 injuries) proceeded directly to nephrectomy and did not require concomitant surgery for non-urological trauma. Emergency explorative laparotomy was conducted in 90 patients, of which 24 went on to have a nephrectomy and one proceeded to angioembolisation. One patient with a grade 5 injury underwent a delayed nephrectomy after failed angioembolisation. Of the patients who underwent exploratory laparotomy, there was no change in the rate of nephrectomy over the 9-year period.
Of the penetrating renal injuries, six underwent laparotomy and only one patient (with a grade 5 injury) went on to require a nephrectomy. The renal injuries in the remaining 12 patients were treated conservatively (Table 2).
|Conservative, n||Laparotomy, n||Nephrectomy, n||Total, n|
There was no significant change in the pattern of renal injury management over the 9-year period (P = 0.07; Fig. 5).
In all, 23 patients died, of which one had active management of their renal injury (discussed later). Renal injury was not attributed to be the cause of death in any of these patients. In all, 17 of these patients died within the first 24 h of admission due to significant multi-trauma. Of these 17 patients, there were three grade 1 (17.6%), one grade 2 (5.9%), five grade 3 (29.4%), two grade 4 (11.8%) and six grade 5 (35.3%) injuries. These patients had a mean (range) ISS of 56 (38–75), and nISS of 67 (50–75). Nine patients were deceased in the Emergency Department, deemed to have unsurvivable head injuries, or unable to be resuscitated after aggressive management and in some cases thoracotomy and cardiac massage in the Emergency Department. Seven patients underwent emergency laparotomy and damage control surgery, of which four were deceased intraoperatively through failure to control extensive intra-thoracic, intra-abdominal and/or head injuries.
Of the six patients who died ≤24 h of admission, there were one grade 1, three grade 2, one grade 3 and one grade 5 injuries. The patient with a grade 5 injury was a 15-year-old multi-trauma patient who underwent nephrectomy at the time of laparotomy but died 3 days later in the Intensive Care Unit (ICU). The other patients' renal injuries were managed conservatively, four of whom died at 8–11 days and one, a 74-year-old patient, died at 65 days after prolonged recovery and pseudomonas sepsis.
There was a significant difference in outcome (P ≤ 0.01) and number of hours in ICU (P = 0.034) of patients aged < 65 years, compared with those aged ≥ 65 years. In all, 18 of 300 (6%) patients aged < 65 years died, compared with five of 38 (13.2%) patients aged ≥ 65 years.
There was a correlation (P < 0.05) between survival and; age, ISS and nISS. There was no correlation between grade of injury and survival or the number of hours in the ICU.
All of the patients with penetrating injuries survived.
One patient (with a grade 1 injury) was transferred to another institution within the first week of admission, so could not be followed up. In all, 14 of the remaining 338 patients developed complications, comprising one grade 2, eight grade 3, two grade 4 and three grade 5 renal injuries, thus resulting in a complication rate of 1.2%, 9.5%, 3.6% and 8.6%, respectively.
|Renal injury grade||Renal injury management||Complication, n||Total, n|
Complications included urinoma and pseudoaneurysm formation, which were treated with urological intervention and angiography ± embolisation, respectively. Other complications included infections that were treated with antibiotics, and persistent haematuria which settled conservatively.
There was a significant difference in complication rate and grade of injury (P < 0.003), and conservative management (P = 0.004), with the lowest complication rate seen in those managed conservatively. The correlation between the complication rate and age of the patient neared statistical significance (P = 0.051). There was no correlation between complication rate and ISS.
The present series shows that conservative management is feasible even for higher grade injuries with blunt and penetrating injuries.
At our institution the decision to treat a renal injury is primarily based on haemodynamic instability whilst considering the grade of renal injury and the presence of concomitant injuries.
Over the 9-year period, conservative management was successful in all grade 1 and 2 renal injuries, and 94.9%, 90.7% and 35.1% of grade 3, 4 and 5 injuries, respectively.
The present study included 54 patients with grade 4 injuries, of which 4 (7.4%) had angioembolisation, one (1.9%) underwent nephrectomy and 90.7% were treated conservatively. This is consistent with published data [5, 14].
Whilst conservative management has been widely accepted for blunt renal injuries in haemodynamically stable patients, the management of penetrating injuries remains controversial. A non-operative approach has been advocated in selected stabbing injuries and to a lesser extent in gunshot injuries . Currently at our institution, penetrating injury does not mandate operative intervention and haemodynamically stable patients will undergo CT imaging. Patients with entry wounds anterior to the mid-axillary line undergo laparoscopy to determine whether the peritoneum has been breached, followed by laparotomy if confirmed. If there is evidence of retroperitoneal injury from an anterior penetrating injury, the retroperitoneum is explored only if there is concern of perforated viscous or there is an expanding haematoma.
At our institution over the 9-year period, 13 patients were seen with penetrating renal injuries. Seven patients underwent emergency laparotomy, of which one patient with a grade 5 injury after a stabbing went on to have a nephrectomy. The renal injuries in the remaining 12 patients were successfully managed conservatively.
There has been variable success of angio-intervention for blunt renal injuries. In one study of 434 patients with blunt renal injuries from 2002 to 2008, six (27.2%) of 22 had failed angioembolisation requiring laparotomy . Another study using the USA National Trauma Data Bank examined the use of diagnostic angiography and angioembolisation, identifying 9002 renal injuries between 2002 and 2007 . Angioembolisation was performed in 77 patients, of whom 68 required successive therapy (commonly repeat angioembolisation), including all patients with grade 4 and 5 renal injuries. The overall renal salvage rate was 92%, including 88% for grade 4 and 5 injuries. The high rate of successive therapy was not experienced in our institution, although overall there was a similar renal salvage rate. Haemostasis was successfully attained during the first procedure in the 10 patients treated with angioembolisation. This suggests that the success of angioembolisation varies amongst institutions, potentially influenced by patient selection, the level of experience of the interventional radiologist and the volume of renal trauma seen at the institution.
The impact of angioembolisation vs nephrectomy on the patient's recovery is difficult to determine in the context of trauma, where renal injury infrequently occurs in isolation. The present study found that the median (IQR) LOS for patients who underwent angioembolisation was 12.7 (10.5–23.4) days, compared with 10.7 (6–18.2) days for those treated conservatively and 20 (13–34.3) days for those who underwent nephrectomy. However, the median renal injury grades were 2, 4, and 5, and the median (IQR) ISS were 25 (16.5–36), 25 (16.3–29) and 38 (29–50) for those treated conservatively, by embolisation and nephrectomy, respectively.
Similar demographic data have been reported in other studies, with most injuries affecting males and being incurred due to blunt trauma. Although higher rates of penetrating injuries have been reported in some regions, e.g. a prospective study conducted in Pakistan between 2005 and 2007, 22% (11/50) of patients had penetrating injuries, most commonly due to firearm, reflective of the civil instability in their region .
The present study found that survival is related to age and ISS, with no correlation to the grade of renal injury. Given that renal injury commonly occurs with concomitant abdominal injuries, the latter is not surprising as ISS will more accurately reflect the patients' overall condition.
However, in an ageing society, the difference in outcomes between patients of different ages is of interest. A USA study examining outcomes in geriatric genitourinary trauma found that patients aged ≥ 65 years had more ICU admissions, longer ICU stays and twice the mortality rate of patients aged 18–65 years; however, there was no difference in mean ISS between the groups . This is consistent with the findings at our institution. Interestingly we also found a correlation between the age of the patient and the grade of injury, with increasing age being associated with lower injury grades, although no associated correlation between age and ISS. This is likely to be attributed to the different pattern of injury mechanism, in that younger people are more frequently involved in higher impact mechanisms (e.g. motor vehicle accidents), as opposed to lower impact mechanisms, e.g. low-level falls in older patients.
The lowest complication rate was seen amongst the patients managed conservatively. Of the high-grade injuries, there was a complication rate of 10 of 136 (7.4%) in the conservatively managed group vs two of eight for embolisation, and three of 24 (12.5%) for nephrectomy patients. This supports the adoption of a non-operative approach in appropriate patients, despite the presence of high-grade injuries.
The strengths of the present study are that it is a large single-centre series, and that the data were collected from a database that is recorded prospectively and maintained by dedicated data registry staff. The present study focuses on the contemporary management of renal injuries.
We accept that there are controversies on how to treat grade 4 and 5 injuries.
A significant limitation of the present study is a lack of data about long-term renal function in the patients managed conservatively. A formal assessment of creatinine clearance as part of the long-term follow-up would help to identify whether there is improved preservation of renal function in this group, and whether it is counterbalanced by an increase in incidence of post-traumatic hypertension and associated nephropathy.
There is potential selection bias in the present study. There are two adult and one paediatric Level 1 Major Trauma Services (MTS) in Victoria, and triage guidelines for rural, metropolitan and air ambulance services direct all patients with major or potentially major trauma to an MTS if one is accessible within 30 min . Non-MTS hospitals are also directed to transfer patients to an MTS if they fulfil major trauma criteria. Major trauma is defined by any of the following criteria; death due to injury, an ISS of >15, injury requiring urgent surgery, or an ICU stay of >24 h requiring mechanical ventilation. The Alfred is one of the two adult MTS in Victoria, and it receives most of the road trauma cases. As such the present results are likely to over represent the proportion of; high-grade renal injuries, multisystem injuries and road trauma.
Patients with haematuria and normal urological studies were excluded from the present study, although they meet AAST grade 1 criteria, thus further underestimating the true incidence of low-grade injuries.
Follow-up data on the effect that renal injury grade and management have on long-term complications, e.g. renal function, hypertension and chronic pain, were beyond the scope of the present study.
In conclusion, road trauma is the greatest cause of renal injury. Renal injury may be a life-threatening event but if managed correctly can be managed safely without the need for nephrectomy in most cases. Blunt trauma accounts for most renal injuries, of which a greater proportion are a less severe injury grade.
We would like to acknowledge the medical and administrative staff that maintain the TraumaNet database.
American Association for the Surgery of Trauma
Focused Assessment with Sonography for Trauma
Intensive Care Unit
(new) injury severity score
length of stay
Victorian State Trauma Registry