Journal of the New Zealand Medical Association, 10-September-2004, Vol 117 No 1201
Impact of concomitant trauma in the management of blunt splenic injuries
Albert Lo, Anne-Marie Matheson, Dave Adams
Major changes have occurred in the management of blunt splenic injuries in the previous 50 years. The concept of universal splenectomy for splenic injuries remained unchallenged until King and Shumacker described (in 1951) life-threatening infections in paediatric patients after splenectomy.1
The incidence of overwhelming post-splenectomy infection (OPSI) following splenectomy in the trauma setting is reported to be 0.5%, and can reach up to 20% in patients with haematological disorders.2 The concept of conservative management of splenic injuries has become increasingly accepted by trauma surgeons, especially in the past two decades because of the development of modern imaging techniques.
However, in the presence of a tender, distended abdomen and other concomitant injuries, a rapid deterioration in a patient's haemodynamic condition presents a major challenge for clinicians trying to pinpoint the source of bleeding. These patients are often taken to theatre for exploratory laparotomy after positive findings on ultrasound (FAST) scan or diagnostic peritoneal lavage (DPL). Alternatively, if the patient's initial haemodynamic status is stable enough to allow for abdominal computed tomography (CT) imaging which shows splenic injuries, clinicians are faced with the dilemma of whether to operate or to monitor these patients with multisystem injuries, whose potential to withstand subsequent major haemorrhage may be compromised.
Conservative management remains the cornerstone of treating haemodynamically stable blunt splenic injuries. This study is undertaken to review our cumulative experience of non-operative management and to identify the risk factors for operative management in patients with blunt splenic injuries.
Patients and Methods
From September 1997 to February 2003, retrospective analysis of medical records of patients with blunt splenic injury treated in Middlemore Hospital, Auckland was carried out.
Data collected included patient demographics, mechanisms of injury, vital signs, blood results on admission, imaging and other diagnostic modalities, transfusion requirement, operative findings, postoperative course, duration of hospital stay, and follow-up. Details of immunisation such as Pneumovax®, Haemophilus influenzae type B vaccine, meningococcal vaccine, and further use of CT scans in re-assessment of the progress of splenic injuries were also recorded.
For each patient, an injury severity score3,4 (ISS) and grading of severity of splenic injury by CT scan (based on the organ injury scale [OIS])5 were calculated. The ISS is an anatomical scoring system that provides an overall score derived by combining squared abbreviated injury scores3 from the three most severely injured body regions.
Patient management was categorised as:
Statistical analysis of the data was performed with the Student's t test for age, ISS, and chi-squared test for comparison of parameters between patient groups.
Eighty patients were identified. The overall mean age at presentation was 34.2 years (range 2 to 84 years). There were 52 males and 28 females patients (male:female ratio of 1.86:1). Seven of the patients were ≤15 years old.
As illustrated in Figure 1, there were 35 patients (24 male, 11 female) in the operative management (OM) group; 43 patients (27 male, 16 female) in the non-operative management (NOM) group; and 2 patients (1 male, 1 female) who required splenectomy after failure of conservative treatment (F/NOM).
The mean age of patients was 34.0 in the OM group and 34.3 in the NOM and F/NOM group. Stated ethnic origins were European: 54%; Maori: 23%; Pacific Island: 11%; and other ethnic groups (including Chinese and Indian) in the remainder.
Mechanism of injuries
Motor vehicle accident (MVA) was the most common (involving 57 patients) mechanism of injury, while 10 patients sustained splenic injuries from sports activities; 8 from falls, and 5 from various other mechanisms (eg, assault, crush injuries).
Figure 1. Distribution of patients
Among the patients involved in MVA, 38 patients were either passengers or drivers in a vehicle, while the remaining patients were pedestrians (n=9), motorcyclists (n=9), and a cyclist (n=1).
Pedestrians were the most severely injured group, with moderate-to-severe head injuries being noted in 78% of this group (Table 1). A higher proportion of pedestrians required operative management, although this did not reach statistical significance.
Table 1. Mechanism of injury in motor vehicle accident (MVA) patients
GCS=Glasgow coma score; ISS=injury severity score; *p<0.05
Utility of CT scans on initial assessment of splenic injuries
All patients (100%) in the NOM group had a CT scan to determine the extent of splenic injuries; while only 48.6% of patients in the OM group had a CT scan. The remaining patients in the OM group were haemodynamically unstable after initial fluid resuscitation. One patient (an 8-year-old male pedestrian hit by a car) was transferred to the operating theatre on clinical grounds alone, while the other subjects had positive findings either on diagnostic peritoneal lavage (34.3%) or focused abdominal sonography for trauma (FAST) exams (14.3%).
Half of those patients with positive diagnostic peritoneal lavage (DPL) were found to have other associated intra-abdominal injuries (eg, liver laceration, mesenteric tear, greater omentum tear).
The distribution of the severity of splenic injury (as determined by CT scan) for the operative management group and conservative treatment group is demonstrated in Figure 2. There is a trend towards increased severity of splenic injuries in the operative management group. For those patients who had successful conservative management of their grade 4 and 5 splenic injuries, associated injuries were limited to thoracic injuries such as rib fractures and haemopneumothorax.
Figure 2. Distribution of severity of splenic injuries by CT grading
OM=operative management; NOM=non-operative management.
Non-operative management (NOM)
Forty-five patients (56%) were initially treated non-operatively, 43 (54%) of these had successful NOM. This is a 96% success rate of non-operative management. Thirteen patients (29% in the NOM group) were initially admitted to the intensive care unit (ICU) for observation, incurring an average length of stay of 2 days (range 1–11 days). There were no deaths in the NOM group.
Four patients (9%) required epidural anaesthesia for pain relief. Red blood cell (RBC) transfusion was given to three NOM patients (average: 3 units). Repeat CT scans (during both inpatient stay and outpatient follow-up) to examine the extent and progress of splenic injuries were performed on five patients (11% in the NOM group).
Operative management (OM)
A total of 28 patients (35% splenectomy rate) had splenectomy (26 of the 35 patients in the OM group and 2 patients in the F/NOM group). The remaining 9 patients in the OM group did not have splenic injuries, which were severe enough to warrant splenectomy. A variety of splenic conservation techniques were performed in these patients including Surgicel® wrap, suturing, and partial excision.
Five out of 9 patients (in OM group) who did not have splenectomy were found to have other intra-abdominal injuries such as small bowel mesenteric tear, liver lacerations, and ruptured bladder which required simultaneous surgical interventions. The mean operative time was 81 minutes.
Twenty-five patients (71.4% in the OM group) required intra-operative blood transfusion, with an average of 6.8 units transfused. Twenty-one patients (60% in the OM group) were admitted into ICU postoperatively for an average of 3 days.
There were no deaths in the NOM group including those two patients with failure of NOM. However, there were 3 intra-operative deaths and seven postoperative deaths. Two patients succumbed to uncontrolled intra-abdominal haemorrhage intra-operatively, and the remaining death was due to the severity of combined intra-abdominal and thoracic injuries (haemopneumothorax and haemopericardium).
Causes of postoperative deaths included extensive brain injury sustained during trauma (three patients), coagulopathy (three patients) and multi-organ failure (one patient). This gives a mortality rate of 28.6% in the OM group (12.5% overall mortality rate).
Failure of non-operative management (NOM)
There were two patients in the F/NOM group. The first patient had re-bleeding from her splenic lacerations 5 days after admission to the general surgical ward. On admission, her GCS was 15, ISS: 13, and haemoglobin: 123 g/L. She had a grade II splenic injury on CT. However, on day 5, she became hypotensive (despite fluid resuscitation) and was found to have a reduced haemoglobin (93 g/L). She was taken to the operating theatre for emergency splenectomy.
The second patient was a 35-year-old female who was anticoagulated with warfarin for Factor V Leiden and previous deep vein thromboses (DVTs) (INR=2.2 on admission). She sustained right radius and left humerus fractures in a motor vehicle accident. She was haemodynamically stable with a haemoglobin level of 143 g/L. CT abdomen on admission did not reveal any splenic injury or other evidence of intra-abdominal bleeding.
Eight days post-injury she became hypotensive with signs of abdominal peritonism and her haemoglobin level dropped to 82 g/L. During emergency laparotomy, it was found that there was a small splenic capsular tear with 2 litres of free blood in the peritoneal cavity. Splenectomy was carried out.
Both patients had an uneventful recovery period. Due to the small number of patients, meaningful analysis of predictors for failure of conservative treatment was not possible. Neither of the patients in the F/NOM group had a repeat CT abdomen prior to emergency splenectomy.
Operative versus non-operative management
Table 2 highlights the differences between those patients who were initially planned for conservative treatment (NOM and F/NOM) and those who had emergency surgery (OM). Thirty-five patients were taken to the operating theatre for emergency laparotomy. The mean ISS of this OM group was 30, significantly higher (p<0.05) than conservative treatment groups, 13.
Table 2. Comparison of patient groups
OM=operative management; NOM=non-operative management; F/NOM=failure of NOM, requiring splenectomy; GCS=Glasgow coma score; ISS=injury severity score; BP=blood pressure; ED=emergency department; *p<0.05.
Put another way, the proportion of patients with severe injuries (as defined by ISS ≥16) was 89% in the OM group compared with 34% in the NOM and F/NOM group. The respective differences between these groups with regards to systolic hypotension (BP <100 mmHg): (37% vs 2%), severity of head injury (GCS ≤13): (40% vs 9%) and requirement of blood transfusion in ED: (40% vs 9%), were also statistically significant (p<0.05).
All the patients who had splenectomy received pneumococcal vaccine. In addition, all except three of these patients were immunised against Haemophilus influenzae type b and meningococcal infections. Twelve patients in the NOM group received pneumococcal vaccine, the majority of whom had high grade (III-V) splenic injuries on their CT scans.
Algorithms delineating the initial management of splenic injury recommend either observation or operation as acceptable initial therapeutic options in patients who are appropriately selected.6
A variety of operative techniques for salvaging injured spleens have been described, and splenorrhaphy is an accepted alternative to splenectomy when clinically and technically feasible.7 Splenectomy is increasingly reserved for the conditions of haemodynamic instability or anatomic injury beyond repair. In this study series, both patients in the F/NOM group had splenectomy because of haemodynamic instability. Their splenic injuries did not preclude repair (grade II injury and capsular tear respectively), but rather their unstable haemodynamic conditions mandated the procedure.
We attempted non-operative management in 56% of patients, and achieved 96% success rate (54% overall rate of NOM). Two large series corroborate these results. Hunt et al,8 in a state-wide analysis of 2258 patients, found that conservative management rate increased from 33.9% to 46.3% over a 5-year period, with a success rate of 94%. A large, single institution study by Pachter et al9 of 190 consecutive patients reported 65% initial rate of NOM and 98% success rate. In an Australasia setting, the reported initial rate of NOM ranged from 46% to 75%.10–12
Patients with blunt splenic injuries often have a variety of concomitant injuries (eg, head injuries, orthopaedic injuries, or other intra-abdominal injuries). There is no specific rigid treatment protocol for treating splenic injuries in our institution. Management is individualised for each patient and based on multiple variables. For persistent haemodynamic instability (or unresolved concerns about other concomitant injuries), operative management is the preferred treatment option. The mean ISS of OM group in our study series was 30, compared to 13 in the conservative treatment group.
Within the limitations of a retrospective study, ISS ≥16, hypotension, GCS ≤13, and requirement for blood transfusion were found to be significantly associated with operative management. Furthermore, a study of 226 patients by Shapiro et al over a 5.5-year period also identified ISS, GCS score, revised trauma score (RTS), and amount of RBC transfused as risk factors for patients sustaining blunt splenic injuries requiring immediate surgery.13
These observations may lead to the suggestion that the responsible trauma surgeons have a selection bias for treating multi-system trauma patients operatively in a bid to decrease the proportion of patients likely to fail non-operative management without compromising the overall rate of successful non-operative management. Indeed, this problem of selecting appropriate treatment option for multi-system trauma patients will continue to exist in the absence of concrete, reliable predictors for failure of non-operative management. Age >55, degree of haemoperitoneum, ISS >15, and American Association for the Surgery of Trauma (AAST) injury scale of grade >III on CT scans are suggested by some authors to be contraindications for non-operative management of blunt splenic injuries.14,15 However, none of these predictors have proven to be completely reliable.
In this study, 9 patients (out of 35) in the operative group did not have splenic injuries which were severe enough to warrant splenectomy—but 5 of these were found to have other intra-abdominal injuries and were treated accordingly. One may argue that the remaining 4 patients (11.4% in OM group) should be given a trial of conservative therapy. However, their haemodynamic instability after initial fluid resuscitation, in whom concomitant intra-abdominal injuries cannot be definitely ruled out, mandated exploratory laparotomy.
The use of CT imaging has become an integral tool for the management of splenic injuries in patients who are haemodynamically stable. Several grading systems for the severity of splenic injuries are available.10,11,16,17 We used the OIS (organ injury score) because of its compatibility with the calculation of the ISS. Sugrue et al12 suggested that CT changes are predictive of outcome for non-operative treatment and that patients with more severe splenic injuries on CT grading should undergo early surgical intervention regardless of the cardiovascular status of the patient. On the other hand, Pachter et al9 reported their algorithmic approach to stable patients which includes a period of nonoperative management for all stable patients, regardless of the grade of injury.
Shapiro et al13 concluded that CT imaging underestimate the severity of injury, especially in the subset of patients who failed non-operative management because of progression of injury and continued bleeding. In the present series, there were 6 patients in the non-operative group with grade 4 and 5 splenic injuries on CT scans with successful conservative treatment.
Although CT imaging is a useful tool, the reported accuracy in determining the severity of splenic injury is variable.18 Therefore, we believe decision-making should not be dictated solely by the CT findings.
Both non-operative (NOM) and operative management (OM) have a role in the current treatment of blunt splenic injuries. The results of this study suggest that non-operative management of appropriately selected patients can be achieved with acceptable outcome. In the presence of concomitant trauma, patients with blunt splenic injuries are more likely to require operative management if they have ISS ≥16, hypotension, GCS ≤13, and requirement for blood transfusion.
Author information: Albert Lo, House Surgeon; Anne-Marie Matheson, Trauma Co-ordinator; Dave Adams, General and Vascular Surgeon, Department of General Surgery, Middlemore Hospital, Otahuhu, Auckland
Correspondence: Mr Dave Adams, Department of General Surgery, Middlemore Hospital, Otahuhu, Auckland. Fax: (09) 623 6451; email: DaveA@middlemore.co.nz
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals