Impact of severe splenic trauma and its management on the occurrence of infections in intensive care unit

Abstract

Purpose

Severe trauma can be responsible for a major systemic inflammatory response followed by post-traumatic immunosuppression. Immune imbalance promotes infections and increases mortality. This risk could be further increased in case of impaired splenic immune function due to splenic trauma. The objective of this study was to evaluate whether severe splenic trauma increased infectious complications in trauma patients during intensive care unit (ICU) stay.

Methods

This was an observational, bicentric, retrospective, case–control study including patients admitted for severe trauma from January 2011 to December 2020 in two level 1 trauma centers. Patients with American Association for the Surgery of Trauma (AAST) ≥III splenic trauma (case) were matched (1:3) with patients without splenic trauma (controls) according to age, sex, Injury Severity Score, initial Glasgow Coma Score, and Simplified Acute Physiology Score II. Demographic, trauma management, and infection data were collected. The primary endpoint was the incidence density of any infectious disease during ICU stay.

Results

Among 7,304 severe trauma patients, 130 patients with AAST ≥III splenic trauma were included. 10 patients could not be matched with controls. 17 patients (14.1%) underwent splenectomy, 56 patients (46.7%) had non-operative management with arterioembolization, and 47 (39.2%) had non-operative management without embolization. There was no difference between cases and controls regarding incidence density of infections (44.1 (34.6–55.5)/1,000 person days vs 43.6 (37.7–50.2)/1,000 person days, incidence rate ratio=1 95% CI (0.77 to 1.3) p=0.94), type of infection, involved microorganisms, or severity (septic shock 12% vs 9.2%, p=0.6; acute respiratory distress syndrome 14% vs 9.2%, p=0.2).

Conclusion

In the present study, AAST ≥III splenic trauma in severe trauma patients was not associated with an increased risk of infection during ICU stay.

Level of evidence III

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Severe trauma induces immunosuppression due to systemic inflammation. As the spleen plays a key role in immune function, high-grade splenic trauma (American Association for the Surgery of Trauma (AAST) ≥III) may exacerbate this effect and increase the risk of intensive care unit (ICU)-acquired infections.

WHAT THIS STUDY ADDS

• High grade of splenic trauma (AAST ≥III) in the context of severe trauma is not associated with a higher risk of ICU infections. The choice of management strategy (splenectomy, non-operative management, or embolization) does not significantly impact infection risk in the ICU.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These findings suggest that routine antibiotic prophylaxis for splenic trauma in the ICU may not be necessary.

Background

Among severe trauma, spleen injury is one of the most frequent injuries and accounts for 30–43% of abdominal trauma.^{1 2} Currently, non-operative management (NOM), with or without splenic arterial embolization (SAE), is the strategy of choice except for uncontrolled hemorrhagic shock in which case hemostatic splenectomy remains the therapy of choice. Apart from operation, SAE as well as trauma itself may reduce functional spleen volume and thereby impair its immune function. Recent data suggest that long-term splenic function remains after SAE. However, this issue is still under discussion and no data exist regarding the risk of early infection in patients admitted with spleen trauma. The spleen is involved in immune response through complement activation, antigen presentation, immunoglobulin production and memory B lymphocyte storage.³ Severe trauma is associated with an increased risk of infection and mortality⁴ due to post-traumatic immunosuppression.^{5 6} Consequently, the summation of post-traumatic immunosuppression and immune dysfunction after spleen trauma with or without SAE could increase the rate of early infections.

This study aimed to evaluate the incidence of infection during intensive care unit (ICU) stay regarding the presence or absence of spleen trauma (defined as an American Association for the Surgery of Trauma (AAST) grade ≥III). The secondary objective was to evaluate the impact of therapeutic strategy (NOM with or without arterial embolization or splenectomy) on the incidence of infections during ICU stay.

Methods

Study design

This study is a retrospective case-control analysis using data from the French trauma registry, TraumaBase. It adheres to the STROBE (STrengthening the Reporting of OBservational studies in Epidemiology) guidelines for observational cohorts (online supplemental file 1).

In Paris area, all patients suspected of severe trauma are necessarily cared for by a physician-lead enhanced care team in a mobile ICU and directly transported to a level 1 trauma center.⁷ All consecutive patients with trauma and admitted to either Beaujon Hospital or European Georges Pompidou Hospital (Level 1 academic trauma centers) between January 2011 and December 2020 were screened for inclusion. Patients who died or were discharged from the ICU within 48 hours of admission were excluded.

Case subjects were characterized by an AAST grade III or higher splenic trauma, as analyzed by initial multiphasic whole-body CT scan interpreted by local senior radiologists and independently reviewed by two experienced abdominal radiologists for the purpose of this study. The surgeon assessed the AAST gradation during the operation of patients undergoing exploratory laparotomy for hemostatic splenectomy. Control subjects were patients admitted in one of the two-participant trauma centers and without any splenic trauma.

We matched one case with three controls based on the following variables: sex (strictly matched), age (within a 10-year range), initial Glasgow Coma Score (GCS) (categorized as GCS <8, GCS 8–12, GCS >12), Injury Severity Score (ISS) (within a 4-point range), and Simplified Acute Physiology Score II (SAPS II) (within a 10-point range). These criteria were defined on the basis of data in the literature⁴ which highlight a decrease in infectious risk by female gender (OR 0.63 (0.49–0.86) p=0.003), whereas alteration of initial GCS and increase in ISS increase infectious risk (respectively, taking as reference GCS >13, GCS <8 OR 6.71 (3.52–12.81) p<0.001; GCS 8–12 OR 3.62 (2.39–5.5) p<0.001 and per ISS point OR 3. 52 (2.41–5.35) p<0.001).

Data collection

Patient data were primarily sourced from a comprehensive, prospective trauma registry and supplemented with information from medical files when necessary. For each included patient, the following variables were documented: age, gender, comorbidities, prehospital characteristics (including accident details and initial clinical data), arrival characteristics (including clinical data and Extended Focused Assessment with Sonography in Trauma results). Additional metrics included hemorrhagic shock (defined as transfusion of at least four blood products during the first 6 hours of admission), ICU length of stay and mortality rate.

Data regarding infectious complications were collected, detailing timing, location, microbial ecology and severity of the event. SAPS II,⁸ Sepsis-related Organ Failure Assessment (SOFA)⁹ and ISS¹⁰ were duly recorded.

Spleen infarction zone on initial CT scans and the type of arterioembolization (proximal or distal) were independently reviewed by two experienced abdominal radiologists.

Outcome measures and definitions

The primary outcome was the incidence density of infection during ICU stay, calculated as the number of infections per 1,000 ICU patient-days. The secondary outcome was the incidence of infectious complications during ICU stay. Each infection counted for one event. Diagnosis of infection was based on a combination of clinical symptoms and microbiological positive culture.^{11 12} Severity of infectious episodes was assessed by reviewing the presence of septic shock, defined according to the third international consensus definitions for sepsis and septic shock,¹³ and acute respiratory distress syndrome (ARDS) according to Berlin criteria.¹⁴

The impact of the initial therapeutic intervention (ie, NOM without SAE, NOM with SAE, splenectomy) on the occurrence of infectious complications was also assessed. Of note, if the treatment strategy changed during follow-up, the patient was recategorized into the appropriate subgroup as of the date the intervention changed.

Splenic trauma management in both trauma centers

As recommended by the French or the World Society of Emergency Surgery guidelines,¹⁵ management of spleen trauma depended on the hemodynamic status of the patient and the AAST grade of the lesion. NOM was the cornerstone of treatment for patients with hemodynamic stability. SAE completed the initial management after multidisciplinary discussion if required (arterial blush, severity of spleen trauma). Patients with hemodynamic instability and intra-abdominal free fluid on sonography underwent emergent laparotomy with splenectomy. According to the French guidelines, only patients with severe splenic trauma that required splenectomy received antibiotic prophylaxis. This prophylaxis was started 24–48 hour before ICU discharge.

Statistical analysis

Categorical variables were described by frequency and percentage, and compared using either the χ² or Fisher’s exact test, as appropriate. Quantitative variables were described using median values and IQRs, and compared using the Wilcoxon test. Negative binomial regression models were performed with calculation of the incidence rate ratios (IRR) along with their 95% CIs. The models were adjusted for variables showing a significant difference between the two populations in the univariate analysis (p<0.05) or those deemed clinically relevant. The time-to-first infection and overall survival rate were compared using log-rank tests.

A subgroup analysis of the impact of splenic trauma management was performed using the Kruskal-Wallis test, Pearson’s χ² test and a univariate negative binomial regression model. All tests were two-tailed and values of p<0.05 were considered significant. Analyses were performed with R Studio (V.4.2).

Results

Patient’s characteristics

Between January 2011 and December 2020, 8,506 patients were screened, 130 patients with severe splenic trauma were included but 10 cases (8%) could not be matched. 120 cases and 347 controls were finally analyzed (figure 1).

Figure 1. Flow chart of TraumaBase admissions and splenic trauma. AAST, American Association for the Surgery of Trauma; NOM, non-operative management; SAE, splenic arterial embolization.

Patients were young (31 (22–46) and 31 (25–45) years respectively; p=0.4), mostly men (82% in both groups) and with few comorbidity (American Society of Anesthesiologists (ASA) score >3 1.7% vs 2%, p>0.9). Road accidents were more frequent in the splenic trauma group (79% vs 56%, p<0.001) whereas more falls were observed in the control group (16% vs 35%, p<0.001). Patient’s severity was similar except for the incidence of hemoperitoneum and hemorrhagic shock (higher in the spleen trauma group) and for the head and neck Abbreviated Injury Scale (AIS) (higher in the non-splenic trauma group). There was no difference in length of stay in ICU between the two groups (7 days (5–17) vs 7 days (4–8), p=0.02). Mortality was lower in the splenic trauma group (table 1).

Table 1. Characteristics of patients with and without splenic trauma.

	Splenic trauman=120	Absence of splenic injuryn=347	P value
Demographics
Age (years)	31 (22–46)	31 (25–44.5)	0.4
Male sex	98 (82)	284 (82)	>0.9
BMI (kg/m²)	24 (21.8–26.8)	24.5 (22.1–26.6)	0.4
Medical history and treatments
ASA score ≥3	2 (1.7)	7 (2)	>0.9
Treated diabetes	2 (1.7)	12 (3.5)	0.5
Neoplasia	0 (0)	2 (0.6)	>0.9
Immunosuppression	0	3	0.6
Prehospital
Glasgow Coma Score	15 (13–15)	15 (13–15)	0.6
Orotracheal intubation	35 (29)	105 (30)	>0.9
Vasopressor support	22 (18)	51 (15)	0.4
Cardiorespiratory arrest	2 (1.7)	11 (3.2)	0.5
Resuscitation room
On arrival
Glasgow Coma Score	15 (15–15)	15 (14–15)	0.2
Mean arterial pressure (mm Hg)	86 (73–99)	89 (76–103)	0.08
Positive fast-echo (hemoperitoneum)	62 (62)	37 (13)	<0.001
Hemorrhagic shock	25 (21)	41 (12)	0.02
Lactate (mmol/L)	2.5 (1.5–3.6)	2.3 (1.5–3.6)	0.6
Care in the resuscitation room
Cardiorespiratory arrest	0 (0)	4 (1.2)	0.6
Vasopressor dose before departure if hemorrhagic shock (mg/hour)	4.7 (3–6.7)	4.9 (3.2–6.2)	0.8
Direct block or arterioembolization	8 (6.7)	7 (2)	0.029
Accidentology
Road accident	95 (79)	194 (56)	<0.001
Motorbike	56 (47)	98 (28)	<0.001
Light/heavy vehicle	25 (21)	51 (15)	0.2
Pedestrian	7 (5.8)	34 (9.8)	0.3
Bike	5 (4.2)	6 (1.7)	0.2
Other	2 (1.7)	5 (1.4)	>0.9
Fall	19 (16)	120 (35)	<0.001
Blunt object	3 (2.5)	12 (3.5)	0.8
Penetrating weapon	1 (0.8)	9 (2.6)	0.5
Ballistic trauma	1 (0.8)	5 (1.4)	<0.9
Unknown	1 (0.8)	7 (2)	0.7
Trauma assessment
Head and neck AIS	0 (0–2)	2 (0–4)	<0.001
Face AIS	0 (0–0)	0 (0–1)	0.13
Chest AIS	3 (2–3)	3 (0–3)	0.3
Abdominal AIS	3 (3–4)	0 (0–2)	<0.001
Spine and pelvis AIS	2 (0–2)	2 (0–3)	0.02
Peripheral AIS	0 (0–0)	0 (0–0)	0.5
Severity scores
ISS	27 (21.5–34)	26 (20–34)	0.6
SAPS II	27 (16–39.2)	28 (17–39)	>0.9
SOFA	4 (1–9)	4 (1–9)	0.7
Characteristic of hospitalization
Death	2 (1.7)	28 (8.1)	0.014
WWLST	3 (2.5)	15 (4.3)	0.6
Length of stay in ICU (days)	7 (5–17)	7 (4–8)	0.2
Length of hospital stay (days)	14.5 (10–28.2)	19 (10–39)	0.2
Orotracheal intubation	65 (54.2)	247 (71.2)	<0.001
Mechanical ventilation duration (days)	7 (2–14)	4 (2–14)	0.2

Values expressed as n (%) or median (IQR 0.25–0.75).

AIS, Abbreviated Injury Scale; ASA, American Society of Anesthesiologists; BMI, body mass index; ISS, Injury Severity Score; SAPS II, Simplified Acute Physiology Score II; SOFA, Sepsis-related Organ Failure Assessment; WWLST, withdrawing or withholding of life-sustaining treatment.

Risk of ICU infections in patients with splenic trauma

The incidence density of ICU infections (from any site) did not differ between patients with splenic trauma and those without before (44.1 (34.6–55.5) vs 43.6 (37.7–50.2) infections per 1,000 ICU patient days, p=0.94) and after adjustment for age, head and neck AIS, and diabetes (IRRa 1 (0.79 to 1.4), p=0.77). (Table 2). Neither did the rate of patients who developed at least one episode of infection during their ICU stay (41 (34%) vs 114 (33%), p=0.8). Median time to first infection was 4 days (2–6) for cases versus 5 days (3–7) for controls (p=0.1) and the proportion of patients with more than one infection during their stay in ICU was also similar (17 patients in the case group (14.2%) vs 55 patients in the control group (15.9%), p=0.7). Furthermore, there was no difference in the distribution of causative bacteria between cases and controls (figure 2). The proportion of patients with a severe complication such as ARDS or septic shock did not differ between the two groups (respectively 14% vs 9.2%, p=0.2, and 12% vs 9.2%, p=0.6).

Table 2. Incidence density of infections in the ICU.

	Splenic trauman=120	Absence of splenic injuryn=347	Incidence rate ratio(95% CI)	P value	Adjusted incidence rate ratio(95% CI)	P value
Incidence density of infections	44.1 (34.6–55.5)	43.6 (37.7–50.2)	1 (0.77 to 1.3)	0.94	1 (0.79 to 1.4)	0.77
Of which respiratory	31.4 (23.5–41.2)	32.7 (27.6–38.5)	0.95 (0.69 to 1.3)	0.77	1 (0.75 to 1.4)	0.77
Of which urinary	2.4 (0.7–6.2)	2 (0.9–3.8)	1.5 (0.39 to 4.5)	0.53	0.69 (0.15 to 3.2)	0.63
Of which abdominal	3.6 (1.3–7.9)	1.8 (0.8–3.5)	0.78 (0.26 to 2.2)	0.65	0.84 (0.21 to 2.9)	0.8
Of which cerebral and meningeal	0 (0–2.2)	0.2 (0–1.2)	–	–	–	–
Of which surgical site infection	3.6 (1.3–7.9)	4.2 (2.5–6.6)	0.69 (0.25 to 1.6)	0.43	0.92 (0.29 to 2.6)	0.88
Of which on equipment	3 (1–7.1)	2.4 (1.2–4.4)	0.69 (0.22 to 1.9)	0.49	0.4 (0.08 to 1.5)	0.2

Values expressed as incidence (95% CI).

The incidence density of infections corresponds to the number of infections per 1,000 ICU patient days.

Adjustment is made on age, head and neck AIS and diabetes.

AIS, Abbreviated Injury Scale; ICU, intensive care unit.

Figure 2. Distribution of germs responsible for infections in the intensive care unit.

Factors associated with infection after spleen trauma were diabetes, ISS, SAPS II and SOFA scores and the presence of coma and hemorrhagic shock (table 3).

Table 3. Factors associated with the occurrence of infections in patients with splenic trauma.

	Incidence rate ratio (95% CI)	P value
Center (Beaujon reference)	1.3 (0.58 to 2.7)	0.47
ASA score ≥3	3.7 (0.85 to 14)	0.03
Age (≥30 years old)	1.1 (0.6 to 1.9)	0.8
Sex (male)	2.2 (0.87 to 6.2)	0.1
Treated diabetes	3.6 (1.2 to 10)	0.01
ISS	1 (1 to 1.1)	<0.001
SAPS II	1 (1 to 1)	<0.001
SOFA	1.2 (1.1 to 1.3)	<0.001
Hemorrhagic shock	2.2 (1.2 to 3.9)	0.004
Prehospital Glasgow	0.94 (0.88 to 1)	0.02
Grade AAST
AAST III	–	–
AAST IV (reference AAST III)	0.69 (0.37 to 1.3)	0.2
AAST V (reference AAST III)	0.64 (0.22 to 1.7)	0.34
NOM without SAE	–	–
NOM with SAE	0.75 (0.39 to 1.5)	0.36
Proximal arterial embolization	0.78 (0.39 to 1.6)	0.44
Distal arterial embolization	0.72 (0.25 to 1.9)	0.51
Splenectomised	1.4 (0.64 to 2.9)	0.37
Spleen infarction zone
<25%	–	–
25–50% (reference <25%)	0.72 (0.29 to 1.6)	0.42
50–75% (reference <25%)	0.19 (0.01 to 1)	0.11
>75% (reference <25%)	0.4 (0.02 to 2.4)	0.39

AAST, American Association for the Surgery of Trauma; ASA, American Society of Anesthesiologists; ISS, Injury Severity Score; NOM, non-operative management; SAE, splenic arterial embolization; SAPS II, Simplified Acute Physiology Score II; SOFA, Sepsis-related Organ Failure Assessment.

Influence of management strategy of splenic trauma on the risk of ICU infections

Among the 120 patients with spleen trauma, 47 (39.2%) received NOM without SAE, 56 (46.7%) NOM with SAE and 17 (14.1%) required a splenectomy (figure 1). Initial strategy of NOM without SAE failed in 14 patients (29.8%): 13 patients (27.7%) required embolization and 1 patient (2.1%) splenectomy. In addition, two patients (3.6%) patients from the NOM with SAE group required splenectomy.

There was no difference in infection incidence density in splenic trauma patients with regards to management strategy (49 (31.4–72.9) infections per 1,000 ICU patient days in patients with NOM without SAE, 37 (24.8–53.1) infections per 1,000 ICU patient days after NOM with SAE, and 52.5 (32.1–81.1) infections per 1,000 ICU patient days after splenectomy, p=0,82) (table 4). The rate of patients with at least one infection during ICU stay did not differ according to initial management with respectively 16 patients (34%) in NOM without SAE and 17 patients (26%) after NOM with SAE and 9 patients (47%) after splenectomy (p=0.2).

Table 4. Incidence density of infections according to the management of splenic trauma.

	NOM without SAEn=47	NOM with SAEn=66	Splenectomyn=19	Incidence rate ratio (95% CI) NOM with vs without SAE	P value	Incidence rate ratio (95% CI) splenectomy vs NOM without SAE	P value
Incidence density of infections	49 (31.4–72.9)	37 (24.8–53.1)	52.5 (32.1–81.1)	0.76 (0.4 to 1.3)	0.31	1.1 (0.6 to 1.9)	0.82
Of which respiratory	38.8 (23.3–60.6)	28.1 (17.6–42.5)	28.9 (14.4–51.7)	0.72 (0.39 to 1.3)	0.3	0.74 (0.34 to 1.5)	0.44
Of which urinary	4.1 (0.5–14.7)	2.6 (0.3–9.2)	0 (0–9.7)	0.62 (0.08 to 5.2)	0.64	–	–
Of which abdominal	4.1 (0.5–14.7)	0 (0–4.7)	10.5 (2.9–26.9)	–	–	2.6 (0.5 to 19)	0.28
Of which cerebral and meningeal	–	–	–	–	–	–	–
Of which surgical site infection	2 (0.1–11.4)	2.6 (0.3–9.2)	7.9 (1.6–23)	1.2 (0.12 to 27)	0.86	3.9 (0.49 to 78)	0.24
Of which on equipment	0 (0–7.5)	3.8 (0.8–11.2)	5.2 (0.6–19)	–	–	–	–

Values expressed as incidence (95% CI).

The incidence density of infections corresponds to the number of infections per 1,000 patient days of resuscitation with the treatment (NOM without or with SAE or splenectomy).

If an infection occurs after the change of therapy, the infection is counted in the group corresponding to this new therapy.

NOM, non-operative management; SAE, splenic arterial embolization.

Discussion

In this retrospective bicentric case-control study, severe splenic trauma (AAST ≥III) was not associated with an increased risk of infectious diseases during ICU stay. Furthermore, this risk did not vary according to spleen trauma management strategy (ie, NOM with or without arterial embolization or splenectomy).

To our knowledge, our study is the first one to examine whether splenic trauma and its management modify the occurrence of infection during the early phase of trauma ICU stay. Although the long-term risk of infection after splenectomy is well known,9,11 few studies have examined the risk of early infections after splenic trauma.

We first hypothesized that the consequences of spleen injury and its treatment on adaptive immunity would occur in the short term and increase the risk of infection in the ICU, a critical phase for post-traumatic immunosuppression. Post-traumatic immunosuppression is due to a systemic inflammatory response syndrome secondary to the release of Damage Associated Molecular Patterns (DAMPS) and reflects an uncontrolled systemic innate immune response caused by the emission of large amounts of DAMPs. This initial intense inflammatory response is followed by a counterbalancing compensatory anti-inflammatory response syndrome.^{5 6} The resultant is associated with an increased risk of inflammation-related complications such as infections, contributing to increased morbidity and mortality.⁴ In this study, the focus on a time window in which traumatic immunosuppression is at its peak (even for non-splenic trauma patients) and hyposplenism at its start might explain why no difference in infection incidence was found. In that the exposure period was relatively short (median 1 week in both groups). Another explanation could be the important level of residual splenic function in this study. In a recent study, Arvieux et al¹⁶ suggested a threshold of 50% residual parenchyma was to be considered critical. In our study, the size of splenic infarction was not associated with the risk of infection but only 14 non-splenectomised patients had an infarction zone above 50%. In the end, early adaptive immunity appears to be preserved in this work, as the predominant microorganisms isolated were gram-negative bacilli (non-encapsulated bacteria). In our cohort, patients with severe splenic trauma who did not require splenectomy did not receive antibiotic prophylaxis during ICU stay. In the splenectomy group, antibiotic prophylaxis was only started 24 hours to 48 hours before ICU discharge. Considering that cases and controls have the same rate of infectious complications and the same type of microorganisms, this study suggests that there is no need for prophylaxis in patient with severe splenic trauma during the ICU stay.

In a study of 155 patients with blunt splenic trauma, Gauer et al¹⁷ found a higher rate of early infectious complications in the emergent splenectomy group compared with patients treated by NOM but there was no control group without splenic trauma. However, we did not find any infectious risk associated with splenectomy (IRR 1.4 (0.64 to 2.9), p=0.37) (table 3) maybe because as discussed above, traumatic immunosuppression overwhelmed the impact of spleen trauma on immune function.

Currently, many patients with blunt splenic injury are considered for NOM. The development of this strategy for the treatment of splenic trauma reduced the need for urgent splenectomies and thereby the risk of post-splenectomy infection. However, to what extent the spleen immune function is impaired after NOM is unclear. A review by Schimmer et al¹⁸ included 12 studies and found no immune impairment after NOM even with SAE. However, this review focused on long-term outcomes. In our study, among patients with AAST ≥III splenic trauma, the rate of patients with at least one infection during ICU stay was the same regardless of initial management: 16 patients (34%) in case of NOM without SAE, 17 patients (26%) after NOM with SAE and 9 patients (47%) after splenectomy (p=0.2). Our results are consistent with the study of Arvieux et al¹⁶ who observed no difference in splenic infarction volume between NOM with or without SAE. Thus, NOM with SAE appears to be a safe and effective technique without any excess infectious rate at an early or late stage. This strategy could therefore be appropriate for all severe splenic trauma grade AAST ≥III if the clinical presentation allows it.

This study has many strengths. First, the study population corresponds to the classic pattern of European trauma cohort (young men with few comorbidities and blunt trauma19,21), and the rate of NOM is congruent with previous literature.²² Second, data quality was ensured by a national prospective database²³ and CT scan proofreading by senior radiologists especially for our study. Third, matching was performed a priori on known infectious risk factors such as sex, age, initial GCS, ISS score, and SAPS II to limit confusion bias. We did not match cases and controls based on hemorrhagic shock but rather on injury pattern and severity. However, considering that transfusion and hemorrhagic shock are risk factors for infection, this choice may have influenced the results.

This study has several biases, primarily inherent in its retrospective design. Patients discharged from the ICU within the first 48 hours were excluded, considering their exposure to infectious risk was too short but this choice carries the risk of selection bias. The long study period also hinders interpretation of the results. Advances in trauma care, including a reduction in the incidence of hemorrhagic shock and improvements in intensive care management, may have reduced the incidence of infectious complications and induced another bias. Altogether our results should be taken cautiously and confirmed prospectively in large multicentric cohorts. The evaluation of spleen immune function, for example through biological markers, could be an interesting topic of further research.

Conclusion

In this study, severe splenic trauma (AAST ≥III) was not associated with the incidence of infections during ICU stay. Moreover, management of splenic trauma (NOM without SAE, NOM with SAE or splenectomy) did not seem to influence the risk of early infection. This work suggests that immunosuppression induced by severe trauma is not influenced by the presence of splenic trauma at its early phase.

Data availability statement

Data are available upon reasonable request.