Invited Commentary: Agreement for Blunt Splenic Injuries Has Room for Improvement

The article by Adams-McGavin et al ¹ evaluates inter-rater agreement for CT grading of blunt splenic trauma using the Revised Organ Injury Scale (OIS) of the American Association for Surgery of Trauma (AAST). ² The OIS of the AAST is the most commonly used classification of splenic trauma. It has undergone several updates since its creation in 1994, most recently in 2018. Importantly, this article is the first study to report inter-rater reliability of splenic injury grades since the 2018 AAST scoring system update. In addition, this article raises awareness of potential pitfalls when applying the AAST scoring system.

Correct classification of blunt splenic injuries is of high clinical importance. The spleen is the most commonly injured solid organ in the setting of blunt trauma, resulting in significant morbidity and mortality for patients. ³ The imaging grade of splenic injury drives clinical decision making, with low-grade injuries (ie, grades I-III) managed non-operatively and high-grade injuries (ie, grades IV-V) undergoing operative management or splenic artery embolization (SAE).^4,5 Recommendations from the Society of Interventional Radiology state that providers should consider SAE for any patient with a high-grade splenic injury. ⁶ Given the impact of imaging grade on patient management, accurately distinguishing between low and high-grade splenic injuries is imperative. However, Adams-McGavin et al found minimal absolute inter-rater agreement in patients with blunt splenic trauma when using the AAST scoring system, which slightly improved when grouped into low-grade vs high-grade injuries.

This finding raises several interesting points that radiologists reading within an emergency department setting should consider. First, the language within the AAST scoring system can lead to differential application of the scoring system, particularly as it relates to patients with multiple splenic injuries. In patients with multiple splenic injuries, radiologists should classify the injury grade by the highest grade of injury present, and the grade can be upgraded by 1 for multiple injuries up to a grade III. ¹ As such, upgrading the splenic trauma grade in the setting of multiple injuries cannot upgrade a low-grade injury (ie, grades I-III), which are often managed conservatively, to a high-grade injury (ie, grades IV-V), which often require intervention.^4,6 This nuanced point, mentioned only briefly in the guidelines, leaves some ambiguity to the wording, and in the absence of case examples can result in inconsistent application of the scoring system, as discussed in more detail in this article.

The second key point this article makes is to draw attention to the largest change within the 2018 AAST revision, the incorporation of vascular injury, as defined as a pseudoaneurysm, arteriovenous fistula, and/or active extravasation into the imaging criteria of injury grading. Specifically, the presence of a vascular injury or active extravasation automatically results in a high-grade injury based on the 2018 AAST revision. ² The authors appropriately investigated inter-rater disagreement about low-grade vs high-grade injuries, the majority of which were due to differences in identification and interpretation of hyperattenuating foci as locations of vascular injury or bleeding.

Third, and related to the above changes in the 2018 update incorporating vascular injury into the grading system, the authors discuss best scanning methods for evaluation of splenic trauma. The CT scanning technique employed by the authors’ institution was a split-bolus single-pass CT of the chest, abdomen, and pelvis. This method allows for combination of arterial and portal-venous phase imaging in 1 acquisition, reducing radiation dose to the patient and resulting in fewer images for the radiologist to interpret. While this is a commonly used method, a large amount of variability exists in trauma CT protocols, with techniques ranging from single to multiple boluses and acquisitions, and clinical data providing support to multiple techniques.^7–9 The AAST recommendation is for dual-phase imaging capturing both the arterial and portal-venous phases when assessing for solid organ trauma to improve sensitivity in detection of vascular injuries. ² As the authors postulate, it is possible that the single acquisition method led to discrepancies between the readers, lowering the inter-reader agreement. The authors’ discussion of scanning methods are worth considering for any trauma center.

A final point inferred from this paper is the use of training sets and windowing as methods to improve reliability as well as sensitivity and specificity in detection of solid organ trauma. Commonly identified sources of disagreement in the study included mistakenly identifying splenic clefts or streak artifacts as lacerations. While both are common mistakes, implementation of training sets aimed at splenic trauma could mitigate these discrepancies, particularly in centers where trainees interpret studies. A related source of disagreement resulted from missing subtle hematomas best visible on narrow imaging windows. There are limited recent studies evaluating the use of narrow windows for detection of solid organ trauma, ¹⁰ raising the possibility for future avenues of study.

Patient care is increasingly reliant on imaging findings, particularly in the setting of trauma, where algorithmic approaches are key for appropriately triaging patients. As the most commonly injured solid organ in blunt trauma, radiologists should describe splenic injuries in a standardized way consistent with the AAST grading system, particularly with the 2018 updates. However, the limitations of this scoring system highlighted by Adams-McGavin et al bring to light common pitfalls that can greatly affect patient care. This is certainly an area in need of future research, improved scoring, and CT scanning techniques to improve inter-rater agreement.