Can We Trust Them? FAST and DPA in Assessing Unstable Patients Following Blunt Abdominal Trauma

Abstract

The diagnosis of intra-abdominal exsanguination in hemodynamically unstable blunt trauma patients is almost universally determined by Focused Assessment with Sonography in Trauma (FAST). However, FAST has been reported to have poor sensitivity and is therefore associated with false-negative examinations. Our institutional practice includes diagnostic peritoneal aspirate (DPA) following two negative FASTs to address the poor sensitivity. We hypothesized that serial FAST alone would be able to exclude clinically significant abdominal bleeding in an unstable blunt trauma patient. A retrospective cohort study was conducted between 2018 and 2020 at a major tertiary trauma referral hospital, including all blunt trauma patients who were hemodynamically unstable. Two groups were analyzed: 1. “FAST+”: those who had a positive FAST scan and proceeded to a trauma laparotomy, and 2. “DPA”: those who had serial negative FAST scans and proceeded to DPA. Of the 12 patients in the FAST+ group, 92% correctly identified the abdomen as the source of instability. Of the seventeen patients in the DPA group, only two (12%) had positive DPA. Both patients underwent laparotomies, but neither identified an abdominal source of hemodynamic instability. The most common cause of hemodynamic instability in the DPA group was pelvic bleeding from major pelvic ring disruption. The sensitivity and specificity of the serial FAST exam for clinically significant abdominal bleeding were 100% and 94%, respectively. These data suggests that two sequential negative FAST scans are adequate for excluding intra-abdominal bleeding as the source of instability, with further investigation with DPA not identifying any clinically significant sources of intra-abdominal bleeding.

1. Introduction

The challenge in managing patients with hemodynamic instability after blunt abdominal trauma is to determine the body cavity of the source of instability (chest/abdomen/pelvis/long bones/external). Focused assessment with sonography in trauma (FAST) and diagnostic peritoneal aspirate/lavage (DPA/DPL) are two ways of diagnosing an intra-abdominal source of instability.

While DPA/DPL is highly sensitive, it has largely been replaced as the standard of care by FAST [1,2]. FAST scans are an operator-dependent imaging modality that is non-invasive, accessible, and has a favorable specificity and sensitivity profile. While FAST is used as a first-line investigation for identifying the source of hemodynamic instability, DPA/DPL is still used as a secondary investigation for the exclusion of an abdominal source in unstable patients where no source of instability has been identified. The reason for this is that some studies have documented low sensitivity of FAST, thus raising the question of whether FAST is able to confidently exclude abdominal sources of instability [3,4,5,6,7,8,9].

Given that FAST is based on the detection of abdominal fluid, its sensitivity is likely to be positively correlated with the volume of fluid present in the abdomen. Thus, in the setting of an unstable patient due to intra-abdominal exsanguination, the sensitivity of FAST scans should be higher than expected. The aim of the study is to assess whether two negative FAST scans are adequate to exclude intra-abdominal sources of bleeding in a hemodynamically unstable blunt abdominal trauma with suspected abdominal injuries.

2. Materials and Methods

2.1. Patient Selection

This is a retrospective observational study conducted at a major tertiary trauma referral center in Sydney, Australia. This study was conducted in accordance with the Declaration of Helsinki and approved by the WSLHD Human Research Ethics Committee. Patients were selected based on the following criteria: (a) blunt trauma mechanism; (b) age > 18 years; (c) Injury Severity Score (ISS) > 15; (d) hemodynamic instability on arrival to the trauma center (defined as systolic blood pressure < 90 mmHg and/or tachycardia > 120 bpm and unresponsive to fluid resuscitation). Patients with penetrating abdominal injuries were excluded from the study. The study period was between January 2018 and June 2020.

At our institution, a formalized trauma management pathway exists, and all blunt abdominal trauma patients with hemodynamic instability were managed using the algorithm shown in Figure 1 [10]. Patients analyzed for this study included two groups: (i) patients who required a trauma laparotomy for an abdominal source of instability (identified by a positive FAST scan); and (ii) those assessed with a DPA following two negative FAST scans. These patient groups are summarized in Figure 2. The latter group of study subjects represented patients who remained hemodynamically unstable, despite having a negative FAST, and had no other source of instability identified.

2.2. Clinical Management and Definitions

FAST scans were performed (or supervised) by a qualified operator, usually an emergency department trainee or specialist. FAST scans were documented as negative or positive. A positive FAST scan was defined as the presence of free intra-peritoneal fluid in Morrison’s pouch, the splenorenal recess, or the pelvis. Positive scans due only to the presence of fluid in the pericardial window were not considered positive in this study.

Patients who were considered persistently hemodynamically unstable, despite initial crystalloid and/or blood product resuscitation and two negative FAST scans, went on to have a DPA performed. DPAs were performed (or supervised) by a qualified operator, usually a general surgical trainee or specialist. The DPA was performed with a sterile, open technique with a pre-packaged kit. An infraumbilical incision through skin and subcutaneous tissue is performed. The fascia is grasped and incised, and a small peritoneal opening is created. The catheter is inserted into the peritoneal cavity, and a 10 mL syringe is used to aspirate fluid. A positive DPA was defined as ≥10 mL of blood aspirated.

A CT scan and/or laparotomy were used as the gold standard to confirm, in this study, instability from an abdominal source. Instability from an abdominal source was defined as a significant intra-abdominal hemorrhage requiring laparotomy to improve hemodynamics through hemorrhage control. A minor intra-abdominal hemorrhage not requiring surgical intervention was not defined as “instability from an abdominal source”.

2.3. Data Collection and Analysis

Data were collected on patient demographics, including age and sex. Initial trauma assessment was documented, with blood pressure and heart rate used to identify hemodynamic instability.

Descriptive statistics were used to analyze outcomes and demographic variables. Continuous variables were expressed as mean ± standard deviation or median [interquartile ranges], while categorical variables were expressed as n (%). Continuous variables were analyzed using an independent sample t-test or Kruskall–Wallis (depending on the normality of the data), whereas categorical variables were analyzed using the Pearson Chi-square test. DPA sensitivity and specificity were calculated.

3. Results

3.1. Basic Characteristics

During the study period, a total of 1291 patients were assessed following major trauma call activation at our institution. Of these patients, seventeen blunt trauma injury patients were hemodynamically unstable with no definite source of instability (after two negative FAST scans). These patients underwent assessment with a DPA. A further 12 patients were hemodynamically unstable patients who required a category 1 or 2 (<15 min—immediate life-threatening or <1 h—life-threatening, respectively) laparotomy following a positive FAST scan. These two groups of patients were analyzed for this study, as shown in Figure 2.

These two groups were compared as they represented persistently hemodynamically unstable patients with positive and negative FAST. The results of this comparison are summarized in

Table 1. Patient demographics and outcomes of DPA patients and FAST-positive laparotomy patients.

	FAST Negative, Leading to DPA (n = 17)	FAST Positive, Leading to Trauma Laparotomy (n = 12)	p-Values
Sex, Male	14 (82%)	8 (67%)	0.403
Age, Mean Years (±SD)	54.5 (±22.2)	38.3 (±14.9)	0.019
Mechanism of Injury, n (%)			0.255
MVC	5 (29%)	6 (50%)
MBC	3 (18%)	4 (33%)
Pedestrian/Bicycle vs. vehicle	6 (35%)	1 (8%)
Other	3 (18%)	1 (8%)
Injury Severity Score, Median [IQR]	29 [17–33.5]	28 [22–42.5]	0.594
Initial systolic blood pressure			0.348
<90	9 (53%)	8 (66%)
90–100	2 (12%)	-
100–140	4 (24%)	4 (33%)
>140	2 (12%)	-
Initial heart rate			0.108
50–100	5 (29%)	1 (8%)
100–120	5 (29%)	1 (8%)
120–140	6 (35%)	5 (42%)
>140	1 (6%)	4 (33%)
Positive DPA	2 (12%)	n/a
Operative Management			<0.001
Laparotomy	3 (18%)	12 (100%)
Pre-peritoneal packing	4 (24%)	-
Other procedure	4 (24%)	-
Interventional Radiology			0.944
No Angiography	10 (59%)	7 (58%)
Angiography + Intervention	5 (29%)	4 (33%)
Angiography (no intervention)	2 (12%)	1 (8%)
Laparotomy Findings			<0.001
Solid organ injury	1 (6%)	6 (50%)
Mesenteric/viscus injury	-	5 (42%)
No significant intra-abdominal injuries	16 (94%)	1 (8%)
Findings on laparotomy	2 (12%)	1 (8%)
Implied by no need for laparotomy	14 (82%)	0 (0%)
Source of Instability			<0.001
Hypovolaemic cause
Abdomen	-	11 (92%)
Pelvis	7 (41%)	-
Long bone/Limb	4 (23%)	-
External	-	-
Chest	2 (12%)	-
Distributive cause	2 (12%)	1 (8%)
Cardiogenic cause	1 (6%)	-
Length of Stay, Median Days [IQR]	22 [4.5–37]	17 [7–62.5]	0.739

Data expressed as: n (%), mean ± standard deviation, median [IQR]. Abbreviations: diagnostic peritoneal aspirate (DPA); interquartile range (IQR); motor vehicle collision (MVC); motorbike collision (MBC); standard deviation (SD).

. There was a male predominance in both groups (82% and 67%, respectively). The mean age was significantly older in the DPA group (55 vs. 38 years, p = 0.019). The FAST-positive group was also more likely to have significant intra-abdominal injuries (92% vs. 6%, p < 0.001). The median injury severity score (ISS) and initial hemodynamics on presentation were not significantly different between the two groups.

3.2. DPA Confirmed No Additional Patients with Abdominal Sources of Instability in the Context of Negative FAST

Of the seventeen patients undergoing DPA, two patients (12%) had a “positive” DPA, as defined by aspiration of >10 mL of blood. However, both of these patients were found to have pelvic sources of bleeding, requiring pre-peritoneal packing and interventional radiology (Table 1). Two independent clinicians (RQ and TP) retrospectively reviewed the case notes and determined the source of instability was the pelvis in both instances and that the DPA result did not change management. These patients were therefore considered to be false-positive patients. All fifteen other patients had negative DPAs. Of these, only one subsequently required a laparotomy, although the indication for this was for concerns regarding duodenal injury (duodenal contusions and intraperitoneal free gas post-DPA) on a CT scan. At the operation, the patient had minor liver lacerations managed with topical hemostatic agents and intestinal serosal tears. A further four (24%) negative DPAs revealed small splenic lacerations on a CT scan, not requiring further interventional or operative management. Therefore, these 15 (88%) cases were considered true negatives. In summary, while the specificity of DPA in our patient cohort was 88.2% (95% CI: 63.5–98.5), none of the seventeen patients (0%) undergoing DPA had an abdominal source of blood loss as the cause of hemodynamic instability (

Table 2. Results of FAST and diagnostic peritoneal aspirate.

	FAST	DPA
True positive	11	0
False negative	0	0
True negative	17	15
False positive	1	2
Sensitivity	-
Specificity	94%	88.2%
PPV	91.6%
NPV	100%

Abbreviations: focused assessment with sonography for trauma (FAST), diagnostic peritoneal aspirate (DPA), positive predictive value (PPV), negative predictive value (NPV).

).

3.3. Non-Hypovolaemic Shock Is Prevalent in FAST Negative Patient Group

Interestingly, almost a fifth of patients (n = 3, 18%) had a non-hypovolaemic cause of shock (Table 1). One patient had cardiogenic shock, and two had distributive shock (secondary to severe head injury and C1/2 cervical vertebra dislocation).

Non-abdominal sources of hypovolaemic shock were also prevalent, with pelvic sources of bleeding being the most common (n = 7, 41%).

4. Discussion

This study aims to confirm that two negative FASTs in the setting of hemodynamically unstable blunt trauma patients are sufficient to exclude the abdomen as a source of instability. Our study shows that none of the patients who had two negative FAST scans had an intra-abdominal source of instability. Furthermore, DPA in this group revealed no additional patients with such a source of instability. Conversely, 92% of hemodynamically unstable patients with FAST-positive scans had hemodynamically significant intra-abdominal injuries. These findings confirm that a positive FAST reliably predicted an abdominal source of instability, and a negative FAST accurately ruled it out. The sensitivity and specificity of the serial FAST exam for clinically significant bleeding in our study were 100% and 94%, respectively. In this discussion, we will consider this result in the context of existing studies on the accuracy of FAST and discuss the modern role of DPA.

In the assessment of blunt abdominal trauma, a FAST scan is recommended as the first-line diagnostic modality according to the Advanced Trauma Life Support (ATLS) and Eastern Association for Surgery in Trauma (EAST) guidelines [11,12]. FAST is a non-invasive bedside investigation that is shown to have a wide range of sensitivity (42.9–98.1%) and high specificity (93.9–100%) to hemoperitoneum [3,4,5,7,8]. However, the majority of literature on FAST scans is based on hemodynamically stable patients. Hsu et al. [6] assessed FAST scans in hemodynamically unstable patients with a sensitivity of 78% and a specificity of 97%. The majority (19/22) of false negatives in the study were clinically insignificant due to trace fluid in the pelvis seen on a CT scan [6]. These findings reflect that a FAST scan requires a minimum amount of free fluid to be detectable on ultrasound. Several studies have quantified this volume and found anywhere between 100 and 300 mL is required [13,14,15]. The corollary of this is that the sensitivity of detection is likely to improve as the volume of free fluid present in the abdomen increases [9]. In an unstable patient from an abdominal source where multiple liters of blood are in the peritoneal space, it is reasonable to expect the sensitivity of FAST to be higher than previously reported. This hypothesis is supported by Cha et al. [16], who found the sensitivity of DPL increased between stable and unstable patients. Our present findings support this hypothesis.

Our finding also suggests that while FAST (and DPAs) are effective at excluding an intra-abdominal source of hemodynamic instability, they are ineffective at excluding intra-abdominal injury. Five patients who were DPA and FAST negative were found to have a solid organ injury on a CT scan. In four patients, the CT findings were clinically insignificant and did not require any invasive intervention. One patient proceeded to a laparotomy due to concurrent small bowel injuries, and subsequently, the liver laceration was managed with hemostatic agents. Cha et al. [16] showed similar findings with laparotomies following negative DPLs, having missed bowel injuries, diaphragmatic injuries, and minor liver lacerations. Thus, after patients have achieved hemodynamic stability, a CT scan is still required to detect additional abdominal injuries. Some of such injuries may still require further surgical intervention.

An important implication of this study is that there should be a high suspicion of a non-hypovolemic cause of shock in hemodynamically unstable blunt trauma patients who are both FAST and DPA negative. In the trauma setting, hypovolemic shock due to hemorrhage is the most common cause of hemodynamic instability [12]. However, our study identified three (18%) cases of instability in FAST-negative patients due to distributive or neurogenic shock. Similarly, Schellenberg et al. [17] reported mortality due to traumatic brain injury or spinal cord injury in 13 (40%) of their cohort. Thus, the pursuit of the abdomen as the source in patients who are FAST negative (with or without being DPA negative) may be a futile exercise; instead, non-hemorrhagic causes should be actively excluded.

DPA/DPL use in modern trauma literature has two indications: as a definitive method to exclude intra-abdominal sources of instability and in stable patients to help identify bowel and mesenteric injuries, which are notoriously rare [2,11,17,18]. Our findings do not support the former indication in the context of two negative FASTs. In contrast, the contemporary study by Schellenberg et al. [17] reported 80% sensitivity, 100% specificity, and 93% accuracy of DPA on 41 patients. These results suggest an ongoing benefit of DPA in the diagnosis of instability in the abdomen. However, Schellenberg et al. [17] involved a heterogeneous cohort. It included patients with penetrating injuries, as well as both hemodynamically stable and unstable patients. In addition, there was no standardized institutional protocol for the indication of DPA. Interestingly, this study included FAST-positive patients with a history of cirrhosis, which highlights another role of DPA—for the assessment of patients with alternative causes of intra-abdominal fluid, such as the cirrhotic trauma patient [17].

Finally, one can consider how the practice of DPA is affected by the emerging practice of pre-peritoneal packing for unstable patients with pelvic fractures. On the one hand, DPA may allow the assessment of the abdomen prior to pre-peritoneal packing in the theater to determine if concurrent laparotomy is needed, but on the other hand, during the procedure of pre-peritoneal packing, it would be simple to create a peritoneal window to assess the abdomen directly. This, in essence, is a modified approach to DPA.

This study has several weaknesses, including the small cohort size and retrospective data collection reliant on documentation at major trauma calls. However, all blunt abdominal trauma patients who presented to our institution were managed as per the trauma algorithm, minimizing the risk of bias in the data.

5. Conclusions

These results suggest that two negative FAST scans are adequate to exclude intra-abdominal bleeding as the source of instability. In such patients with two negative FASTs, an alternate source of instability should be sought, including consideration of non-hypovolaemic causes of shock.