Article Text

Download PDFPDF
Emergency focussed assessment with sonography in trauma (FAST) and haemodynamic stability
  1. Zoë A Smith,
  2. Darryl Wood
  1. Department of Emergency Medicine, Ngwelezana Hospital, Kwa-Zulu Natal, South Africa
  1. Correspondence to Dr Zoë A Smith, Ngwelezana Hospital Emergency Department, Empangeni, KwaZulu Natal, South Africa; drzoesmith{at}gmail.com

Abstract

Background Focussed assessment with sonography in trauma (FAST) has assumed a key role in the rapid non-invasive assessment of thoracoabdominal trauma and assists in decreasing disposition time. This study evaluates FAST's efficacy with respect to haemodynamic stability in a South African emergency department (ED).

Methods Data were collected prospectively by four emergency medicine doctors trained in emergency ultrasonography. FAST scans were performed by one ED doctor and timings, scan result and disposition were recorded. Patient haemodynamic stability was assessed by the emergency doctor performing the scan; subjectively at the time of scanning and objectively using calculation of the shock index. All scan results were subsequently verified by a second ED doctor in a blinded fashion and by CT scanning or operative intervention when clinically indicated.

Results 166 FAST scans were conducted of which 36 (21.7%) were positive. Mean age was 30.6 years (SD 12.8). 74.1% of patients sustained blunt traumatic injury. Doctors’ subjective haemodynamic stability assessments had higher specificity, sensitivity and predictive values than shock index alone. Haemodynamic instability and a positive FAST result were significantly related (p=0.004). Sensitivities and specificities of FAST scans for blunt and penetrating trauma were 93.1% and 100%, and 90.0% and 100%, respectively. Corresponding values for pneumothoraces were 84.6% and 100%.

Discussion This study showed a valuable role for FAST in all traumas, particularly in haemodynamic compromise. As an addition to the physician's repertoire of bedside assessment tools, it improves diagnostic capabilities in comparison with simple haemodynamic assessments alone.

  • Trauma
  • ultrasound
  • emergency department
  • clinical assessment
  • non invasive

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Since its inception in the 1990s1 the use of focussed assessment with sonography in trauma (FAST) scanning has assumed a key role in the rapid non-invasive assessment and subsequent management of patients suffering thoracoabdominal trauma. Considerable advantages over other imaging modalities available for trauma assessment are well recognised. The evolution of FAST has seen its incorporation into American College of Surgeons Advanced Trauma Life Support (ATLS) training2 and it is currently used by emergency physicians in resuscitation rooms worldwide to assist timely decision-making.3–5

Following a Cochrane review in 2005, some have criticised the rapid promotion of emergency FAST scanning, stating there is insufficient evidence to support the adoption of ultrasonography-based clinical pathways in the assessment of blunt trauma.6–8 However, this has been contested by others, which include a randomised controlled trial comparing the use of FAST scanning with clinical assessment alone.9 Such studies have shown FAST decreases time to disposition or operative intervention by 64–76%, reduces the requirement for CT scanning and decreases complication rates and length of hospital stay.9–11

The detection of haemoperitoneum and haemopericardium by FAST scanning relies on appropriate training and experience, with studies estimating the learning curve levels after a minimum of 30 scans.12 ,13 While it remains operator dependent, in skilled hands it is a consistently reliable ‘rule-in’ tool in the case of a positive scan.14 ,15 Some studies have shown it is less reliable in penetrating trauma16 but in blunt abdominal trauma, FAST sensitivities and specificities range between 75–99% and 90–100%, respectively.6 ,17–19 Assessment of pneumothoraces with ultrasound has been added to the classical four views obtained in the FAST scan as part of an extended-FAST (eFAST) protocol, with encouraging results.20 ,21

According to ATLS protocol, FAST scanning should be performed immediately after the primary survey.2 It is considered most useful in haemodynamically unstable patients whose cause for hypotension is unclear.22 The diagnosis of haemodynamic instability itself presents many challenges.23 In this study we aimed to evaluate the efficacy of FAST scanning with respect to haemodynamically stable and unstable patients presenting to our emergency department (ED), which experiences a high burden of trauma. In addition we aimed to assess the usefulness of FAST scanning in predicting the presence of internal haemorrhage as compared with that afforded by simple haemodynamic assessments alone.

Methods

Ethical approval for the study was obtained from Ngwelezana Hospital Ethics Committee. Data were collected prospectively from FAST scans conducted in all trauma patients during a 9-month study period at Ngwelezana Hospital, a 554-bedded district government hospital in Kwa-Zulu Natal, South Africa. FAST scans were performed following the primary survey. Performance of scans was restricted to one emergency medicine (EM) consultant, one senior EM trainee and two EM medical officers. All had a minimum level of training of the UK College of EM Level 1 Emergency Ultrasound and were experienced in performing unsupervised scans in the ED. The method for FAST scanning is well described in the literature.24 The classical four views (figure 1) were obtained using a GE Logiq 5 ultrasound machine (GE Healthcare, UK) and a 3.5 Hz abdominal probe. Those doctors trained in eFAST also scanned for pneumothoraces using a vascular probe and the results for each view were documented along with subsequent radiological confirmation.

Figure 1

The classical four probe positions used in focussed assessment with sonography in trauma scanning to detect fluid in dependent areas: hepatorenal (Morrison's pouch), splenorenal, pericardial and pelvic (pouch of Douglas).

By means of confirmation of FAST scan results, scans were repeated by a second credentialed emergency doctor in a blinded fashion. This was followed by a minimum observation period of 24 h. The timing of the second scan was dependent on the availability of a second trained ED doctor. Alternatively, when clinically indicated, initial FAST scanning was followed and confirmed by CT scanning, diagnostic peritoneal lavage or operative intervention. The gold standard employed was determined according to the clinical status of the patient at clinicians’ discretion. Cases in which confirmation was not possible by either of these means were excluded from analysis. There were no additional exclusion criteria. Demographic data and information regarding the time of the scan relative to admission, mechanism of injury, scan result, mode of confirmation and final destination of the patient were documented.

Two separate assessments of the patient's haemodynamic status were made. The doctor performing the scan made a subjective clinical assessment of stability in terms of ‘stable’ or ‘unstable’ at the same time as the scan was performed. Heart rate (beats per minute) and blood pressure (mm Hg) were also recorded at the time of the scan, which enabled calculation of the ‘shock index (SI)’, a simple objective contemporaneous measure of haemodynamic stability. This method has been shown to be more reliable than either heart rate or systolic blood pressure alone but clearly has limitations.23 ,25_ENREF_25 It is calculated as described in the literature as a ratio of the heart rate to systolic blood pressure to produce the SI (<1=not shocked/‘stable’ and >1=shocked/‘unstable’).

Data were recorded in a departmental logbook, anonymised and stored in a Microsoft Excel 2010 file before export to SPSS for Windows V.16.1 for statistical analyses (significance level p=0.05).

Results

One hundred and sixty-six FAST scans were conducted during the 9-month study period. The male to female ratio was four to one and the mean age was 30.6 years (SD 12.8). Ninety-one scans (54.8%) were conducted within normal working hours and the remainder when staffing levels were reduced. There were a total of 36 positive scans (21.7%). Of the 166 patients undergoing FAST, 74.1% had sustained blunt traumatic injury (figures 2 and 3). Of those presenting with penetrating trauma, two-thirds were stab wounds and the remainder gunshot injuries.

Figure 2

Flow diagram illustrating scans performed in the emergency department during the study period.

Figure 3

Distribution of mechanism of injuries in patients scanned using focussed assessment with sonography in trauma in the resuscitation room.

Haemodynamic stability

Emergency doctors classified 31 patients (18.7%) as being haemodynamically ‘unstable’ at the time of the scan, whereas according to the SI calculations, 30.1% fell into this category (table 1). When relating haemodynamic stability to the true presence of free intraperitoneal fluid or haemopericardium, doctors’ assessments had higher specificity, sensitivity, negative and positive predictive values, but sensitivities and positive predictive values for either method alone were poor (table 2).

Table 1

FAST scan results related to stability of the patients at the time of scan as reported by physicians (‘Stable’/‘Unstable’) and according to the shock index (SI=ratio of heart rate to systolic blood pressure)

Table 2

Accuracy of physician assessment of stability as compared with the calculated shock index (SI) for predicting haemoperitoneum or pericardium according to the true outcomes

A χ2 test showed a significant relationship between haemodynamically unstable patients according to the SI and a positive FAST result (p=0.004). When this data is separated into blunt and penetrating trauma, the association remains between blunt trauma (p=0.003) but not penetrating trauma (p=0.417). There was no significant statistical difference in stability or FAST scan result of patients suffering blunt as compared with penetrating trauma.

Timings of scans

Two-thirds of patients (n=109) were scanned within 30 min of arrival in the ED. Those falling outside of this window included 40.0% (n=20) of the ‘unstable’ patients according to SI calculations. There was no statistical difference between the time interval prior to scanning (more or less than 30 min) and either the stability of the patient or their FAST scan result.

Scan confirmation

FAST scans were confirmed by more invasive diagnostic means in 47.6% of cases; CT (n=56), laparotomy (n=21), diagnostic peritoneal lavage (n=2). The remainder were all confirmed by a second blinded qualified ED ultrasonographer (n=87), followed by a minimum 24h period of clinical evaluation. The accuracy of FAST scanning is shown in table 3. One hundred twenty-six scans for pneumothoraces were conducted as part of eFAST scans of which 12 (9.5%) were positive. Two of these positive scans were in patients with suspected tension pneumothoraces.

Table 3

Accuracy of FAST and pneumothorax scans

Discussion

As was noted in a previous preliminary study in our ED,4 the positive scan rate in this study (21.7%) was higher than has been reported in other studies elsewhere in the world which include positive rates ranging from 5.2–13.6%.14 ,15 ,19_ENREF_15 This reflects the high trauma load experienced in South African EDs and the importance of rapid decision-making. In addition, in many of these centres, limited imaging capabilities lead to significant therapeutic and diagnostic challenges.

This study reported 100% specificity and positive predictive values in FAST scans conducted in all types of trauma. Consistent with other studies, the sensitivity was better in blunt abdominal trauma (93.1%) than in penetrating trauma (90.0%). Possible reasons for false negative FAST results are that clotted blood can generate various degrees of echogenicity and may be mistaken for normal surrounding soft tissue, and not all abdominal injuries produce free fluid. Bowel injury and solid organ injury (as often occurs in penetrating trauma) without significant bleeding may be undetectable by ultrasound. In two of the false negative cases seen in penetrating trauma in this study, bowel or other visceral injury was found at laparotomy. The sensitivity of penetrating injury in our study was higher than others, which have reported figures as low as 48%.26 Despite its critics, some physicians advocate the use of FAST in penetrating abdominal trauma to assist identification of significant haemoperitoneum and to help prioritise management when treating multiple penetrating injuries.27 ,28 This study offers support for this rationale, particularly since 25% of the penetrating trauma patients were haemodynamically unstable and a timely assessment is thus critical.

When considering different methods of assessment of haemodynamic stability, physicians’ subliminal clinical acumen was a superior predictor of subsequent findings of haemoperitoneum or haemopericardium than the SI score calculated directly from vital signs. Clinicians are able to take into account factors including examination findings and initial response to fluid resuscitation, whereas normal vital signs at the time of assessment paint a distorted picture in a transiently responsive patient. However, it is important to note that neither method was a good determinant of internal haemorrhage with very poor specificities (44–46%) and positive predictive value (38–55%). This stresses the need for additional diagnostic modalities, whether this is obtained from repeated clinical examination, serial FAST scans or by other radiological means. The more recent development of a focussed rapid echocardiographic evaluation scan allows effective assessment of cardiac function and inferior vena caval filling, providing feedback regarding haemodynamic status and guiding further management in critically ill trauma patients.29

This study demonstrated a significant statistical relationship between haemodynamically unstable blunt trauma patients and those with positive FAST scans. Failure to demonstrate the same relationship in penetrating trauma is likely to be due to thoracic causes for haemodynamic compromise rather than an intra-abdominal cause for instability.

When considering the timing of FAST scans, the majority of patients were scanned in the first 30 min of arrival in the ED. This is likely to represent adherence to ATLS guidelines stating that FAST should be performed immediately following completion of the primary survey. However, when broken down according to haemodynamic stability, our data shows that a proportion of unstable patients were not scanned until after 30 min of arrival in the ED. Thus this is unlikely to have occurred within the ATLS timeframe. A possible reason for this is that such patients may have arrived in the ED and initially been haemodynamically stable and subsequently deteriorated, prompting transfer to the resuscitation area. In some cases there may also have been some delay prior to scanning due to the limited number of emergency doctors currently trained in emergency ultrasound. In order to address these issues, there is a need for accurate triage of trauma patients and further training in FAST scanning technique as part of ED training in South Africa.

The specificity and sensitivity of ultrasound assessment of pneumothorax as compared with radiological means in this study were 84.6% and 100%, respectively. This is comparable with other studies, which have shown respective ranges of 86.2–98.1% and 97.2–100%.21 ,25 In two cases of suspected tension pneumothorax a rapid assessment with sonography added weight to the clinical assessment and allowed for confident rapid decompression and subsequent intercostal drain insertion. In one case a pneumothorax was detected with ultrasound but was not visible on chest x-ray. Previous studies comparing ultrasound assessment of pneumothorax with diagnosis on chest x-ray have shown ultrasound assessment may be superior. Sensitivities and specificities for chest x-ray diagnosis are 27.6–75.5% and 100%, respectively.21 ,30 In addition it seems to be an easily acquirable skill that has successfully been incorporated into prehospital care in Europe.31

The strengths of this study include a healthy number of subjects and a robust method of confirming all scan results in a blinded fashion. The departmental purchase of a new ultrasound machine made it possible to save and review any equivocal scans. Conversely, scan acquisition was limited to periods when the four trained emergency ultrasonographers were present in the ED, and the timing of the second confirmatory scan was variable for the same reason. Future research might ensure blinding of the scanning doctor to the stability status of the patient. In addition, times to definitive surgery or disposition and data regarding subsequent morbidity and mortality would further enhance the assessment of the benefit of the timing of FAST scans.

Conclusion

FAST scanning in our ED shows accuracy at the top end of the spectrum in comparison with previous studies in blunt and penetrating trauma. We would caution against disregarding FAST in the assessment of penetrating trauma, particularly in haemodynamic compromise. Our study has also shown promising results for the ultrasonographic assessment of pneumothoraces and its use in this respect should be encouraged. There is a need to improve the timeliness of eFAST scans, particularly in presence of haemodynamic instability where FAST is most useful. In support of previous studies, FAST has proven a valuable addition to the physician's repertoire of bedside assessment tools, showing considerable improvement in diagnostic capabilities in comparison with simple haemodynamic assessments alone.

References

Footnotes

  • Contributors Both authors contributed to the study design and data collection. ZAS was responsible for data analysis and the writing of the manuscript which was revised by DW.

  • Competing interests None.

  • Ethics approval Ethics Committee Ngwelezana Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.