Authors: Emre Özlüer, Asst. Prof.
Editor: Betül Gülalp, Prof.
Date: 20 August 2017
DOI: 10.1097/SHK.0b013e31826d1eaf


Since hemorrhagic shock is the most frequent cause of the death in polytrauma patients, early detection plays the key role in management of the resuscitation. Several parameters are used for the assessment of the blood loss such as; Injury Severity Score (ISS), inferior vena cava (IVC) diameter and collapsibility index, lactate and base deficit, vital signs, hemoglobin/hematocrit ratio. However, none of these parameters has been found to be sensitive in hemodynamically stable patients, adequately. (1).

The authors based their study on previous animal models (2) (3) (4). A reduction in spleen size was shown in experimental hemorrhagic shock. Also in a case report (5), it was postulated that splenic vasoconstruction due to adrenergic overstimulation might be the leading cause of the reduction in spleen size.

Dynamic changes in the hepatic arterial waveform and RI are clues for the splenic steal syndrome with Doppler sonography (6). The increase of RI in renal, hepatic, and splenic artery were reported in patients with liver cirrhosis in literature. (7) 

In this study, the authors hypothesized that a more direct and less variable index of spleen circulation may be represented by the splenic Doppler resistive index (RI). Thus, they set up  a study investigating whether  splenic Doppler RI was a noninvasive method for early detection of occult hemorrhage.



This prospective study was conducted in the emergency department of San Martino Hospital, Genova which is a urban level I trauma center between  April 30, 2010, and February  28, 2011.

The exclusion criteria were as follows;

  • Patients with ISS≤16
  • Younger than 18 or older than 65 years
  • Hb≤10 g/dL
  • SBP<90 mmHg
  • Penetrating trauma
  • Vasoactive drug support
  • History of chronic liver disease
  • Diabetes
  • Free abdominal fluid diagnosed by FAST
  • Infeasibility of a splanchnic Doppler assessment including splenic vessels
  • Splenic injuries at computed tomography (CT) scan
  • Ultrasonographic signs of hypovolemia, assumed with antero-posterior diameter of VCI less than 9 mm during expiratory phase

Out of 159 patients 49 of them did not match any exclusion criteria and became eligible for the study. Patients were further divided into two groups: those who developed hemorrhagic shock within the first 24 h (n=22) and those who did not (n=27). Hemorrhagic shock was defined as low blood pressure  (SBP<90 mmHg) with low urine output (<30 ml/s) and blood lactate greater than 2 mmol/L with transfusion requirements of two or more units of red blood cells or need for interventions, including angiography with transcatheter arterial emobolization or surgical hemostasis.  

All of the enrolled patients were examined following the ATLS recommendations. Arterial and  venous blood samples were obtained for laboratory testing and cross-match. FAST was performed. In FAST- negative patients, splenic Doppler RI measurements were performed prior a CT scanning. FAST-positive patients were sent to obtain an immediate abdominal CT scanning or OR for surgery depending on the hemodynamic status were excluded.

Patients with haemorrhagic shock were transfused with a ratio of 1:1:1 between red blood cells, fresh frozen plasma and platelets. At 24 h an abdominal ultrasound examination was performed to measure splenic  Doppler RI and spleen size.

Doppler measurements were obtained with a 3.5 MHz convex probe in pulsewave mode and with a 2-4 MHz sector transducer  for the study of VCI. After initial assessment, the patient was placed in supine or right-sided position until complete spleen visualization was achieved. Splenic size (area) was measured by obtaining a coronal view that included the hilum. The measurements then were compared to the sizes in abdominal CT scan.

Splenic artery and vein were identified using color Doppler at 1 cm away from the hilum. With a 30-60 degrees angle peak systolic and end-diastolic flow velocities were measured.  The measurements were performed over the upper, mid and lower poles of the spleen and averaged. These were used to calculate intraobserver variability. Also, interobserver variability was calculated via performing the ultrasonographic measurements by two ultrasonographers which were unaware of the other one’s results.

The volume status of the patients was assessed using conventional VCI measurement technique and collapsibility index of VCI was calculated.

A priori power analysis revealed that the minimum sample size required in each group ranged 19 to 8 patients. Statistical significance was assumed for  p<0,05.

All variables were expressed as SD or percentage. D’Agostino-Pearson omnibus test was used to evaluate the normal distribution of each continuous variable. Only splenic Doppler RI was found to be normally distributed. All continuous variables were tested by Mann Whitney U test and categorical data were tested by  Fisher exact  test. Correlations between splenic Doppler RI, US area, CT area and CT volume were evaluated using Spearman correlation coefficient.  

The Hosmer-Lemeshow test was carried out to check goodness of fit of the logistic regression models that reached statistical significance. The diagnostic accuracy, cut-off values and the sensitivity and the specifity of the measurements were defined using ROC curve analysis. Intraobserver and interobserver variability was tested by intraclass correlation coefficient (ICC). In addition, The Cronbach α reliability coefficient was provided as a further measurement of internal consistency.

Statistical calculations revealed intraobserver ICC 0.986 (95% CI, 0.974-0.993; P<0.001), and The Cronbach α reliability coefficient 0.973, respectively. The interobserver ICC was 0.962 (95% CI, 0.922-0.982; P<0.001) and the Cronbach α reliability coefficient was found 0.927.



In 22 of 49 patients hemorrhagic shock were developed within the first 24 h. Three of these patients underwent to an explorative laparotomy for suspected intra-abdominal injuries, 11 underwent angiography for arterial hemorrhage requiring therapeutic intervention, and 8 were managed conservatively. The mean red blood cell requirement by patients who developed hemorrhagic shock was 6 units. Five patients died due to multisystem organ failure.

The sensitivity of Splenic Doppler RI was %73, the specificity %74, PPV %70, NPV %77 with a cut-off value of 0.65, respectively.

Splenic Doppler RI value was significantly higher in patients with hemorrhagic shock. Also, ISS was greater and the standard base excess was lower in this group of patients. Splenic Doppler RI values after effective resuscitation at 24h in this group decreased. This value was not changed in the other group. Considering other clinical of biochemical parameters including spleen size, the authors did not find any significant difference between two groups. In hemorrhagic shock group, the authors showed the enlargement of the spleen but this was not statistically significant difference between two groups.

Logistic regression analysis showed that splenic Doppler RI (OR, 1.37; P=0.001), ISS (OR, 1.15; P=0.002), and the standart base excess (OR, 0.77; P=0.032) parameters were significantly associated with the risk of hemorrhagic shock.  

IVC measurements did not reveal any statistically significant difference between two groups. The authors suggested that this parameter might be useful when patients are hypovolemic and could predict recurrence of shock after resuscitation but not to detect ongoing hemorrhage in hemodynamically stable patients in the early phase.


The authors concluded the main findings of their study under four main topics:

  1. About half of polytrauma patients with stable hemodynamic conditions at ED admission developed shock due to occult bleeding
  2. At ED admission of the patients with occult bleeding who subsequently developed hemorrhagic shock within 24 h splenic Doppler RI and ISS values were higher and the Standard base excess was lower when compared to the other group.
  3. By multivariate analysis, it was suggested that splenic Doppler RI was the only independent significant predictor of occult hemorrhagic shock with high sensitivity and specificity
  4. Persistently high splenic Doppler RI after resuscitation was associated with mortality due to persistent occult hypoperfusion



Considering the limitations of this study, splenic Doppler RI value cannot be reliable in patients with splenic lacerations, patients who cannot change their posture and unable to take a deep breath. It is possible to assume that the ultrasonographic and Doppler measurements requires great experience to avoid interfering the immediate resuscitative efforts in polytrauma patients. Since the elder and the patients with comorbidities such as chronic liver failure were excluded, the accuracy of the Doppler measurement is at stake in this population. These causes are limiting the population of the study into a little group. Thus, it is impossible to make a generalization inthe light of the data obtained. In addition, the authors did not evaluate the patients according to the shock classification. This leads to a confusion whether the accuracy of the measurements were consistent between a patient with stage 3 and stage 2 hemorrhagic shock.

Because the authors excluded the clinical condition of the patients, calculation of the sensitivity and specificity of the cut-off value of the RI assuming hemorrhagic shock, the clinical significance of the results are stil vague. 

Another ostentatious point is about the performers of the ultrasonographic measurements.  It was emphasized that the first named author, FC, had 15 years of Doppler experience but we do not know the level of the experience of the other ultrasonograph, CB. Furthermore, the duration of time consumed while performing the measurements is not clear.

Since the splenic Doppler RI values were averaged of three different measurements, this led the data getting close to the average. Thus, the SD of the groups was minimized and the more significant results were obtained with a rather small population.

In the literature, splenic Doppler RI measurement was investigated in splenik steal syndrome (6), and the measurement of splanchnic arterial flow velocities in cirrhotic patients (7). The authors of this study we ‘dissected’ had tested this method in the ED patients for the first time. This is why I think that this measurement method may lead to further studies designed in emergency departments.  



  1. Spahn DR, Cerny V, Coats TJ, Duranteau J, Fernández-Mondéjar E, Gordini G, Stahel PF, Hunt BJ, Komadina R, Neugebauer E, et al.: Task Force for Advanced Bleeding Care in Trauma. Management of bleeding following major trauma: a European guideline [published correction appears in Crit Care 24;11(2):414, 2007]. Crit Care 11 (1): R17, 2007
  2. Shen TC: The effect of experimental anaemia on the size of the spleen. J Physiol 3: 74–80, 2009.
  3. Barcroft J, Stephens JG: Observations upon the size of the spleen. J Physiol 12: 1–22, 2010.
  4. Hamza SM, Kaufman S: Role of spleen in integrated control of splanchnic vascular tone: physiology and pathophysiology. J Physiol Pharmacol 87: 1–7, 2009.
  5. Goodman LR, Aprahamian C: Changes in splenic size after abdominal trauma. Radiology 176: 629, 1990.
  6. Li C, Quintini C, Hashimoto K, Fung J, Obuchowski NA, Sands MJ, Wang W. Role of Doppler Sonography in EarlyDetection of Splenic StealSyndrome. J Ultrasound Med. 2016 Jul;35(7):1393-400.
  7. Glišić TM, Perišić MD, Dimitrijevic S, Jurišić V. Doppler assessment of splanchnicarterialflow in patientswithliver cirrhosis: correlationwithammoniaplasmalevelsand MELD score. J Clin Ultrasound. 2014 Jun;42(5):264-9.

Emre Özlüer, Asst. Prof.

Betül Gülalp, Prof.

EMATUS web coordinator. Anyone...Behind the passion and heart...


No comments found. Be a first.

Write a Comment

Course Calendar

Authors: Mehmet Ali Aslaner, Asst. Prof. · 24 January 2018