Introduction

Trauma injury remains the leading cause of death among people aged less than 35 years, with 40% of trauma deaths imputable to uncontrolled hemorrhagic shock or its consequences [1, 2].

Currently, fluid resuscitation is the first step in the hemodynamic management of traumatic hemorrhagic shock [3]. The rapid vascular access and stabilization of the cardiovascular system can protect the patient from the severe consequences of hypovolemic shock. The origins of fluid resuscitation can be traced back to the thirties of the nineteenth century, when Thomas Latta performed an attempt of intravenous fluid resuscitation for the first time [4]. In the period 1879–1881 Kroneecker and Landerer stated that in cases of blood loss the most valuable thing is to rapidly restore the vascular bed volume. For this purpose, they proposed using a normal saline solution with the addition of sugar [5, 6]. The development of fluid therapy was in the 1920s, when Alfred Blalock experimented with incremental hemorrhage to induce shock in dogs [7, 8]. In his research, Blalock used blood pressure (BP), cardiac output as well as blood oxygen content from left and right ventricles to evaluate the effect of three types of treatment: saline, transfusion and pharmacological treatment.

Applying an appropriate fluid therapy strategy may restore tissue perfusion and consequently oxygenation of the body. Fluid resuscitation can be carried out based on changes in hemodynamics, diuresis, serum lactate levels or alkaline deficit. However, excessive fluid resuscitation could contribute to the development of coagulopathy of trauma [3, 9] as well as tissue edema [10], which can lead to alterations of tissue perfusion and complications such as abdominal compartment syndrome or adult respiratory distress syndrome [11, 12]. The optimal level of BP during resuscitation of hemorrhagic shock patients is still debated.

The present work aimed to assess the effect of hypotensive versus conventional resuscitation strategies in traumatic hemorrhagic shock patients on mortality, need for blood transfusion and adverse events (specifically: acute myocardial or renal failure or acute respiratory distress syndrome).

Methods

This systematic review and meta-analysis adhere to the reporting guidelines of the Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) statement (Suppl. Table 1).

Search strategy

Available literature databases including EMBASE, MEDLINE, Google Scholar, Cochrane Central Register of Controlled Trials (CENTRAL) were searched from the inception of the databases until 18 June 2020. Searches were conducted independently by two persons (K.S. and L.S.). The papers were restricted to those with the English language. Reference lists of eligible articles were reviewed and content experts were consulted with (K.J.F. and M.J.J.) to identify additional published reports. Incomplete data were dealt with by contacting the principal authors, when possible, to ask for missing or unclear information.

The search strategy was comprised of MESH terms and keywords such as: “shock”OR, “hemorrhagic” OR, “trauma” OR, “injury” OR, “hypotensive resuscitation” OR, “limited resuscitation” OR, “fluid resuscitation” OR, “limited fluid”. To identify in-progress or terminated studies, clinicaltrials.gov registry was also searched.

Study selection

This study included randomized controlled trials (RCTs) and quasi-randomized trials. Observational studies, case reports, studies not based on original research and studies not involving patients, conference papers as well as letters to the editor were excluded from the present study.

Data extraction

Using a standardized data extraction sheet, two authors (K.S. and L.S.) independently extracted data from each report included. Any discrepancies were resolved by consensus with the third author (J.S.). When necessary for data or article clarification, personal communication was made with select study authors. Baseline patient characteristics were extracted as well as data about each trial’s intervention, inclusion and exclusion criteria, mortality and adverse events. For all clinical outcomes, the number of events that occurred in each arm of each trial were tabulated.

Quality assessment

Two reviewers (J.S. and A.S.) independently assessed the methodological quality of each eligible article using the “risk of bias” assessment tool of the Cochrane Handbook [13]. The following domains were evaluated for RCTs: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias) and other bias [12]. Each was graded “yes”, “no”, or “unclear”, which reflected a high risk of bias, low risk of bias, and uncertain bias, respectively (Suppl. Fig. S1). The review authors’ judgments about each risk of bias item are provided in Supplementary Figure S2. The overall risk of bias for the study was rated ‘low’ if 7 or more domains were rated low, ‘moderate’ if 4 to 6 domains were rated low, and ‘high’ if 1 to 3 domains were rated low.

Statistical analysis

Statistical analysis was done by two authors (A.S. and L.S.) independently and was cross-validated. For continuous outcomes, mean difference (MD), and for dichotomous outcomes were used, and risk ratios (RR), were calculated. All continuous data with either means with standard deviations (SD) or medians with interquartile ranges (IQR) as reported in the primary study are presented. When the continuous outcome was reported in a study as median, range, and IQR, means and SD using the formula described by Hozo et al. [14] were estimated. For descriptive purposes, absolute and relative frequencies are reported for categoric variates.

Statistical heterogeneity and inconsistency were measured by using the Cochran Q test and I2, respectively [14]. Odds ratios (OR) with 95% confidence intervals (CI) were calculated as summary statistics. The pooled OR was calculated with the Mantel-Haenszel method. Weighted mean differences and 95% CIs were computed for continuous variables, again using a fixed-effect method in cases of low statistical inconsistency (I2 ≤ 50%) and using a random-effect method in cases of moderate or high statistical inconsistency (I2 > 50%) [15]. Results were considered statistically significant at p < 0.05. Statistical analyses were performed with the Review Manager (version 5.4; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark).

Results

Eligible studies and their characteristics

The flowchart of the literature search is presented in Figure 1. The initial search returned 432 records from all the databases. Two more studies were identified from the references of the relevant articles. Two hundred and ninety-six records were further screened by titles and abstracts and 138 duplicate articles were removed. Unrelated articles, comments, reviews, letters, and duplicate articles were excluded. Then 47 articles were assessed by accessing the full-text. Nineteen studies were excluded because of unavailable data, duplicates, and unrelated topics. Finally, 28 studies were included in the analysis [16–43].

Figure 1. Flow diagram showing stages of database searching and study selection as per PRISMA guidelines.

Assessment of quality

The quality assessment is represented in Supplementary Figures S1 and S2. The method of random sequence generation was perfect in all the studies. There were high risks of attrition bias, lack of intention-to-treat analysis, and selective reporting. The detailed information about blinding and allocation concealment was insufficient in most studies. None of the trials included was at low risk of bias across all domains.

Mortality

Twenty-eight studies reported overall mortality [16–23]. Mortality with hypotension fluid resuscitation was 12.5% and was statistically significant, being smaller than with the conventional fluid resuscitation group – 21.4% (RR = 0.58; 95% CI: 0.51–0.66; I2 = 37%; p < 0.001; Fig. 2). In contrast, only one study by Morrison et al. [28] indicated mortality rates during the first 24 hours. According to this study, mortality for hypotension versus conventional fluid resuscitation varied and amounted to 13.6% vs. 21.7% respectively (RR = 0.63; 95% CI: 0.25–1.58; p = 0.32).

Figure 2. Forest plot of hypotension versus. conventional fluid resuscitation, relative to mortality. The center of each square represents the relative risk for individual trials, and the corresponding horizontal line stands for a 95% confidence interval (CI). The diamonds represent pooled results.

Adverse events

The polled analysis showed that hypotension fluid resuscitation compared to conventional fluid resuscitation was associated with a lower risk of adverse events (10.8% vs. 13.4%, respectively; RR = 0.70; 95% CI: 0.59–0.83; I2 = 52%; p < 0.001).

The use of hypotension versus conventional fluid resuscitation showed a higher incidence of anemia (74.3% vs. 68.6%), thrombocytopenia (33.6% vs. 29.4%) and acute renal failure (8.8% vs. 8.1%). In other types of adverse events the relationship was reversed, and the use of hypotension fluid resuscitation was associated with a lower risk of complications (Table 1).

Fluid balance and transfusion requirements

Additional analysis showed that patients who were treated with hypotension fluid resuscitation required smaller volumes of fluids than the conventional fluid resuscitation group (MD = –1.02; 95% CI: –1.33 to –0.71; I2 = 99%; p < 0.001; Fig. 3).

Figure 3. Forest plot of hypotension versus conventional fluid resuscitation, relative to fluid balance and transfusion requirements. The center of each square represents the mean difference for individual trials, and the corresponding horizontal line stands for 95% confidence interval (CI). The diamonds represent pooled results.

Length of stay ICU/hospital

The length of stay in the intensive care unit (ICU) was reported by two studies [16, 32]. The polled analysis did not show significant differences in the length of stay in ICU between the groups (MD = 0.38; 95% CI: –1.83–2.59; I2 = 73%; p = = 0.74; Suppl. Fig. S3). Three studies indicated length of stay in hospital [16, 29, 32]. The difference between therapeutic groups was not statistically significant (MD = –0.82; 95% CI: –2.43–0.78; I2 = 0%; p = 0.32; Suppl. Fig. S4).

Discussion

The purpose of this research was to compare the effects of hypovolemic and conventional fluid resuscitation on the mortality rate among patients with traumatic hemorrhagic shock. Meta-analysis for overall mortality showed that hypovolemic fluid resuscitation offered benefit in comparison with conventional fluid resuscitation for patients with traumatic hemorrhagic shock at the final follow-up (p < 0.001).

Obtaining intravascular access in hypovolemic patients (especially trauma patients) should be done as soon as possible. In patients with hemorrhage, the most important part of the procedure is to stop the hemorrhage. In such a patient’s hospital setting, transfusion of blood substitutes should be limited in favor of the transfusion of blood components. It is recommended to transfuse the red blood cell concentrate in a volume that maintains the hemoglobin concentration at 7–9 g/dL. Fresh frozen plasma should be transfused immediately at a dose of 10– –15 mL/kg b.w. and further replenishment of fresh frozen plasma should depend on the volume of red blood cells transfused and the coagulogram. In the case of platelets, they should be transfused in sufficient quantities to maintain a concentration of 50,000/ /mL. To mitigate the effects of hypovolemic shock caused by the injury, transfusion of cryoprecipitate or fibrinogen concentrate may also be considered.

Fluid therapy aims not only to maintain and restore the intravascular volume but also, by optimizing the preload, to increase cardiac output and improve tissue perfusion. Discussions are still ongoing as to whether crystalloids (i.e. 0.9% saline or Ringer’s Lactate) or colloidal solutions (i.e. dextran’s, gelatins, HAES) should be used in the initial phase of fluid resuscitation. De Crescenzo et al. [44] indicates no beneficial effect of hypertonic saline with or without dextran in general trauma patients. In turn, Martin et al. [45] in his meta-analysis indicated that crystalloids were less efficient than colloids at stabilizing resuscitation endpoints. The application of an appropriate fluid resuscitation strategy is, therefore, more important than the type of fluid administered. Malbrain et al. [46] showed that a positive cumulative fluid balance is associated with intra-abdominal hypertension and worse outcomes. Hypotensive fluid resuscitation as shown by numerous studies may offer a survival benefit over conventional fluid resuscitation for trauma patients. It can also further reduce blood loss and thus blood product utilization. It can, therefore, be concluded that fluid resuscitation in trauma patients should be based on a specific compromise between too small a volume leading to hypoperfusion and too much hydration in patients, which may result in increased bleeding due to increased BP. The use of too large a volume of fluid can lead to “dilutive” hemorrhagic bleeding.

As indicated by numerous randomized studies — as also confirmed by this meta-analysis — hypovolemic fluid resuscitation can bring benefits in the management of the trauma patient [28, 29, 47]. However, it should be remembered that the hypotension should not last longer than 1 hour [22, 28]. It should be kept in mind that the rules of fluid resuscitation in trauma patients with concomitant craniocerebral trauma are different. Thereafter, as recommended by the National Institute for Health and Clinical Excellence (NICE), it is important to increase BP (systolic BP 110–120 mmHg) as quickly as possible to secure proper brain perfusion and prevent secondary brain changes.

In this meta-analysis, hypovolemic fluid resuscitation was associated with a higher incidence of thrombocytopenia, renal failure or anemia in comparison with conventional management. The above symptoms are closely related. Thrombocytopenia is the most frequently diagnosed hemorrhagic flaw and may lead to anemia. As many authors indicate, acute kidney injury is a common feature in patients with thrombocytopenia-associated multiple organ failure with incidences as high as 42% in disseminated intravascular coagulation, 58% in thrombocytopenic purpura and 100% in hemolytic uremic syndromes [48–50]. On the other hand, in the case of conventional fluid resuscitation, a statistically significantly higher incidence of acute respiratory distress syndrome (ARDS) or multiple organ dysfunction syndrome (MODS) was observed. Jiang et al. [51] indicate that hypovolemic/restricted fluid resuscitation can effectively eliminate inflammatory factors, improve immune function, maintain the stability of blood components, and reduce the incidences of ARDS and MODS. In the case of hypotension fluid resuscitation, there was also a higher incidence of acute myocardial infarction, which may be caused by lower myocardial overload.

Limitations of the study

The present study has some limitations. First, only in several articles study groups are appropriate, in other articles the sample size is relatively small, which led to a wide 95% CI. Second, two studies referred to mean arterial pressure, the others to systolic blood pressure.

Conclusions

The present study findings show significant associations between hypotensive fluid resuscitation and a decreased risk of adverse events and cardiovascular mortality in hypovolemic shock trauma patients. There is an urgent need for a large multicenter randomized trial to confirm these findings.

Acknowledgments

The study was supported by the ERC Research Net and by the Polish Society of Disaster Medicine.

Conflict of interest: None declared

References