A flow diagram of the included studies is shown in Fig. 1. After the literature search, 691 records were identified. A total of 109 duplicates and 554 irrelevant articles, were removed after screening the titles and abstracts, resulting to 28 reports for full-text review that were considered potentially relevant. Overall, 7 articles met the inclusion criteria and were included in our study [8, 10,11,12,13, 22, 23]. All studies included in our analysis were nonrandomized cohort studies conducted in the North America between 2014 and 2022, comprising six studies from the United States [8, 10,11,12,13, 23] and one study from Canada [22]. The reasons for the exclusion of studies were as follows: no comparison group (n = 14); for military purpose (n = 3); lacks outcome of interest (n = 2); and no related intervention (n = 2).

The flow diagram of the systematic review

The study characteristics are summarized in Table 1. The analyzed studies involved 4,095 patients, of whom 1,692 and 2,403 received and did not receive PH-TQ, respectively. The studies compared the effect of PH-TQ, with no TQ in five studies [8, 11, 13, 22, 23]; PH-TQ vs. late-TQ in one study [10]; and PH-TQ vs. (late-TQ + no-TQ) in one study [12]. The study participants were mostly males with an average age < 45 years. The participant characteristics in each study are shown in Supplementary 2. All analyzed studies provided crude numbers of mortality, whereas only four [8, 11, 13, 23] controlled for confounders using matching or regression approaches. Three types of blood products were analyzed for infusion: packed red blood cells (pRBCs), platelets, and plasma. Blood transfusion in the first 24 h was reported in five studies [8, 11, 12, 22, 23], blood transfusion in the first hour was reported in one study [10], and the total amount of blood products required in another study [13]. All the analyzed studies documented various side effects following TQ use. These included amputation in all studies [8, 10,11,12,13, 22, 23], compartment syndrome in four [10, 11, 13, 22], thromboembolic complication in two [13, 23], nerve palsy in two [12, 13], and infection in two [13, 23]. Individual studies have reported other complications in the pulmonary, cardiac, musculoskeletal, and renal system [11, 13, 23]. The outcomes and potential complications associated with TQ use reported in each study are shown in Supplementary 3. The overall risk of bias varied from moderate to critical, and most of the included studies were subject to a serious to critical risk of bias owing to the risk of confounding factors. The results of the risk of bias assessment are shown in Fig. 2.

Results of risk of bias assessment for included studies using the ROBINS-I tool

The results are summarized in Fig. 3, and detailed information on each outcome is provided in the Supplementary Appendix. For the primary outcome, the pooled results demonstrated applying PH-TQ (compared to those who did not receive PH-TQ) was associated with a significantly lower mortality rate (OR, 0.48; 95% CI 0.27–0.86; I2, 47%) (Fig. 3 and Supplementary 4). In the sensitivity analysis of primary outcome, the results remained statistically significant when using only adjusted outcomes in four studies (aOR, 0.34, 95% CI 0.19–0.59, I2, 0%) (Supplementary 5). For secondary outcomes, the use of PH-TQ is not associated with reduced use of blood products such as pRBC (MD, -2.09 [unit]; 95% CI -5.00–0.82, I2, 99%) or FFP (MD, -1.0 [unit]; 95% CI -4.0–2.0; I2, 98%) (Fig. 3, Supplementary 6–7) No significant difference was observed regarding the length of hospital stay (MD, -0.80 [day]; 95% CI -2.90–1.30, I2, 66%) or intensive care unit length of stay (MD, -0.51 [day]; 95% CI -2.08–1.06; I2, 92%) (Fig. 3, Supplementary 8–9).

Forest plot of study outcome

For the safety outcomes, the use of PH-TQ is not associated with increased amputation (OR, 0.85; 95% CI 0.43–1.68; I2, 60%) and compartment syndrome (OR, 0.94; 95% CI, 0.37–2.35; I2, 0%) (Fig. 3, Supplementary 10–11). The sensitivity analysis utilizing data from studies comparing PH-TQ use and no TQ use revealed that, despite a reduction in heterogeneity among studies, the utilization of PH-TQ is not significantly associated with secondary outcomes, including safety outcomes (Supplementary 12–17). The funnel plot, which exhibit asymmetry, is shown in Supplementary 18. Because of the limited number of studies included (n < 10), the Egger’s test was not performed. The certainty of evidence (CoE) using the GRADE method for the treatment recommendations of individual endpoints is summarized in Table 2. The CoE was downgraded by domains in the risk of bias and inconsistency owing to the high proportion of studies with a serious risk of bias and high heterogeneity across studies. An asymmetric funnel plot for mortality outcomes was observed. Although only seven studies were included in the review, the publication bias domain was downgraded as concerns arose. In summary, the overall CoE showed very low quality of evidence across all reported endpoints.