Predictors of massive transfusion protocols activation in patients with trauma in Korea: a systematic review

Background

Trauma is a major cause of death in Korea, and hemorrhage is a major cause of early death in trauma. Various studies have investigated blood transfusion strategies in patients with trauma [1,2]. In particular, because of the importance of developing a strategy to control hemorrhagic shock, many guidelines and studies have recommended that each institution should implement massive transfusion protocols (MTPs) [3,4]. The implementation of MTP reduces mortality and overall blood product usage in patients with trauma. Although predicting the need for MT is difficult, various predictive scores have been developed and validated, such as the Assessment of Blood Consumption (ABC) score, Trauma-Associated Severe Hemorrhage (TASH) score, and shock index (SI) [5–7]. However, these predictive scores are not often used in Korea. In Korea, regional trauma centers were designated in 2012 to organize a national trauma system for the first time, and it can be said that the government is currently designing and operating 17 regional trauma centers. Our institution is one of these trauma centers, and we have implemented and developed our own MTP [8]. Except for a few trauma centers, emergency medical institutions that apply MTPs are scarce. In an observational study conducted in Korea, only 14% of patients received MT due to trauma [9]. Recently, an increasing number of institutions have been started MTPs in patients with trauma, and related studies are investigated; However, these studies are lacking in Korea. Therefore, it is necessary to analyze the current influence of MTPs in patients with trauma. In particular, as MTPs activation at the appropriate time improves outcomes in patients with trauma, it is essential to investigate the related studies in our region.

Objectives

The aim of this systematic review was to identify and assess the predictors of MTPs activation in patients with trauma in Korea.

Search strategy

The PubMed, Embase, Cochrane Library, Research Information Sharing Service databases, KoreaMed, and KMbase were searched from November 2022. Key terms such as “trauma,” “Korea,” “transfusion” and “massive transfusion” were sorted. The full search strategy is outlined in Material S1.

All studies conducted in Korea that utilized predictors of MTPs activation in adult patients with trauma were included. MT was defined as the presence of 10 or more packed red blood cells in the first 24 hours. Included studies were to require to have complete information: total number of patients, information on predictors, and methodology for developing the prediction score. The following studies were excluded: pediatric studies, small case studies (<20 patients), gray literature, and articles mentioning MT for nontraumatic injuries, and articles that only predicted transfusion but not MT.

Screening process and data extraction

All articles identified in the search were screened independently by two investigators. Full texts were reviewed and assessed for eligibility based on the defined inclusion and exclusion criteria. Disagreements were resolved independently by a third reviewer at all stages. Data were extracted through discussions with the third reviewer. The details collected from the data included year of publication, number of patients, study design (prospective/retrospective case-control or cohort study, randomized controlled trial), and the inclusion and exclusion criteria. The outcomes of this systematic review were the predictors of MTPs activation in patients with trauma in Korea. The parameters included for predictions were odds ratios with 95% confidence intervals, sensitivities, specificities, and areas under the receiver operating characteristic curve (AUROC).

Quality assessment and analysis

The quality of the observational studies was assessed using the Newcastle-Ottawa scale. A detailed assessment of the risk of bias is available in Table 1 [10–19]. The data analysis used was qualitative, allowing for an overall interpretation of the data based on a qualitative summary. It would not be appropriate to summarize the overall predictors because of the heterogeneity in patient numbers and predictors. The results were reported in accordance with the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines (Table S1).

Table 1.

Newcastle-Ottawa scale

Included studies

The literature search yielded 2,132 articles. After removing duplicated, 1,934 articles underwent title and abstract screening for eligibility, of which 24 articles underwent full-text assessment for eligibility. Fifteen more articles were excluded, one article was added through database of Korea, and 10 articles were finally eligible for inclusion. The agreement for the selection of articles after a full-text review by the two independent investigators (DS and JP) was reached before resolution by the third reviewer (KJ). Disagreement in selection was due to articles being excluded because they involved pediatric patients and not being chosen for MT but for transfusion. The PRISMA flowchart for this study is shown in Fig. 1.

Fig. 1.

PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) flowchart of literature search.

Study description

The characteristics and predictive abilities of the included articles are shown in Tables 2 and 3 [10–19], respectively. Ten retrospective observational studies were identified by the search. Clinical assessments such as prehospital systolic blood pressure (SBP), shock index (SI), prehospital SI, modified SI (MSI), modified early warning system (MEWS), diastolic blood pressure (DBP), and reverse SI multiplied by the Glasgow Coma Scale (rSIG) were most frequently used, reported in five articles. Laboratory values, such as lactate level, fibrinogen degradation product (FDP) to fibrinogen ratio, and rotational thromboelastometry (ROTEM), were reported in four articles. Imaging examinations, such as evaluation of pelvic bleeding scores, were reported in one article.

Table 2.

Characteristics of studies in the systematic review

Table 3.

Performance of predictors to activate massive transfusion protocol

Park et al. [10] validated the utility of prehospital SBP and SI in predicting the need for MT in a retrospective observational study. In this study, the patients were stratified by age and use of antihypertensive drugs. MT was predicted based on a prehospital SBP of 110 mmHg in patients older than 65 years who were using antihypertensive drugs (P=0.024). In addition, MT was predicted based on an SI of 1.0 in patients older than 65 years who were not using antihypertensive drugs (P=0.002) and those younger than 65 years who were not using antihypertensive drugs (P=0.044). Lee et al. [11] evaluated all patients retrospectively utilizing DBP for predicting MT. The study reported that an initially decreased DBP was an independent predictor for MT (P<0.001); the AUROC value for the prediction of MT was 0.777. Wang et al. [12] conducted a retrospective study. In this study, prehospital SI and prehospital MSI (heart rate [HR] / mean arterial blood pressure at the prehospital stage) were used as predictors of MT, with AUROC values of 0.773 and 0.765, respectively. Lee et al. [13] validated the utility of the rSIG (SBP / HR × Glasgow Coma Scale) for predicting the need for MT in a retrospective observational study. This study reported that the rSIG was an independent predictor of MT, and the AUROC value for the prediction of MT was 0.842. Chae et al. [14] retrospectively evaluated patients with traumatic brain injury (TBI), utilizing MEWS to predict MT (P<0.001); the AUROC value was 0.808. Kyoung et al. [15] conducted a retrospective study in which lactate levels were used as a predictor of MT (P=0.039); the AUROC value was 0.790. Lee et al. [16] retrospectively evaluated all patients using the FDP/fibrinogen ratio to predict MT (P=0.029). Baik et al. [17] validated the utility of ROTEM for predicting the need for MT in a retrospective observational study, and the AUROC value of ROTEM was 0.860. Park et al. [18] retrospectively evaluated all patients, TBI patients (Abbreviated Injury Scale ≥3), and non-TBI patients using lactate levels to predict MT. This study reported that lactate levels were independent predictor of MT in all patients and non-TBI patients (P<0.001), and the AUROC values for the prediction of MT were 0.779 and 0.842, respectively. Lee et al. [19] retrospectively evaluated patients with pelvic bone fractures utilizing pelvic bleeding score for predicting the need for MT; the AUROC value was 0.718.

Through our search, we identified studies conducted in Korea on predictors of MT and which used tools that requires clinical assessments, laboratory values, or imaging examinations only. ROTEM had the highest AUROC value of 0.86, followed by rSIG, MEWS, SI, and lactate level.

Limitations

This systematic review has some limitations. First, the heterogeneity of the studies led to a qualitative, not quantitative, summation of the results. Second, all studies were retrospective and observational. Third, each study had different inclusion criteria, such as age, presence of TBI, and presence of pelvic bone fracture.

Conclusions

Our systematic review identified predictors of MTPs activation in patients with trauma in Korea; predictions were performed using tools that requires clinical assessments only, laboratory values only, or imaging examinations only. Among them, ROTEM, rSIG, MEWS, SI, and lactate level showed good effects for predicting the need for MT. The use of predictors for MTPs activation should be individualized based on hospital resource and skill set and should act as clinical decision support tools. In future studies, a comprehensive study on MT in patients with and without trauma in Korea is required. Furthermore, it would be helpful to establish a cohort of MT in Korea.