Effectiveness of Preoperative Red Cell Preparation and Intraoperative Massive Transfusion in Brain Tumor Operation

Keywords

massive transfusion - intraoperative transfusions - brain tumor - blood transfusion

Introduction

In the case of a brain tumor, surgical management is the primary therapeutic option for tissue diagnosis, tumor removal, or intracranial pressure reduction. Blood component preparation is one of the preoperative processes that may be ordered before surgery. However, an overabundance of requests for blood preparation has been observed in prior studies, particularly packed red cells (PRCs).[1] [2] [3] [4] Chotisukarat et al found that the crossmatch to transfusion (C/T) ratio was 4.3% in 1,018 individuals who had elective neurosurgery operations.[3] Additionally, Saringcarinkul and Chuasuwan also studied 377 patients who had undergone neurosurgical operations and reported a high C/T ratio of 6.6.[4]

The unexpected vigorous bleeding during tumor resection was a concern for neurosurgeons because the event has been associated with mortality.[5] [6] Therefore, the request for more units of preoperative blood products for the patient's safety has been observed.[3] [4] Nevertheless, the unnecessary blood products that were prepared led to the loss of resources and an increase in the amount of labor done in blood banks. Hence, the preoperative preparation of blood ought to optimize the potential benefit from the tradeoff between unexpectedly massive blood losses and blood waste.

According to the literature review, a few publications have addressed the risk factors of intraoperative transfusions in brain tumor surgery. Skull base tumor, meningioma, children with an age younger than 4 years, operative time longer than 270 minutes, and preoperative hemoglobin lower than 12.2 g/dL were associated with intraoperative PRC transfusion from prior studies.[7] [8] [9] However, various preoperative factors need to be further investigated to enable the establishment of a guideline or the blood component preparation protocol for balancing between preoperative crossmatch blood products and utilization.[4] Hence, the objectives of the present study were as follows: (1) the primary objective was to assess the effectiveness index of blood preparation and utilization; (2) the secondary objective was to explore factors associated with intraoperative transfusion that could be considered to set blood preparation protocol for brain tumor operation in the future; and finally, (3) the third objective was to identify the prevalence and analyze risk factors of massive transfusion.

Methods

Study Design and Study Population

A retrospective cohort study was done by reviewing medical records among brain tumor patients who had undergone cranial operations between January 2014 and January 2019. Exclusion criteria were unavailable crossmatch and transfusion data, unavailable preoperative imaging, unavailable preoperative details, and no definite diagnosis from the pathological report. Preoperative clinical characteristics, laboratory results, and treatment outcome were collected from electronics-based medical records. Preoperative magnetic resonance imaging scans were reviewed for tumor characteristics as follows: tumor size, tumor volume, number of tumors, lateralization, location, and midline shift. In addition, the tumor classification and the World Health Organization (WHO) grading were collected based on the official reports by the pathologist.

The main objective was to describe the effectiveness of blood utilization according to the C/T ratio, transfusion probability (Tp), and transfusion index (Ti) as follows:

A C/T ratio was defined as the number of units cross-matched/number of units transfused[10] [11] and a C/T ratio of 2.0 or below suggested that blood utilization was effective.[12]

Tp was defined as the number of patients transfused/number of patients cross-matched × 100. A Tp of 30% or higher indicated successful blood utilization.

Ti was defined as the number of units transfused/number of patients cross-matched. A value of 0.5 or more was thought to show the effectiveness of blood being used.

For the third objective, the massive transfusion was defined as a patient who received more than 4 units of PRC within 1 hour or more than 10 units of PRC within 24 hours.[13] [14]

Ethical Considerations

A human research ethics committee of the Faculty of Medicine, Prince of Songkla University approved the present study (REC 64–477–10–1). Because of the retrospective study design, patients were not required to provide informed consent. However, patients' identity numbers were encoded before analysis.

Statistical Analysis

Following the study objectives, proportion and percent were used to describe the results of the categorical variables, whereas mean and standard deviation (SD) were used to define continuous variables. Moreover, the C/T ratio, Tp, and Ti were calculated according to the definitions. Therefore, binary logistic regression was used for estimating factors associated with intraoperative transfusion. In addition, factors affecting massive transfusion were analyzed using binary logistic regression with univariate and multivariable analysis. In detail, the predictors were explored using binary logistic regression analysis, and the candidate variables with p-values of 0.10 were identified for multivariable analysis to generate the final model. Hence, multivariable analysis was performed with a backward elimination procedure. Finally, the model that had the lowest Akaike information criterion (AIC) was chosen as the final model. All p-values less than 0.05 were considered statistically significant, and the variance inflation factor (VIF) was used to detect multicollinearity in the final model, with a VIF value of 10 or above indicating multicollinearity.[15] Statistical analysis was performed using R version 4.4.0 software (R Foundation, Vienna, Austria).

Results

Baseline Clinical Characteristics

A total of 1,719 patients underwent screening; however, 11 individuals were excluded according to the exclusion criteria. As a result of this, the remaining 1,708 patients were examined. [Table 1] presents the clinical features and there was a male dominance in the study population. The mean age was 47.6 (SD 17.2) years and the mean body mass index (BMI) was 23.5 (SD 4.4) kg/m². The majority of the American Society of Anesthesiologists (ASA) classification was ASA class 3 in 85.6%, while the emergency operation was observed in 6.8% of the present cohort.

Table 1 Baseline characteristics of the present cohort (N = 1,708)

Characteristics	N (%)
Sex
Male	733 (42.9)
Female	975 (57.1)
Mean age, y (SD)	47.6 (17.2)
Age, y
0–15	137 (8.0)
> 15–30	108 (6.3)
> 30–40	194 (11.4)
> 40–50	450 (26.3)
> 50–50	451 (26.4)
> 60	368 (21.5)
Underlying disease
Hypertension	270 (15.8)
Diabetes mellitus	182 (10.7)
Dyslipidemia	188 (11.0)
Liver disease	28 (1.6)
Renal failure	37 (2.2)
Preoperative seizure	157 (9.2)
Preoperative current medication
Antiplatelet	21 (1.2)
Clexane	7 (0.4)
Warfarin	2 (0.1)
Mean body mass index, kg/m2	23.5 (4.4)
American Society of Anesthesiologists classification
1	3 (0.2)
2	238 (13.9)
3	1,462 (85.6)
4	5 (0.3)
Preoperative laboratory (SD)
Mean hematocrit, %	12.8 (1.6)
Mean hemoglobin, g/dL	39.0 (10.1)
Mean white blood cell count, ×103/µL	10.1 (5.0)
Mean neutrophil, %	67.2 (16.1)
Mean lymphocyte, %	24.6 (12.5)
Mean neutrophil-to-lymphocyte ratio	5.2 (8.0)
Mean platelet count, ×103/µL	290.7 (90.5)
Mean prothrombin time ratio	0.98 (1.87)
Mean international normalized ratio	1.08 (1.23)
Tumor characteristics
Mean diameter of tumor, cm (SD)	3.7 (1.6)
Mean tumor volume, mL (SD)	34.8 (40.4)
Mean preoperative midline shift, cm (SD)	0.31 (0.48)
Midline shift group, cm
< 0.5	1,229 (72.0)
≥ 0.5	479 (28.0)
Tumor location
Supratentorial location	1,451 (85.0)
Infratentorial location	257 (15.0)
Intraventricular tumor	47 (2.8)
Pineal tumor	33 (1.9)
Intraosseous/Skull-based tumor	44 (2.6)
Neurosurgical operation
Craniotomy	965 (56.5)
Craniectomy	113 (6.6)
Suboccipital or rectosigmoid approach	228 (13.3)
Endoscopic transsphenoidal approach	250 (14.6)
Burr hole with biopsy	152 (8.9)
Emergency operation	116 (6.8)
Estimated blood loss, mL	773.2 (1137.0)
Tumor classification
Meningioma	550 (32.2)
Glioma	377 (22.1)
Pituitary adenoma	241 (14.1)
Schwannoma	81 (4.7)
Metastasis	141 (8.3)
Lymphoma	111 (6.5)
Medulloblastoma	21 (1.2)
Craniopharyngioma	38 (2.2)
Neuroblastoma	10 (0.6)
Germinoma	22 (1.3)
Other	111 (6.5)
WHO grade
I	963 (56.4)
II	161 (9.4)
III	91 (5.3)
IV	493 (28.9)
Outcome
Intraoperative transfusion	766 (44.8)
Massive transfusion	79 (4.6)

Abbreviations: SD, standard deviation; WHO, World Health Organization.

For preoperative hematologic laboratories, anemia (hemoglobin less than 10 g/dL) was found at 5.0%, and the mean neutrophil-lymphocyte (NL) ratio was 5.2 (SD 8.0). Craniotomy was the main operation in 56.5%, whereas decompressive craniectomy with tumor removal was found in 6.6% of total cases. In addition, an endoscopic transsphenoidal approach and burr hole with biopsy was performed in 14.6 and 8.9%. For pathological diagnosis, meningioma was the most common brain tumor that was resected in 32.2%. In detail, 90.4% of meningiomas were WHO grade I, while WHO grade II and III meningiomas were found in 8.4 and 1.3%. For gliomas, WHO grade IV gliomas (glioblastoma) were found in 45.6%, whereas WHO grade III, II, and I gliomas were found in 21.5, 24.4, and 8.5%, respectively.

Effectiveness Index of Preoperative Blood Preparation

Almost all patients (98.4%) had preoperative crossmatch preparation ordered for a total of 6,068 PRC units, but 45.5% (766/1,682) of total preparations were used during the operation. [Table 2] shows the C/T ratio, Tp, and Ti of PRC by tumor classification and operation. Overall, C/T ratio, Tp, and Ti were 3.27, 45.54%, and 1.10, respectively. According to tumor classification, all tumors had a C/T ratio greater than 2.0, but meningioma had nearly the effective threshold of this indicator. Surgery of pituitary adenoma and lymphoma had a high C/T ratio and Tp less than 30%, which means that the blood preparations for these tumors were ineffective.

Table 2 Crossmatch to transfusion ratio, transfusion probability, and transfusion index of packed red cells by tumor classification and operation

Tumor classification/ operation	Preoperative preparation		Intraoperative utilization		C/T ratio	Tp (%)	Ti
	Patient with crossmatch (n)	Total crossmatch (units)	Patient received transfusion (n)	Total transfusion (units)
Total	1,682	6,068	766	1,855	3.27	45.54	1.10
Tumor classification
Meningioma	550	2,166	360	1,073	2.02	65.45	1.95
Glioma	368	1,374	138	272	4.95	37.50	0.74
Pituitary adenoma	240	728	60	90	8.09	25.00	0.38
Schwannoma	81	306	36	64	4.78	44.44	0.79
Metastasis	141	509	52	64	7.95	36.88	0.45
Lymphoma	97	279	12	18	15.50	12.37	0.19
Medulloblastoma	20	69	14	24	2.88	70.00	1.20
Craniopharyngioma	38	142	19	31	4.58	50.00	0.82
Neuroblastoma	10	30	4	13	2.31	40	1.30
Germinoma	21	73	10	13	5.62	47.62	0.62
Other	116	420	61	181	2.32	52.59	1.56
Operation
Craniotomy	958	3,701	510	1,251	2.96	53.24	1.31
Craniectomy	112	426	69	207	2.06	61.61	1.85
Suboccipital/retrosigmoid approach	227	824	184	217	3.80	81.06	0.96
Endoscopic transsphenoidal approach	249	745	58	104	7.16	23.29	0.42
Burr hole with biopsy	136	373	11	16	23.31	8.09	0.12

Abbreviations: C/T ratio, crossmatch to transfusion ratio; Ti, transfusion index; Tp, transfusion probability.

All of the operations had a C/T ratio that was greater than 2.0, and almost all of them, with the exception of the endoscopic transsphenoidal and tumor biopsy operations, had a Tp that was lower than 30%. This demonstrated that the preoperative PRC preparations for these procedures were unsuccessful.

Factors Associated with Intraoperative Transfusion

The prevalence of intraoperative PRC transfusion was 44.8% in the present study. According to the secondary objective, factors significantly related to intraoperative PRC transfusion were being female, younger age, lower BMI, ASA classification, preoperative hematocrit, hemoglobin, platelet count, NL ratio, tumor diameter, tumor volume, tumor classification, WHO grade, intraventricular tumor, intraosseous/skull-based tumor, type of operation, and estimated blood loss by univariate analysis. By multivariable analysis with backward elimination method, age, BMI, ASA classification, estimated blood loss, and type of operations are significantly associated with intraoperative PRC transfusion, as shown in [Table 3]. Additionally, the final model's factors all had VIF values under 10.

Table 3 Binary logistic regression analysis for intraoperative transfusion

	Univariate analysis		Multivariable analysis
Factor	Odds ratio (95%CI)	p-Value	Odds ratio (95%CI)	p-Value
Gender
Male	Ref
Female	1.94 (1.59–2.36)	< 0.001
Age, y	0.98 (0.97–0.99)	< 0.001	0.98 (0.97–0.99)	< 0.001
Body mass index, kg/m2	0.94 (0.92–0.96)	< 0.001	0.91 (0.88–0.94)	< 0.001
Underlying disease
Hypertension[a]	1.05 (0.81–1.36)	0.69
Diabetes mellitus[a]	1.00 (0.74–1.37)	0.95
Dyslipidemia[a]	0.96 (0.71–1.31)	0.83
Liver disease[a]	0.92 (0.43–1.95)	0.83
Renal failure[a]	1.16 (0.60–2.24)	0.63
Preoperative seizure[a]	0.96 (0.69–1.33)	0.81
Preoperative current medication
Antiplatelet[a]	0.92 (0.38–2.19)	0.85
Warfarin[a]	1.23 (0.77–19.69)	0.88
Clexane	0.49 (0.09–2.53)	0.39
American Society of Anesthesiologists classification
1–2	Ref		Ref
3–4	0.60 (0.45–0.80)	< 0.001	1.87 (1.28–2.38)	< 0.001
Preoperative hematologic laboratory
Hematocrit, %	0.88 (0.86–0.90)	< 0.001
Hemoglobin, g/dL	0.66 (0.61–0.70)	< 0.001
Platelet count, ×103/µL	1.002 (1.001–1.003)	< 0.001
White blood cell count, ×103/µL	1.01 (0.99–1.03)	0.11
Neutrophil/lymphocyte ratio	1.01 (1.003–1.028)	0.01
Partial thromboplastin time ratio	0.86 (0.60–1.24)	0.44
International normalized ratio	0.96 (0.83–1.10)	0.58
Preoperative hemoglobin level, g/dL
≥ 10	Ref		Ref
< 10	5.82 (3.35–10.10)	< 0.001	13.46 (6.99–25.91)	< 0.001
Tumor location
Supratentorial tumor	Ref
Infratentorial tumor	1.15 (0.88–1.50)	0.29
Intraventricular tumor[a]	2.02 (1.11–3.66)	0.02
Pineal tumor[a]	1.68 (0.84–3.38)	0.14
Intraosseous/Skull-based tumor[a]	2.19 (1.17–4.08)	0.01
Tumor volume, mL	1.016 (1.012–1.019)	< 0.001
Diameter of tumor, cm	1.41 (1.32–1.50)	< 0.001
Preoperative midline shift, cm	1.62 (1.31–2.01)	< 0.001
Emergency operation[a]	0.96 (0.65–1.40)	0.84
Estimated blood loss, ml	1.003 (1.002–1.004)	< 0.001
Estimated blood loss level-, mL
< 500	Ref		Ref
500–1,000	4.59 (3.60–5.85)	< 0.001	4.44 (3.36–5.87)	< 0.001
> 1,000	53.93 (34.55–84.32)	< 0.001	56.04 (34.55–90.89)	< 0.001
Neurosurgical operation
Craniotomy	Ref		Ref
Craniectomy	1.39 (0.93–2.08)	0.09	1.02 (0.61–1.71)	0.92
Suboccipital/rectosigmoid approach	0.95 (0.71–1.27)	0.76	1.43 (0.99–2.06)	0.053
Endoscopic transsphenoidal approach	0.27 (0.19–0.37)	< 0.001	0.72 (0.47–1.09)	0.12
Burr hole with biopsy	0.07 (0.03–0.13)	< 0.001	0.23 (0.11–0.47)	< 0.001
Tumor classification
Meningioma	Ref
Glioma	0.30 (0.23–0.40)	< 0.001
Pituitary adenoma	0.17 (0.12–0.24)	< 0.001
Schwannoma	0.42 (0.26–0.67)	< 0.001
Metastasis	0.30 (0.21–0.45)	< 0.001
Lymphoma	0.06 (0.03–0.11)	<0.001
Medulloblastoma	1.05 (0.41–2.66)	0.90
Craniopharyngioma	0.52 (0.27–1.02)	0.058
Neuroblastoma	0.35 (0.09–1.26)	0.10
Germinoma	0.44 (0.18–1.03)	0.06
Other	0.58 (0.39–0.87)	0.009
Meningioma[a]	3.50 (2.83–4.34)	< 0.001	1.75 (1.28–2.38)	< 0.001
Glioma[a]	0.64 (0.51–0.81)	< 0.001
Pituitary adenoma[a]	0.35 (0.26–0.48)	< 0.001
Schwannoma[a]	0.98 (0.62–1.54)	0.98
Metastasis[a]	0.69 (0.48–0.99)	0.04
Lymphoma[a]	0.13 (0.07–0.24)	< 0.001
Medulloblastoma[a]	2.48 (0.99–6.19)	0.05
Craniopharyngioma[a]	1.23 (0.64–2.35)	0.51
Neuroblastoma[a]	0.81 (0.23–2.91)	0.75
Germinoma[a]	1.02 (0.44–2.38)	0.95
WHO grade
IV	Ref
III	0.65 (0.39–1.09)	0.10
II	1.49 (1.03–2.14)	0.03
I	2.14 (1.71–2.68)	< 0.001

Abbreviations: CI, confidence interval; WHO, World Health Organization.

^a Data show only “yes group” while reference groups (no group) are hidden.

Prevalence and Factors Associated with Massive Transfusion

In this study, massive transfusion was observed in 79.4% of total cases and meningioma, increased diameter of tumor, intraosseous/skull-based tumor, and craniotomy were candidate factors significantly related to the event of the massive transfusion by univariate analysis, as shown in [Fig. 1]. Therefore, multivariable analysis with the backward elimination procedure was performed and found that meningioma, intraosseous/skull-based tumor, and diameter of tumor were all strongly linked to intraoperative PRC transfusion with lowest AIC, as shown in [Fig. 2]. Furthermore, the VIF for every factor included in the final model was less than 10.

Fig. 1 The odds ratio plot of various factors using univariate analysis.

Fig. 2 The odds ratio plot of factors associated with massive transfusion using multivariable analysis.

Discussion

Preoperative PRC preparation was overrequested in the present study, according to various indicators. As a result, more than half of all preparations were not employed that preferred the unnecessary crossmatch and over workload in the routine clinical practice. These findings were consistent with those of previous studies. Based on operation, the C/T ratio, Tp, and Ti of craniotomy with tumor removal were 5, 20%, and 0.5, respectively. Moreover, the endoscopic transsphenoidal approach had the C/T ratio, Tp, and Ti of 11, 7%, and 0.4, respectively, whereas those for the tumor biopsy had a C/T ratio, Tp, and Ti of 12, 8%, and 0.2, respectively.[3] According to the findings of tumor classification, surgery of meningiomas had effective indexes. The concordance results were similar to what had been shown in the Saringcarinkul and Chuasuwan study, which reported the Tp of patients with meningiomas was 49%.[4] However, pituitary adenoma had an imbalance between PRC preparations and utilization in the present study, which could be explained by concerns about the operation being close to internal carotid injury and can result in unexpected massive bleeding during the operation. The endoscopic transsphenoidal approach is the common operation for pituitary adenoma, while tumor biopsy is usually performed for cases of lymphoma. However, vascular complication is uncommon. From the literature review, intraoperative internal carotid injury has been reported in 0.12 to 1.1%,[16] [17] and intraoperative bleeding was reported in 12.3% of neuronavigation-guided biopsy patients.[18] Therefore, the type and screen procedure processes check patient blood for ABO-Rh groups and unusual antibodies that might make donor blood incompatible may be an alternative resolution to reduce unnecessary cross-matches in low probability cases of requiring blood products.[19]

As a result, the ineffectiveness of preoperative PRC preparation and utilization was observed that potentially led to unnecessary costs and the workload of a blood bank. Balancing between preoperative crossmatch and actual transfusion has been challenged. Currently, no standard guideline exists for crossmatch protocol or the Maximum Surgical Blood Order Schedule (MSBOS) in brain tumor operation. There are several methods for contributing to the guideline or MSBOS, for example, nomogram and machine learning (ML).[20] [21] [22] In the past, prior studies calculated MSBOS by the following equation (1.5*Ti)[3] [4] or consensus according to the procedures from prior studies.[23] [24] In addition, Hu et al developed a nomogram predicting a transfusion in patients undergoing total knee arthroplasty from various predictors with multivariable analysis and reported the area under the curve ranged from 0.839 to 0.884 for the predictability.[25] ML is a sophisticated computer technology that learns from data to discover patterns and make predictions.[26] [27] Liu et al used ML to predict PRC transfusion in mitral valve surgery and found that the accuracy of prediction was 86.8%.[26] In addition, Huang et al predicted PRC transfusion using various algorithms of ML and reported that the random forest algorithm had the best performance of prediction with 82.35%.[27] Therefore, predicting the PRC transfusion in patients with brain tumors by novel methods has been challenged. To create a predictive model in clinical prediction tools, feature selection is a critical step, and one technique of feature selection may be to investigate the significant factors associated with PRC transfusion using multivariable analysis.[28] [29]

Younger age and low BMI were the significant factors related to intraoperative PRC transfusion in the present study. Similarly, Vassal et al found that in brain tumor patients who were younger than 4 years the risk of intraoperative transfusion was explained by tolerance blood loss in children less than adults. Hemorrhagic shock in children was more common than in adults from prior studies.[30] Additionally, previous studies reported that skull-based surgery and meningioma are potential factors linked with blood transfusion.[8] [9] These were in concordance with our findings which shows that intraosseous/skull-based tumor and meningioma were associated with both intraoperative transfusion and massive blood loss. Although meningioma is a benign tumor, hypervascularity and numerous feeding vessels are common findings of this tumor.[31] The sunburst flow void was found in 96.5% of the cases, whereas the serpentine flow void was found in just 3.5% of meningiomas.[32] Intraosseous tumor removal and skull-based surgery are complex procedures that frequently bleed from various vessels, including branches of the carotid artery in the basilar skull, the diploic vein in the cranial vault, and bridging veins near to the superior sagittal sinus during craniotomy.[1] [33] [34]

As per the authors' knowledge, the present study is the first study that mentioned predictors linked to intraoperative transfusion for brain tumor surgery that may be used to create the clinical prediction tools and MSBOS in the future. However, there were certain limitations in the present study that should be acknowledged. The current study was a retrospective cohort analysis, which might have resulted in bias from confounding variables.[35] Nevertheless, we attempted to adjust and control bias using multivariable analysis in the present study. Additionally, the incidence of intraoperative massive bleeding and transfusion has been reported in the range of 3 to 8% for cranial operations. Multicenter trials should be conducted in the future to address the increased occurrence of this complication for testing the predictive model's performance and will be useful in making guidelines. Finally, our hospital did not follow an autologous blood transfusion protocol during routine practice.[36] The present study's findings may help physicians identify high-risk operations and plan for autologous blood transfusions during surgery, which will reduce PRC transfusion and utilization ratios.[37] [38]

Conclusion

Unnecessary preoperative blood component preparation for brain tumor surgery was noticed in routine practice. Exploring factors that are strongly associated with intraoperative transfusion and massive bleeding has posed a challenge in optimizing between crossmatch and actual use; moreover, those will be developed into a crossmatch guideline in the future.

• References

• 1 Kisilevsky A, Gelb AW, Bustillo M, Flexman AM. Anaemia and red blood cell transfusion in intracranial neurosurgery: a comprehensive review. Br J Anaesth 2018; 120 (05) 988-998
• 2 Boutin A, Chassé M, Shemilt M. et al. Red blood cell transfusion in patients with traumatic brain injury: a systematic review and meta-analysis. Transfus Med Rev 2016; 30 (01) 15-24
• 3 Chotisukarat H, Akavipat P, Sookplung P. et al. Effectiveness index of preoperative blood preparation for elective neurosurgery at Prasat Neurological Institute. Thai J Anesthesiol 2017; 43: 232-240
• 4 Saringcarinkul A, Chuasuwan S. Maximum surgical blood order schedule for elective neurosurgery in a university teaching hospital in Northern Thailand. Asian J Neurosurg 2018; 13 (02) 329-335
• 5 Barbosa RR, Rowell SE, Sambasivan CN. et al; Trauma Outcomes Group. A predictive model for mortality in massively transfused trauma patients. J Trauma 2011; 71 (2, Suppl 3): S370-S374
• 6 Akaraborworn O, Chaiwat O, Chatmongkolchart S. et al. Prediction of massive transfusion in trauma patients in the surgical intensive care units (THAI-SICU study). Chin J Traumatol 2019; 22 (04) 219-222
• 7 Vassal O, Desgranges FP, Tosetti S. et al. Risk factors for intraoperative allogeneic blood transfusion during craniotomy for brain tumor removal in children. Paediatr Anaesth 2016; 26 (02) 199-206
• 8 Lagman C, Sheppard JP, Beckett JS. et al. Red blood cell transfusions following resection of skull base meningiomas: risk factors and clinical outcomes. J Neurol Surg B Skull Base 2018; 79 (06) 599-605
• 9 Neef V, König S, Monden D. et al. Clinical outcome and risk factors of red blood cell transfusion in patients undergoing elective primary meningioma resection. Cancers (Basel) 2021; 13 (14) 3601
• 10 Raghuwanshi B, Pehlajani NK, Sinha MK, Tripathy S. A retrospective study of transfusion practices in a Tertiary Care Institute. Indian J Anaesth 2017; 61 (01) 24-28
• 11 Zewdie K, Genetu A, Mekonnen Y, Worku T, Sahlu A, Gulilalt D. Efficiency of blood utilization in elective surgical patients. BMC Health Serv Res 2019; 19 (01) 804
• 12 Ejaz A, Frank SM, Spolverato G, Mavros M, Kim Y, Pawlik TM. Variation in the use of type and crossmatch blood ordering among patients undergoing hepatic and pancreatic resections. Surgery 2016; 159 (03) 908-918
• 13 Patil V, Shetmahajan M. Massive transfusion and massive transfusion protocol. Indian J Anaesth 2014; 58 (05) 590-595
• 14 Chidester SJ, Williams N, Wang W, Groner JI. A pediatric massive transfusion protocol. J Trauma Acute Care Surg 2012; 73 (05) 1273-1277
• 15 Vatcheva KP, Lee M, McCormick JB, Rahbar MH. Multicollinearity in regression analyses conducted in epidemiologic studies. Epidemiology (Sunnyvale) 2016; 6 (02) 227
• 16 Usachev D, Sharipov O, Abdali A. et al. Internal carotid artery injury in transsphenoidal surgery: tenets for its avoidance and refit-a clinical study. Brain Sci 2021; 11 (01) 99
• 17 Al-Shami H, Alnemare AK. Inadvertent internal carotid artery (ICA) injury during transsphenoidal surgery: review of literature. Egypt J Neurosurg 2021; 36: 4
• 18 Taweesomboonyat C, Tunthanathip T, Sae-Heng S, Oearsakul T. Diagnostic yield and complication of frameless stereotactic brain biopsy. J Neurosci Rural Pract 2019; 10 (01) 78-84
• 19 Chaudhary R, Agarwal N. Safety of type and screen method compared to conventional antiglobulin crossmatch procedures for compatibility testing in Indian setting. Asian J Transfus Sci 2011; 5 (02) 157-159
• 20 Tunthanathip T. Translational Medicine in Neurosurgery. Bangkok, Thailand: Sahamit Pattana Printing; 2022
• 21 Wang J, Zhao Y, Jiang B, Huang X. Development of a nomogram to predict postoperative transfusion in the elderly after intramedullary nail fixation of femoral intertrochanteric fractures. Clin Interv Aging 2021; 16: 1-7
• 22 Mitterecker A, Hofmann A, Trentino KM. et al. Machine learning-based prediction of transfusion. Transfusion 2020; 60 (09) 1977-1986
• 23 Mcpherson RA, Pincus MR. Guidelines for ordering blood for elective surgery also referred to as maximum surgical blood order schedule (MSBOS). In: Mcpherson RA, Pincus MR. eds. Henry's Clinical Diagnosis and Management by Laboratory Methods. 24 ed. Missouri: Elsevier; 2021: 1664-1665
• 24 Universiti Kebangsaan Malaysia. Guidelines on Blood and Blood Components Transfusion: Maximum Surgical Blood Order Schedule (MSBOS) For Elective Surgery. [Internet]. 2021 . Accessed September 28, 2021, at: https://hctm.ukm.my/makmal/wp-content/uploads/2020/10/2-MSBOS-ed.5.pdf
• 25 Hu C, Wang YH, Shen R. et al. Development and validation of a nomogram to predict perioperative blood transfusion in patients undergoing total knee arthroplasty. BMC Musculoskelet Disord 2020; 21 (01) 315
• 26 Liu S, Zhou R, Xia XQ. et al. Machine learning models to predict red blood cell transfusion in patients undergoing mitral valve surgery. Ann Transl Med 2021; 9 (07) 530
• 27 Huang X, Wang Y, Chen B. et al. Ability of a machine learning algorithm to predict the need for perioperative red blood cells transfusion in pelvic fracture patients: a multicenter cohort study in China. Front Med (Lausanne) 2021; 8: 694733
• 28 Tunthanathip T, Oearsakul T. Application of machine learning to predict the outcome of pediatric traumatic brain injury. Chin J Traumatol 2021; 24 (06) 350-355
• 29 Taweesomboonyat C, Kaewborisutsakul A, Tunthanathip T, Saeheng S, Oearsakul T. Necessity of in-hospital neurological observation for mild traumatic brain injury patients with negative computed tomography brain scans. JHSMR 2020; 28: 267-274
• 30 Gonzalez KW, Desai AA, Dalton BG, Juang D. Hemorrhagic shock. J Pediatr Intensive Care 2015; 4 (01) 4-9
• 31 Lagman C, Ong V, Nguyen T. et al. The meningioma vascularity index: a volumetric analysis of flow voids to predict intraoperative blood loss in nonembolized meningiomas. J Neurosurg 2018; 9 (11) 1-6
• 32 Wang C, Xu Y, Xiao X, Zhang J, Zhou F, Zhao X. Role of intratumoral flow void signs in the differential diagnosis of intracranial solitary fibrous tumors and meningiomas. J Neuroradiol 2016; 43 (05) 325-330
• 33 Sasaki K, Saito A, Nishijima Y. et al. Giant intraosseous meningioma associated with calvarial hyperostosis and subcutaneous invasion: case reports and literature review. Asian J Neurosurg 2021; 16 (03) 589-594
• 34 Ishihara H, Ishihara S, Niimi J. et al. The safety and efficacy of preoperative embolization of meningioma with N-butyl cyanoacrylate. Interv Neuroradiol 2015; 21 (05) 624-630
• 35 Kahlert J, Gribsholt SB, Gammelager H, Dekkers OM, Luta G. Control of confounding in the analysis phase - an overview for clinicians. Clin Epidemiol 2017; 9: 195-204
• 36 Tunthanathip T, Sae-Heng S, Oearsakul T, Kaewborisutsakul A, Taweesomboonyat C. Economic impact of a machine learning-based strategy for preparation of blood products in brain tumor surgery. PLoS One 2022; 17 (07) e0270916
• 37 Forgie MA, Wells PS, Laupacis A, Fergusson D. International Study of Perioperative Transfusion (ISPOT) Investigators. Preoperative autologous donation decreases allogeneic transfusion but increases exposure to all red blood cell transfusion: results of a meta-analysis. Arch Intern Med 1998; 158 (06) 610-616
• 38 Tunthanathip T. Malignant transformation in low-grade astrocytoma for long-term monitoring. J Cancer Res Ther 2022; 18 (06) 1616-1622