A retrospective analysis was conducted on data from 369 patients with posterior pelvic ring injuries treated with PMISS at our hospital from January 2015 to January 2022. This sample size was calculated using Solvin’s formula. Inclusion criteria are as follows: (1) patients with posterior pelvic ring injury and sacroiliac screw fixation; (2) AO/OTA classification: 61b, 61c1.2, 61C1. 3 [18]; (3) blood examination was performed before and after operation; (4) no history of thrombosis, abnormal coagulation function, and other high-risk hemorrhagic diseases; (5) the clinical data were complete; (6) signed informed consent. Exclusion criteria are as follows: (1) accompanied by anterior fixation (except staged operation with external fixator); (2) old fracture and pathological fracture; (3) AO/OTA classification: 61A, 61c1. 1, 61C2, 61C3; (4) history of preoperative blood transfusion and long-term use of anticoagulants; (5) chronic hemorrhagic diseases, hematological diseases, and malignant tumors; (6) incomplete clinical data; (7) the rehydration volume was more than 2000 m1 [19] within 24 h after operation. According to the exclusion criteria, a total of 291 patients were excluded from the study. Among them, 276 patients underwent plate fixation of the anterior pelvic ring, 7 cases were rehydrated more than 2000 ml within 24 h after operation, 4 cases had incomplete clinical data, and 4 cases received a preoperative blood transfusion. The CONSORT flow chart of the included patients is shown in Fig. 1. Therefore, a total of 78 patients were included in this study with an average of (47.35 ± 5.23) years. There were 42 males and 36 females. The study protocol was approved by Ningbo NO.6 Hospital Institutional Review Board (No. L2022090). The CONSORT checklist of this study protocol are shown in Supplementary Material 1. All methods were performed in accordance with the relevant guidelines and regulations. All patients and their families signed the informed consent.

CONSORT flow chart for patient recruitment

All patients needed a clean enema the night before the operation. Under general anesthesia, the patient took the flat position and underwent C-arm fluoroscopy positioning (pelvic entrance position, exit position, and lateral position). An incision was made about 1 cm on the side of the posterior superior iliac spine, separated the soft tissue to the iliac bone, and then placed the positioning sleeve. The guide needle with a diameter of 2 mm was vertically inserted into the posterior side of the iliac wing. and the S1 vertebral body was inserted between the superior sacral notch and the S1 anterior fissure through the sacroiliac joint space, the sacral auricular surface, and the sacral wing. Through the sacroiliac joint space, sacral auricular surface, and sacral wing, the guide needle entered the S1 vertebral body. During the placement of the guide needle, we need to keep the guide needle parallel to the superior sacral notch and tilt from the rear outside to the front inside at an angle (25°) to the coronal plane of the pelvis. The C-arm fluoroscopy should be repeated to observe and adjust the position of the guide needle in the process of needle threading. According to the need of fixation, another screw can be fixed in the S1 vertebral body repeatedly or screwed at the S2 level. All hollow lag screws were 6.5 mm in diameter and were produced by China company (Weigao).

Antibiotics were applied twice within 24 h after operation. Recheck blood routine and give blood transfusion according to anemia. Lower limb activities began on the second day after the operation. After lying in bed for 4–6 weeks, the patient started to move out of bed and gradually walk with weight on the affected limb. The X-ray would be checked regularly.

The basic information of patients from the electronic medical record system of our hospital was collected, including the patient’s age, gender, body mass index (BMI), surgical duration, time to surgery, anesthesia, anemia, blood transfusion history, smoking history, alcohol, hypertension, diabetes, cardiovascular and cerebrovascular diseases, HB, HCT, albumin, type of posterior pelvic ring injury, fixed position, and depth (Fig. 2). The injury of the posterior pelvic ring was divided into three types. Type I was an unstable fracture of the iliac wing which can be combined with partial injury of the sacroiliac joint. Type II was an unstable injury of the sacroiliac joint and type III was an unstable fracture of the sacrum. All indicators were measured by the same researcher 3 times, and then the average value was taken to reduce the error. Hemoglobin was used to assess the patient’s anemia. According to the definition of the World Health Organization, men < 130 g/L and women < 120 g/L were defined as anemia [20].

Definition of screw depth: make a connecting line at the front edge of the sacroiliac joint on both sides at the pelvic entrance position and divide it into four equal parts. The sacroiliac screw head can reach area AB, area BC, area CD and area DE, which are I-degree, II-degree, III-degree, and IV-degree, respectively

The blood volume (BV) of the patient was calculated according to the equation proposed by Nadler et al. [21]:

$$}=}1\times }3+}2\times }+}3$$

(BV: the preoperative blood volume (L); H: height (m); W: body mass (kg); male: k1 = 0.3669, k2 = 0.03219, K3 = 0.6041; female: k1 = 0.3561, k2 = 0.03308, k3 = 0.1833).

VBL = blood volume in the suction bottle + increased blood volume such as gauze − flushing fluid volume.

The loss of red blood cells in postoperative blood can be calculated by the difference between preoperative and postoperative HCT.

Red blood cell loss = BV × (HCT before and HCT after) + blood transfusion.

TBL = red blood cell loss (L)/HCT mean [22].

(HCT mean = 1/2 × (HCT before + HCT after), the results of the HCT examination on the morning of the second day after admission were HCT before, and the results of the HCT examination on the morning of the third day were HCTafter).

HBL = TBL − VBL.

HBL ratio = HBL/TBL.

All data were described and analyzed by SPSS 22.0 (IBM, USA). The normality test of measurement data was carried out by Kolmogorov Smirnov. The description of measurement data adopted xˉ ± s or median (interquartile range) according to the data distribution type. The Pearson correlation test was used to analyze measurement data. Rate or percentage was used to describe counting data. At the same time, the Spearman correlation was used for statistical analysis. The variables with high correlation were selected, and the independent risk factors related to hidden blood loss were determined by hierarchical regression analysis. A positive coefficient indicated a positive effect on the dependent variable (hidden blood loss), while a negative coefficient indicated a negative effect on the dependent variable. P < 0.05 was considered statistically significant.