2.1 Patients
All protocols and procedures used in this study were reviewed and approved by the Ethics Committee of Huzhou Hospital. The patients enrolled in this work provided informed consent. The inclusion criteria for enrollment were as follows: unilateral shoulder arthroscopy, degenerative injury, and complete medical data, while exclusion criteria included: traumatic rotator cuff injury, long-term bleeding disorders, blood transfusion throughout the assessment period, and multiple injuries. We retrospectively reviewed a total of 59 patients (23 males and 36 females) who had undergone shoulder arthroscopy between December 2019 and June 2020. Afterward, we recorded data which comprised of gender, age, weight, height, body mass index (BMI), the level of SIS, preoperative and postoperative hematocrit (Hct) and hemoglobin (Hb), operative time, hypertension, and diabetes mellitus.
2.2 Surgical technique and postoperative therapy
Based on the American Society of Anesthesiologists (ASA) physical score, all individuals received general anesthesia. The operation was performed with the patient lying in a lateral decubitus position by experienced surgeons (SZ). Hypotensive anesthesia, irrigated with epinephrine saline solution, and intra-articular tranexamic acid was applied to reduce bleeding and to provide a clear operative field. After the operation, the patients routinely received short-term postoperative analgesia.
2.3 Management of blood loss
Intraoperative visible blood loss was too little to be assessed. In addition, no patient underwent wound drainage. Therefore, we ignored it, and TBL was approximately considered as HBL. Notably, it is generally believed that fluid shifts have been largely completed and the hemodynamics of patients become stable on the second or third days after surgery [11]. Thus, a complete blood count including Hct and Hb of all patients were recorded before the operation and on the third postoperative day.
2.4 Calculation of the hidden blood loss
We computed the patient’s blood volume (PBV) as described by Nadler et al. as follows [12]:
PBV (L) = k1 × height (m)3 + k2 × weight (kg) + k3,
where k1 = 0.3669, k2 = 0.03219, and k3 = 0.6041 are for males, whereas k1 = 0.3561, k2 = 0.03308, and k3 = 0.1833 are for females.
In this study, the TBL was estimated using the formula developed by Gross et al. as stated below [13]:
TBL (L)=PBV(L)×((Hctpre-Hctpost))/Hctave
Where Hctpre is the initial preoperative Hct, Hctpost is the Hct on the second or third day postoperatively, and Hctave is the average of Hctpre and Hctpost.
We recorded perioperative Hb so as to calculate Hbloss according to the following formula:
Hbloss(g/L) = Hbpre-Hbpost
2.5 Level of SIS
Here, we classified SIS into three stages as described by Neer et al.. In particular, patients with stageⅡ impingement underwent arthroscopic surgery after failure of conservative treatment. Subsequently, those patients with stage Ⅲ impingement were also included.
2.6 Additional measurements
A hemoglobin level of < 130 g/L for men and < 120 g/L for women was defined as anemia based on the World Health Organization (WHO) criteria [14].
2.7 Statistical analysis
All data analyses were implemented using SPSS software, version 26.0 (SPSS, Chicago, IL). A level of P < 0.05 was considered statistically significant. Mean ± standard deviations were used as descriptive statistics. First, we used the Shapiro-Wilk test to assess the normality of variables, while the paired sample t-test was used to compare the differences between pre- and post-operative Hb levels and Hct values. We then examined the significant differences between the two stages using the independent sample’ Student t-test and χ2 test. Additionally, we adopted the χ2 test to compare the pre- and post-operative incidence of anemia. Finally, multivariate linear regression analysis was performed to identify the influencing factors associated with HBL, including three quantitative variables (age, BMI, and operative time) and four qualitative variables (sex, stage of SIS, diabetes, and hypertension). A positive coefficient indicates a positive influence on the dependent variable (HBL), whereas a negative coefficient reflects a negative influence.