Population
From August 2015 to August 2019, patients with intertrochanteric fractures undergoing PFNA (short) were retrospectively enrolled in this study. Before February 2018, patients with PFNA were not treated with TXA. After February 2018, all the patients with PFNA were treated with topical TXA. The inclusion criteria were (1) age ≥70 years at the time of injury; (2) a confirmed diagnosis of intertrochanteric fractures classified according to AO type by X-ray or CT; and (3) eligibility for intertrochanteric fracture surgery using the PFNA procedure, as determined by the senior orthopaedic surgeon. The exclusion criteria were (1) allergy to TXA or low-molecular-weight heparin; (2) old, multiple or pathological fractures; (3) severe dysfunction of the heart, lung, liver, or kidney or coagulation dysfunction; (4) anticoagulant therapy such as antiplatelet drugs or warfarin before the operation; (5) recent or ongoing thromboembolic events including deep venous thrombosis, pulmonary embolism, arterial thrombosis, cerebral thrombosis, or stroke; and (6) follow-up of less than 1 month. A total of 218 patients were enrolled, 82 patients were administered topical TXA (study group), and 136 patients were not administered topical TXA during PFNA.
Intraoperative and postoperative procedures
All the patients underwent a standard surgical procedure by a single senior surgeon who specialized in hip and trauma surgery and had 8 years of experience. All patients received spinal or general anaesthesia. In the study group, the wound was bathed in 3 g/100 ml TXA solution for 5 min after greater trochanter exposure and before wound closure. For all patients, one drain was placed in the wound. The drain was clamped for 6 h and then released. The limb pneumatic pump was used for all patients on the first day after the operation (pneumatic pump treatment should be stopped if lower extremity deep vein thrombosis occurs after surgery). All patients received standard thromboprophylaxis with low-molecular-weight heparin from the second day after admission to 24 h prior to the operation and for 12 h after the operation. The drain was removed when the drainage volume was <20 ml. When patients’ haemoglobin (Hb) concentration was <70 g/L, allogeneic blood transfusion was administered. Routine follow-up visits were scheduled at 1, 3, 6 and 12 months postoperatively.
Data collection
Data were collected from medical records. Demographic and clinical characteristic data included age, sex, height, weight, and time from injury to surgery; AO type of fracture, American Society of Anesthesiologists (ASA) score and anaesthesia method; and preoperative Hb and haematocrit (Hct) levels. The intraoperative and postoperative clinical data included the operation time and intraoperative blood loss (IBL); postoperative Hb and Hct levels; postoperative coagulation indicators including prothrombin time (PT), international normalized ratio (INR), activated partial thromboplastin time (APTT) and D-dimer; postoperative blood loss (PBL), which was evaluated by the drainage volume; and transfusion volume. Postoperative complications included deep venous thrombosis and pulmonary embolism; wound haematoma and deep or superficial infection; myocardial infarction; and cerebrovascular accidents. Mortality and readmission rates one month after discharge were also collected.
Calculation methods
Blood loss was determined on the basis of millilitres. IBL=weight of surgical sponges+volume of blood in suction canisters-the volume of irrigation fluids. PBL was evaluated by wound drainage. The Hb levels in the blood were measured preoperatively and on postoperative day 1 (POD1) and 3 (POD3). Total blood loss (TBL)=patient’s blood volume (PBV)×(Hctpre-Hctpod3)/Hctave[13]. PBV=k1×height (m)3+k2×weight (kg)+k3 (k1=0.3669, k2=0.03219, and k3=0.6041 for men; k1=0.3561, k2=0.03308, and k3=0.1833 for women)[14]; Hctpre=the preoperative Hct level; Hctpod3=the Hct level on postoperative day 3; Hctave =the average of the Hctpre and Hctpod3. Hidden blood loss (HBL)=TBL-IBL-PBL+transfusion.
Statistical methods
To minimize selection bias, propensity score matching was performed prior to analysis. All patients in the study group (n=82) were included in the final analysis and matched with patients who did not receive topical TXA (control group, n=82), as shown in Fig. 1. Due to the propensity matching process, the groups did not differ significantly in terms of age, sex, height, weight, time from injury to surgery, AO type of fracture, ASA score, anaesthesia method, or preoperative Hb and Hct level (Table 1). Once the two groups of patients were matched, further statistical analyses were conducted. Numerical data are presented as the mean±standard deviation (mean±SD) and were compared with independent t-tests. Categorical data were compared with chi-square and Fisher’s exact tests. P<0.05 was considered statistically significant. SPSS 23.0 statistical software (Chicago, IL) was used to analyse the data in this study.