To explore the techniques and precautions for preventing and controlling periprosthetic femoral splitting during artificial hip arthroplasty. Methods We retrospectively analyzed 581 cases of artificial hip arthroplasty (including 171 cases of femoral neck fracture, 357 cases of patients with femoral head necrosis, and 53 cases of patients with osteoarthritis) completed in our department during the 5-year period of 2009~2013, with 421 cases of total hip replacement (all of the metal socket cups were fixed with biotype, 395 cases of femoral stem prosthesis fixed with biotype, and 26 cases of cemented fixation), 160 cases of femoral head replacement (83 cases of biological fixation, and 77 cases of cemented fixation). There were 160 cases of femoral head replacement (83 cases of biologic fixation and 77 cases of cemented fixation). Intraoperative femoral splitting occurred in 11 cases, with an incidence rate of 1.9%. Results: 493 of 581 patients were followed up for 1-5 years, and 88 patients were lost to follow-up; the Harris Hip Efficacy Scoring System evaluated an excellent rate of 90.1%. The 11 patients (12 hips) who had femoral splitting were all followed up for 1~5 years, and the femoral splitting fracture healed in 3~6 months without any adverse effect on the patients; the Harris Hip Efficacy Scoring System evaluated the excellence rate of 90.9%. Conclusion 1, intraoperative femoral splitting almost always occurs in patients with relatively short stature and relatively small marrow cavity of the femur, while the distal split of the prosthesis handle (MayoII type) almost always occurs in patients with funnel-type marrow cavity (marrow cavity opening index CFI>4.7) and patients with biologically based fixation of tapered handle. 2.The occurrence of femoral splitting is related to surgical operations: insufficient removal of bone from the greater trochanter when expanding the medulla, insufficient expansion of the medulla by a straight expander in the case of a narrow medullary cavity, inappropriate selection of the size and type of femoral stem prosthesis, and violent penetration of the femoral stem prosthesis when it is installed too tightly without reexpanding the medulla. Keywords: arthroplasty, replacement, hip; femoral fracture; intraoperative complications, periprosthetic femoral fracture is a common surgical complication in artificial hip arthroplasty, and the incidence rate is reported to be about 1% in initial artificial hip arthroplasty, 5% in uncemented prosthesis, and less than 1% in cemented prosthesis [1]; and even the incidence rate of periprosthetic fracture is 3%~20% in uncemented prosthesis [1]; even the incidence rate of periprosthetic fracture is 3%~20% in uncemented prosthesis [2]. The incidence of periprosthetic fracture in non-cemented prosthesis ranges from 3% to 20% [2]. 581 cases of artificial hip arthroplasty, 421 cases of total hip arthroplasty, and 160 cases of femoral head replacement were accomplished in our department during the 5 years from 2009 to 2013. Intraoperative femoral splitting occurred in 11 cases, with an incidence rate of 1.9%. It is now reported as follows. 1 .Materials and methods 1.1 Case data 581 cases in this group, 301 male cases, 280 female cases, age 34~96 years old, average 57 years old, disease duration 1d~15 years, average 2.3 years. Surgical etiology: 171 cases of femoral neck fracture, 357 cases of patients with femoral head necrosis, 53 cases of patients with osteoarthritis. Surgical methods: 421 cases of total hip replacement (all metal socket cups were fixed with biological type, 395 cases of femoral stem prosthesis with biological type fixation, 26 cases of cemented type fixation), 160 cases of femoral head replacement (83 cases of biological type fixation, 77 cases of cemented type fixation). There were 11 cases of intraoperative femoral fracture, with an incidence rate of 1.9%. The fracture sites were: 6 cases of fracture corresponding to the proximal end of the prosthesis handle (Mayo type I), and 5 cases of fracture corresponding to the body of the prosthesis handle (Mayo type II), and all of these cases were biologically fixed with conical prosthesis handles. 1.2 Surgical methods All cases were anesthetized with continuous epidural block. The Moore approach to the hip was used to enter the hip joint layer by layer, cut off the short external rotator muscles, reveal the joint capsule at the back of the hip joint, make a longitudinal incision along the long axis of the femoral neck and peel the femoral neck upward and downward to reveal the femoral neck, and then cut off the femoral neck with a pendulum saw or trim the stump of the femoral neck to take out the head of the femur. The acetabulum is firstly exposed and cleaned, polished, externally turned 40~45° anteriorly tilted 10~25° to place the acetabular cup, loaded with liner and checked for stability. Then use the cotter to remove part of the cancellous bone in the femoral neck medulla, expand the medulla appropriately with a straight medulla expander, file the medulla repeatedly, implant the femoral stem prosthesis with an anterior tilt of about 15° (103 cases of cemented type, 478 cases of biologic type), reset the hip joint, and check the stability of the joint, the degree of mobility, the length of the lower limb, and whether there was any impingement at the limit of the activity. One plasma drain was placed. Three to four small bone holes were drilled in the inter-rotator crest, and the short external rotator muscles were sutured to the inter-rotator crest, and the gluteus maximus fascia and iliac radial fascia were firmly sutured to strengthen the “supportive force” of the soft tissues at the back of the joint, and the incision was closed, and the operation was completed. There were 11 cases of femur splitting during the operation, 2 cases occurred when grinding and filing the medullary cavity, and 9 cases occurred in the process of prosthesis insertion. Treatment: 6 cases of proximal femur splitting: 4 cases were only locally split, the width of the crack was less than 1mm, the length of the crack was less than 2cm, which didn’t affect the firmness of the fixation of the prosthesis without any special treatment; 2 cases of the width of the crack was more than 1mm, the length of the crack was more than 2cm, and the fixation of prosthesis was already obviously loosened. In 2 cases, the width of the crack was more than 1mm, the length was more than 2cm, and there was obvious loosening of the prosthesis fixation, the patients were fixed with 2~3 wires, and the patients were told that they could only stand and walk with weight bearing after 2~3 months after the operation; 5 cases of femoral stem (corresponding to the body of the prosthesis handle) were split, because only one side of the cortical bone was split, and the continuity of the femoral stem was still present, and the firmness of the fixation of the prosthesis was still acceptable, so the patients did not have any special treatment, and the patients were told that they could stand and walk with weight bearing after 2~3 months after the operation only. 1.3 Postoperative treatment After the operation, antimicrobial agents were given to prevent infection for 2~3d, and low molecular heparin sodium was given to prevent deep vein thrombosis for 7~10d, and the patients were given double lower limb blood circulation driver and CPM rehabilitation training. According to the amount of plasma drainage (≤30ml), the plasma drainage tube was removed 2~3d after surgery. After the operation, the lower limbs of the operated side were kept in the neutral position of abduction and rotation to prevent dislocation of the prosthesis. Turning, sitting, getting out of bed, standing and walking activities were gradually practiced in 3~14d postoperatively, and the healthy side went first when walking. However, 2 cases of proximal femoral split were fixed with steel wire ring and 5 cases of femoral stem split (corresponding to the prosthesis stem) were not weight-bearing standing and walking until 2~3 months after surgery. After discharge from the hospital, patients were instructed not to sit on short stools and not to stilt their legs as much as possible, and not to flex their hips more than 90° within 6 weeks and 120° after 6 weeks. The patients were followed up at 3 months, 6 months, 1 year and every year thereafter and X-rays were taken to know the condition of the prosthesis. 2. Results 581 patients 493 patients were followed up for 1 to 5 years. 88 patients were lost to follow-up. Among them, 1 case of acetabular cup loosening was revised out of hospital, 1 case of prosthesis handle loosening was revised in our department, the treatment effect was satisfactory, and 1 case of infection was found to be cured at an early stage in 23d after operation. The Harris Hip Efficacy Scoring System (1969) was used to evaluate the postoperative efficacy, which was categorized into four aspects: pain, function, deformity and joint mobility. Our group scored 66-100 points, with an average of 89.3 points, of which 395 cases were excellent, 49 cases were good, 36 cases were moderate, and 13 cases were poor, with an excellent rate of 90.1%. Because we were relatively more concerned about the 11 patients (12 hips) who had femoral cleavage fracture, all of these 11 patients were followed up for 1~5 years. The femoral split fracture healed in 3~6 months and did not affect the patient adversely. 1 case of bilateral total hip replacement with a hidden split in the left femoral stem was not detected, and because the patient loaded weight too early, the femoral prosthesis stem was found to be sinking in the X-ray film at 4 months after the operation, and the patient reported that there was no discomfort in the hip, but there was weakness in the knee joint (see Fig. 1). 11 patients with Harris Hip Efficacy Score of 66-100 points, with an average of 89.3 points, of which the excellent was 89.3 points, with a mean of 1 to 5 points. 89.3 points, of which 7 cases were excellent, 3 cases were good, 1 case was in the middle and 0 cases were poor, with an excellent rate of 90.9%. 3, Discussion 3.1 Analysis of the causes of femoral splitting and how to prevent and control By comparing and analyzing the various aspects of patients with no femoral splitting and those with femoral splitting in this group of cases, we found that intraoperative splitting of two parts of the femur almost always occurs in patients with relatively short stature and relatively small marrow cavity of the femur, and those with distal splitting of prosthesis stem (MayoII type) almost always occur in funnel-type marrow cavity (marrow cavity opening). index CFI > 4.7) and in patients with biologically based fixed tapered stems. At the same time, the occurrence of femoral splitting is also related to surgical operation, for example, the proximal femur is prone to femoral splitting if insufficient bone is removed from the greater trochanter during marrow expansion, the femoral stem is prone to femoral splitting if the marrow expander is insufficiently expanded in a narrow medullary cavity, the femoral prosthesis is not of the appropriate size and type, and the femoral prosthesis needs to be taken out and reexpanded if the femoral stem prosthesis is too tightly fitted, and the likelihood of femoral splitting occurs when violence is directly struck into the femoral prosthesis is increased. After summarizing the lessons learned from the above, we are highly vigilant for patients with relatively short stature, relatively small marrow cavity and funnel-type marrow cavity, fully removing the medial bone of the greater trochanter when expanding the medulla, expanding the medullary expander to the minimum diameter size of the prosthesis handle to be used or choosing the appropriate model size and type of femoral prosthesis handle, and removing and expanding the medulla again as much as possible when the prosthesis handle is not placed smoothly, and forbidding the violence to be struck into it directly, and so on. Subsequently, intraoperative femoral fracture rarely occurred in our operation. 3.2 Treatment of femoral fracture and whether it affects the operation Once femoral fracture occurs during the operation, do not be in a hurry, try to avoid further aggravation of the fracture, analyze the cause and specific situation, and take different methods to deal with the situation. According to our experience: (1) Generally speaking, if the proximal femoral split is limited to localized cracking, the width of the crack is less than 1mm, the length is less than 2cm, if the prosthesis has been placed in place, it does not affect the solidity of the prosthesis fixation of the patient does not need special treatment, the case is almost equal to the normal situation, and will not affect the patient in any way; if the prosthesis has not been placed in place, in order to prevent the splitting of the split from further widening and extending to the distal end, then the prosthesis should be removed and the patient should be treated with a wire. In order to prevent further widening and extension of the split to the distal end, the prosthesis wire ring should be removed and the medulla should be expanded again and/or the medullary cavity should be filed or replaced with a smaller prosthesis. (2) If the width of the cleft is greater than 1 mm, the length is greater than 2 cm, and the prosthesis fixation has been obvious looseness, the prosthesis needs to be taken out, and then fixed with steel wire ringing for 2~3 passes, the prosthesis should be re-inserted, and the patient is told that the patient can only weight-bearing standing and walking in 2~3 months after the operation [4]; in addition to causing the patient to be able to weight-bearing later than 2~3 months, this case will not have a major impact on the patient’s functional recovery in general, whether it will lead to subluxation in the long term. Whether it will lead to prosthesis subsidence and shorten the service life in the long term needs to be further observed and studied. (3) If the femoral stem (corresponding to the stem of the prosthesis) is split, only part of the cortical bone is split, the continuity of the femoral stem still exists, and the solidity of the prosthesis fixation is still acceptable, no special treatment is needed, and the patient is only told that he/she can stand and walk with weight bearing only 2~3 months after the operation (see Fig. 1); in addition to the fact that the patient can not bear weight until 2~3 months later, it does not have any impact on the patient’s functional recovery. (4) If the femoral stem (corresponding to the body of the prosthesis handle and far from the tip) has been fractured, i.e., all the cortical bone has lost continuity, it is necessary to carry out internal fixation of fracture reduction (according to the situation, using the ring embracing device, the half side of the cortical fixation steel plate or adding wires, etc.), and at the same time, use the lengthening type of the prosthesis handle; the occurrence of such a case is actually an accident of the surgery, which is equivalent to the revision surgery that has already been carried out on the spot.