Nowadays, anterior and posterior surgical approaches are used for lumbar spinal stenosis. To reduce the variation resulting from different surgical methods, only patients receiving posterior lumbar surgery were selected for the prediction of intraoperative blood use10.
Aged patients are less tolerant to anemia, so the probability of transfusion of allogeneic red blood cells increases during surgery9. This study demonstrated that the risk of blood use during surgery increases for patients over 75 years old with lumbar spinal stenosis or femoral fracture. The transmission risk of infectious diseases increases from blood transfusion, but usually has a long latency, so the risk of transfusion-related infection is not as important in the evaluation of allogeneic red blood cell transfusion in the older population. The short-term benefits obtained from allogeneic red blood cell transfusion will be more important to the condition of aged patients11-13.
Hb concentration is the most common indication for judging whether red blood cells need to be transfused clinically14. Existing transfusion guidelines tend to set an upper threshold for hemoglobin of 10 g/dL, above which transfusion is not indicated, and a lower hemoglobin threshold of between 6 and 7 g/dL, below which transfusion is recommended15-18. It is difficult to obtain this indicator in time during surgery. Otherwise, the amount of blood used can be greatly affected by intraoperative bleeding and anesthesia. This analysis showed that preoperative anemia in patients with Hb≤110g/L increased the probability of blood use during surgery, which is consistent with the conclusions of some previous studies9,19. However, Hb concentration itself does not fully represent the tissue oxygen supply of patients20-22. Increasingly more studies have illustrated that different patients have a different tolerance to anemia, which also causes a significant difference in the probability of clinical blood use. The results of this study illustrated that the decrease of platelet count preoperatively does not affect blood use during surgery. In this study, patients' platelet data were normally distributed according to the Kolmogorov-Smirnov test(P>0.05). However, there were few samples with decreased platelet counts (23 in 1151 patients), which may lead to the statistical deviation. In addition, it has been discovered through clinical work that many patients who have no abnormality or a slight abnormality of prothrombin time or activated partial thromboplastin time in a routine coagulation examination actually have coagulation dysfunction23. In recent years, it has been found that a thromboelastogram examination may be more diagnostic than a coagulation function examination. A thromboelastogram can predict the obstruction of coagulation function and the need for massive blood transfusion through dynamic evaluation of blood clot formation, blood clot strength and the blood clot dissolution process24-27.
BMI creates a different calculation for intraoperative blood use in different conditions. Low body weight increases the probability of intraoperative blood use in aged patients with femoral fracture, which may be due to the relatively small whole body blood volume of low body weight patients. Perioperative bleeding has a great influence on whole body blood volume. Low-weight patients have less subcutaneous fat, which is not conducive to local hemostasis and will lead to increased hidden blood loss9,28-29. On the contrary, overweight patients have a greater probability of blood use during surgery for lumbar spinal stenosis in aged patients. Being overweight or obesity increases the burden on the lumbar spine, so overweight patients have serious lumbar lesions or multi-level lesions that indirectly lead to a rise in the probability of blood use during surgery. Further, the central and peripheral venous pressures of patients are usually reduced by anesthesia during spinal surgery, which reduces the bleeding of surgical wounds. The cardiopulmonary compliance of obese patients decreases, and higher ventilation pressure is needed to overcome the low cardiopulmonary compliance, which gives rise to an increase of venous pressure and more bleeding30-31.
Aged patients increasingly are taking aspirin, clopidogrel or other anti-platelet aggregation drugs to reduce the risk of cardiovascular and thrombotic events, and many patients will take them during the perioperative period32-33. It is necessary to weigh the increased risk of perioperative bleeding caused by continuous use of these drugs and the increased risk of thromboembolic events after withdrawal34-37. According to this analysis, patients with femoral fracture who took these antiplatelet aggregation drugs for a long time before surgery had an increased probability of blood transfusion during surgery, but the relative risk was not high. However, most surgeries for lumbar spinal stenosis are elective. Patients who have taken antiplatelet aggregation drugs for a long time should stop taking the drugs for more than seven days before surgery, thus reducing the influence of these drugs on intraoperative bleeding38-40.
Interestingly, the results of this study illustrated that patients with chronic diseases such as coronary heart disease, diabetes, hypertension or renal insufficiency have no increased risk of blood use during lumbar spinal stenosis or femoral fracture surgery. This may be because these blood transfusion decisions are determined by the experience of clinicians, who generally set a higher threshold for perioperative blood transfusion in the elderly and are more inclined to use red blood cell products during surgery. Currently, the internationally recognized restrictive blood transfusion strategy usually sets the blood transfusion threshold at 80 g/L for perioperative patients41-42, and suggests that there is no need to relax the blood transfusion threshold for patients with a history of coronary heart disease2,43-44. However, some studies have established that the survival time of patients with existing heart diseases tends to decrease in the restricted blood transfusion strategy group, implying that critically ill patients with heart and vascular diseases may benefit from higher Hb levels45.
Different factors affecting blood use during surgery must be considered for different conditions46. Lumbar spinal stenosis in most aged patients is caused by degenerative diseases or long-term strain, so there is a high probability that two or more vertebral segments will be affected30. If multi-segment surgery is performed at the same time, the wound will be larger than that of patients with single-segment lesions, the amount of bleeding will increase, and the probability of red blood cell transfusion will increase. Previous studies have shown that if multiple segments at the same site are operated separately, this will increase intraoperative blood loss, which is consistent with our analysis47-49. Femoral shaft fractures are mostly caused by severe trauma, and the amount of bleeding can reach 1000-1500 mL. If it is an open or comminuted fracture or a fracture of the distal 1/3 the of femur, it is easy to puncture the popliteal artery and vein. The amount of bleeding may be greater because the blood vessel is located posterior to the fracture site, and the distal end of the fracture often is angled backward towards these vessels. If it is a closed fracture of the femur, the recessive blood loss may far exceed the dominant blood loss. Therefore, if there is a long time between the fracture and the surgery,, the Hb concentration will decline gradually, which will markedly increase the probability for red blood cell infusion during surgery50. Unfortunately, there was no detailed record of the amount of time between fracture occurring and the patient being treated in the emergency room, as this could translate to dehydration and affect the initial Hb as well. Therefore, the clinical information of the patients we analyzed was the one closest to the time of surgery to minimize the impact of the above factors on the results of the study.
This research has several limitations. The data analyzed in this study were retrospectively extracted from clinical medical records, so the detection time points of some test results were not uniform. The records of past medical history of some patients may not be completely consistent with the actual situation. Moreover, the blood transfusion treatment decisions made for these patients were not based on fixed guidelines, but on the judgment of different surgeons, so there is some variation44,51.In addition, the population analyzed in this study was from a single department of a single medical institution, and the number of cases with complete clinical case data was small, so the statistical power is limited. This predictive model needs to be tested and improved in a larger and more diverse population.