Posterior spinal fusion is the procedure of choice for patients with trauma, spinal infection, spinal deformities and degenerative spinal diseases(16). The total number of operations in PSL has been rising annually, especially in China(17, 18). Significant blood loss is the most frequent surgical complication and receiving a great deal of attention in the spinal surgeon, as well as anesthesiologist(19). There was no consensus among the definition of major blood loss; it is generally accepted that one volume of blood loss reaching the total blood loss (60 mL/kg adult) within 24 hours was defined as major blood loss(20). Now evolving techniques have helped to the clinician to aid treatment decisions. Meanwhile, blood transfusion is by far the most effective way to treat the major blood loss in spinal surgeries(21). However, blood shortage is an increasing problem within developing countries, especially in rural, underdeveloped areas of China(22). Thus, early and accurate identification of the risk of blood transfusion is not only to save blood resources but also for better clinical outcome of patients. Although it has been recognized that PSL closely related to significant blood loss and transfusion, the related risk factors for transfusion were still unclear(10, 23). In our cohort, 289 (32.7%) required blood transfusion. We identified independent risk factors associated with transfusion as follows: increased levels of fusion, prolonged operative time, longer time to surgery, total intraoperative EBL, and a low preoperative Hb level.
The preoperative Hb was 115.5 [107.9, 126.2] g/L in the no transfused group and 109.9 [104.2, 114.8] g/L in the transfused group (p < 0.001), respectively. Previous studies have reported that low preoperative Hb was the risk factor with a longer LOS, increased complications, higher costs, and increased mortality(24, 25). Similarly, consistent with what we reported before, Josiah et al. preoperative Hb was a critical predictor in the complex spine surgery(26). These results corroborate the ideas of Adunsky et al., who suggested that patients with Hb < 120 g/L faced a significant risk of transfusion, with a risk that was approximately five times higher than the patients with Hb > 120 g/L(27). It can thus be suggested that low preoperative Hb lead to poor immune and a poor tolerance had unfortunately resulted in low compensative ability to surgical and anaesthetic trauma, as well as blood loss.
However, it was interesting that time to surgery was an independent risk factor of transfusion. Previous studies confirmed that the duration from admission to surgery was correlated with the incidence of complication, morality, and clinical outcome(28). Through the statistical analysis, we confirmed these conclusions; patients in this cohort receive transfusion has a length of stay, which could be explained by poor preoperative general conditions. In no transfused group, the meantime to surgery was 5.1 [4.3, 5.8] days compared to 6.5 [5.4, 7.7] days in patients who received transfusion (p < 0.001). We cautiously assume that this may be because patients with delayed surgery have a series of reasons such as poor general condition, relatively complex surgery or more complications, which might lead to challenging bleeding and need for transfusion.
In this present study, total intraoperative EBL was the strongest independent risk factor for transfusion. Although there is no consistent conclusion on the evaluation of blood loss and the indication of transfusion in spinal surgery, a considerable proportion of studies have reported that in cases requiring blood transfusion, the range of intraoperative blood loss is 650 ml to 2839 ml(20). Moreover, even in the same diseases, we observed a significant positive relationship between blood loss and different procedures, including open surgery or minimally invasive surgery, and unilateral laminar fenestration decompression and fusion, laminar fenestration decompression and fusion, and total laminectomy and decompression. Yang et al.(29) conducted a randomized controlled study on blood loss during lumbar minimal invasive transforaminal lumbar interbody fusion (MIS-TLIF) and open transforaminal lumbar interbody fusion (TLIF), and reported that the total operative blood loss was 355 ml for MIS-TLIF and 538 ml for open TLIF. Zhang et al.(30) reported that the total intraoperative blood loss was 602 ml for lumbar MIS-TLIF and 42 ml for oblique lumbar interbody fusion (OLIF). Morcos et al.(10) reported after a retrospective analysis of transfusion risk factors in lumbar fusion surgery in Canada that perioperative blood transfusion was 18% and intraoperative blood loss was 1018 ml (transfusion group) and 477 ml (non-transfusion group) in posterior lumbar fusion, respectively, and the difference in intraoperative blood loss between the two groups was statistically significant, but multivariate analysis showed that intraoperative blood loss was not a risk factor for perioperative blood transfusion. Therefore, intraoperative judgments of the operators were critical to the assessment of the risk of transfusion.
What attracted our attention is that operative time and numbers of levels of fusion were independent risk factors of transfusion, which was consistent with previous studies(10). The risk of blood transfusion in 2 segments of posterior lumbar fusion was 1.5 times higher than that in 1 fusion, and the risk of blood transfusion in 3 or more segments of fusion was 3 times higher than that in 1 fusion. This confirms that there is a correlation between prolonged posterior spinal fusion surgery and increased operative blood loss(31). Long fusion requires extensive exposure of the spine for pedicle screw placement and intraspinal decompression, which means that a large number of muscles and soft tissues behind the spine need to be dissected from bone tissue, and the more exposed the muscles, soft tissues, and bone surfaces, the increased blood loss during the operation period. Morcos et al.(10) found that the increase of operative segment would prolong the operative time, and then increase the risk of blood transfusion. Therefore, in the face of complex, more difficult, or more fused segments, we believe that good communication within the surgical team, between surgeons, operating room nurses and anaesthetists could reduce the operative time and incidence of transfusion.
Here, we develop a novel predictive nomogram for predicting the risk of transfusion in patients receiving lumbar PSL based on a single high-volume centre for the first time in northeast China. Our nomogram can demonstrate all the key factors graphically and can individually evaluate the incidence of blood transfusions after lumbar PSF. This model can assistant contribute to clinical decision making and identity the patients with a high risk of transfusion(32). Additionally, it provides references for blood transfusion and saves blood resources and hospitalization costs(32, 33).
Several limitations of this study should not be ignored. First, this was a retrospective, single high-volume centre study with possible bias, which limits it's a generality and weak some statistical analyses. Second, external validation, especially in other regions and countries, in the future research is needed. Third, specific data was could not be obtained from the medical records or were missing, including particular procedure, transfusion-related complications, intraoperative fluid infusion volume and intraoperative urine volume.