Predicting Operative Blood Loss During Spinal Fusion for Adolescent Idiopathic Scoliosis

         Multilevel spinal fusion has been well established as the primary approach for the surgical correction of adolescent idiopathic scoliosis (AIS).1–7 Significant intraoperative and postoperative blood loss has typically been associated with spinal fusion surgery.2,8,9 Excessive intraoperative and postoperative blood loss increases the transfusion burden on the patient, causing increased exposure to blood products and its associated risks. These risks include an increased rate of infections, hemodynamic instability, decreased cardiac, pulmonary, and renal function secondary to fluid shifts, and death.10–12 To develop targeted strategies to limit intraoperative and postoperative blood loss, risk factors for blood loss must be clearly delineated. Recent studies have specified patient-specific and surgery-specific factors affecting intraoperative blood loss in all patients undergoing multilevel spinal fusion. Significant patient factors affecting operative blood loss include the severity and type of spinal deformity, as well as patient height and weight.9,13–19 Surgery-dependent factors include operative time, procedure performed, surgical approach, number of vertebrae fused, number of anchors placed, average mean arterial pressure (MAP) during surgery, blood salvage techniques, and the use of antifibrinolytic medications.9,13–22 Additional factors have also been indicated as possible predictors of operative blood loss; however, their predictive value is unclear.

          These factors include, but are not limited to: preoperative hemoglobin, preoperative hematocrit, autologous donation, and history of coagulopathy.15,17,18,23 Although certain factors have been shown to be associated with increased blood loss during multilevel spinal fusion, it has yet to be determined if blood loss can be reliably predicted. From a large, single surgeon cohort of AIS patients, this study aims to provide a comprehensive assessment of factors leading to increased intraoperative blood loss. In addition, predictors of increased blood loss are used to develop a formula to preoperatively predict intraoperative blood loss. This information could enable targeted use of available blood salvage techniques and be useful for counseling patients and their families before surgery. METHODS IRB approval for this study was obtained at the NYU Hospital for Joint Diseases, New York, NY. From the *Department of Orthopaedic Surgery, NYU Hospital for Joint Diseases, New York, NY; and wStevens Institute of Technology, Hoboken, NJ. IRB Approved at New York University—Langone Medical Center. None of the authors received financial support for this study. The authors declare no conflict of interest. Reprints: Baron S. Lonner, MD, Department of Orthopaedic Surgery, NYU Hospital for Joint Diseases, 820 2nd Avenue, #7A, New York, NY 10017. E-mail: blonner@nyc.rr.com. Copyright r 2013 by Lippincott Williams & Wilkins ORIGINAL ARTICLE 372 | www.pedorthopaedics.com J Pediatr Orthop  Volume 33, Number 4, June 2013 Appropriate measures were taken to ensure that patient confidentiality was protected.

          A retrospective chart review of consecutive patients who had undergone primary spinal fusion for AIS by a single surgeon (blinded surgeon) from January 2000 to January 2008 was completed. The inclusion criteria for this study were: a diagnosis of AIS, age of 10 to 21 years, and primary spinal fusion surgery of at least 4 vertebral levels by anterior, posterior, or combined anterior/posterior approach. Patients treated with video-assisted thoracoscopic surgery with anterior instrumentation were excluded from this study. Exclusion criteria for this study also included prior history of spine surgery at the level(s) of fusion, refusal to receive blood products for any reason (ie, Jehovah’s Witness), or abnormalities in PT/PTT/INR. Demographic data collected included sex, age, preoperative height, preoperative weight, preoperative BMI, and medical comorbidities.

           Preoperative full scoliosis radiographs were obtained and radiographic measurements were performed by the operative surgeon. Preoperative major coronal curves, as well as T2-T12 kyphosis, were assessed. Preoperative white blood cell count, hemoglobin, hematocrit, platelet count, PT, PTT, and INR were also recorded. Intraoperative data collected included procedures performed, approach, estimated blood loss (EBL), total operative time, total levels fused, cell saver volume, intravenous fluids given, units transfused, and complications. The volume of blood transfused both intraoperatively and postoperatively was recorded. Transfusion criteria used were: symptomatic anemia including hypotension, orthostatic hypotension, tachycardia, and hemoglobin <8 g/dL. In addition, the patient’s MAP was noted at incision and at 15-minute intervals throughout the procedure. Postoperative data collected included major coronal curves, T2-T12 kyphosis, white blood cell count, hemoglobin, hematocrit, platelet count, PT, PTT, INR, drain output, systolic and diastolic blood pressures, and early complications.

          Statistical Analysis Demographic data was assessed using descriptive statistics. The relationship between EBL (mL) and independent variables was initially assessed using Pearson correlation. Variables with a significant correlation were then analyzed using a stepwise multivariate linear regression. In these regressions, the dependent variable was EBL. A P value of 0.05 was set to indicate statistical significance. RESULTS A total of 340 patients [259 females, 81 males; mean age, 15.2 y (range, 10 to 21 y)] underwent spinal arthrodesis for AIS from 2000 to 2008 by the second author (B.L.). Preoperative data can be found in Table 1. The most prevalent surgical approach was a posterior-only approach (55%). The median (interquartile range) EBL for posterior approach was 700 mL (range, 450 to 1000 mL), for anterior approach was 300 mL (range, 200 to 400 mL), and for combined approach was 1000 mL (range, 788 to 1500 mL). Distribution of EBL by approach is shown in Figure 1. Complications were reported in 4 of the procedures.

           Complications included change in approach (3 patients) and hemodynamic instability (1 patient). The distribution of operative parameters by surgical approach is shown in Table 2. Preoperative variables that were highly correlated with blood loss were: age (y), height (cm), weight (kg), BMI (kg/m2 ), coronal curve magnitude (degrees), T2-T12 kyphosis (degrees), largest structural curve (degrees), hemoglobin (g/dL), hematocrit (%), platelet count, PT (s), PTT (s), INR, and blood pressure. The intraoperative variables that were highly correlated with blood loss were: MAP at time of incision, average MAP during procedure, total levels fused, total operative time, number of implant anchors (pedicle screws, hooks, and wires) used, complications, cell saver volume (mL), volume of allogenic blood transfused (units), and IV fluids given (mL). Results of Pearson correlation is shown in Table 3. Multivariate linear regression identified male sex, decreased preoperative kyphosis, and increased operative time as predictors of increased blood loss for posterior spinal fusion (PSF).

          Independent of height and weight, male sex was an independent predictor of increased blood loss for PSF. For anterior spinal fusion, increased MAP at the time of incision and increased operative time were identified as predictors of increased blood loss. Using the results of this analysis, the following formula was developed to predict blood loss during PSF: Blood lossðmLÞ ¼ CþðOpTime½min 6:4Þ ðPreOp Kyphosis½degrees 8:7Þ C ¼ 233 if male; 270 if female TABLE 1. Preoperative Data Male n = 81 Female n= 259 Demographics Age [median (range), (y)] 15.8 (10.1-21.0) 15.0 (10.2-21.75) Height [mean (SD), cm] 171 (9) 158 (16) Weight [mean (SD), kg] 60.2 (20.8) 51.2 (16.0) BMI [mean (SD), kg/m2 ] 21.8 (4.9) 21.2 (4.3) Curve parameters Kyphosis [median (range), deg.] 28 [(22) to 72] 33 [(17) to 65] Major structural curve [mean (SD), deg.] 60 (11) 52 (13) Lenke curve type Type 1 [n (%)] 42 (52) 121 (47) Type 2 [n (%)] 11 (14) 19 (7) Type 3 [n (%)] 5 (6) 13 (5) Type 4 [n (%)] 1 (1) 5 (2) Type 5 [n (%)] 13 (16) 70 (27) Type 6 [n (%)] 3 (4) 16 (6) Laboratory values Hemoglobin [mean (SD)] 13.6 (1.5) 12.4 (1.2) Hematocrit [mean (SD)] 40.6 (4.0) 36.9 (3.7) Platelets [mean (SD)] 258 (56) 282 (65) PT [mean (SD)] 11.7 (1.8) 11.3 (1.64) PTT [mean (SD)] 30.8 (4.4) 30.7 (4.1) INR [mean (SD)] 1.21 (0.99) 1.04 (0.10) J Pediatr Orthop  Volume 33, Number 4, June 2013 Predicting Operative Blood Loss r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | 373 DISCUSSION Multilevel spinal fusion is the standard surgical option for the treatment of AIS.1–7

           Operative treatment of AIS is typically associated with lower amounts of mean blood loss than surgical correction of scoliosis secondary to an underlying neuromuscular disorder.8,9,15,17,18 Despite this fact, excessive blood loss remains a risk of spinal fusion for AIS.2,8,9 Most published studies report a mean EBL ranging between 750 and 1500 mL in this population.8 Autologous blood predonation, controlled hypotensive anesthesia, intraoperative blood salvage (cell saver), normovolemic hemodilution, and perioperative antifibrinolytics have been used to limit intraoperative blood loss and the need for allogenic transfusion during spinal fusion surgery for AIS.11,19–22,24–27 Two recent studies have attempted to identify patient and operative factors predictive of blood loss and transfusion requirements in adolescents and children undergoing spinal fusion for scoliosis. From a large cohort of AIS patients (n = 262), Shah et al28 reported the number of levels fused, male sex, duration of surgery, use of pedicle screws, major Cobb angle, and age were predictors of intraoperative blood loss.

          Similarly, Meert et al18 reported underlying neuromuscular disease, lower body weight, and number of levels fused as predictors of transfusion requirements in a heterogenous cohort of adolescents and children. Both of these studies included multiple surgeons, leading to intersurgeon variability. In addition, the effect of various surgical procedures was not reported in either study. In the adult population, blood losses are often increased due to medical comorbidites and rigid curves that require multiple corrective osteotomies and greater surgical manipulation to achieve adequate correction.29 In FIGURE 1. Estimated blood loss by approach. The dark line represents the median. The bottom of box represents the 25th percentile. The top of box represents the 75th percentile. T bars extend to 1.5 times height of box. Dots represent outliers; asterisks represent extreme outliers. TABLE 2. Operative Data Posterior Approach n = 188 Anterior Approach n = 124 Combined Approach n= 28 Rib resection [n (%)] 13 (7) 2 (2) 6 (22) Thoracoplasty [n (%)] 14 (7) 12 (10) 9 (32) ICGB harvested [n (%)] 7 (4) 0 0 VATS release [n (%)] 4 (2) 50 (40) 21 (75) Incision map [mean (SD)] 68.2 (8.9) 66.7 (7.2) 71.6 (8.5) Posterior levels fused [mean (SD)] 9.5 (2.3) N/A 11.1 (1.7) Anterior levels fused [mean (SD)] N/A 4.5 (1.2) 5.6 (1.2) Anchors used [mean (SD)] 13 (4) 8 (2) 14 (4) Operative time [mean (SD), min] 213 (60) 232 (72) 394 (78) Units transfused [mean (SD), units] 0.7 (1.4) 0.3 (0.5) 1.7 (1.1) Cell saver [mean (SD), mL] 320 (264) 34 (77) 406 (258) Estimated blood loss [mean (SD), mL] 907 (775) 323 (171) 1277 (821) Ialenti et al J Pediatr Orthop  Volume 33, Number 4, June 2013 374 | www.pedorthopaedics.com r 2013 Lippincott Williams & Wilkins 112 adults, Zheng et al17 reported preoperative hemoglobin, levels fused, and body weight as independent predictors of blood loss during revision posterior spinal decompression, fusion, and segmental instrumentation for degenerative lumbar spinal stenosis.

          Nuttall et al15 found low preoperative hemoglobin, tumor surgery, increased number of levels fused, history of pulmonary disease, decreased amount of autologous blood available, and no use of Jackson table as predictors of increased allogenic transfusion requirements in 244 adults undergoing multilevel spine surgery. Generally, cohorts of adult patients are comprised of a heterogenous population, making the prediction of operative blood loss difficult. Furthermore, neither of these studies attempted to develop a formula to predict blood loss based on patient and surgical factors. From a large cohort of AIS patients, we delineated numerous patient and surgical factors strongly associated with increased blood loss for any surgical approach. These include male sex, height, weight, preoperative T2-T12 kyphosis, preoperative major curve magnitude, preoperative lumbar Lenke modifier, preoperative hematocrit, predonation of autologous blood, MAP at incision, surgical approach, rib resection, video-assisted thoracoscopic surgery release, bone graft harvesting, number of levels fused, number of anchors placed, and operative time (Table 3).

         Of these factors, male sex, decreased preoperative kyphosis, and increased operative time were independently predictive of increased blood loss for posterior fusion. Male patients had more blood loss than female patients even when controlling for height and weight. To our knowledge, this is the first study to identify preoperative kyphosis as an independent predictor of blood loss during spinal fusion for AIS. We speculate that preoperative hypokyphosis may be associated with increased blood loss because more Ponte osteotomies are typically performed to lengthen the posterior column. It is possible that male sex is associated with increased blood loss due to increased muscle mass compared with females. Perhaps increased muscle mass requires greater manipulation to achieve correction. Although it has been well known that increased operative times lead to increased blood loss, we now counsel male patients and those with decreased kyphosis regarding the possibility of increased blood loss, and advise them that blood donation may be warranted. For adolescents treated anteriorly, increased MAP at incision and longer operative times were independent predictors. Identification of MAP at incision as an independent predictor advocates for the use of controlled hypotension during anterior surgical approach.

            We now typically only start a case if the MAP is between 65 and 75 mm Hg. AIS patients typically have flexible curve types with relatively little medical comorbidity. Limiting the analysis further to a single surgeon cohort allows for a comprehensive analysis of only patient, radiographic, and surgery-related factors. We believe that this enables the identification of predictors not previously identified. Predictors of blood loss were used to develop a predictive formula for PSF. Although this model was developed from a single surgeon database, only patient and surgical factors are included. This allows for numerous independent surgeons to utilize the model to predict blood loss, while making small adjustments based on surgeon and institution variability. One potential limitation of the above stated formula is that operative time is an independent variable. Although estimations of operative time before surgery may be inaccurate, a reasonable estimation of operative time, based on patient and surgical characteristics, can be accomplished by an experienced surgeon.

           As operative time is a known predictor of blood loss, models that do not include it may be inaccurate.13 In addition, the above stated formula can be recalculated intraoperatively, with an adjusted operative time. Future studies should evaluate the benefit of this model and other, similar models for their predictive value.28 Further studies may also benefit from evaluating postoperative blood loss after the different surgical approaches, and if this is significant, incorporating it into the predictive formula. Excessive intraoperative blood loss continues to be one of the most common complications of spinal surgery. Although recent studies have delineated factors predictive of increased blood loss and transfusion, to date, no study has provided a comprehensive analysis of a large cohort of AIS patients treated by a single surgeon. Our results indicate that male sex, operative time, and preoperative kyphosis are predictive of increased blood loss during PSF, whereas MAP at incision and operative time are predictive of blood loss during anterior spinal fusion. Using these predictors, we have developed a formula to preoperatively predict intraoperative blood loss during TABLE 3.

           Variables Correlated With Increased Blood Loss Pearson Correlation r Value Significance P Sex 0.3 < 0.0001 Height 0.24 < 0.0001 Weight 0.15 < 0.01 Hematocrit 0.13 < 0.02 MAP at incision 0.16 < 0.01 Donated units 0.14 < 0.05 Kyphosis T2-T12 0.2 < 0.0001 Major curve 0.22 < 0.0001 Lumbar modifier A 0.21 < 0.0001 Lumbar modifier C 0.17 < 0.01 Sagittal modifier negative 0.12 < 0.05 Anterior levels fused 0.25 < 0.0001 Posterior levels fused 0.28 < 0.0001 Posterior approach 0.44 < 0.0001 Anterior approach 0.63 < 0.0001 Combined approach 0.3 < 0.0001 No. anchors used 0.35 < 0.0001 Rib resection 0.16 < 0.01 VATS procedure 0.18 < 0.0001 Bone graft harvest 0.12 < 0.05 Total operation time 0.34 < 0.0001 MAP indicates mean arterial pressure; VATS, video-assisted thoracoscopic surgery. J Pediatr Orthop  Volume 33, Number 4, June 2013 Predicting Operative Blood Loss r 2013 Lippincott Williams & Wilkins www.pedorthopaedics.com | 375 PSF. Preoperative estimation of blood loss enables surgeons to better counsel patients and their families, as well as guide the use of antifibrinolytics and other blood salvage techniques.

REFERENCES

1. Bjerkreim I, Steen H, Brox JI. Idiopathic scoliosis treated with Cotrel-Dubousset instrumentation: evaluation 10 years after surgery. Spine (Phila Pa 1976). 2007;32:2103–2110.

2. Guigui P, Blamoutier A. Complications of surgical treatment of spinal deformities: a prospective multicentric study of 3311 patients. Rev Chir Orthop Reparatrice Appar Mot. 2005;91:314–327.

3. Helenius I, Remes V, Yrjonen T, et al. Harrington and CotrelDubousset instrumentation in adolescent idiopathic scoliosis. Longterm functional and radiographic outcomes. J Bone Joint Surg Am. 2003;85-A:2303–2309.

4. Lenke LG, Bridwell KH, Blanke K, et al. Radiographic results of arthrodesis with Cotrel-Dubousset instrumentation for the treatment of adolescent idiopathic scoliosis. A five to ten-year followup study. J Bone Joint Surg Am. 1998;80:807–814.

5. Luque ER. Segmental spinal instrumentation for correction of scoliosis. Clin Orthop Relat Res. 1982;163:192–198.

6. Storer SK, Vitale MG, Hyman JE, et al. Correction of adolescent idiopathic scoliosis using thoracic pedicle screw fixation versus hook constructs. J Pediatr Orthop. 2005;25:415–419.

7. Wang Y, Fei Q, Qiu G, et al. Anterior spinal fusion versus posterior spinal fusion for moderate lumbar/thoracolumbar adolescent idiopathic scoliosis: a prospective study. Spine (Phila Pa 1976). 2008;33:2166–2172.

8. Shapiro F, Sethna N. Blood loss in pediatric spine surgery. Eur Spine J. 2004;13(suppl 1):S6–17.

9. Edler A, Murray DJ, Forbes RB. Blood loss during posterior spinal fusion surgery in patients with neuromuscular disease: is there an increased risk? Paediatr Anaesth. 2003;13:818–822.

10. Guay J, de Moerloose P, Lasne D. Minimizing perioperative blood loss and transfusions in children. Can J Anaesth. 2006;53:S59–S67.

11. Kuklo TR, Owens BD, Polly DW Jr. Perioperative blood and blood product management for spinal deformity surgery. Spine J. 2003; 3:388–393.

12. Moran MM, Kroon D, Tredwell SJ, et al. The role of autologous blood transfusion in adolescents undergoing spinal surgery. Spine (Phila Pa 1976). 1995;20:532–536.

13. Guay J, Haig M, Lortie L, et al. Predicting blood loss in surgery for idiopathic scoliosis. Can J Anaesth. 1994;41:775–781.

14. Marks M, Petcharaporn M, Betz RR, et al. Outcomes of surgical treatment in male versus female adolescent idiopathic scoliosis patients. Spine (Phila Pa 1976). 2007;32:544–549.

15. Nuttall GA, Horlocker TT, Santrach PJ, et al. Predictors of blood transfusions in spinal instrumentation and fusion surgery. Spine (Phila Pa 1976). 2000;25:596–601.

16. Vitale MG, Levy DE, Park MC, et al. Quantifying risk of transfusion in children undergoing spine surgery. Spine J. 2002;2: 166–172.

17. Zheng F, Cammisa FP Jr., Sandhu HS, et al. Factors predicting hospital stay, operative time, blood loss, and transfusion in patients undergoing revision posterior lumbar spine decompression, fusion, and segmental instrumentation. Spine (Phila Pa 1976). 2002;27: 818–824.

18. Meert KL, Kannan S, Mooney JF. Predictors of red cell transfusion in children and adolescents undergoing spinal fusion surgery. Spine (Phila Pa 1976). 2002;27:2137–2142.

19. Florentino-Pineda I, Thompson GH, Poe-Kochert C, et al. The effect of amicar on perioperative blood loss in idiopathic scoliosis: the results of a prospective, randomized double-blind study. Spine (Phila Pa 1976). 2004;29:233–238.

20. Grant JA, Howard J, Luntley J, et al. Perioperative blood transfusion requirements in pediatric scoliosis surgery: the efficacy of tranexamic acid. J Pediatr Orthop. 2009;29:300–304.

21. Thompson GH, Florentino-Pineda I, Poe-Kochert C. The role of amicar in decreasing perioperative blood loss in idiopathic scoliosis. Spine (Phila Pa 1976). 2005;30:S94–S99.

22. Tzortzopoulou A, Cepeda MS, Schumann R, et al. Antifibrinolytic agents for reducing blood loss in scoliosis surgery in children. Cochrane Database Syst Rev. 2008;3:CD006883.

23. Kannan S, Meert KL, Mooney JF, et al. Bleeding and coagulation changes during spinal fusion surgery: a comparison of neuromuscular and idiopathic scoliosis patients. Pediatr Crit Care Med. 2002;3:364–369.

24. Anand N, Idio FG Jr., Remer S, et al. The effects of perioperative blood salvage and autologous blood donation on transfusion requirements in scoliosis surgery. J Spinal Disord. 1998;11:532–534.

25. Copley LA, Richards BS, Safavi FZ, et al. Hemodilution as a method to reduce transfusion requirements in adolescent spine fusion surgery. Spine (Phila Pa 1976). 1999;24:219–222; discussion 23-4.

26. Ridgeway S, Tai C, Alton P, et al. Pre-donated autologous blood transfusion in scoliosis surgery. J Bone Joint Surg Br. 2003;85: 1032–1036.

27. Sethna NF, Zurakowski D, Brustowicz RM, et al. Tranexamic acid reduces intraoperative blood loss in pediatric patients undergoing scoliosis surgery. Anesthesiology. 2005;102:727–732.

28. Shah S, Celerin L, Stanton R, In: AAOS Annual Meeting. What Factors Are Important in Determining Intraoperative Blood Loss During Scoliosis Surgery?. Las Vegas, NV; 2009.

29. Aebi M. The adult scoliosis. Eur Spine J. 2005;14:925–948.