Association between perioperative surgical home implementation and transfusion patterns in adolescents with idiopathic scoliosis undergoing spinal fusion

Blood transfusions in patients with adolescent idiopathic scoliosis after fusion have been associated with increased morbidity, mortality, and cost.


| INTRODUC TI ON
Posterior spine fusion is a complex surgery with the potential for significant blood loss and perioperative transfusion. Perioperative bleeding risk is high because this procedure involves extensive surgical exposure and potentially lengthy operative time. Blood loss in patients with adolescent idiopathic scoliosis who undergo this procedure has been associated with increased morbidity, mortality, and cost. Complications include acute lung injury, immunologic dysfunction, and transmitted infection exposure. Further, studies have shown that transfusion has been associated with longer hospital stays and higher mortality. 1,2 Various approaches have been used to reduce transfusions in the pediatric population, including autologous blood predonation, anti-fibrinolytics, controlled hypotension, intraoperative blood salvage (cell saver), normovolemic hemodilution, restrictive transfusion strategies, and preoperative hemoglobin optimization, and each approach yielded varying levels of success. 5,6 Significant work has been undertaken in the adult population to develop comprehensive blood-management programs, which have reduced transfusions effectively across one hospital system. 7 Preliminary data have shown that implementation of a perioperative blood-management program within the supporting structure of a perioperative surgical home model is linked to a reduction in perioperative transfusion in patients with adolescent idiopathic scoliosis. 3 However, studies have not thoroughly described the association of a perioperative surgical home on blood management in this patient population. The intent of this report is to further define the factors associated with a decrease in blood transfusion after the implementation of a perioperative surgical home for patients with adolescent idiopathic scoliosis.

| Data acquisition
The authors performed a single-center case-control study. After approval by the institution's Institutional Review Board, we

What this article adds
• Implementation of the perioperative surgical home model in this patient population was associated with a significant decrease in perioperative blood transfusion.
• Adopting the perioperative surgical home model at our institution contributed to an effective approach toward minimizing unnecessary transfusion.

4.
Compensated blood loss = sum of various sources of transfusion including autologous, allogeneic, and cell saver volumes. Hematocrit levels of autologous and allogeneic volumes were assumed to be 60% based on the average hematocrit as per our institution's blood bank. Cell saver hematocrit levels were assumed to be 50% based on the average determined by our institution's transfusion program.

| Perioperative surgical home program description
The ASA defines perioperative surgical home as "a patient-centric, team-based model of care created by leaders within the ASA to help meet the demands of a rapidly approaching health care paradigm that will emphasize gratified providers, improved population health, reduced care costs and satisfied patients." 9 At the Children's National Health System, multidisciplinary teams were created with providers from orthopedic surgery, anesthesiology, pain management, case management, nursing, and physical therapy. These committees created a standardized clinical pathway with preoperative, intraoperative, postoperative, and postdischarge components. Aspects of this pathway were based on evidence documented in the literature, and when clear evidence was not available, standardized care was determined by consensus among providers. The various phases of the pathway are outlined in Table 1 and were previously published in detail. 3 Clarification of certain elements of the pathway is warranted.
The perioperative surgical home employed intraoperative goaldirected IV fluid therapy in which the fluid volumes infused were limited to crystalloid infusions <5 mL kg −1 h −1 , with additional fluid volume given only as dictated by physiologic variables and labs indicating inadequate resuscitation including heart rate or blood pressure abnormalities, increasing base deficits, pulse pressure variation, or lactates. 10 A restrictive transfusion strategy was also part of the perioperative surgical home. The decision to transfuse in this pathway is determined based on the overall condition of the patient including the severity of anemia, anticipated or ongoing blood loss, and general health of the patient. Further, a hemoglobin trigger of 7 was suggested, however the decision to transfuse blood products remained at the discretion of the anesthesiologist. No hemoglobin trigger or transfusion strategy was specified in the prePSH group. As part of the

| Statistical analysis
Normality was assessed for continuous outcomes using the  In terms of the primary outcome, after introduction of the perioperative surgical home program, the odds of red blood cell transfusion during the intraoperative period were 0.29 (95% CI 0.10-0.80; P = 0.017), as compared to preperioperative surgical home ( Table 4).

| RE SULTS
The odds of an overall transfusion during either period were less after implementation of the perioperative surgical home (OR = 0.30; 95% CI 0.13-0.70; P = 0.005). As for secondary outcomes, when patients were transfused red blood cells, the volume of transfusion when adjusted for weight was decreased postperioperative surgical home. Analysis of perioperative fluid management shows no difference between the two groups in volumes of cell saver or albumin given ( P < 0.001) ( Table 3). Further, intrathecal morphine was more frequently given to postperioperative surgical home than to prepe- P < 0.001, respectively) ( Table 5).  a role as all providers were aware that they were being evaluated.

| D ISCUSS I ON
A higher average temperature nadir is associated with a reduction in intraoperative blood loss and may decrease the need for blood transfusion. 11,12 Hypothermia is known to impair both platelet function and coagulation cascade enzyme function. 13 Rajagopalan et al showed that mild hypothermia (34-36°C) significantly increases blood loss (by 16%) and increases the relative risk for transfusion (by 22%). 12 Although we only recorded the lowest intraoperative temperature, one can speculate that the higher temperature nadir in the postperioperative surgical home patients may have led to reduction in blood loss and transfusion because normothermia was reached more quickly.
The use of aminocaproic acid has been shown to significantly reduce perioperative blood loss and the need for transfusion in this patient population. 14  Nonetheless, literature suggests that decreased crystalloid use and packed red blood cell transfusion is associated with decrease in length of stay in surgical patients. 21,22 This finding may have contributed to the improvement in length of stay for postperioperative surgical home patients.
We acknowledge several weaknesses in this study. First, inherent limitations exist in the estimation of blood loss. Visual estimates of blood loss by the anesthesiologist and the proceduralist have been proven to be inaccurate. 25 The gravimetric approach to blood loss can also be imprecise. 26 We believe that calculated blood loss is a reasonable method by which to estimate perioperative blood loss. 8 However, although other innovative approaches using image-processing algorithms to estimate blood loss in suction canisters could provide more accurate estimations, they were not used in this study. 27 The second weakness in this study is our reliance on a retrospective preperioperative surgical home control group that is quite small and may limit the conclusions made from the comparison to the larger prospective postperioperative surgical home group. A third weakness of this study is that the analysis does not determine to what extent the effects seen can be attributed to specific anesthesiologists or surgeons. It has been shown that increased anesthesiologist experience is associated to improved perioperative outcomes in multiple complex surgeries. 28,29 Another important consideration of this analysis is that while the changes in outcomes we identify are associated with implementation of the perioperative surgical home, we cannot confirm the cause of these changes. These results could be caused by change in practice over time or another cause. One other limitation in this analysis is the lack of data regarding any change in use of preoperative iron supplementation with implementation of the perioperative surgical home, as this may have influenced the starting hemoglobin levels of the patients and impacted transfusion rates. Finally, the data set is incomplete. Missing data may have biased the results; however, it is worth noting that the most important data points, including calculated blood loss and volume of crystalloid, had datasets that were largely intact.

| CON CLUS ION
Implementation of blood-conservation strategies as part of a perioperative surgical home for patients with adolescent idiopathic scoliosis who undergo posterior spinal fusion is associated with a significant decrease in perioperative blood transfusion. Reasons for this change appear to be multifactorial. Data indicate that the most likely cause is the reduction in hemodilution associated with a more restrictive approach to fluid management in conjunction with a more restrictive red blood cell transfusion strategy.
Improved temperature control, standardized administration of anti-fibrinolytics, and shorter surgical time may have also contributed to decreased need for perioperative transfusion in this patient population. Adopting the perioperative surgical home model at our institution contributed to an effective approach to minimizing unnecessary transfusions.

E TH I C A L A PPROVA L
The protocol for this study was approved by the Children's National Institutional Review Board.