Identifying Variations in Blood Use Based on Hemoglobin Transfusion Trigger and Target among Hepatopancreaticobiliary Surgeons

Intraoperative Allogeneic Blood Transfusion Has No Impact on Postoperative Short-Term Outcomes After Pancreatoduodenectomy for Periampullary Malignancies: A Propensity Score Matching Analysis and Mediation Analysis

Background/Objectives: The aim of this study was to investigate the association between intraoperative blood transfusion (BT) and the short-term outcomes of pancreatoduodenectomy (PD) for patients with periampullary malignancies. Methods: In a retrospective two-center cohort analysis, we utilized a logistic and mixed-effects ordinal regression, nonparametric partial correlation, and mediation analysis, complemented by propensity score matching (PSM) and weighting. Results: A total of 491 patients were included. Of these, 18 (3.7%) received an intraoperative BT. An intraoperative BT was associated with blood loss (odds ratio (OR) per 100 mL = 1.42; 95% CI 1.27 to 1.62; p < 0.001) and relatively high ASA classes (OR = 3.75; 95% CI 1.05 to 17.74; p = 0.041). Intraoperative blood loss (r = 0.27; p < 0.001) but not intraoperative BT (r = 0.015; p = 0.698) was associated with postoperative complications. Intraoperative BT was associated with postoperative complications according to the unadjusted regression (OR = 1.95; 95% CI 1.38-2.42, p < 0.001) but not the multivariable ordinal regression. In the mediation analysis for relative risk (RR), intraoperative BT was beneficial (RR = 0.51; 95% CI: 0.01-0.78), and blood loss (RR = 2.49; 95% CI: 1.75-177.34) contributed to the occurrence of major postoperative complications. After PSM, analyses revealed that an intraoperative BT did not have a significant impact on the rates of postoperative major complications (OR = 1.048; p = 0.919), clinically relevant postoperative pancreatic fistula (OR = 0.573; p = 0.439) or postoperative 90-day mortality (OR = 0.714; p = 0.439). Conclusions: When adjusting for intraoperative blood loss, intraoperative BT is not associated with postoperative complications.

National Multi-Institutional Validation of a Surgical Transfusion Risk Prediction Model

BACKGROUND

Accurate estimation of surgical transfusion risk is important for many aspects of surgical planning, yet few methods for estimating are available for estimating such risk. There is a need for reliable validated methods for transfusion risk stratification to support effective perioperative planning and resource stewardship.

STUDY DESIGN

This study was conducted using the American College of Surgeons NSQIP datafile from 2019. S-PATH performance was evaluated at each contributing hospital, with and without hospital-specific model tuning. Linear regression was used to assess the relationship between hospital characteristics and area under the receiver operating characteristic (AUROC) curve.

RESULTS

A total of 1,000,927 surgical cases from 414 hospitals were evaluated. Aggregate AUROC was 0.910 (95% CI 0.904 to 0.916) without model tuning and 0.925 (95% CI 0.919 to 0.931) with model tuning. AUROC varied across individual hospitals (median 0.900, interquartile range 0.849 to 0.944), but no statistically significant relationships were found between hospital-level characteristics studied and model AUROC.

CONCLUSIONS

S-PATH demonstrated excellent discriminative performance, although there was variation across hospitals that was not well-explained by hospital-level characteristics. These results highlight the S-PATH's viability as a generalizable surgical transfusion risk prediction tool.

Point-of-care haemoglobin accuracy and transfusion outcomes in non-cardiac surgery at a Canadian tertiary academic hospital: protocol for the PREMISE observational study

INTRODUCTION

Transfusions in surgery can be life-saving interventions, but inappropriate transfusions may lack clinical benefit and cause harm. Transfusion decision-making in surgery is complex and frequently informed by haemoglobin (Hgb) measurement in the operating room. Point-of-care testing for haemoglobin (POCT-Hgb) is increasingly relied on given its simplicity and rapid provision of results. POCT-Hgb devices lack adequate validation in the operative setting, particularly for Hgb values within the transfusion zone (60-100 g/L). This study aims to examine the accuracy of intraoperative POCT-Hgb instruments in non-cardiac surgery, and the association between POCT-Hgb measurements and transfusion decision-making.

METHODS AND ANALYSIS

PREMISE is an observational prospective method comparison study. Enrolment will occur when adult patients undergoing major non-cardiac surgery require POCT-Hgb, as determined by the treating team. Three concurrent POCT-Hgb results, considered as index tests, will be compared with a laboratory analysis of Hgb (lab-Hgb), considered the gold standard. Participants may have multiple POCT-Hgb measurements during surgery. The primary outcome is the difference in individual Hgb measurements between POCT-Hgb and lab-Hgb, primarily among measurements that are within the transfusion zone. Secondary outcomes include POCT-Hgb accuracy within the entire cohort, postoperative morbidity, mortality and transfusion rates. The sample size is 1750 POCT-Hgb measurements to obtain a minimum of 652 Hgb measurements <100 g/L, based on an estimated incidence of 38%. The sample size was calculated to fit a logistic regression model to predict instances when POCT-Hgb are inaccurate, using 4 g/L as an acceptable margin of error.

ETHICS AND DISSEMINATION

Institutional ethics approval has been obtained by the Ottawa Health Science Network-Research Ethics Board prior to initiating the study. Findings from this study will be published in peer-reviewed journals and presented at relevant scientific conferences. Social media will be leveraged to further disseminate the study results and engage with clinicians.

Development of a machine learning model to predict intraoperative transfusion and guide type and screen ordering

STUDY OBJECTIVE

To develop an algorithm to predict intraoperative Red Blood Cell (RBC) transfusion from preoperative variables contained in the electronic medical record of our institution, with the goal of guiding type and screen ordering.

DESIGN

Machine Learning model development on retrospective single-center hospital data.

SETTING

Preoperative period and operating room.

PATIENTS

The study included patients ≥18 years old who underwent surgery during 2019-2022 and excluded those who refused transfusion, underwent emergency surgery, or surgery for organ donation after cardiac or brain death.

INTERVENTION

Prediction of intraoperative transfusion vs. no intraoperative transfusion.

MEASUREMENTS

The outcome variable was intraoperative transfusion of RBCs. Predictive variables were surgery, surgeon, anesthesiologist, age, sex, body mass index, race or ethnicity, preoperative hemoglobin (g/dL), partial thromboplastin time (s), platelet count x 109 per liter, and prothrombin time. We compared the performances of seven machine learning algorithms. After training and optimization on the 2019-2021 dataset, model thresholds were set to the current institutional performance level of sensitivity (93%). To qualify for comparison, models had to maintain clinically relevant sensitivity (>90%) when predicting on 2022 data; overall accuracy was the comparative metric.

MAIN RESULTS

Out of 100,813 cases that met study criteria from 2019 to 2021, intraoperative transfusion occurred in 5488 (5.4%) of cases. The LightGBM model was the highest performing algorithm in external temporal validity experiments, with overall accuracy of (76.1%) [95% confidence interval (CI), 75.6-76.5], while maintaining clinically relevant sensitivity of (91.2%) [95% CI, 89.8-92.5]. If type and screens were ordered based upon the LightGBM model, the predicted type and screen to transfusion ratio would improve from 8.4 to 5.1.

CONCLUSIONS

Machine learning approaches are feasible in predicting intraoperative transfusion from preoperative variables and may improve preoperative type and screen ordering practices when incorporated into the electronic health record.

Contemporary practice and perception of autologous blood salvage in hepato-pancreatico-biliary operations: an international survey

BACKGROUND

This study aimed to assess contemporary knowledge, attitudes and behaviors around transfusion of intraoperative salvaged blood (sRBCt) during hepato-pancreatico-biliary (HPB) operations. Findings are meant to inform the design of future studies that address provider concerns to change behaviors and improve patient outcomes.

METHODS

A survey was designed and assessed for relevance, readability and content, and distributed to an international audience of surgeons performing HPB operations.

RESULTS

The 237 respondents were predominantly distributed across North America (37.55%), Europe (27.43%) and Asia (19.83%). Roughly one-half (52.74%) of respondents had used sRBCt in HPB surgery before. Transplantation surgeons were more likely than HPB surgeons to have previously used sRBCt [odds ratio = 5.18 (95% CI 1.89-14.20)]. More respondents believed sRBCt was safe for non-cancer versus cancer operations (68.57% vs. 24.17%, p < 0.0001). Less than half (45.71%) of respondents believed that sRBCt was safe in clean-contaminated fields. Most did not utilize preoperative strategies to avoid donor transfusion.

CONCLUSION

Practices related to sRBCt in HPB operations vary widely and there is no consensus on its use. Concerns seem primarily related to cancer-specific and infectious outcomes. While further studies are pursued, surgeons may increase their utilization of preoperative strategies to boost hemoglobin levels for at risk patients.

Implementation of a patient blood management in an Italian City Hospital: is it effective in reducing the use of red blood cells?

To evaluate the effect of patient blood management (PBM) since its introduction, we analyzed the need for transfusion and the outcomes in patients undergoing abdominal surgery for different types of tumor pre- and post-PBM. Patients undergoing elective gastric, liver, pancreatic, and colorectal surgery between 2017 and 2020 were included. The implementation of the PBM program was completed on May 1, 2018. The patients were grouped as follows: those who underwent surgery before the implementation of the program (pre-PBM) versus after the implementation (post-PBM). A total of 1302 patients were included in the analysis (445 pre-PBM vs. 857 post-PBM). The number of transfused patients per year decreased significantly after the introduction of PBM. A strong tendency for a decreased incidence of transfusion was evident in gastric and pancreatic surgery and a similar decrease was statistically significant in liver surgery. With regard to gastric surgery, a single-unit transfusion scheme was used more frequently in the post-PBM group (7.7% vs. 55% after PBM; p = 0.049); this was similar in liver surgery (17.6% vs. 58.3% after PBM; p = 0.04). Within the subgroup of patients undergoing liver surgery, a significant reduction in the use of blood transfusion (20.5% vs. 6.7%; p = 0.002) and a decrease in the Hb trigger for transfusion (8.5, 8.2-9.5 vs. 8.2, 7.7-8.4 g/dl; p = 0.039) was reported after the PBM introduction. After the implementation of a PBM protocol, a significant reduction in the number of patients receiving blood transfusion was demonstrated, with a strong tendency to minimize the use of blood products for most types of oncologic surgery.

Endoscopic Combined Intrarenal Surgery vs Percutaneous Nephrolithotomy for Large and Complex Renal Stone: A Systematic Review and Meta-Analysis

Background: Managing complex and large renal stones with percutaneous nephrolithotomy (PCNL) is difficult because of the likelihood of residual stones and multiple access. Endoscopic combined intrarenal surgery (ECIRS) is introduced as an improvement to the procedure to manage stones in one session. The objective of this systematic review and meta-analysis is to compare the efficacy and safety between ECIRS and PCNL for treating large and complex renal stones. Materials and Methods: We conducted a systematic review in the Embase, Scopus, and MEDLINE databases based on the 2020 Preferred Reporting Items for Systematic Review and Meta-Analyses guideline. Eligible studies comprised both randomized and nonrandomized studies comparing ECIRS and PCNL. Results: A total of five nonrandomized studies and one randomized controlled trial were included. The analysis was divided into two subgroups based on the PCNL type, a conventional PCNL (cPCNL) and a mini-PCNL (mPCNL). The one-step stone-free rate (SFR) of ECIRS were significantly higher compared with both the cPCNL (odds ratio [OR] 5.14, 95% confidence interval [CI] 2.54 to 10.4, p < 0.001) and mPCNL (OR 4.27, 95% CI 2.57-7.1, p < 0.001). There were no significant differences in mean operative time and hemoglobin drop between both groups (p > 0.05). The use of auxiliary procedures was significantly higher in both PCNL groups compared with the ECIRS group (OR 0.19, 95% CI 0.13-0.30, p < 0.001). The overall complication rate of ECIRS was lower compared with PCNL (OR 0.43, 95% CI 0.21-0.85, p = 0.02), especially urosepsis, in which the incidence was lower compared with cPCNL (OR 0.14, 95% CI 0.02-0.78, p = 0.02), but not mPCNL (p > 0.05). Conclusion: ECIRS is an effective and safe treatment particularly for large and complex nephrolithiasis, with significantly higher one-step SFR, a lower necessity for auxiliary procedures, and a lower complication rate compared with PCNL.

Models to predict the probability for intraoperative RBC transfusion during lumbar spinal stenosis and femoral fracture surgeries in aged patients

OBJECTIVE

The goal of this study was to predict the probability of transfusion of red blood cells and the volume of blood consumption based on the clinical characteristics of patients before surgery.

METHODS

The medical records of 565 patients over 65 years old who underwent posterior lumbar surgery and 586 patients over 65 years old receiving femoral fracture surgery were reviewed. The clinical characteristics of the patients were subjected to multivariate regression analysis. The scores of these factors' influences on intraoperative red blood cells infusion were based on the odds ratio of each multivariate risk factor. Non-linear regression was performed to predict the probability of intraoperative blood transfusion and the volume of blood used for patients with different scores.

RESULTS

The factors that significantly influenced blood use during lumbar spinal stenosis and femoral fracture surgery in aged patients(P < 0.05) included age, body mass index, abnormal coagulation function, preoperative hemoglobin, administration of antithrombotic drugs, multisegmental lesions of the lumbar spine, femoral shaft fracture, secondary lumbar surgery and the time from fracture to surgery exceeding 48 h. According to our risk scoring system, patients of posterior lumbar surgery scored 0-10 and patients of femoral fracture had a score of 0-12. More than 50 % of patients receiving an intraoperative red blood cells transfusion during surgery scored>1.

CONCLUSION

The scoring system can be used as a predictive model for the probability of red blood cells transfusion and the blood volume in aged patients undergoing lumbar spinal stenosis and femoral fracture surgeries.

Mortality and clinical outcome of Italian patients undergoing orthopaedic surgery: effect of peri-operative blood transfusion

BACKGROUND

Blood transfusion is a relevant issue for elderly and frail patients, as they are often anaemic and have chronic diseases. Transfusion of red blood cells (RBC) can potentially affect morbidity and mortality of elderly patients undergoing major orthopaedic surgery.

MATERIALS AND METHODS

We carried out a retrospective analysis of 2,593 patients undergoing major orthopaedic surgery between 2013 and 2017 in a single research institution in the Region of Apulia. The aims of the study were: 1) to describe the characteristics of transfused patients according to a restrictive or liberal strategy of transfusion and haemoglobin (Hb) triggers and targets; 2) to investigate the effect of RBC transfusion on mortality and complications.

RESULTS

Older, women and patients with American Society of Anesthesiologists (ASA) score 3-4 were more often transfused. Those with lower admission Hb level had a higher risk of being transfused. Hb triggers were associated with the patients' age. A restrictive transfusion strategy was significantly more frequent in patients undergoing primary knee replacement and in those with higher estimated blood loss. We did not observe any significant difference of complications in patients transfused with a liberal vs restrictive strategy. Logistic regression correcting for potential confounders revealed that sex (males more than females), duration of stay in hospital, hip fracture and Charlson score >4 were good predictors of complications and/or mortality. Mortality was significantly higher in males and in older patients with ASA score 3-4.

DISCUSSION

In this large cohort of Italian patients undergoing major orthopaedic surgery males were significantly more exposed than women to complications and in-hospital mortality. Furthermore, those undergoing urgent surgery because of hip fracture had a 3-fold higher chance of complications. Charlson score >4 and ASA 3-4 are good predictors of complications and mortality, respectively.

Postoperative Hematocrit and Adverse Outcomes in Pediatric Cardiac Surgery Patients: A Cross-Sectional Study From the Society of Thoracic Surgeons and Congenital Cardiac Anesthesia Society Database Collaboration

BACKGROUND

We sought to examine potential associations between pediatric postcardiac surgical hematocrit values and postoperative complications or mortality.

METHODS

A retrospective, cross-sectional study from the Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) and Congenital Cardiac Anesthesia Society Database Module (2014-2019) was completed. Multivariable logistic regression models, adjusting for covariates in the STS-CHSD mortality risk model, were used to assess the relationship between postoperative hematocrit and the primary outcomes of operative mortality or any major complication. Hematocrit was assessed as a continuous variable using linear splines to account for nonlinear relationships with outcomes. Operations after which the oxygen saturation is typically observed to be <92% were classified as cyanotic and ≥92% as acyanotic.

RESULTS

In total, 27,462 index operations were included, with 4909 (17.9%) being cyanotic and 22,553 (82.1%) acyanotic. For cyanotic patients, each 5% incremental increase in hematocrit over 42% was associated with a 1.31-fold (95% confidence interval [CI], 1.10-1.55; P = .003) increase in the odds of operative mortality and a 1.22-fold (95% CI, 1.10-1.36; P < .001) increase in the odds of a major complication. For acyanotic patients, each 5% incremental increase in hematocrit >38% was associated with a 1.45-fold (95% CI, 1.28-1.65; P < .001) increase in the odds of operative mortality and a 1.21-fold (95% CI, 1.14-1.29; P < .001) increase in the odds of a major complication.

CONCLUSIONS

High hematocrit on arrival to the intensive care unit (ICU) is associated with increased operative mortality and major complications in pediatric patients following cardiac surgery.

Declining Use of Red Blood Cell Transfusions for Gastrointestinal Cancer Surgery: A Population-Based Analysis

BACKGROUND

Gastrointestinal cancer surgery patients often develop perioperative anemia commonly treated with red blood cell (RBC) transfusions. Given the potential associated risks, evidence published over the past 10 years supports restrictive transfusion practices and blood conservation programs. Whether transfusion practices have changed remains unclear. We describe temporal RBC transfusion trends in a large North American population who underwent gastrointestinal cancer surgery.

METHODS

We conducted a population-based retrospective cohort study of patients who underwent gastrointestinal cancer resection between 2007 and 2018 using health administrative datasets. The outcome was RBC transfusion during hospitalization. Temporal transfusion trends were analyzed with Cochran-Armitage tests. Multivariable regression assessed the association between year of diagnosis and likelihood of RBC transfusion while controlling for confounding.

RESULTS

Of 79,764 patients undergoing gastrointestinal cancer resection, the median age was 69 years old (interquartile range (IQR) 60-78 years) and 55.5% were male. The most frequent procedures were colectomy (52.8%) and proctectomy (23.0%). A total of 18,175 patients (23%) received RBC transfusion. The proportion of patients transfused decreased from 26.5% in 2007 to 18.9% in 2018 (p < 0.001). After adjusting for patient, procedure, and hospital factors, the most recent time period (2015-2018) was associated with a reduced likelihood of receiving RBC transfusion [relative risk 0.86 (95% confidence interval: 0.83-0.89)] relative to the intermediate time period (2011-2014).

CONCLUSION

Over 11 years, we observed decreased RBC transfusion use and reduced likelihood of transfusion in patients undergoing gastrointestinal cancer resection. This information provides a foundation to further examine transfusion appropriateness or explore if additional transfusion minimization in surgical patients can be achieved.

Evaluation for Perioperative Blood Transfusion during Major Abdominal Procedures in a Pediatric Population: A Retrospective Observation Cohort Study

Objectives

The purpose of this study was to examine modifiable factors and their impact on perioperative blood transfusion for pediatric patients with major abdominal procedures.

Methods

This is a retrospective review of 1,506 patients who underwent major abdominal surgical procedures in a tertiary medical center from January 2008 to June 2018. Clinical data about blood administration including triggers and targets for intra- or postoperative transfusion were collected and analyzed. The inappropriate transfusion (transfusion > 8.0 g/dL of hemoglobin [Hb] trigger) and overtrans-fusion criteria (target transfusion > 10.0 g/dL or > 2 g/dL of target minus trigger level) were applied to examine the intraoperative factors with the intraoperative transfusion practice. Perioperative morbidity was further assessed based on the inappropriate transfusion and overtransfusion status.

Results

Intraoperative transfusion was used in 468 (31.1%) of the 1,506 patients included in the study. Among them, 212 (45.3%) intraoperative transfusion episodes were classified as inappropriate, and 135 cases (28.8%) were confirmed as overtransfusion. On univariate analysis, inappropriate transfusions were observed more commonly among patients with younger age (p < 0.001) and who underwent hepatic resection (p < 0.001) or intestinal resection (p < 0.001). Overtransfusion was also associated with elevated trigger of 8.0 g/dL Hb (p = 0.006) and younger age (p = 0.003). No perioperative complications were associated with inappropriate transfusions and overtransfusion under multivariate analysis.

Conclusions

Overtransfusion was common in hepatic resection and younger age, but to definitely prove this hypothesis, a prospective randomized trial needs to be performed.

Perioperative blood transfusions following hepatic lobectomy: A national analysis of academic medical centers in the modern era

BACKGROUND

The purpose of the study was to characterize the prevalence and impact of perioperative blood use for patients undergoing hepatic lobectomy at academic medical centers.

METHODS

The University HealthSystem Consortium (UHC) database was queried for hepatic lobectomies performed between 2011 and 2014 (n = 6476). Patients were grouped according to transfusion requirements into high (>5 units, 7%), medium (2-5 units, 6%), low (1 unit, 8%), and none (0 units, 79%) during hospital stay for comparison of outcomes.

RESULTS

Over 20% of patients undergoing hepatic lobectomy received blood perioperatively, of which 35% required more than 5 units. Patients with high transfusion requirements had increased severity of illness (p < 0.01). High transfusion requirements correlated with increased readmission rates (23.4% vs. 19.2% vs. 16.6% vs. 13.5%), total direct costs ($31,982 vs. $20,859 vs. $19,457 vs. $16,934), length of stay (9 days vs. 8 vs. 7 vs. 6), and in-hospital mortality (10.8% vs. 2.0% vs. 0.9% vs. 2.0%) compared to medium, low, and no transfusion amounts (all p < 0.01). Neither center nor surgeon volume were associated with transfusion use.

CONCLUSION

High transfusion requirements after hepatic lobectomy in the United States are associated with worse perioperative quality measures, but may not be influenced by center or surgeon volume.

Patient blood management for liver resection: consensus statements using Delphi methodology

BACKGROUND

Blood loss and transfusion remain a significant concern in liver resection (LR). Patient blood management (PBM) programs reduce use of transfusions and improve outcomes and costs, but are not standardized for LR. This study sought to create an expert consensus statement on PBM for LR using modified Delphi methodology.

METHODS

An expert panel representing hepato-biliary surgery, anesthesiology, and transfusion medicine was invited to participate. 28 statements addressing the 3 pillars of PBM were created. Panelists were asked to rate statements on a 7-point Likert scale. Three-rounds of iterative rating and feedback were completed anonymously, followed by an in-person meeting. Consensus was reached with at least 70% agreement.

RESULTS

The 35 experts panel recommended routine pre-operative transfusion risk assessment, and investigation and management of anemia with iron supplementation. Intra-operatively, restrictive fluid administration without routine central line insertion was recommended, along with intermittent hepatic pedicle occlusion and surgical techniques considerations. Specific criteria for restrictive intra-operative and post-operative transfusion strategy were recommended.

CONCLUSIONS

PBM for LR included medical and technical interventions throughout the perioperative continuum, addressing specificities of LR. Diffusion and adoption of these recommendations can standardize PBM for LR to improve patient outcomes and resource utilization.

Transfusion Practices Among General Surgeons at a Tertiary Care Center: a Survey Based Study

INTRODUCTION

Blood transfusion practices affect both patient's outcomes and utilization of institutional resources. Evidence shows that liberal blood transfusion has a detrimental effect on patient's outcome. A restrictive approach of blood transfusion is recommended by current clinical guidelines.

AIM

The aim of this study was to evaluate the attitudes, knowledge, and practices of general surgery (GS) staff and residents regarding peri-operative blood transfusion and anemia management.

MATERIAL AND METHODS

A self-administered, web-based questionnaire was developed, and its link was sent to the emails of all general surgeons at King Abdul-Aziz University Hospital (KAUH), Jeddah city, Saudi Arabia. The questionnaire included four parts: 1) background of surgeons; 2) preoperative assessment and management of anemia; 3) post-operative blood transfusion and alternatives; and 4) enablers and barriers.

RESULTS

56 surgeons responded to the questionnaire. We found variations in blood transfusion practices, notably the hemoglobin threshold. For stable non-cardiac cases, 7 g/dL was considered the threshold by 50% of respondents. For stable patients with past cardiac disease, a higher threshold was chosen by most (9 g/dL by 43% and 10 gm/dL by 21%). Most respondents believed that transfusion had no effect on the risk of survival (73%) and on the risk of cancer recurrence (55%) after oncologic surgical resection. Recognized facilitators were the availability of scientific evidence (84%), medicolegal concerns (57%), preference (52%), and institutional protocols (50%).

CONCLUSION

Although current clinical guidelines recommend a restrictive transfusion practice, most respondents tended to over-order blood for elective procedures and were not aware of the potential complications of liberal blood transfusion. To implement the restrictive transfusion policies, health institutions should improve the awareness of surgeons and incorporate a strong supporting evidence in formulating local institutional guidelines.

Blood Transfusions in Major Pancreatic Surgery: A 10-Year Cohort Study Including 1404 Patients Undergoing Pancreatic Resections in Finland

Background:

Despite guidelines on blood transfusion (TF) thresholds, there seems to be great variation in transfusion policies between hospitals and surgeons. In order to improve and unify blood transfusion policies, the Finnish Red Cross Blood Service carried out a project concerning the optimal use of blood products (Verivalmisteiden optimaalinen käyttö) between 2002 and 2011. In this study, we determined the blood transfusion trends in major pancreatic surgery in Finland.

Methods:

Initially, 1337 patients who underwent major pancreatic resections between 2002 and 2011 were classified into the TF+ or TF− groups. Centers were divided into high-, medium-, and low-volume centers. The blood transfusion trends and the trigger points for blood transfusions in these patients were determined.

Results:

There were no differences between high-, medium- and low-volume centers in blood usage, trigger points or the use of reserved blood units after pancreatoduodenectomy or total pancreatectomy. However, the trigger points were lowered significantly during the study period at high-volume centers (p = 0.003), and a better use of reserved blood units was found in high- (p < 0.001) and medium-volume (p = 0.043) centers. In addition, a better use of reserved blood units was found in high-volume centers after distal pancreatectomy (p = 0.020)

Conclusion:

Although only minor changes in blood transfusion trends after pancreatoduodenectomy or total pancreatectomy were found generally, the lowering of the transfusion trigger point and the best use of reserved blood units during the study period occurred in high-volume centers.

Percentage change in hemoglobin level and morbidity in spine surgery patients

OBJECTIVE

The aim of this study was to characterize the association between percentage change in hemoglobin (ΔHb)—i.e., the difference between preoperative Hb and in-hospital nadir Hb concentration—and perioperative adverse events among spine surgery patients.

METHODS

Patients who underwent spine surgery at the authors’ institution between December 4, 2008, and June 26, 2015, were eligible for this retrospective study. Patients who underwent the following procedures were included: atlantoaxial fusion, subaxial anterior cervical fusion, subaxial posterior cervical fusion, anterior lumbar fusion, posterior lumbar fusion, lateral lumbar fusion, excision of intervertebral disc, and excision of spinal cord lesion. Data on intraoperative transfusion were obtained from an automated, prospectively collected, anesthesia data management system. Data on postoperative hospital transfusions were obtained through an Internet-based intelligence portal. Percentage ΔHb was defined as: ([preoperative Hb − nadir Hb]/preoperative Hb) × 100. Clinical outcomes included in-hospital morbidity and length of stay associated with percentage ΔHb.

RESULTS

A total of 3949 patients who underwent spine surgery were identified. Of these, 1204 patients (30.5%) received at least 1 unit of packed red blood cells. The median nadir Hb level was 10.6 g/dl (interquartile range 8.7–12.4 g/dl), yielding a mean percentage ΔHb of 23.6% (SD 15.4%). Perioperative complications occurred in 234 patients (5.9%) and were more common in patients with a larger percentage ΔHb (p = 0.017). Hospital-related infection, which occurred in 60 patients (1.5%), was also more common in patients with greater percentage ΔHb (p = 0.001).

CONCLUSIONS

Percentage ΔHb is independently associated with a higher risk of developing any perioperative complication and hospital-related infection. The authors’ results suggest that percentage ΔHb may be a useful measure for identifying patients at risk for adverse perioperative events.

The impact of perioperative blood transfusions on short-term outcomes following hepatectomy

Background

Bleeding and need for red blood cell transfusions (RBCT) remain a significant concern with hepatectomy. RBCT carry risk of transfusion-related immunomodulation that may impact post-operative recovery. This study soughs to assess the association between RBCT and post-hepatectomy morbidity.

Methods

Using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) registry, we identified all adult patients undergoing elective hepatectomy over 2007-2012. Two exposure groups were created based on RBCT. Primary outcomes were 30-day major morbidity and mortality. Secondary outcomes included 30-day system-specific morbidity and length of stay (LOS). Relative risks (RR) with 95% confidence interval (95% CI) were computed using regression analyses. Sensitivity analyses were conducted to understand how missing data might have impacted the results.

Results

A total of 12,180 patients were identified. Of those, 11,712 met inclusion criteria, 2,951 (25.2%) of whom received RBCT. Major morbidity occurred in 14.9% of patients and was strongly associated with RBCT (25.3% vs. 11.3%; P<0.001). Transfused patients had higher rates of 30-day mortality (5.6% vs. 1.0%; P<0.0001). After adjustment for baseline and clinical characteristics, RBCT was independently associated with increased major morbidity (RR 1.80; 95% CI: 1.61-1.99), mortality (RR 3.62; 95% CI: 2.68-4.89), and 1.29 times greater LOS (RR 1.29; 95% CI: 1.25-1.32). Results were robust to a number of sensitivity analyses for missing data.

Conclusions

Perioperative RBCT for hepatectomy was independently associated with worse short-term outcomes and prolonged LOS. These findings further the rationale to focus on minimizing RBCT for hepatectomy, when they can be avoided.

Restrictive Transfusion Strategy Does Not Affect Clinical Prognosis in Patients with Ectopic Pregnancy

To assess the effects of restrictive transfusion strategy on hemoglobin (Hb) levels and prognosis in patients with ectopic pregnancy and severe hemorrhage undergoing emergency surgery, patient data were collected from 2012 to 2016. Following transfusion guidelines, restrictive transfusion was performed; at Hb levels of 60-70 to 100 g/L, transfusion was continued or not based on disease status. The patients were divided into four groups: blood loss < 400 ml (N1), 400-799 ml (N2), 800-1199 ml (N3), and ≥1200 ml (N4). Several prognosis parameters were assessed. Group N4 was further divided based on blood loss amounts (1200-1999, 2000-2999, 3000-3999, and 4000-5000 ml) for subgroup analyses. Blood loss, hemoglobin levels at discharge, and American Society of Anesthesiologists (ASA) scores were not associated with patient prognostic parameters, including intensive care unit (ICU) occupancy, cure, and healing rates, and surgical complications and hospital stay. No statistically significant difference was obtained in hospital stay among N1, N2, and N3 groups. Compared with N1 patients, cases with blood loss ≥ 1200 ml had significantly longer hospital stay. Interestingly, hospital stay was correlated with surgical approach, location of pregnancy, and operation time. Restrictive transfusion strategy could be safely used for emergency surgery in ectopic pregnancy with acute blood loss.

Effect of liberal blood transfusion on clinical outcomes and cost in spine surgery patients

BACKGROUND CONTEXT

Blood transfusions in spine surgery are shown to be associated with increased patient morbidity. The association between transfusion performed using a liberal hemoglobin (Hb) trigger-defined as an intraoperative Hb level of ≥10 g/dL, a postoperative level of ≥8 g/dL, or a whole hospital nadir between 8 and 10 g/dL-and perioperative morbidity and cost in spine surgery patients is unknown and thus was investigated in this study.

PURPOSE

This study aimed to describe the perioperative outcomes and economic cost associated with liberal Hb trigger transfusion among spine surgery patients.

STUDY DESIGN/SETTING

This is a retrospective study.

PATIENT SAMPLE

The surgical billing database at our institution was queried for inpatients discharged between 2008 and 2015 after the following procedures: atlantoaxial fusion, anterior cervical fusion, posterior cervical fusion, anterior lumbar fusion, posterior lumbar fusion, lateral lumbar fusion, other procedures, and tumor-related surgeries. In total, 6,931 patients were included for analysis.

OUTCOME MEASURES

The primary outcome was composite morbidity, which was composed of (1) infection (sepsis, surgical-site infection, Clostridium difficile infection, or drug-resistant infection); (2) thrombotic event (pulmonary embolus, deep venous thrombosis, or disseminated intravascular coagulation); (3) kidney injury; (4) respiratory event; and (5) ischemic event (transient ischemic attack, myocardial infarction, or cerebrovascular accident).

MATERIALS AND METHODS

Data on intraoperative transfusion were obtained from an automated, prospectively collected anesthesia data management system. Data on postoperative hospital transfusion were obtained through a Web-based intelligence portal. Based on previous research, we analyzed the data using three definitions of a liberal transfusion trigger in patients who underwent red blood cell transfusion: a liberal intraoperative Hb trigger as a nadir Hb level of 10 g/dL or greater, a liberal postoperative Hb trigger as a nadir Hb level of 8 g/dL or greater, or a whole hospital nadir Hb level of 8-10 g/dL. Variables analyzed included in-hospital morbidity, mortality, length of stay, and total costs associated with a liberal transfusion strategy.

RESULTS

Among patients with a whole hospital stay nadir Hb between 8 and 10 g/dL, transfused patients demonstrated a longer in-hospital stay (median [interquartile range], 6 [5-9] vs. 4 [3-6] days; p<.0001) and a higher perioperative morbidity (n=145 [11.5%] vs. n=74 [6.1%], p<.0001) than those not transfused. Even after adjusting for age, gender, race, American Society of Anesthesiologists class, Charlson Comorbidity Index score, estimated blood loss, baseline Hb value, and surgery type, logistic regression analysis revealed that patients with a nadir Hb of 8-10 g/dL who were transfused had an independently higher risk of perioperative morbidity (odds ratio=2.11, 95% confidence interval, 1.44-3.09; p<.0001). Estimated additional costs associated with liberal trigger use, defined as a transfusion occurring in patients with a whole hospital stay nadir Hb of 8-10 g/dL, ranged from $202,675 to $700,151 annually.

CONCLUSIONS

Transfusion using a liberal trigger is associated with increased morbidity, even after controlling for possible confounders. Our results suggest that modification of transfusion practice may be a potential area for improving patient outcomes and reducing costs.

Physiologic correlates of intraoperative blood transfusion among patients undergoing major gastrointestinal operations

BACKGROUND

Guidelines for transfusion focus on nadir levels of hemoglobin. Hemoglobin triggers may not be helpful, however, in defining appropriate intraoperative use of packed red blood cells. We sought to define the use of intraoperative packed red blood cells relative to quantitative physiologic factors at the time of operation.

METHODS

Prospective intraoperative data on patients undergoing a major gastrointestinal operation between 2010 and 2014 were analyzed. Risk of intraoperative transfusion was assessed with multivariable extended Cox models using baseline clinical covariates and time-varying intraoperative covariates.

RESULTS

Among 2,316 patients, the mean preoperative hemoglobin was 12.6 g/dL (standard deviation = 2.0 g/dL), while the median estimated blood loss was 200 mL (interquartile range: 100-55 mL). Overall, 357 (15.4%) patients received a transfusion intraoperatively. A greater hazard of transfusion was associated with a greater American Society of Anesthesiologists class (ref: American Society of Anesthesiologists class I-II; American Society of Anesthesiologists class III-IV; hazard ratio = 1.44, 95% confidence interval, 1.18-1.77, P < .001), and a lesser preoperative hemoglobin level (per 1 g/dL increase; hazard ratio = 0.70, 95% confidence interval, 0.65-0.74, P < .001). In addition, an increase in heart rate of 10 beats/min above the cumulative average at any measurement was associated with up to a 30% increased probability of transfusion (hazard ratio = 1.30, 95% confidence interval, 1.15-1.47, P < .001); similarly, an increase in mean arterial pressure of 10 mm Hg was associated with an 8% decreased likelihood of transfusion (hazard ratio = 0.92, 95% confidence interval, 87-0.99, P = .017). In contrast, nadir hemoglobin was not associated with the risk of receiving a transfusion (hazard ratio = 1.10, 95% confidence interval, 0.97-1.23, P = .129). Among patients who received an intraoperative transfusion, 9.2% (n = 33) never had a hemoglobin nadir below 10 g/dL, nor an average mean arterial pressure less than 65 mm Hg or a heart rate greater than 100 beats/min around the time of transfusion.

CONCLUSION

Among the intraoperative factors, heart rate, and mean arterial pressure were strongly associated with the likelihood of receiving a transfusion, despite the observation that 9.2% of patients never had a physiologic indicator for transfusion or a nadir hemoglobin below 10 g/dL, suggesting a subset of patients could benefit from a decrease in intraoperative rate of transfusion.

Perioperative blood transfusion is associated with post-operative infectious complications in patients with Crohn's disease

Background

We have previously demonstrated that blood transfusion (BT) was associated with post-operative complications in patients undergoing surgery for Crohn’s disease (CD), based on our institutional data registry. The aim of this study was to verify the association between perioperative BT and infectious complications in CD patients enrolled in the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database.

Methods

All CD patients undergoing surgery between 2005 and 2013 were identified from NSQIP. Variables were defined according to the ACS NSQIP guidelines. The primary outcome was infectious complications, including superficial, deep and organ/space surgical site infection, wound dehiscence, urinary tract infection, pneumonia, systemic sepsis and septic shock. Multivariate analyses were performed to assess the risk factors for post-operative infections.

Results

All 10 100 eligible patients were included and 611 (6.0%) received perioperative BT. BT patients were older, lighter in weight and more likely to be functionally dependent. BT patients were more likely to have post-operative infectious complications than those without BT, including superficial surgical site infection (SSI) (10.8% vs 7.4%, p=0.002), deep SSI (3.3% vs 1.6%, p=0.003), organ/space SSI (14.2% vs 5.4%, p<0.001), pneumonia (3.8% vs 1.3%, p<0.001), urinary tract infection (3.9% vs 2.2%, p=0.006), sepsis (11.5% vs 4.5%, p<0.001) and sepsis shock (3.1% vs 0.8%, p<0.001). Multivariate analysis showed that intra- and/or post-operative BT was an independent risk factor for post-operative infectious complications (odds ratio [OR] = 2.2; 95% confidence interval [CI]: 1.8–2.7; p<0.001) and the risk increased with each administered unit of red blood cell (OR = 1.3, 95% CI: 1.2–1.5). Other independent factors were history of smoking, chronic heart disease, diabetes, hypertension and the use of corticosteroids. Pre-operative BT, however, was not found to be a risk factor to post-operative infections.

Conclusions

Intra- and/or post-operative, not pre-operative, BT was found to be associated with an increased risk for post-operative infectious complications in this CD cohort. Therefore, the timing and risks and benefits of BT should be carefully balanced.

Laparoscopic sleeve gastrectomy as first-line surgical treatment for morbid obesity among adolescents

BACKGROUND

The increasing prevalence of obesity has necessitated the increasing use of bariatric surgery in the adolescent population. Outcomes following laparoscopic sleeve gastrectomy (LSG) among adolescents, however, have not been well-studied. We report outcomes following LSG as a first-line surgical therapy in patients under 21years of age.

METHODS

All patients who underwent LSG as a primary surgical option for morbid obesity were identified at the University of Illinois at Chicago between 2006 and 2014. Standard clinicopathologic and outcomes data were recorded.

RESULTS

We identified 18 patients (13 females, 5 males) who underwent LSG. Mean patient age was 17.8±1.7years. Mean BMI among all patients was 48.6±7.2kg/m² and did not differ by gender (P=0.68). One patient (5.6%) experienced a 30-day perioperative complication (pulmonary embolism). Median LOS following LSG was 3days (IQR: 2, 3). 2 patients (11.1%) were readmitted within 30-days because of feeding intolerance that resolved without invasive intervention. At a median follow-up of 10.6 (range: 0-38) months, percent excess weight loss (%EWL) among all patients was 35.6%. Among patients with at least 2years follow-up (n=3), %EWL was 50.2%.

CONCLUSIONS

Laparoscopic sleeve gastrectomy in morbidly obese adolescents is a safe and feasible option. Short- and long-term weight loss appears to be successful following LSG. As such, LSG should be strongly considered as a primary surgical treatment option for all morbidly obese adolescents.

LEVEL OF EVIDENCE

Level IV.

Impact of Delta Hemoglobin on Provider Transfusion Practices and Post-operative Morbidity Among Patients Undergoing Liver and Pancreatic Surgery

BACKGROUND

Delta hemoglobin (ΔHb) is defined as the difference between the preoperative Hb and the lowest post-operative Hb level. We sought to define the impact of ΔHb relative to nadir Hb levels on the likelihood of transfusion, as well as characterize the impact of ΔHb and nadir Hb on morbidity among a large cohort of patients undergoing complex hepatopancreatobiliary (HPB) surgery.

METHODS

Patients who underwent pancreatic or hepatic resection between January 1, 2009 and June 30, 2015 at Johns Hopkins Hospital were identified. Data on the perioperative ΔHb, nadir Hb, as well as blood utilization were obtained and analyzed. Multivariable logistic regression models were used to identify the factors associated with ΔHb and the impact of ΔHb on perioperative morbidity. A Bayesian model was used to evaluate the correlation of ΔHb and nadir Hb with the likelihood of transfusion, as well as the impact on morbidity.

RESULTS

A total of 4363 patients who underwent hepatobiliary (n = 2200, 50.4 %) or pancreatic (n = 2163, 49.6 %) surgery were identified. More than one quarter of patients received at least one unit of packed red blood cells (PRBC) (n = 1187, 27.2 %). The median nadir Hb was 9.2 (IQR 7.9-10.5) g/dL resulting in an average ΔHb of 3.4 mg/dL (IQR 2.2-4.7) corresponding to 26.3 %. Both ΔHb and nadir Hb strongly influenced provider behavior with regards to use of transfusion. Among patients with the same nadir Hb, ΔHb was strongly associated with use of transfusion; among patients who had a nadir Hb ≤6 g/dL, the use of transfusion was only 17.9 % when the ΔHb = 10 % versus 49.1 and 80.9 % when the ΔHb was 30 or 50 %, respectively. Perioperative complications occurred in 584 patients (13.4 %) and were more common among patients with a higher value of ΔHb, as well as patients who received PRBC (both P < 0.001).

CONCLUSIONS

The combination of the Hb trigger with ΔHb was associated with transfusion practices among providers. Larger ΔHb values, as well as receipt of transfusion, were strongly associated with risk of perioperative complication following HPB surgery.

Improvement of the Surgical Apgar Score by Addition of Intraoperative Blood Transfusion Among Patients Undergoing Major Gastrointestinal Surgery

BACKGROUND

The surgical Apgar score (SAS) has been shown to correlate with postoperative outcomes. A key component of the SAS is estimated blood loss (EBL), which has been shown to be inaccurate and discordant with intraoperative blood transfusion. Given this, the objective of the current study was to assess the added predictive value of the including receipt of intraoperative transfusion to the SAS.

METHODS

We identified 1833 patients undergoing major gastrointestinal surgery (pancreatic, hepato-biliary, and colorectal) between January 1, 2010 and August 31, 2013 at Johns Hopkins Hospital. The primary outcome was postoperative complications or death. A modified SAS was created by assigning a "0" EBL score for every patient who received an intraoperative blood transfusion, regardless of the actual EBL. Model performance was tested using logistic regression and c-statistic.

RESULTS

Mean EBL of the entire cohort was 250 mL. Two hundred ninety-two patients (15.9 %) received at least 1 unit of blood intraoperatively. Approximately, one half of patients (55.1 %) who had an EBL <1000 mL received an intraoperative transfusion. Patients who received an intraoperative transfusion (transfusion n = 94, 32.2 % vs. no transfusion n = 221, 14.3 %; P < 0.001) and those with increasing EBL had a higher incidence of postoperative morbidity and/or death (≤100 mL: 11.6 %, 101-600 mL: 16.9 %, 601-1000 mL: 24.5 %, >1000 mL: 29.2 %; P < 0.001). The variance inflation factor between EBL and intraoperative transfusion was 1.23 for postoperative morbidity/mortality, suggesting that the multicollinearity between the two variables was low. With the inclusion of intraoperative transfusion in the modified SAS, the modified model (c-statistic 0.6552) had an improved discrimination of predicting postoperative morbidity and mortality as compared to the original SAS (c-statistic 0.6391) (P = 0.01). The modified SAS demonstrated improvement in predicting raw differences in the incidence of postoperative morbidity/mortality based on the overall score (P < 0.05).

CONCLUSIONS

The inclusion of intraoperative transfusion in a modified SAS significantly improves the risk-stratifying ability of the score with regard to postoperative morbidity and mortality. Given the variability of intraoperative transfusion, its discordance with EBL, and its strong negative impact on postoperative outcomes, we strongly support the inclusion of this factor in a modified SAS.

Variation in inpatient hospital and physician payments among patients undergoing general versus orthopedic operations

BACKGROUND

Comparative data on surgeon payments for operative procedures are not well documented. We sought to assess variations in surgeon payments after common general and orthopedic operations using a nationally representative sample of privately insured patients.

METHODS

A total of 486,506 patients who underwent a general (appendectomy, cholecystectomy, colectomy) or orthopedic (total knee replacement, total hip replacement) operation between 2010-2012 were identified from the Truven Health MarketScan database.

RESULTS

Median age was 54 years (general operation, 44 years vs orthopedic operation, 58 years; P < .001). Patients had an average Charlson Comorbidity Index of 0 (interquartile range [IQR]: 0, 1). Median duration of stay was 3 days (IQR: 2, 4) (general operation, 3 days [IQR: 1, 5] vs orthopedic operation, 3 days [IQR: 2, 3]; P < .001). Total hospital payments averaged $18,209 (IQR: $11,751, $26,598) (general operation: $12,744 [IQR: $8,402, $19,896] vs orthopedic operation: $22,386 [IQR: $16,888, $30,100]; P < .001). Median surgeon reimbursement was $1,923 (IQR: $1,146, $2,676), with orthopedic surgeon payments being on average twice as high as general surgeon payments ($2,349 vs $1,191; P < .001). Median surgeon payment varied among both general (appendectomy: $903 vs cholecystectomy: $1,125 vs colectomy: $2,209; P < .001) and orthopedic operations (total knee replacement: $2,282 vs total hip replacement: $2,392; P < .001). The presence of a postoperative complication resulted in an increase in hospital payments by 25% and surgeon payments by 11%.

CONCLUSIONS

Hospital and surgeon payments following routine general and orthopedic operations vary greatly. Patients with comorbid conditions and those who experienced a postoperative complication resulted in higher overall payments. Though significant variability existed at the surgeon level, hospital payments were responsible for the highest source of variability.

Blood transfusion is an independent predictor of morbidity and mortality after hepatectomy

BACKGROUND

Previous studies have indicated that blood transfusion is associated with increased risk of worse outcomes among patients selected for hepatectomy. However, the independent effect of transfusion has not been confirmed. We hypothesize that blood transfusion is an independent factor that affects outcomes in patients undergoing hepatectomy.

MATERIALS AND METHODS

Patients at tertiary care center who underwent hepatectomy between 2006 and 2013 were identified and linked with the American College of Surgeons National Surgical Quality Improvement Program PUF data set. Multivariable logistic regression analysis was used to estimate the effect of blood transfusion on 30-d mortality and morbidity, adjusted for differences in extent of resection and estimated probabilities of morbidity and mortality.

RESULTS

Among 522 patients in the study, 48 (9.2%) patients required perioperative blood transfusion within 72 h of resection, and 172 (33%) underwent major hepatectomy. Indications for hepatectomy included metastatic neoplasm (n = 229, 44%), primary hepatic neoplasm (n = 108, 21%), primary extra-hepatic biliary neoplasm (n = 23, 4%), and nonmalignant indications (n = 162, 31%). Eighty-eight (17%) patients had a postoperative morbidity. Blood transfusion was significantly associated with postoperative morbidity (odds ratio [OR] = 4.18, 95% CI = 2.18-8.02, P = 0.0001) and mortality (OR = 14.5, 95% CI = 3.08-67.8, P = 001), after adjustment for the concurrent effect of National Surgical Quality Improvement Program estimated probability of morbidity (OR = 1.15, 95% CI = 0.11-12.2, P = 0.042). The extent of resection was not significantly associated with morbidity (OR = 1.30, 95% CI, 0.74-2.28, P = 0.366) or mortality (OR = 1.14, 95% CI = 0.24-5.50, P = 0.870).

CONCLUSIONS

Blood transfusion is a highly statistically significant independent predictor of morbidity and mortality after hepatectomy. Judicious use of perioperative transfusion is indicated in patients with benign and malignant indications for liver resection.

Understanding Perioperative Transfusion Practices in Gastrointestinal Surgery-a Practice Survey of General Surgeons

BACKGROUND

Despite guidelines recommending restrictive red blood cell transfusion (RBCT) strategies, perioperative transfusion practices still vary significantly. To understand the underlying mechanisms that lead to gaps in practice, we sought to assess the attitudes of surgeons regarding the perioperative management of anemia and use of RBCT in patients having gastrointestinal surgery.

METHODS

We conducted a self-administered Web-based survey of general surgery staff and residents, in a network of eight academic institutions at the University of Toronto. We developed a questionnaire using a systematic approach of items generation and reduction. We tested face and content validity and test-retest reliability. We administered the survey via emails, with planned reminders.

RESULTS

Total response rate was 48.1 % (62/125). Half (51.0 %) of respondents stated that they were unlikely to conduct a preoperative anemia work-up. About 54.0 % reported ordering preoperative oral iron supplementation for anemia. Most respondents indicated using a 70 g/L hemoglobin trigger (92.0 %) for transfusion. Factors increasing thresholds above 70 g/L included cardiac comorbidity (58.0 %), acute cardiac disease (94.0 %), symptomatic anemia (68.0 %), and suspected bleeding (58.0 %). With those factors, the transfusion threshold often increased above 90 g/L. Respondents perceived RBCTs to increase the postoperative morbidity (62 %), but not to impact the mortality (48 %) and cancer recurrence (52 %). Institutional protocols (68.0 %), blood conservation clinics (44.0 %), and clinical practice guidelines (84.0 %) were believed to encourage restrictive use of RBCTs.

CONCLUSION

Self-reported perioperative transfusion practices for GI surgery are heterogeneous. Few respondents investigated preoperative anemia. Stated use of RBCT indications varied from recommendations in published guidelines for patients with symptomatic anemia. Establishing team consensus and implementing local blood management guidelines appear necessary to improve uptake of evidence-based recommendations.

Nomogram to predict perioperative blood transfusion for hepatopancreaticobiliary and colorectal surgery

Background

Predictive tools assessing risk of transfusion have not been evaluated extensively among patients undergoing complex gastrointestinal surgery. In this study preoperative variables associated with blood transfusion were incorporated into a nomogram to predict transfusion following hepatopancreaticobiliary (HPB) or colorectal surgery.

Methods

A nomogram to predict receipt of perioperative transfusion was developed using a cohort of patients who underwent HPB or colorectal surgery between January 2009 and December 2014. The discriminatory ability of the nomogram was tested using the area under the receiver operating characteristic (ROC) curve and internal validation performed via bootstrap resampling.

Results

Among 4961 patients undergoing either a HPB (56·3 per cent) or colorectal (43·7 per cent) resection, a total of 1549 received at least 1 unit of packed red blood cells, giving a perioperative transfusion rate of 31·2 per cent. On multivariable analysis, age 65 years and over (odds ratio (OR) 1·52), race (versus white: black, OR 1·58; Asian, OR 1·86), preoperative haemoglobin 8·0 g/dl or less (versus over 12·0 g/dl: OR 26·79), preoperative international normalized ratio more than 1·2 (OR 2·44), Charlson co-morbidity index score over 3 (OR 1·86) and procedure type (versus colonic surgery: major hepatectomy, OR 1·71; other pancreatectomy, OR 2·12; rectal surgery, OR 1·39; duodenopancreatectomy, OR 2·65) were associated with a significantly higher risk of transfusion and were included in the nomogram. A nomogram was constructed to predict transfusion using these seven variables. Discrimination and calibration of the nomogram revealed good predictive abilities (area under ROC curve 0·756).

Conclusion

The nomogram predicted blood transfusion in major HPB and colorectal surgery.

Impact of Transfusion on Cancer Growth and Outcome

For many years, transfusion of allogeneic red blood cells, platelet concentrates, and plasma units has been part of the standard therapeutic arsenal used along the surgical and nonsurgical treatment of patients with malignancies. Although the benefits of these blood products are not a matter of debate in specific pathological conditions associated with life-threatening low blood cell counts or bleeding, increasing clinical evidence is nevertheless suggesting that deliberate transfusion of these blood components may actually lead to negative clinical outcomes by affecting patient’s immune defense, stimulating tumor growth, tethering, and dissemination. Rigorous preclinical and clinical studies are needed to dimension the clinical relevance, benefits, and risks of transfusion of blood components in cancer patients and understand the amplitude of problems. There is also a need to consider validating preparation methods of blood components for so far ignored biological markers, such as microparticles and biological response modifiers. Meanwhile, blood component transfusions should be regarded as a personalized medicine, taking into careful consideration the status and specificities of the patient, rather than as a routine hospital procedure.

Packed red blood cell transfusion after surgery: are we "overtranfusing" our patients?

BACKGROUND

Data on the hemoglobin (Hb) after transfusion, or the "target," which reflects the "dose" of blood given are not well studied. We sought to examine the incidence and causes of "over transfusion" of red blood cells after surgery.

METHODS

Data on blood utilization including Hb triggers and targets were obtained for patients undergoing colorectal, pancreas, or liver surgery between 2010 and 2013.

RESULTS

A total of 2,905 patients were identified, of which 895 (31%) were transfused (median age 64, interquartile range: 53 to 72; 51% men; median American Society of Anesthesiologists class 3, interquartile range: 3-3; 51% pancreatic, 14% hepatobiliary, 21% colorectal, and 14% other). Among these, 512 (57%) were overtransfused (final Hb target after transfusion ≥9.0 g/dL). Among patients who were overtransfused, 171 (33%) were transfused at too high an initial trigger (>8.0 g/dL), whereas 304 (59%) had an appropriate trigger but received ≥2 packed red blood cell (PRBC) units, suggesting an opportunity to have transfused fewer units. There was significant variation in overtransfusion among surgeons (range 0% to 80%, P = .003).

CONCLUSIONS

Excess use of blood transfusion is common and was due to PRBC utilization for too high a transfusion trigger, as well as too many units transfused.

Advances in hepatectomy technique: Toward zero transfusions in the modern era of liver surgery

BACKGROUND

Perioperative blood transfusions suppress immunity and increase hospital costs. Despite multiple improvements in perioperative care, rates of transfusion during/after hepatectomy are reported to range from 25 to 50%. The purpose of this study was to determine the current risk factors for perihepatectomy transfusion by assessing the impact of recent technical advances in liver surgery on transfusion rates.

METHODS

Using our prospectively maintained hepatobiliary tumor database from a high-volume center, a modern cohort of 2,249 hepatectomies (2004-2013) were identified. Patient and operative characteristics were compared between 2 time periods, 2004-2008 (n = 1,139) and 2009-2013 (n = 1,110). Throughout the study interval, transfusions were given based on clinical assessment and not triggered by laboratory thresholds.

RESULTS

Compared with the early cohort, the recent cohort had more patients with an American Society of Anesthesiologists score of ≥ 3 (79 vs 74%), preoperative chemotherapy (73 vs 68%), and a lesser median preoperative hemoglobin (12.9 vs 13.1 mg/dL) and platelet (215,000 vs 243,000) values (all P < .001). Despite these adverse risk factors, with an increasing use of the 2-surgeon resection technique (63 vs 50%), estimated blood loss (309 vs 394 mL), transfusion rates (6 vs 15%), and duration of stay (7.0 vs 8.4 days) were decreased (all P < .001) with no change in overall morbidity or mortality. Multivariate analysis of the recent cohort determined that the independent risk factors associated with transfusion were preoperative anemia and >350 mL of blood loss. The only independent factor associated with less transfusion was use of the 2-surgeon technique for hepatic parenchymal transection.

CONCLUSION

With the exception of patients with moderate to severe preoperative anemia requiring major hepatectomy, recent technical advances have decreased significantly the need for transfusion in liver surgery.

Red Cell Transfusion Triggers and Postoperative Outcomes After Major Surgery

BACKGROUND

The effect of packed red blood cell (PRBC) transfusion on postoperative outcomes of patients undergoing major surgery remains unclear. We sought to determine the impact of blood utilization, as well as transfusion practices, on perioperative outcomes of patients undergoing cardiothoracic-vascular (CT-V) and gastrointestinal (GI) procedures.

METHODS

Patients who underwent major surgical procedures at Johns Hopkins Hospital between 2009 and 2014 were identified. Data on perioperative hemoglobin (Hb) and blood utilization were obtained; transfusion strategy was categorized as liberal (Hb trigger ≥7 g/dL) vs. restrictive (Hb trigger <7 g/dL). Risk-adjusted logistic regression models and propensity score matching were used to assess the association between transfusion triggers and perioperative morbidity.

RESULTS

Among 10,163 patients undergoing either CT-V (50.9 %) or GI (49.1 %) surgery, 4401 (43.3 %) patients received PRBCs. Of the 4401 patients transfused, 71.2 % were transfused using a liberal trigger (≥7 g/dL hemoglobin), while 28.8 % had a restrictive trigger (<7 g/dL). The median number of PRBCs transfused was 3 (restrictive 5 vs. liberal 2 units). While ischemic adverse events were more common among patients undergoing CT-V surgery (17.3 %), infection was the more common complication among patients undergoing GI surgery (11.9 %). American Society of Anesthesiologist (ASA) class 3-4, Charlson score ≥3, and total units of transfused PRBCs were independently associated with overall complications (all P < 0.05). Patients in the restrictive transfusion group did not have increased risk of complications compared with the liberal transfusion group on multivariable analysis (odds ratio (OR) 1.16, 95 % confidence interval (CI) 0.98-1.38; P = 0.08) or after propensity score matching (OR 1.04, 95 % CI 0.88-1.22; P = 0.65).

CONCLUSIONS

Liberal transfusion triggers after major surgery were more common than restrictive practice. Patients with restrictive transfusion trigger did not have increased risk for complications compared with patients transfused with a liberal trigger.

Understanding drivers of hospital charge variation for episodes of care among patients undergoing hepatopancreatobiliary surgery

BACKGROUND

Understanding factors associated with variation in hospital charges may help identify means to increase savings. The aim of the present study was to define potential variation in hospital charges associated with hepatopancreatobiliary(HPB) surgery.

METHODS

Patients who underwent an HPB procedure between 2009-2013 were identified. Total hospital charges were tabulated for room and board, surgical/anaesthesia services, medications, laboratory/radiology services and other miscellaneous charges.

RESULTS

Approximately 2545 patients underwent either a pancreas (66.8%) or liver/biliary (33.2%) resection. The mean total charges for all patients were $42,357 ± 33,745 (pancreas: $46,352 ± 34,932 versus the liver: $34,303 ± 29,639; P < 0.001). Morbidity (pancreas, range: 7-18%; liver, range: 9-18%) and observed:expected (O:E) length of stay (LOS)(pancreas, range: 0.67-1.64; liver, range: 1.06-3.35) varied among providers (both P < 0.001). While a peri-operative complication resulted in increased total hospital charges (complication: $66,401 ± 55,124 versus no complication: $39,668 ± 29,250; P < 0.001), total charges remained variable even among patients who did not experience a complication (P < 0.001). Surgeons within the lowest quartile of O:E LOS had lower total charges ($33 879 ± $27 398) versus surgeons in the highest quartile ($49,498 ± 40 971) (P < 0.001). Surgeons with the highest O:E LOS had higher across-the-board charges (operating room, highest quartile: $10,514 ± $4496 versus lowest quartile: $7842 ± $3706; medication, highest quartile: $1796 ± $3799 versus lowest quartile: $925 ± $2211; radiology, highest quartile: $2494 ± $4683 versus lowest quartile: $1424 ± $3247; P = 0.001; laboratory, highest quartile: $4236 ± $5991 versus lowest quartile: $3028 ± $3804; all P < 0.001).

CONCLUSIONS

After accounting for in-hospital complications, the total mean hospital charges for HPB surgery remained variable by case type and provider. While the variation in charges was associated with LOS, provider-level differences in across-the-board charges were also noted.

Significant Variation in Blood Transfusion Practice Persists following Upper GI Cancer Resection

PURPOSE

Perioperative blood transfusions are costly and linked to adverse clinical outcomes. We investigated the factors associated with variation in blood transfusion utilization following upper gastrointestinal cancer resection and its association with infectious complications.

METHODS

The Statewide Planning and Research Cooperative System was queried for elective esophagectomy, gastrectomy, and pancreatectomy for malignancy in NY State from 2001 to 2013. Bivariate and hierarchical logistic regression analyses were performed to assess the factors associated with receiving a perioperative allogeneic red blood cell transfusion. Additional multivariable analysis examined the relationship between transfusion and infectious complications.

RESULTS

Among 14,875 patients who underwent upper GI cancer resection, 32 % of patients received a perioperative blood transfusion. After controlling for patient, surgeon, and hospital-level factors, significant variation in transfusion rates was present across both surgeons (p < 0.0001) and hospitals (p < 0.0001). Receipt of a blood transfusion was also independently associated with wound infection (OR = 1.68, 95% CI = 1.47 and 1.91), pneumonia (OR = 1.98, 95% CI = 1.74 and 2.26), and sepsis (OR = 2.49, 95% CI = 2.11 and 2.94).

CONCLUSION

Significant variation in perioperative blood transfusion utilization is present at both the surgeon and hospital level. These findings are unexplained by patient-level factors and other known hospital characteristics, suggesting that variation is due to provider preferences and/or lack of standardized transfusion protocols. Implementing institutional transfusion guidelines is necessary to limit unwarranted variation and reduce infectious complication rates.

The impact of peri-operative blood transfusions on post-pancreatectomy short-term outcomes: an analysis from the American College of Surgeons National Surgical Quality Improvement Program

BACKGROUND

Peri-operative red blood cell transfusions (RBCT) may induce transfusion-related immunomodulation and impact post-operative recovery. This study examined the association between RBCT and post-pancreatectomy morbidity.

METHODS

Using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) registry, patients undergoing an elective pancreatectomy (2007-2012) were identified. Patients with missing data on key variables were excluded. Primary outcomes were 30-day post-operative major morbidity, mortality, and length of stay (LOS). Unadjusted and adjusted relative risks (RR) with a 95% confidence interval (95%CI) were computed using modified Poisson, logistic, or negative binomial regression, to estimate the association between RBCT and outcomes.

RESULTS

The database included 21 132 patients who had a pancreatectomy during the study period. Seventeen thousand five hundred and twenty-three patients were included, and 4672 (26.7%) received RBCT. After adjustment for baseline and clinical characteristics, including comorbidities, malignant diagnosis, procedure and operative time, RBCT was independently associated with increased major morbidity (RR 1.49; 95% CI: 1.39-1.60), mortality (RR 2.19; 95%CI: 1.76-2.73) and LOS (RR 1.27; 95%CI 1.24-1.29).

CONCLUSION

Peri-operative RBCT for a pancreatectomy was independently associated with worse short-term outcomes and prolonged LOS. Future studies should focus on the impact of interventions to minimize the use of RBCT after an elective pancreatectomy.

Variation in the use of type and crossmatch blood ordering among patients undergoing hepatic and pancreatic resections

BACKGROUND

The use of preoperative blood orders involved in major gastrointestinal surgery has been poorly studied. The objective of the current study was to analyze compliance with guidelines and factors associated with crossmatch and blood ordering among patients who underwent a hepatic or pancreatic resection.

METHODS

All patients who underwent a hepatic or pancreatic resection between 2010 and 2013 at Johns Hopkins Hospital were identified. Crossmatch to transfusion (C/T) ratios were calculated based on transfusion rates and total units used. A C/T ratio of >2.0 was considered excessive.

RESULTS

Among the 2,629 patients, 11,574 units of packed red blood cells (PRBCs) were crossmatched (mean number of PRBC units crossmatched: 5.9 ± 7.3). Of the 2,629 patients, 34.1% of patients received ≥ 1 PRBCs and 3,611 total units of PRBC were transfused, resulting in an institutional C/T ratio of 2.17 based on the proportion of patients transfused and 3.21 based on the overall number of PRBC units transfused. Using our criterion of excessive C/T ratio of >2.0, we found that 411 patients (45.9%) were "excessively" crossmatched among transfused patients. Among nontransfused patients, 41.0% (n = 711) of patients were crossmatched for >2 units PRBC. Factors associated with receipt of "excess" crossmatch included the presence of multiple comorbidities and an estimated blood loss >450 mL (both P < .05). Provider-level variation among surgeons (C/T ratio range 1.22-3.81) also was associated strongly with "excess" crossmatching ratios (odds ratio 2.56, 95% confidence interval 2.09-3.13; P < .001).

CONCLUSION

More than 1 in 4 patients received crossmatch orders that exceeded institutional and national guidelines. Provider variation both among surgeons and anesthesiologists was an important factor associated with crossmatch variation and excessive ordering of blood for crossmatch in patients undergoing pancreatic and hepatic surgery.

Perioperative Blood Transfusion and the Prognosis of Pancreatic Cancer Surgery: Systematic Review and Meta-analysis

BACKGROUND

Perioperative blood transfusion (PBT) is common in pancreatic surgery. Recent studies have suggested that PBT may be associated with worse long-term outcomes.

METHODS

A systematic review and meta-analysis of studies comparing long-term clinical outcomes of cancer patients undergoing curative-intent pancreatic surgery with regard to occurrence of PBT was performed.

RESULTS

A total of 23 studies (4339 patients) were included in the systematic review, and 19 studies (3646 patients) were included in the meta-analysis. Nearly half (45.8 %) of all patients were female (range 25-60 %), and median age ranged from 59 to 72 years. About half (46.5 %, range 19-72 %) of the patients were transfused. Most had pancreatic ductal adenocarcinoma (69.5 %), while others had ampullary carcinoma (15.0 %), cholangiocarcinoma (7.4 %), or exocrine tumors of pancreas (8.1 %). Most (91.1 %) underwent pancreaticoduodenectomy, while the remaining patients underwent a total or distal pancreatectomy. The 5-year overall survival for all patients ranged from 0 to 65 %. Thirteen and nine of 19 studies reported a detrimental effect of PBT on survival on univariable and multivariable analysis, respectively. Overall, PBT was associated with shorter overall survival (pooled odds ratio 2.43, 95 % confidence interval 1.90-3.10); this finding was reproduced in sensitivity analysis.

CONCLUSIONS

Patients receiving PBT had significantly lower 5-year survival after curative-intent pancreatic surgery. Further research should focus on implementing guidelines for and discerning factors associated with the poor outcomes after PBT.