Siobhan McGurk

Assessment of education and computerized decision support interventions for improving transfusion practice

BACKGROUND: Overuse of blood products is common, but prior efforts to improve transfusion decisions have met with limited success.

ABO incompatibility as an adverse risk factor for survival after allogeneic bone marrow transplantation

BACKGROUND: Graft ABO incompatibility has not been thought to aflect patient survival after allogeneic bone marrow transplantation, although it may be associated with prolonged erythroid aplasia and immediate or delayed hemolysis.

Determinants of red cell, platelet, plasma, and cryoprecipitate transfusions during coronary artery bypass graft surgery: the Collaborative Hospital Transfusion Study

BACKGROUND: Very little is known about the determinants of blood transfusions in patients undergoing coronary artery bypass graft surgery. STUDY DESIGN AND METHODS: To identify factors that influenced the transfusion of red cells, platelets, plasma, and cryoprecipitate, statistical methods were used to study 2476 consecutive diagnosis‐ related group 106 and 107 patients in five teaching hospitals who underwent coronary artery bypass surgery between January 1, 1992, and June 30, 1993. RESULTS: The likelihood of red cell transfusion was significantly associated with 10 preoperative factors: 1) admission hematocrit, 2) the patients age, 3) the patients gender, 4) previous coronary artery bypass surgery, 5) active tobacco use, 6) catheterization during the same admission, 7) coagulation defects, 8) insulin‐dependent diabetes with renal or circulatory manifestations, 9) first treatment of new episode of transmural myocardial infarction, and 10) severe clinical complications. Platelet and/or plasma transfusions were strongly associated with the dose of red cells transfused. Transfusion requirements and other in‐hospital outcomes were associated with patient characteristics, surgical procedure (reoperation vs. primary procedure), and the conduits used for revascularization (venous graft only, venous and internal mammary artery graft, or internal mammary artery graft only). Blood resource use and donor exposures were evaluated with respect to the risk to patients of contracting hepatitis C virus and human immunodeficiency virus infections. CONCLUSION: The classification of coronary artery bypass graft patients on the basis of attributes known preoperatively and by conduits used yields subsets of patients with distinctly different transfusion requirements and in‐ hospital outcomes.

Reduction in incidence of deep sternal wound infections: Random or real?

OBJECTIVE Comorbidities predisposing cardiac surgical patients toward deep sternal wound infection, such as diabetes and obesity, are rising in the United States. Longitudinal analysis of risk factors, morbidity, and mortality was performed to assessed effects of these health trends on deep sternal wound infection rates. METHODS In this retrospective analysis of all median sternotomies performed at a single institution from 1991 through 2006, demographic and surgical characteristics were identified from a prospective database. The cohort was separated into periods from 1992 through 2001 and 2002 through 2006 to identify longitudinal trends in risk factors for deep sternal wound infection. Univariate and matched multivariable analyses were performed. RESULTS Overall, study population had increased comorbidities associated with deep sternal wound infection such as obesity, diabetes, and advanced age. Deep sternal wound infections were treated in 285 of 21,000 sternotomies performed during study period (1.35%). Deep sternal wound infection rates decreased from 1.57% to 0.88% in last 5 years. Rate of deep sternal wound infection was reduced among patients with diabetes from 3.2% to 1.0%. Multivariable analysis showed diabetes and smoking to be eliminated as risk factors in last 5 years. Prolonged bypass time was the only variable independently associated with deep sternal wound infection for the entire period. Thirty-day and 1-year mortalities for deep sternal wound infection did not change significantly. CONCLUSIONS Analysis of a large series of cardiac surgical patients demonstrates significant reduction in deep sternal wound infection incidence in 15 years. Introduction of perioperative intravenous insulin may explain some observed risk reduction. Efforts should focus on prevention, because mortality remains elevated.

Long and short-term outcomes following coronary artery bypass grafting in patients with and without chronic kidney disease

BACKGROUND Improved understanding of the incidence and risk factors for operative complications and long-term mortality following coronary artery bypass grafting (CABG) is needed to better define the optimal role for CABG in patients with chronic kidney disease (CKD). METHODS We analysed 2438 patients who underwent CABG at a single centre between 2005 and 2008. Multivariable regression was used to analyse associations and to generate a CKD-specific predictive tool. RESULTS Operative mortality was 4.8% in individuals with stage 3 CKD, 7.1% in individuals with stage 4-5 CKD and 2.2% in those without significant CKD (P < 0.001). CKD was associated with post-operative blood transfusion, acute kidney injury, myocardial injury and cardiac arrest, and use of exogenous blood and acute kidney injury were strongly associated with in-hospital death in CKD patients. Patients with stage 3 (HR 1.64, 95% CI 1.30-45.94) and stage 4-5 CKD (HR 2.77, 95% CI 1.00-2.68) were more likely to die during follow-up than those without CKD, but mortality rates were low among patients who survived to discharge-stage 3 (0.006 deaths/year) and stage 4-5 CKD (0.009/year). A scoring system including urgent or emergent surgery (OR 2.30), prior cardiac surgery (OR 3.06), concurrent valve surgery (OR 2.06), preoperative shock (OR 6.18), and prior stroke (OR 1.98) had 96.4% percent specificity for the detection of in-hospital death in patients with CKD. CONCLUSIONS Perioperative mortality and morbidity remain more frequent in patients with stage 3-5 CKD than patients with preserved renal function, but long-term outcomes in patients surviving hospitalization are favourable. We have developed a predictive tool that holds promise as a means of identifying CKD patients most likely to survive surgery and benefit from CABG.

The Collaborative Hospital Transfusion Study: variations in use of autologous blood account for hospital differences in red cell use during primary hip and knee surgery

BACKGROUND: Red cell use in patients undergoing Diagnosis Related Group (DRG) 209 procedures (major joint and limb reconstruction procedures of the lower extremities) has been shown to have large, unexplained interhospital variations.

Late outcomes comparison of nonelderly patients with stented bioprosthetic and mechanical valves in the aortic position: A propensity-matched analysis

OBJECTIVE Our study compares late mortality and valve-related morbidities between nonelderly patients (aged <65 years) undergoing stented bioprosthetic or mechanical valve replacement in the aortic position. METHODS We identified 1701 consecutive patients aged <65 years who underwent aortic valve replacement between 1992 and 2011. A stented bioprosthetic valve was used in 769 patients (45%) and a mechanical valve was used in 932 patients (55%). A stepwise logistic regression propensity score identified a subset of 361 evenly matched patient-pairs. Late outcomes of death, reoperation, major bleeding, and stroke were assessed. RESULTS Follow-up was 99% complete. The mean age in the matched cohort was 53.9 years (bioprosthetic valve) and 53.2 years (mechanical valve) (P=.30). Fifteen additional measurable variables were statistically similar for the matched cohort. Thirty-day mortality was 1.9% (bioprosthetic valve) and 1.4% (mechanical valve) (P=.77). Survival at 5, 10, 15, and 18 years was 89%, 78%, 65%, and 60% for patients with bioprosthetic valves versus 88%, 79%, 75%, and 51% for patients with mechanical valves (P=.75). At 18 years, freedom from reoperation was 95% for patients with mechanical valves and 55% for patients with bioprosthetic valves (P=.002), whereas freedom from a major bleeding event favored patients with bioprosthetic valves (98%) versus mechanical valves (78%; P=.002). There was no difference in stroke between the 2 matched groups. CONCLUSIONS In patients aged <65 years, despite an increase in the rate of reoperation with stented bioprosthetic valves and an increase in major bleeding events with mechanical valves, there is no significant difference in mortality at late follow-up.

Minimally invasive aortic valve replacement versus aortic valve replacement through full sternotomy: the Brigham and Women’s Hospital experience

BACKGROUND Minimally invasive aortic valve surgery (mini AVR) is a safe and effective treatment option at many hospital centers, but there has not been widespread adoption of the procedure. Critics of mini AVR have called for additional evidence with direct comparison to aortic valve replacement (AVR) via full sternotomy (FS). METHODS Our mini AVR approach is through a hemi-sternotomy (HS). We performed a propensity-score matched analysis of all patients undergoing isolated AVR via FS or HS at our institution since 2002, resulting in 552 matched pairs. Baseline characteristics were similar. Operative characteristics, transfusion rates, in-hospital outcomes as well as short and long term survival were compared between groups. RESULTS Median cardiopulmonary bypass and cross clamp times were shorter in the HS group: 106 minutes [inter-quartile ranges (IQR) 87-135] vs. 124 minutes (IQR 90-169), P≤0.001, and 76 minutes (IQR 63-97) vs. 80 minutes (IQR 62-114), P≤0.005, respectively. HS patients had shorter ventilation times (median 5.7 hours, IQR 3.5-10.3 vs. 6.3 hours, IQR 3.9-11.2, P≤0.022), shorter intensive care unit stay (median 42 hours, IQR 24-71 vs. 45 hours, IQR 24-87, P≤0.039), and shorter hospital length of stay (median 6 days, IQR 5-8 vs. 7 days, IQR 5-10, P≤0.001) compared with the FS group. Intraoperative transfusions were more common in FS group: 27.9% vs. 20.0%, P≤0.003. No differences were seen in short or long term survival, or time to aortic valve re-intervention. CONCLUSIONS Our study confirms the clinical benefits of minimally invasive AVR via HS, which includes decreased transfusion requirements, ventilation times, intensive care unit and hospital length of stay without compromising short and long term survival compared to conventional AVR via FS.

Nonelective cardiac surgery in the elderly: Is it justified?

OBJECTIVE Elderly patients might be denied nonelective cardiac surgery because of the perception of poor outcomes and an unacceptable quality of life. In this study we evaluate long-term survival and quality of life in these patients. METHODS From 1994 to 1999, 262 consecutive patients older than 80 years underwent urgent (n = 223) or emergent (n = 39) cardiac surgery. Of these patients, 160 (61%) underwent coronary artery bypass grafting, 64 (24%) underwent coronary artery bypass grafting plus valve surgery, 17 (7%) underwent valve surgery, and 21 (8%) underwent aortic surgery. Kaplan-Meier survival analysis and quality-of-life assessment were performed, and result were compared with age-adjusted population data. Risk factors for mortality were determined by using Cox regression. The utility of Society of Thoracic Surgeons and EuroSCORE risk scoring were assessed by using area under receiver operating curves. RESULTS Early mortality was 11% (n = 29) overall, 7% (n = 16) in urgent cases, and 33% (n = 13) in emergent cases. Five-year survival was 50% (n = 132) overall, 53% (n = 105) in urgent cases, and 36% (n = 18) in emergent cases. There was no difference in 10-year survival between patients undergoing urgent surgical intervention and age-adjusted population data. Among survivors, quality-of-life measures were equivalent to those of the general elderly population. Risk factors for early mortality were age, emergent procedure, aortic procedure, bypass time, and postoperative complication (renal failure, myocardial infarction, cerebrovascular accident, pneumonia, and reoperation for bleeding). Risk factors for late mortality were peripheral vascular disease, emergent procedure, bypass time, and new renal failure. The EuroSCORE and Society of Thoracic Surgeons risk scores were equivalent but only moderately predictive of mortality. CONCLUSIONS Long-term survival and quality of life after nonelective cardiac surgery can equal that of the general elderly population. Age alone should not disqualify a patient for urgent or emergent cardiac surgery.

Postoperative Recurrence of Mitral Regurgitation After Annuloplasty for Functional Mitral Regurgitation

BACKGROUND We investigated predictive factors of postoperative recurrence of mitral regurgitation (MR) after mitral valve annuloplasty for functional ischemic MR. METHODS This study was a retrospective review of patients with functional MR who underwent mitral ring annuloplasty at our institution from 1998 to 2008. Records were reviewed for perioperative variables, including echocardiographic measurements. Recurrence was defined as MR grade moderate (3+) or greater on any postoperative echocardiogram. RESULTS Of 548 patients who underwent mitral annuloplasty for functional MR, echocardiogram reports were available for review for 250. These patients comprised the study cohort. There were 154 patients with concomitant coronary artery bypass grafting. The left ventricular (LV) end-diastolic (LVED) index was calculated by dividing the LVED dimension by body surface area (BSA). A high LVED index (>3.5 cm/m2) was predictive of MR recurrence (p=0.047): the recurrence rate was 20.3% (13 of 64) in high-index patients and 10.2% (19 of 186) in low-index (<3.5 cm/m2) patients. Median recurrence-free survival was 35 months (range, 19 to 59 months) in high-LVED-index patients and 46.5 months (range, 22 to 75 months) in low-LVED-index patients (p=0.048). Preoperative MR severity and LV ejection fraction were not predictive of MR recurrence. CONCLUSIONS Careful analysis of the preoperative echocardiogram using a novel, easily calculated variable, the LVED index, may help identify patients who are at greater risk for MR recurrence and reduced survival. Such information could prompt consideration of alternative or additional interventions in these patients.