Nousjka P A Vranken1, Patrick W Weerwind2, Paul J C Barenbrug1, Steven Teerenstra3, Yuri M Ganushchak1, Jos G Maessen1. 1. Department of Cardiothoracic Surgery, Maastricht University Medical Centre and Cardiovascular Research Institute Maastricht, Maastricht, Netherlands. 2. Department of Cardiothoracic Surgery, Maastricht University Medical Centre and Cardiovascular Research Institute Maastricht, Maastricht, Netherlands p.weerwind@maastrichtuniversity.nl. 3. Department for Health Evidence, Section Biostatistics, Radboud University Medical Centre, Nijmegen, Netherlands.
Abstract
OBJECTIVES: We aimed to investigate the association of patient characteristics and allogeneic blood transfusion products in development of post-cardiac surgery nosocomial infections. METHODS: This retrospective study was conducted in 7888 patients undergoing cardiac surgery with median sternotomy and cardiopulmonary bypass. Multivariable logistic regression analysis was used for independent effect of variables on infections. RESULTS: A total of 970 (12.3%) patients developed one or several types of postoperative infections. Urinary (n = 351, 4.4%) and pulmonary tract infections (n = 478, 6.1%) occurred more frequently than sternal wound infections (superficial: n = 102, 1.3%, deep: n = 72, 0.9%) and donor site infections (n = 61, 0.8%). Interventions, including valve replacement (P = 0.002) and coronary artery bypass grafting combined with valve replacement (P = 0.012), were associated with increased risk of several types of postoperative infections. Patients' profiles changed substantially over the years; morbid obesity (P = 0.019), smoking (P = 0.001) and diabetes mellitus (P = 0.001) occur more frequently nowadays. Furthermore, surgical site infections showed to be related to morbid obesity (P < 0.001) and higher risk stratification (P = 0.031). Smoking (P < 0.001) and chronic obstructive pulmonary disease (P < 0.001) were related to pulmonary tract infections. In addition, diabetic patients developed more sepsis (P = 0.003) and advanced age was associated with development of urinary tract infections (P < 0.001). Even after correcting for other factors, blood transfusion was associated with all types of postoperative infection (P < 0.001). This effect remained present in both leucocyte-depleted and non-leucocyte-depleted transfusion. CONCLUSIONS: Our data showed that post-cardiac surgery infections occur more frequently in patients with predetermined risk factors. The amount of blood transfusions was integrally related to every type of postoperative infection.
OBJECTIVES: We aimed to investigate the association of patient characteristics and allogeneic blood transfusion products in development of post-cardiac surgery nosocomial infections. METHODS: This retrospective study was conducted in 7888 patients undergoing cardiac surgery with median sternotomy and cardiopulmonary bypass. Multivariable logistic regression analysis was used for independent effect of variables on infections. RESULTS: A total of 970 (12.3%) patients developed one or several types of postoperative infections. Urinary (n = 351, 4.4%) and pulmonary tract infections (n = 478, 6.1%) occurred more frequently than sternal wound infections (superficial: n = 102, 1.3%, deep: n = 72, 0.9%) and donor site infections (n = 61, 0.8%). Interventions, including valve replacement (P = 0.002) and coronary artery bypass grafting combined with valve replacement (P = 0.012), were associated with increased risk of several types of postoperative infections. Patients' profiles changed substantially over the years; morbid obesity (P = 0.019), smoking (P = 0.001) and diabetes mellitus (P = 0.001) occur more frequently nowadays. Furthermore, surgical site infections showed to be related to morbid obesity (P < 0.001) and higher risk stratification (P = 0.031). Smoking (P < 0.001) and chronic obstructive pulmonary disease (P < 0.001) were related to pulmonary tract infections. In addition, diabeticpatients developed more sepsis (P = 0.003) and advanced age was associated with development of urinary tract infections (P < 0.001). Even after correcting for other factors, blood transfusion was associated with all types of postoperative infection (P < 0.001). This effect remained present in both leucocyte-depleted and non-leucocyte-depleted transfusion. CONCLUSIONS: Our data showed that post-cardiac surgery infections occur more frequently in patients with predetermined risk factors. The amount of blood transfusions was integrally related to every type of postoperative infection.