Faculty Bibliography

BACKGROUND:Infective endocarditis (IE) requiring surgical intervention in patients who actively inject drugs poses treatment challenges. Decisions regarding the need for operation are affected by concern for relapse of IE from ongoing injection drug use (IDU). The purpose of this study was to evaluate the effect of active IDU on outcomes after operation for IE. METHODS:All patients with IE surgically treated at Cleveland Clinic from July 1, 2007 to July 1, 2012 were identified from the Cleveland Clinic Infective Endocarditis Registry and the Cardiovascular Information Registry. Of 536 patients operated on for IE during the study period, 41 (8%) actively injected drugs. The primary outcome of the study was death or reoperation for IE. RESULTS:Patients who injected drugs had poorer survival free of reoperation, and the risk of events varied with time. In a multivariable Cox proportional hazards model, using time-dependent covariates, IDU was associated with a higher hazard of death or reoperation between 90 and 180 days (hazard ratio [HR], 9.8; 95% confidence interval [CI], 2.7-35.3) but not before 90 days (HR, 0.38; 95% CI, 0.05-3.1) or after 180 days (HR, 1.8; 95% CI, 0.8-3.8). Among patients who injected drugs, reoperation and death contributed equally to the outcome, whereas among patients who did not inject drugs, reoperation for IE was far less common. CONCLUSIONS:Between 3 and 6 months after operation for IE, patients who inject drugs have a hazard of death or reoperation that is about 10 times that of patients who do not inject drugs. Before and after, the HRs are much smaller and not statistically significant.

BACKGROUND:We sought to determine whether publication of blood conservation guidelines by the Society of Thoracic Surgeons in 2007 influenced transfusion rates and to understand how patient- and hospital-level factors influenced blood product usage. STUDY DESIGN AND METHODS/METHODS:We identified 4,465,016 patients in the Nationwide Inpatient Sample database who underwent cardiac operations between 1999 and 2010 (3,202,404 before the guidelines and 1,262,612 after). Hierarchical linear modeling was used to account for hospital- and patient-level clustering. RESULTS:Transfusion rates of blood products increased from 13% in 1999 to a peak of 34% in 2010. Use of all blood components increased over the study period. Aortic aneurysm repair had the highest transfusion rate with 54% of patients receiving products in 2010. In coronary artery bypass grafting, the number of patients receiving blood products increased from 12% in 1999 to 32% in 2010. Patients undergoing valvular operations had a transfusion rate of 15% in 1999, increasing to 36% in 2010. Patients undergoing combined operations had an increase from 13% to 40% over 11 years. Risk factors for transfusion were anemia (odds ratio [OR], 2.05; 95% confidence interval [CI], 2.01-2.09), coagulopathy (OR, 1.54; 95% CI, 1.51-1.57), diabetes (OR, 1.32; 95% CI, 1.28-1.36), renal failure (OR, 1.29; 95% CI, 1.26-1.32), and liver disease (OR, 1.23; 95% CI, 1.16-1.31). Compared to the Northeast, the risk for transfusion was significantly lower in the Midwest; higher-volume hospitals used fewer blood products than lower-volume centers. Cell salvage usage remained below 5% across all years. CONCLUSION/CONCLUSIONS:Independent of patient- and hospital-level factors, blood product utilization continues to increase for all cardiac operations despite publication of blood conservation guidelines in 2007.

BACKGROUND:Testing excised valves in surgically treated infective endocarditis (IE) patients provides an opportunity to identify the microbial etiology of IE. Microbial sequencing (universal bacterial, mycobacterial, or fungal polymerase chain reaction followed by DNA sequencing) of valves can identify microorganisms accurately, but the value it adds beyond information provided by blood and valve cultures has not been adequately explored. METHODS:Three hundred fifty-six patients who underwent surgery for active IE from January 1, 2010, to January 1, 2013, were identified from our cardiovascular information registry and outpatient parenteral antibiotic therapy registry. Their records were reviewed to identify 174 patients whose valves were sent for sequencing. The microbial etiology of IE was defined using comprehensive clinical, pathologic, and microbiological criteria. Blood culture, valve culture, and valve sequencing were examined to determine how frequently they identified the definitive cause of IE. RESULTS:Of the 174 patients, 162 (93%) had acute inflammation on histopathologic examination of their valves. Valve sequencing was significantly more sensitive than valve culture in identifying the causative pathogen (90% versus 31%, p < 0.001), and yielded fewer false positive results (3% versus 33%, p <0.001). The pathogen would not have been identified in 25 patients (15%) had it not been for valve sequencing. All the value provided by sequencing was attributable to bacterial DNA sequencing; mycobacterial and fungal sequencing provided no additional information beyond that provided by blood culture, histopathology, and valve culture. CONCLUSIONS:Valve sequencing, not valve culture, should be considered the primary test for identifying bacteria in excised cardiac valves.