Smith, Bradford B.

The continued citation of retracted publications from the medical literature is a well-known and persistent problem. We describe the contexts of ongoing citations to manuscripts that have been retracted from a selection of anesthesiology journals. We also examine how bibliographic databases and publisher websites document the retracted status of these manuscripts. The authors performed an analysis of retracted publications from anesthesiology journals using the Retraction Watch database. We then examined how the retraction information was displayed on bibliographic databases, search engines, and publisher websites. The primary outcome was the context of continued citation after retraction of flawed publications within the specialty of anesthesiology. Secondary outcomes included comparison of the documentation, bibliographic databases, search engines, and publisher websites used in identifying the retracted status of these publications and provision of access to the respective retraction notices. A total of 245 original publications were retracted over a 28-year period from 9 anesthesiology journals. PubMed, compared to the other databases and search engines, was the most consistent (98.8%) in documenting the retracted status of the publications examined, as well as providing a direct link to the retraction notice. From the 211 publications retracted before January 2020, there were 1307 postretraction citations accessed from Scopus. The median number of postretraction citations was 3.5 (range, 0–88, with at least 1 citation in 164 publications) in Scopus. Of the postretraction citations, 80% affirmed the validity of the retracted publications, while only 5.2% of citations acknowledged the retraction or misconduct. In 10.2% of the citations from original research studies, retracted manuscripts appeared to influence the decision to pursue or the methods used in subsequent original research studies. The frequency of citation of the 15 most cited retracted publications declined in a similar pattern during the 10 years after retraction. Citation of manuscripts retracted from anesthesiology journals remains a common occurrence. Technological innovations and application of standards for handling retracted publications, as suggested by coalitions of researchers across the spectrum of scientific investigation, may serve to reduce the persistence of this error.

Magnetic gastroesophageal reflux devices are becoming a common treatment option for reflux refractory to medical therapy. These devices are inserted laparoscopically with successful outcomes; however, patients may still complain of dysphagia after implantation. Echocardiographers may be hesitant to perform transesophageal echocardiography (TEE) in these patients as esophageal surgery and dysphagia represent relative contraindications to performing TEE. However, we present 2 cases where intraoperative TEE was performed in patients with reflux devices without complication or image degradation. The described cases, in addition to a review of the perioperative management of these devices, support the use of TEE in this patient population.

BACKGROUND: Patients with existing coronary artery stents are at an increased risk for major adverse cardiac events (MACEs) when undergoing noncardiac surgery (NCS). Although the use of antifibrinolytic (AF) therapy in NCS has significantly increased in the past decade, the relationship between perioperative AF use and its association with MACEs among patients with existing coronary artery stents has yet to be assessed. In this study, we aim to evaluate the association of MACEs in patients with existing coronary artery stents who receive perioperative AF therapy during orthopedic surgery. METHODS: A single-center retrospective cohort study was conducted in adult patients with existing coronary artery stents who underwent orthopedic surgery from 2008 to 2018. Two cohorts were established: patients with existing coronary artery stents who did not receive perioperative AF and patients with coronary artery stents who received perioperative AF. Associations between AF use and the primary outcome of MACEs within 30 days postoperatively and the secondary outcomes of thrombotic complications, excessive surgical bleeding, and intensive care unit (ICU) admissions were analyzed using logistic regression models. Inverse probability of treatment weighting was used to control for confounding. Secondary analyses examining the association between coronary stent type/timing and the outcomes of interest were performed using unadjusted logistic regression models. RESULTS: A total of 473 patients met study criteria, including 294 who did not receive AF and 179 patients who received AF. MACEs occurred in 15 (5.1%) patients who did not receive AF and 1 (0.6%) who received AF ( P = .007). In weighted analyses, no significant difference was found in patients who received AF with regard to MACEs (odds ratio [OR] = 0.13, 95% confidence interval [CI], 0.01-1.74, P = .12), thrombotic complications (OR = 1.19, 95% CI, 0.53-2.68, P = .68), or excessive surgical bleeding (OR = 0.13, 95% CI, 0.01-2.23, P = .16) compared to patients who did not receive AF. CONCLUSIONS: The results of this study are inconclusive whether an association exists between perioperative AF use in patients with coronary artery stents and the outcome of MACEs compared to patients who did not receive perioperative AF therapy. The authors acknowledge that the imprecise CI hinders the ability to definitively determine whether an association exists in the study population. Further large prospective studies, powered to detect differences in MACEs, are needed to assess the safety of perioperative AF in patients with existing coronary artery stents and to clarify the mechanism of perioperative MACEs in this high-risk population.

Abstract Prothrombin complex concentrate (PCC) administration has increased among cardiac surgery patients in recent years; however, use in LVAD implantation/exchange is not widespread due to the fear of thrombotic complications. The purpose of this study was to compare the clinical outcomes of patients undergoing LVAD implantation/exchange with intraoperative PCC administration versus traditional transfusion practices alone. Adult LVAD implants/exchanges at our institution between 2015 and 2018 were included. Patients were categorized as receiving intraoperative PCC or no‐PCC (traditional). The primary outcome was the need for allogenic transfusion and transfusion volume at 48 hours after initial intensive care unit (ICU) admission. Secondary outcomes included metrics of morbidity and mortality. A total of 160 patients (39 PCC, 121 traditional) were analyzed. In unadjusted analysis, patients in the PCC group received lower intraoperative transfusion volumes compared to the traditional group although not statistically significant (1464 mL [IQR 796, 4876] vs. 2568 mL [IQR 1292, 3606]; P value .37). In the fully adjusted analysis, patients in the PCC group had increased odds of transfusion within 48 hours of ICU admission (OR 4.06, 95% CI: 1.35‐12.20; P < .01); however, there was no significant difference in transfusion volumes ( P = .09). Patients receiving PCCs had higher incidence of deep vein thrombosis (10.3% vs. 0%; P < .01) and 30‐day mortality (17.9% vs. 4.1%; P < .01). LVAD pump thrombosis occurred in 2.6% versus 0.8% in the PCC and traditional groups, respectively; P = .98. Patients undergoing LVAD implantation and exchange represent a complex surgical cohort. The results of this study suggest that the intraoperative PCC use during LVAD implant/exchange was associated with reduced intraoperative transfusions. Intraoperative PCC use was, however, associated with higher odds of postoperative transfusion, although transfusion volumes were not significantly different. While the deep vein thrombosis and 30‐day mortality rates were higher in the PCC group, these results are likely related to the degree of surgical and patient complexity rather than PCC use itself. Further studies are needed to assess PCC use in this surgical cohort.

Abstract Background Heart transplant (HTx) patients can develop graft dysfunction (GD) without biopsy evidence of cell or antibody mediated rejection. Cardiac MRI (CMR) can detect inflammatory or infiltrative causes of cardiomyopathy however CMR findings in HTx recipients with GD have not been previously described. Purpose We sought to identify CMR characteristics of HTx patients with GD, and evaluate its additive value in its diagnosis and prognosis. Methods CMR has been performed routinely to evaluate GD at our institutions. There were 37 HTx recipients who presented with acute decline in left ventricular ejection fraction (LVEF) of &lt;50% and &gt;10% from baseline, with no biopsy evidence of rejection between 2007 and 2018. Coronary angiogram with IVUS was done to rule out allograft vasculopathy. Treatment of GD was per discretion of the treating clinician. Responders were defined as those with LVEF improvement &gt;10% at 180 days or greater post-presentation. LV and RV indices, the presence and pattern of late gadolinium enhancement (LGE) were determined by CMR. Results There were 34% females and mean age at transplant was 49±13 years. Median time from HTx to GD was 1.2 years. Presenting symptoms were heart failure (n=25), cardiogenic shock (n=1) and 11 patients were asymptomatic. Mean LVEF at presentation was 37±12% and donor specific antibodies were detected in 38% patients. Most patients were treated with steroid bolus (n=29), and/or plasmapheresis (n=23). There were no major changes made in immunosuppression in 6 patients. Delayed enhancement MRI was performed with standard inversion-recovery (IR) gradient echo sequences, between 5 and 20 minutes after institutional-standard protocol administration of IV gadolinium contrast. Biventricular LGE was present in 18/37 (49%) patients with GD and was more prevalent in responders (57%, 13 of 23) than non-responders (35%, 5 of 14), although not statistically significant (p=0.21). There were two patterns of LGE noted: 1) Diffuse epicardial (10 patients out which 9 were responders) and 2) Patchy pattern with non-specific distribution (8 patients out of which 4 were responders). Myocardial edema by triple IR sequencing was seen in 6 patients, all having diffuse epicardial pattern of enhancement matching the delayed enhancement pattern. When comparing different treatment groups, among those treated for GD (n=31), 12 of 21 (57%) responders had LGE and 4 of 10 (40%) non-responders had LGE (p=0.37), a pattern similar to the broader population. Among those not treated for GD (n=6), 1 of 2 responders had LGE and 1 of 4 non-responders had LGE (p=0.5). Conclusion CMR identified inflammation as a potential cause of GD in approximately 50% of HTx patients. There are 2 distinct patterns of LGE observed in GD, diffuse epicardial (56%) and patchy (44%). Although the presence of LGE in itself is not associated with myocardial recovery, 90% of patients with a diffuse epicardial pattern have recovery of GD. Funding Acknowledgement Type of funding source: None

AbstractBackgroundTransesophageal echocardiography (TEE) is a specialized form of cardiac ultrasound and has been associated with rare but serious complications. In patients with prior esophageal surgery, the risk of esophageal damage or the inability to perform a comprehensive and successful TEE warrants further evaluation.MethodsRetrospective study of patients with prior esophageal surgery who underwent TEE between June 21, 2002 and October 15, 2019. Medical and echocardiographic records were reviewed for image quality and procedural complications. Post‐procedure complications and 30‐day all‐cause mortality were collected. Evaluation by gastroenterology (GI), otolaryngology/ear, nose, throat (ENT), or thoracic surgery (TS) within 30 days of TEE was reviewed in detail.ResultsNinety‐five patients with prior esophageal surgery underwent 145 TEEs. The most commonly performed esophageal procedures were anti‐reflux operations (89%). TEE image quality was degraded in 16% while transgastric imaging was not completed in 37% of cases. A comprehensive TEE was completed in 57% of patients with diagnostic study goals achieved in 96% of cases. Comments describing procedural difficulty were reported in 6% while comments on procedural complications occurred in 1% of cases. Post‐procedure complications occurred in 1% of patients including hypotension and unplanned hospital/ICU admission.ConclusionsConcern for esophageal damage and the inability to perform a comprehensive and diagnostic TEE may limit the usefulness of TEE in patients with prior esophageal surgery. While TEE‐associated complications were rare in this series, a conservative approach with a thorough pre‐procedure assessment, including multi‐disciplinary evaluation when appropriate, is prudent in this complex cohort of patients.