Long Ngo

Implicit Bias among Physicians and its Prediction of Thrombolysis Decisions for Black and White Patients

ContextStudies documenting racial/ethnic disparities in health care frequently implicate physicians’ unconscious biases. No study to date has measured physicians’ unconscious racial bias to test whether this predicts physicians’ clinical decisions.ObjectiveTo test whether physicians show implicit race bias and whether the magnitude of such bias predicts thrombolysis recommendations for black and white patients with acute coronary syndromes.Design, Setting, and ParticipantsAn internet-based tool comprising a clinical vignette of a patient presenting to the emergency department with an acute coronary syndrome, followed by a questionnaire and three Implicit Association Tests (IATs). Study invitations were e-mailed to all internal medicine and emergency medicine residents at four academic medical centers in Atlanta and Boston; 287 completed the study, met inclusion criteria, and were randomized to either a black or white vignette patient.Main Outcome MeasuresIAT scores (normal continuous variable) measuring physicians’ implicit race preference and perceptions of cooperativeness. Physicians’ attribution of symptoms to coronary artery disease for vignette patients with randomly assigned race, and their decisions about thrombolysis. Assessment of physicians’ explicit racial biases by questionnaire.ResultsPhysicians reported no explicit preference for white versus black patients or differences in perceived cooperativeness. In contrast, IATs revealed implicit preference favoring white Americans (mean IAT score = 0.36, P < .001, one-sample t test) and implicit stereotypes of black Americans as less cooperative with medical procedures (mean IAT score 0.22, P < .001), and less cooperative generally (mean IAT score 0.30, P < .001). As physicians’ prowhite implicit bias increased, so did their likelihood of treating white patients and not treating black patients with thrombolysis (P = .009).ConclusionsThis study represents the first evidence of unconscious (implicit) race bias among physicians, its dissociation from conscious (explicit) bias, and its predictive validity. Results suggest that physicians’ unconscious biases may contribute to racial/ethnic disparities in use of medical procedures such as thrombolysis for myocardial infarction.

Implementation and outcomes of the Multiple Urgent Sepsis Therapies (MUST) protocol.

Objectives:To describe the effectiveness of a comprehensive, interdisciplinary sepsis treatment protocol with regard to both implementation and outcomes and to compare the mortality rates and therapies of patients with septic shock with similar historical controls. Design:Prospective, interventional cohort study with a historical control comparison group. Setting:Urban, tertiary care, university hospital with 46,000 emergency department visits and 4,100 intensive care unit admissions annually. Patients:Inclusion criteria were a) emergency department patients aged ≥18 yrs, b) suspected infection, and c) lactate of >4 mmol/L or septic shock. Exclusion criteria were a) emergent operation, b) prehospital cardiac arrest, and c) comfort measures only. Time period: protocol, November 10, 2003, through November 9, 2004; historical controls, February 1, 2000, through January 31, 2001. Intervention:A sepsis treatment pathway incorporating empirical antibiotics, early goal-directed therapy, drotrecogin alfa, steroids, intensive insulin therapy, and lung-protective ventilation. Measurements and Main Results:There were 116 protocol patients, with a mortality rate of 18% (11–25%), of which 79 patients had septic shock. Comparing these patients with 51 historical controls, protocol patients received more fluid (4.0 vs. 2.5 L crystalloid, p < .001), earlier antibiotics (90 vs. 120 mins, p < .013), more appropriate empirical coverage (97% vs. 88%, p < .05), more vasopressors in the first 6 hrs (80% vs. 45%, p < .001), tighter glucose control (mean morning glucose, 123 vs. 140, p < .001), and more frequent assessment of adrenal function (82% vs. 10%, p < .001), with a nonstatistically significant increase in dobutamine use (14% vs. 4%, p = .06) and red blood cell transfusions (30% vs. 18%, p = .07) in the first 24 hrs. For protocol patients with septic shock, 28-day in-hospital mortality was 20.3% compared with 29.4% for historical controls (p = .3). Conclusions:Clinical implementation of a comprehensive sepsis treatment protocol is feasible and is associated with changes in therapies such as time to antibiotics, intravenous fluid delivery, and vasopressor use in the first 6 hrs. No statistically significant decrease in mortality was demonstrated, as this trial was not sufficiently powered to assess mortality benefits.

Breast Cancer Among the Oldest Old: Tumor Characteristics, Treatment Choices, and Survival

PURPOSE Few data are available on breast cancer characteristics, treatment, and survival for women age 80 years or older. PATIENTS AND METHODS We used the linked Surveillance, Epidemiology and End Results-Medicare data set from 1992 to 2003 to examine tumor characteristics, treatments (mastectomy, breast-conserving surgery [BCS] with radiation therapy or alone, or no surgery), and outcomes of women age 80 years or older (80 to 84, 85 to 89, > or = 90 years) with stage I/II breast cancer compared with younger women (age 67 to 79 years). We used Cox proportional hazard models to examine the impact of age on breast cancer-related and other causes of death. Analyses were performed within stage, adjusted for tumor and sociodemographic characteristics, treatments received, and comorbidities. Results In total, 49,616 women age 67 years or older with stage I/II disease were included. Tumor characteristics (grade, hormone receptivity) were similar across age groups. Treatment with BCS alone increased with age, especially after age 80. The risk of dying from breast cancer increased with age, significantly after age 80. For stage I disease, the adjusted hazard ratio of dying from breast cancer for women age > or = 90 years compared with women age 67 to 69 years was 2.6 (range, 2.0 to 3.4). Types of treatments received were significantly associated with age and comorbidity, with age as the stronger predictor (26% of women age > or = 80 years without comorbidity received BCS alone or no surgery compared with 6% of women age 67 to 79 years). CONCLUSION Women age > or = 80 years have breast cancer characteristics similar to those of younger women yet receive less aggressive treatment and experience higher mortality from early-stage breast cancer. Future studies should focus on identifying tumor and patient characteristics to help target treatments to the oldest women most likely to benefit.

Competing Risk of Death: An Important Consideration in Studies of Older Adults

Clinical studies often face the difficult problem of how to account for participants who die without experiencing the study outcome of interest. In a geriatric population with considerable comorbidities, the competing risk of death is especially high. Traditional approaches to describe risk of disease include Kaplan‐Meier survival analysis and Cox proportional hazards regression, but these methods can overestimate risk of disease by failing to account for the competing risk of death. This report discusses traditional survival analysis and competing risk analysis as used to estimate risk of disease in geriatric studies. Furthermore, it illustrates a competing risk approach to estimate risk of second hip fracture in the Framingham Osteoporosis Study and compares the results with traditional survival analysis. In this example, survival analysis overestimated the 5‐year risk of second hip fracture by 37% and the 10‐year risk by 75% compared with competing risk estimates. In studies of older individuals in which a substantial number of participants die during a long follow‐up, the cumulative incidence competing risk estimate and competing risk regression should be used to determine incidence and effect estimates. Use of a competing risk approach is critical to accurately determining disease risk for elderly individuals and therefore best inform clinical decision‐making.

Renal Cell Carcinoma: Dynamic Contrast-enhanced MR Imaging for Differentiation of Tumor Subtypes—Correlation with Pathologic Findings

PURPOSE To retrospectively evaluate whether the enhancement patterns of pathologically proved clear cell, papillary, and chromophobe renal cell carcinomas (RCCs) measured on clinical dynamic contrast agent-enhanced magnetic resonance (MR) images permit accurate diagnosis of RCC subtype. MATERIALS AND METHODS This study was Institutional Review Board approved and HIPAA compliant; informed consent was waived. One hundred twelve patients (76 men, 36 women; age range, 25-88 years; mean age, 58.1 years) underwent MR imaging of 113 renal masses (mean diameter, 5.4 cm) with pathologic diagnoses of clear cell (n = 75), papillary (n = 28), or chromophobe (n = 10) RCC. A 1.5-T clinical MR protocol was used before and after (corticomedullary and nephrographic phases) intravenous administration of contrast agent. Region-of-interest measurements within tumor and uninvolved renal cortex were used to calculate percentage signal intensity change and tumor-to-cortex enhancement index. Subtype groups were compared by using linear mixed-effects models. Receiver operating characteristic (ROC) curve analysis was performed for the comparison of clear cell and papillary RCCs. RESULTS On both the corticomedullary and nephrographic phase images, clear cell RCCs showed greater signal intensity change (205.6% and 247.1%, respectively) than did papillary RCCs (32.1% and 96.6%, respectively) (P < .001). Chromophobe RCCs showed intermediate change (109.9% and 192.5%, respectively). The tumor-to-cortex enhancement indexes at corticomedullary and nephrographic phases were largest for clear cell RCCs (1.4 and 1.2, respectively), smallest for papillary RCCs (0.2 and 0.4, respectively), and intermediate for chromophobe RCCs (0.6 and 0.8, respectively). Signal intensity changes on corticomedullary phase images were the most effective parameter for distinguishing clear cell and papillary RCC (area under ROC curve, 0.99); a threshold value of 84% permitted distinction with 93% sensitivity and 96% specificity. CONCLUSION Clear cell, papillary, and chromophobe RCCs demonstrate different patterns of enhancement on two-time point clinical dynamic contrast-enhanced MR images, allowing their differentiation with high sensitivity and specificity.

Asymptomatic Reactivation of JC Virus in Patients Treated with Natalizumab

BACKGROUND Progressive multifocal leukoencephalopathy (PML) occurs in a fraction of patients with multiple sclerosis who were treated with natalizumab. Most adults who are infected with the JC virus, the etiologic agent in PML, do not have symptoms. We sought to determine whether exposure to natalizumab causes subclinical reactivation and neurotropic transformation of JC virus. METHODS We followed 19 consecutive patients with multiple sclerosis who were treated with natalizumab over an 18-month period, performing quantitative polymerase-chain-reaction assays in blood and urine for JC virus reactivation; BK virus, a JC virus-related polyomavirus, was used as a control. We determined JC virus-specific T-cell responses by means of an enzyme-linked immunospot assay and antibody responses by means of an enzyme-linked immunosorbent assay and analyzed JC virus regulatory-region sequences. RESULTS After 12 months of natalizumab therapy, the prevalence of JC virus in the urine of the 19 patients increased from a baseline value of 19% to 63% (P=0.02). After 18 months of treatment, JC virus was detectable in 3 of 15 available plasma samples (20%) and in 9 of 15 available samples of peripheral-blood mononuclear cells (60%) (P=0.02). JC virus regulatory-region sequences in blood samples and in most of the urine samples were similar to those usually found in PML. Conversely, BK virus remained stable in urine and was undetectable in blood. The JC virus-specific cellular immune response dropped significantly between 6 and 12 months of treatment, and variations in the cellular immune response over time tended to be greater in patients in whom JC viremia developed. None of the patients had clinical or radiologic signs of PML. CONCLUSIONS Subclinical reactivation of JC virus occurs frequently in natalizumab-treated patients with multiple sclerosis. Viral shedding is associated with a transient drop in the JC virus-specific cellular immune response.

A prospective, multicenter derivation of a biomarker panel to assess risk of organ dysfunction, shock, and death in emergency department patients with suspected sepsis

Objective:To define a biomarker panel to predict organ dysfunction, shock, and in-hospital mortality in emergency department (ED) patients with suspected sepsis. Design:Prospective observational study. Setting:EDs of ten academic medical centers. Patients:There were 971 patients enrolled. Inclusion criteria: 1) ED patients age > 18; 2) suspected infection or a serum lactate level > 2.5 mmol/L; and 3) two or more systemic inflammatory response syndrome criteria. Exclusion criteria: pregnancy, do-not-resuscitate status, or cardiac arrest. Measurements and Main Results:Nine biomarkers were assayed from blood draws obtained on ED presentation. Multivariable logistic regression was used to identify an optimal combination of biomarkers to create a panel. The derived formula for weighting biomarker values was used to calculate a “sepsis score,” which was the predicted probability of the primary outcome of severe sepsis (sepsis plus organ dysfunction) within 72 hrs. We also assessed the ability of the sepsis score to predict secondary outcome measures of septic shock within 72 hrs and in-hospital mortality. The overall rates of each outcome were severe sepsis, 52%; septic shock, 39%; and in-hospital mortality 7%. Among the nine biomarkers tested, the optimal 3-marker panel was neutrophil gelatinase-associated lipocalin, protein C, and interleukin−1 receptor antagonist. The area under the curve for the accuracy of the sepsis score derived from these three biomarkers was 0.80 for severe sepsis, 0.77 for septic shock, and 0.79 for death. When included in multivariate models with clinical variables, the sepsis score remained highly significant (p < 0.001) for all the three outcomes. Conclusions:A biomarker panel of neutrophil gelatinase-associated lipocalin, interleukin-1ra, and Protein C was predictive of severe sepsis, septic shock, and death in ED patients with suspected sepsis. Further study is warranted to prospectively validate the clinical utility of these biomarkers and the sepsis score in risk-stratifying patients with suspected sepsis.

Diagnostic utility of abbreviated fluency measures in Alzheimer disease and vascular dementia

Background: Several studies indicate semantic fluency more sensitively discriminates patients with Alzheimer disease (AD) from normal elderly persons, with disproportionate impairment of semantic over phonemic fluency. Objective: To determine the ability of abbreviated fluency measures in the clinic setting (1-minute letter F and animal fluency tests) to detect AD, and to assess whether difference scores between these measures discriminate patients with AD and vascular dementia (VaD) from normal elderly persons. Methods: The authors studied patients with AD (n = 98) meeting National Institute of Neurological Communicative Disorders and Stroke–Alzheimer’s Disease and Related Disorders Association criteria, VaD patients (n = 18) meeting National Institute of Neurological Disorders and Stroke–Association Internationale pour la Recherche et l’Enseignement en Neurosciences criteria, cognitively impaired but not demented patients (CIND; n = 25), vascular CIND patients (VCIND; n = 24), and normal control subjects (NCs; n = 46). Results: Analysis of covariance controlling for age, education, and overall impairment indicated all groups generated fewer animal names compared with NCs, whereas only VaD patients generated fewer letter F words compared with NCs. On standardized scores, patients with AD and CIND, unlike those with VCIND and VaD, scored significantly worse on the animal fluency test than on the letter F fluency test. The animal fluency test was superior in discriminating all patient groups from NCs. Positive likelihood ratios (PLRs) revealed animal fluency scores <15 were 20 times more likely in a patient with AD than in an NC (sensitivity = 0.88; specificity = 0.96). Letter F scores <4 discriminated VaD from AD patients (PLR = 4.0; sensitivity = 0.44; specificity = 0.90). Difference scores <0 (i.e., fewer animal than letter F words) discriminated patients with VCIND from those with CIND (PLR = 2.5; sensitivity = 0.32; specificity = 1.00). Conclusions: A 1-minute semantic fluency test can assist in early detection of dementia in the memory clinic setting.

The association of endothelial cell signaling, severity of illness, and organ dysfunction in sepsis

IntroductionPrevious reports suggest that endothelial activation is an important process in sepsis pathogenesis. We investigated the association between biomarkers of endothelial cell activation and sepsis severity, organ dysfunction sequential organ failure assessment (SOFA) score, and death.MethodsThis is a prospective, observational study including adult patients (age 18 years or older) presenting with clinical suspicion of infection to the emergency department (ED) of an urban, academic medical center between February 2005 and November 2008. Blood was sampled during the ED visit and biomarkers of endothelial cell activation, namely soluble fms-like tyrosine kinase-1 (sFlt-1), plasminogen activator inhibitors -1 (PAI-1), sE-selectin, soluble intercellular adhesion molecule (sICAM-1), and soluble vascular cell adhesion molecule (sVCAM-1), were assayed. The association between biomarkers and the outcomes of sepsis severity, organ dysfunction, and in-hospital mortality were analyzed.ResultsA total of 221 patients were included: sepsis without organ dysfunction was present in 32%, severe sepsis without shock in 30%, septic shock in 32%, and 6% were non-infected control ED patients. There was a relationship between all target biomarkers (sFlt-1, PAI-1, sE-selectin, sICAM-1, and sVCAM-1) and sepsis severity, P < 0.05. We found a significant inter-correlation between all biomarkers, including the strongest correlations between sFlt-1 and sE-selectin (r = 0.55, P < 0.001), and between sFlt-1 and PAI-1 (0.56, P < 0.001). Among the endothelial cell activation biomarkers, sFlt-1 had the strongest association with SOFA score (r = 0.66, P < 0.001), the highest area under the receiver operator characteristic curve for severe sepsis of 0.82, and for mortality of 0.91.ConclusionsMarkers of endothelial cell activation are associated with sepsis severity, organ dysfunction and mortality. An improved understanding of endothelial response and associated biomarkers may lead to strategies to more accurately predict outcome and develop novel endothelium-directed therapies in sepsis.

Effect of Institution-Wide Policy of Colonoscopy Withdrawal Time ≥7 Minutes on Polyp Detection

BACKGROUND & AIMS Practice guidelines recommend that endoscopists spend at least 7 minutes examining the colonic mucosa during colonoscopy withdrawal to optimize polyp yield. The aim of this study was to determine if the implementation of an institution-wide policy of colonoscopy withdrawal time > or = 7 minutes was associated with an increase in colon polyp detection. METHODS All 42 endoscopists at our institute were asked to attain a colonoscopy withdrawal time of at least 7 minutes. Compliance with 7-minute withdrawal time was recorded for all nontherapeutic colonoscopies. Polyp detection ratio (number of polyps detected divided by number of colonoscopies performed) was computed. Regression models were used to assess the association between compliance with 7-minute withdrawal time and polyp detection. RESULTS During the study period, 23,910 colonoscopies were performed. The average age of patients was 56.8 years, and 54% were female. Colon cancer screening or surveillance was the indication for 42.5% of colonoscopies. At the beginning of the study, the polyp detection ratio was 0.48. Compliance with 7-minute withdrawal time for nontherapeutic procedures increased from 65% at the beginning of the initiative to almost 100%. However, no increase in polyp detection ratio was noted over the same period for all polyps (slope, 0.0006; P = .45) or for polyps 1-5 mm (slope, 0.001; P = .26), 6-9 mm (slope, 0.002; P = .43), or > or = 10 mm (slope, 0.006; P = .13). No association was detected when only colonoscopies performed for screening or surveillance were analyzed. CONCLUSIONS An institution-wide policy of colonoscopy withdrawal time > or = 7 minutes had no effect on colon polyp detection.