A prospective cohort study in children with acute lymphoblastic leukemia demonstrates a significant quality of life burden during therapy by showing that many of these patients experience impairment in physical and emotional functioning. A retrospective cohort study in neuroblastoma survivors finds significant cognitive and behavioral impairment, highlighting the prevalence and severity of long-term effects from treatment. Two studies examining data from Yale’s childhood cancer survivorship clinic show that health services for these patients is a major challenge, as the majority of patients do not attend survivorship clinic even up to 10 years post-diagnosis, and the primary care providers of survivors generally feel uncomfortable being responsible for survivorship care. Lastly, a pilot study utilizing activity trackers in pediatric cancer survivors demonstrates the potential feasibility and efficacy of such an intervention to increase exercise in hopes of mitigating late effects. This study represents the type of future research that may help lessen the burden of therapy-related chronic conditions.

Moving forward, successful cancer treatment in children must address quality of life costs and therapy-related chronic conditions, in addition to managing the disease itself. This evolving paradigm in pediatric oncology requires future research in both optimizing health services for this patient population, as well as interventions during and after treatment to help minimize long-term toxicities.

Methods: Study subjects were selected from HIV/HCV co-infected patients receiving care at Yale-New Haven Hospital from February 2014 through April 2016 without advanced fibrosis 5 years prior to study entry. Annual FIB-4 scores dating back 5 years were calculated from the most recent FIB-4 score or the last FIB-4 prior to DAA treatment initiation. Baseline demographics and clinical characteristics including HCV genotype, antiretroviral regimen, HIV viral loads and CD4 counts were collected. Patients were further categorized based on FIB-4 progression over the course of 5 years. Univariate and multivariable logistic regression models were used to examine factors associated with FIB-4 progression and a p-value of 0.05 was chosen as the threshold for statistical significance.

Results: There were 93 patients evaluated including 65 men and 28 women; mean age of 56.7 years; 32.3% were white, 53.8% were black. Injection drug use (IDU) was the major risk factor for HCV acquisition (63.4%) and the most common genotype was genotype 1 (81.2%). The median CD4+ count was 564 cells/mm3, and the majority (88.4%) had HIV viral loads <50 copies/mL. Over 5 years, 25 (26.9%) had FIB-4 progress to >3.25 and 68 (73.1%) had FIB-4 remain <3.25. Demographic variables (age, gender, race/ethnicity, BMI, substance use), clinical variables (HIV viral load, CD4 count, ART use, HIV duration, HCV duration and viral load) and co-morbid conditions such as diabetes and hyperlipidemia did not differ significantly between those whose FIB-4 stayed below 3.25 and those whose FIB-4 progressed to above 3.25 in univariate and multivariable logistic regression models.

Conclusion: In this study of 93 HIV/HCV co-infected patients without baseline advanced fibrosis, 26.9% progressed to advanced fibrosis over the course of 5 years. We did not identify any statistically significant factors that predicted those who were more likely to progress, although clinically relevant factors such as absence of HIV virologic control, low CD4 count, and lack of statin use showed a trend towards significance and should be assessed in future studies in a larger cohort. Our findings highlight the importance of prioritizing all patients with HCV/HIV co-infection for HCV treatment.

Male Harlan Sprague-Dawley rats weighing 100-125g underwent the induction of cirrhosis via inhalation of carbon tetrachloride (CCl4) three times a week until the development of ascites. Animals with ascites were randomized to placebo (methylcellulose solution) or obeticholic acid, administered by orogastric tube at a dose of 5mg/kg/day for 14 days. Stool samples were collected at weeks 1, 2, 4, and 7; at the development of ascites; 1 week into treatment; and at the time of sacrifice. At the end of the 14-day treatment period, animals underwent non-survival surgery with general anesthesia (isoflurane open-drop, followed by ketamine (75mg/kg)/xylazine (5mg/kg) intramuscularly) and strict aseptic conditions. Ascites fluid, blood, and mesenteric lymph nodes were cultured to assess for bacterial growth, a sign of bacterial translocation. Samples from the intestine and colon underwent 16S amplification and sequencing to assess the microbiome. The liver was fixed for histopathological analysis.

A first group of 20 rats was used to establish the experimental model of cirrhosis. These animals completed the induction process with 20-22 weeks of exposure to carbon tetrachloride by inhalation per Dr. Iwakiri’s lab, Lopez-Novoa1 and Garcia-Tsao et al2. However, the dose of CCl4 administered was too low to yield meaningful ascites; the rats consistently gained weight and did not exhibit any of the physical or behavioral hallmarks of animals with decompensated cirrhosis. Timepoint stool microbiome studies were completed on these animals, showing decreases in some beneficial species, such as Bacteroides, but not others, such as Lactobacillus and Bifidobacterium; and insignificant changes in Proteobacteria. After some troubleshooting, we pursued a more aggressive approach to exposing the animals to carbon tetrachloride by inhalation, involving exposing them to enough gas to anesthetize them starting week 4 of the induction process. At the time of this writing, 11 of 40 animals had developed ascites, 6 were undergoing treatment, and 1 had undergone laparotomy and sample collection.

We have so far established a more efficient way of inducing cirrhosis with ascites compared to the one recommended by Dr. Iwakiri’s lab. The induction of cirrhosis by carbon tetrachloride inhalation in rats requires escalating doses of the gas to reach the amount necessary to cause narcosis, in order to produce ascites. The effect of obeticholic acid on bacterial translocation has yet to be completed.

Daniel X. Yang and James B. Yu, Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT

Effective cancer screening detects cancer at an earlier and more curable stage, and thus reduces incident late-stage disease. However, in recent years, there has been sustained controversy over the net benefit of cancer screening due to concerns of overdiagnosis and overtreatment. Therefore, we sought to demonstrate the impact of cancer screening efforts at the population level, specifically examining incidence trends in colorectal, cervical, and prostate cancers. We sought to estimate the number of cervical and colorectal cancer cases prevented during an era of widespread screening, and to examine temporal associations between the 2012 United States Preventative Services Task Force (USPSTF) recommendation against PSA screening and emerging trends in metastatic prostate cancer. We hypothesize that screening efforts would be associated with characteristic shifts in cancer incidence: (1) Increased colonoscopy and sigmoidoscopy utilization would drive increased detection of left-sided colorectal cancers; (2) Widespread Pap smear use would be most effective in higher-risk populations; and (3) Decreased PSA screening would lead to increased late-stage disease. Colorectal, cervical, and prostate cancer incidence spanning over three decades from 1973 to 2013 were collected from the Surveillance, Epidemiology, and End Result (SEER) database. Screening utilization trends were collected from the National Cancer Institute Progress Reports, the National Health Interview Survey, and literature review. We demonstrate that widespread colorectal and cervical cancer screening were temporally associated with 236,000 to 550,000 cases of colorectal cancer and 105,000 to 492,000 cases of cervical cancer, respectively, over the past three decades from 1979 to 2009. Over this time period, the incidence of late-stage colorectal cancer decreased from 118 to 74 cases per 100,000 people (p < .001), and the incidence of late-stage cervical cancer decreased from 5.3 to 3.7 cases per 100,000 women (p < .001). Since the 2008 to 2012 pullback in PSA screening, metastatic prostate cancer incidence among United States men age 70 years or older increased from 54.2 per 100,000 men in 2011, to 56.6 per 100,000 men in 2012 and 59.5 per 100,000 men in 2013. Moreover, we address our hypotheses by: (1) Demonstrating a more prominent decrease in left-sided compared to right-sided colorectal cancer incidence; (2) Demonstrating a convergence of race-specific cervical cancer incidence over time; and (3) Characterizing the recent increase in metastatic prostate cancer incidence. Taken together, our studies document the successes of screening efforts as well as underscore the challenges of balancing the benefits of early detection against the risks of overdiagnosis and overtreatment.

I describe what I term “American medical fundamentalism”: a belief system promoted by the medical profession, with “patient care above all else” as its central tenet and self-sacrifice as a means of achieving that goal. I assert that this latter concept is tied deeply to the identity of the medical profession, and that opposition to residency duty hour regulation is an external manifestation of a deeper existential crisis among physicians, as they confront a shifting landscape that threatens their previously elevated professional position and authority.

I also note that the language used to describe professionalism in medicine draws from traditionally “masculine” values, and that the arguments against stronger residency work hour restrictions uncomfortably parallel those used in the mid-twentieth century to advocate against training women physicians. Finally, in the context of growing awareness of an epidemic of burnout in the medical profession, I suggest that equating self-sacrifice with compassion and competence is not only inaccurate, but also dangerous, and necessitates a reexamination of the values underlying the medical profession’s conception of what it means to be a “good doctor.”

To examine outcomes after postoperative radiation (PORT) and the impact of chemoradiation facility case volume in various populations of patients with non-small cell lung cancer (NSCLC).

Methods

We used patients within the National Cancer Data Base. In our analysis of outcomes after PORT, we examined two populations: 1) incompletely-resected patients with pathologic stage N0-N2, overall American Joint Committee on Cancer stage II-III NSCLC who had undergone a lobectomy or pneumonectomy with positive surgical margins coded as receiving external beam PORT at 50-74 Gy or observation were included and 2) completely-resected patients with stage II-III N0-N1 margin-negative NSCLC who received a total radiotherapy dose within the range of 45.0-74.0 Gy given within the initial treatment course. To examine the impact of chemoradiation (CRT) facility case volume, we identified clinical stage III NSCLC patients diagnosed in 2004-2006 who were treated with definitive concurrent CRT to 59.4-74.0 Gy. High-volume facilities (HVF) were defined as those in the 90th percentile of annual CRT volume (≥12 cases/year). For all populations, multivariable logistic regression was used to determine factors associated with PORT receipt and independent predictors of receiving CRT at HVF. Cox proportional hazards regression was performed for multivariable analyses of overall survival for all populations.

Results

Among 3,395 included incompletely-resected patients, 1,207 (35.6%) received PORT. On multivariable analysis adjusting for demographic and clinicopathologic covariates, PORT (HR=0.80; 95%CI: 0.70-092) was associated with improved survival among margin-positive patients. Subset analysis by nodal stage showed PORT improved survival across all nodal stages.

We then identified 2,167 (6.7%) and 30,269 (93.3%) patients with completely-resected stage II or III N0-N1 NSCLC who were treated with and without PORT, respectively. Overall, completely-resected patients who received PORT had worse survival (HR=1.30; 95% CI 1.20-1.40) compared to those not receiving PORT. This association was unchanged when limited to patients receiving modern treatment with 3-CRT or IMRT (HR=1.35; 95% CI: 1.10-1.65].

Among 10,072 included chemoradiation patients, 1,207 (12.0%) were treated at HVF. When controlling for demographic and clinical covariates including academic affiliation, treatment at HVF was independently associated with a significantly decreased risk of death (HR=0.93; 95% CI: 0.87-0.99; p=0.03). Propensity score matching showed that these findings were robust (HR=0.91; 95% CI: 0.84-0.99; p=0.04).

Conclusions

Our findings suggest that PORT is associated with improved overall survival in incompletely-resected Stage II-III N0-N2 NSCLC patients. However, we found no evidence of benefit from PORT for completely-resected N0-N1 NSCLC, regardless of dose or technique. We also found that treatment at high-volume facilities is associated with improved overall survival among stage III NSCLC patients receiving definitive concurrent CRT. Thus, delivery of PORT in incompletely-resected early stage NSCLC and centralization of care to HVF for stage III NSCLC patients receiving definitive concurrent CRT may be beneficial strategies to optimize patient outcomes.

Some guidelines have emerged in recent years to help distinguish headache due to SAH from more benign causes. However, their generalizability is limited due to inclusion and exclusion criteria of individual studies, as well as unclear definitions of terminology used in criteria. This study aims to address these shortcomings by creating a generalizable clinical decision tool using a broader patient population, and providing clear definitions in a standardized questionnaire.

In this prospective observational study, 158 patients were interviewed using a standardized 15-item questionnaire. Patients eligible were alert, able to communicate and answer the questionnaires in English, and had a headache presentation unrelated to trauma. Data was used to identify differential features of headache secondary to SAH, and these features were used to create a 5-item clinical decision tool.

A total of 583 patients were eligible. Of those, 158 (27%) were enrolled, provided consent, and completed our questionnaire. Of these, 20 had SAH. After adjusting for confounders, patients with SAH were more likely to be ≥ 50 years old, experienced the “worst headache of [their] life,” had headache onset during exertion, had peak intensity instantaneously (<1 second), or had associated neck stiffness/pain. We proposed two clinical decision tools using these 5 features, which had a sensitivity of 100%, negative predictive value of 100%, and negative likelihood ratio of 0, comparable to existing rules.

Thus, our decision rules agree with existing ones and perform similarly, but they have clearer definition for terminology used and can be more generalizable after external validation.

In this study we established the biocompatibility of iPSC derived retinal progenitor cells (RPCs) and human fetal retinal pigment epithelium (hfRPE) with a novel scaffold composed of gelatin, chondroitin sulfate, and hyaluronic acid (GCH). iPSC-RPCs seeded onto GCH scaffolds either as clusters or after dissociation into single cells attached to, proliferated, and differentiated towards a neural retinal fate as evidenced by gene expression and immunohistochemistry. Dissociation of RPC clusters before seeding, however, resulted in a significant decrease in expression of Recoverin, a key photoreceptor marker.

This study also established a model of outer retinal damage in pigs. Argon laser induced outer retinal damage in the porcine visual streak and was detected on histology and multifocal electroretinography (mfERG) at time points immediately after and 10 days after injury. This study laid the groundwork for a pilot study of subretinal GCH scaffold implantation in pigs.

As of October 2016, we have enrolled 173 ICH patients, with an average response rate (i.e. the ability to complete follow-up evaluations) of 85-89% across all follow-up time-points. Using either the mRS or BI, measures of motor disability, there was a significant improvement in scores (p values < 0.001) between discharge and three, six, and 12 months. The mRS scores failed to show a significant difference between follow-up time-points (i.e. three to six months, six to 12 months, etc.). However the BI showed significant improvement in scores between three and 12 months (p = 0.013), as well as between six and 12 months (p = 0.025). A subsequent analysis, comparing BI scores between three to 12 months, showed a significant relationship of time by age (p = 0.047; i.e. differences in improvement depending on the age of the patient), time by admission GCS (p = 0.010; i.e. differences in improvement depending on the admission GCS of the patient), time by ICH volume (p = 0.004; i.e. differences in improvement depending on the initial ICH volume of the patient), and time by location (p = 0.005; i.e. differences in improvement depending on the initial ICH location of the patient). Follow-up plots suggest patients with characteristically more disabling injury as traditionally measured through metrics like the ICH score (i.e. older patients, lower admission GCS, larger ICH volume, deeper location) showed the most improvement between three and 12 months. These results would suggest that ICH motor recovery will improve up to at least 12 months post-injury and patients with more debilitating injury will show improved recovery at the late-stage.

Across quality of life measures (i.e. EQ-5D, EQ-VAS, SS-QoL), patients showed the greatest recovery in motor domains (i.e. Mobility, Self-Care, Activities, Upper Extremities). Despite this improvement, patients self-reported health scores (EQ-VAS, a subjective 0 to 100 point score that patients use to describe their overall health) declined across time-points. There were additionally decreases in scores associated with the “Thinking” domain as well as increases in “Pain” symptoms. Since decreases in self-reported health score matched trends in the “Thinking” and “Pain” domains rather than the increases observed in motor-domains, patients may value non-motor domains when assessing their overall health compared to motor recovery. Finally, while there are improvements of cognitive impairment (measured via the MoCA, total score < 18) across time-points (discharge: 72%, three months: 65%, six months: 53%, 12 months: 48%), these results are more profound for deep compared to lobar ICH.

Preliminary analyses of the cohort enrolled thus far suggest ICH patients improve in motor recovery up to at least 12 months post-injury. However despite robust motor recovery, patients still exhibit high rates of cognitive impairment and impairment in quality of life. These results necessitate the need for long-term end-points in ICH interventional trials with further emphasis on non-motor domains. Continued enrollment and prospective evaluation of this cohort promises further insight into the recovery process.

For tissue-based assays, samples from melanoma patients that received nivolumab, pembrolizumab, or combination ipilimumab/nivolumab at Yale New Haven Hospital from May 2013 to March 2016 were collected. Expression of IRF-1 and PD-L1 in archival pre-treatment formalin-fixed, paraffin-embedded tumor samples were assessed by the AQUA method of quantitative immunofluorescence. Objective radiographic response (ORR) and progression-free survival (PFS) were assessed using modified RECIST v1.1 criteria. For pilot studies of sPD-L1, plasma from 62 patients with non-small cell lung cancer and 10 cancer-free controls were accessed from pre-existing de-identified tissue banks at Yale School of Medicine.

Nuclear IRF-1 expression was higher in patients with partial or complete response (PR/CR) than in patients with stable or progressive disease (SD/PD) (p = 0.044). There was an insignificant trend toward higher PD-L1 expression in patients with PR/CR (p = 0.085). PFS was higher in the IRF-1-high group than the IRF-1-low group (p = 0.017), while PD-L1 expression had no effect on PFS (p = 0.83). In a subset analysis, a strong association between IRF-1 and PFS is seen in patients treated with combination ipilimumab and nivolumab (p = 0.0051). Higher CD3 infiltrates were more likely to be associated with PR/CR (p = 0.0067) and with improved PFS (p = 0.017). Conversely, higher expression of Granzyme B within CD3+ cells was associated with SD/PD (p = 0.023) and a trend toward inferior PFS (p = 0.066). Soluble PD-L1 in human plasma was detected by ELISA, and was elevated in NSCLC cases compared to controls (p < 0.0001).

As a measure of PD-L1 expression capability, IRF-1 expression may be a more valuable predictive biomarker for anti-PD-1 therapy than PD-L1 itself. Additionally, patients with quiescent immune infiltrates may benefit more from anti-PD-1 therapy than those with immune-active tumors. The viability of plasma-based predictive biomarkers for immunotherapies warrants additional investigation.

In this study, I examine oxygen concentration and two small-molecule bone morphogenic protein (BMP) inhibitors, DMH-1 and LDN193189, as potential avenues of increased control over mesenchymal stem cell (MSC) differentiation of into vascular smooth muscle cells (SMCs). Applying BMP-inhibitor concentrations from 0.01 to 10 µM at oxygen tensions of 2 & 20% over two weeks of growth, I use reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) to evaluate resultant expression levels of smooth muscle (SM22-α), bone (OCN), and cartilage (Col2a) marker genes. Via multiple linear regression, I demonstrate that low oxygen growth causes a statistically significant SM22-α downregulation (∆∆Cq = 1.77 ± 0.22, mean ± standard error, p <0.0%) coupled with increased Col2a (∆∆Cq = -2.85 ± 0.73, p =0.1%) and type-I collagen expression (∆∆Cq = -1.75 ± 0.65, p = 1.2%), suggesting that physiological oxygen tensions increase the incidence of chondrogenic differentiation. In contrast, LDN193189 increases SM22-α expression (∆∆Cq = -0.78 ± 0.27 per µM, p =0.8%) and reduces Col2a expression (∆∆Cq = -2.17 ± 0.89, p =2.2%), seeming to usefully suppress chondrogenesis.

Additionally, I evaluate the effects of pulsatile vessel growth conditions on an attractive new cell source: MSCs derived from induced pluripotent stem cells (iPSCs). Using PicoGreen and modified Bradford assays, I demonstrate that pulsatile growth conditions significantly increase dry weights of double-stranded DNA (dsDNA) from 0.0780 ± 0.002% to 0.1414 ± 0.008% (mean ± standard error, p = 0.015%) and collagen from 25 ± 2% to 46 ± 2% (p = 2.3%), approaching a sample of native aorta at 55% collagen.

Overall, results suggest normoxic growth conditions remain superior for SMC differentiation, but that the BMP-inhibitor LDN193189 may have a future role in suppressing cartilage production during TEVG growth. Furthermore, iPSC-derived MSCs demonstrate similar responses to traditional bone marrow MSCs, and may well represent an attractive cell source in future TEVG production.

We first defined the ASCVD 10-year risk for patient cohorts from four different primary care clinics in New Haven, CT by normalizing and aggregating individual patient 10- year ASCVD risk scores. We then calculated changes to this normalized aggregate risk by modeling the effects of evidence-based interventions found in Cochrane Reviews of different efficacy to each of four modifiable risk factors used in the 10-year ASCVD risk calculator. The four different modifiable risk factors include systolic blood pressure, total cholesterol, HDL cholesterol and smoking status. A resulting change in each cohort’s normalized aggregate risk was calculated.

The clinic cohorts had different levels of modeled risk reduction from the same interventions. The magnitude of reduction was dependent on baseline normalized aggregate risk and prevalence of risk factor(s) targeted in the interventions. The three clinics where the baseline normalized aggregate risk was above 100 events per 1,000, patients had a greater risk reduction from an organizational intervention aimed at improving the quality of treatment for hypertensive patients compared to all other evidence-based interventions found in Cochrane Reviews. In the clinic that had a lower baseline normalized aggregate risk, the highest yield intervention was dietary advice by providers. Our data demonstrate that the highest yield lifestyle intervention for any clinic may vary depending on the makeup of the populations and its risk factors.

The tool created in this study can be used by clinic providers and administrators to estimate the effects of various interventions on the ASCVD risk of their clinic cohort. The models generated by this tool can be used to guide strategy and prioritize clinic resources based on the extrapolated effects of evidence based interventions to specific clinic populations. Furthermore, it may also guide interventions planned to address needs identified by community health assessments. Because the tool predicts outcomes for specific patient populations it has the potential to foster the application of evidence-based practices to population health management.

This thesis describes four studies that examine various aspects of the care of non-small cell lung cancer (NSCLC) patients in the United States. All studies were performed utilizing data from the National Cancer Database (NCDB), a clinical tumor registry that captures approximately 70% of all incident cancer cases in the United States.

The first study characterizes the natural history of operable NSCLC, with the hypothesis that it would be superior to non-operable NSCLC. Cohorts of operable and non-operable patients were selected from the NCDB and their overall survival was analyzed using the Kaplan-Meier (KM) technique. The 5-year overall survival for untreated operable NSCLC was 10.1%, 7.3% and 4.9% for stage I, II, and IIIA disease respectively. Untreated operable NSCLC was associated with a superior survival compared to untreated non-operable NSCLC at all stages of disease (P<0.001). The natural history of operable NSCLC is still poor, but varies with stage and is superior to that of non-operable NSCLC.

The second study compared the efficacy of surgical lobectomy and stereotactic body radiotherapy (SBRT) for the treatment of stage I NSCLC in healthy patients, with the hypothesis that lobectomy would be associated with superior survival. Cohorts of “healthy” (i.e. no comorbidities, not considered medically inoperable) patients were selected from the NCDB and propensity-matched cohorts of SBRT and lobectomy patients were compared using the KM technique. Lobectomy was associated with a superior 5-year survival compared to SBRT (59% vs 29%, P < 0.001). These findings suggest that lobectomy leads to superior long-term survival over SBRT in patients with clinical stage I NSCLC whose health does not prohibit the use of surgery.

The third study examined risk factors for perioperative mortality and extended length of stay (eLOS, > 14 days) in NSCLC patients undergoing surgery. The NCDB was queried for NSCLC patients undergoing surgery, and those that experienced an adverse outcome were compared to those who did not using multivariable logistic regression modeling. Overall 30-day mortality rate was 3.4% and varied by surgical procedure. The frequency of adverse events after lung cancer surgery in the NCDB was found to be in line with that of other large databases.

The final study examines the relationship between facility discharge practices and readmission rates following lobectomy for NSCLC with the hypothesis that facilities that have a systematic practice of rapid discharge will have higher readmission rates. Facilities discharge practices were characterized by their median LOS relative to the median LOS for all patients in the same diagnosis year. Risk-standardized readmission rates (RSRRs) were calculated for each level of facility discharge practices (from rapid to slow) and it was found that facilities with a practice of rapid discharge had a significantly lower mean RSRR than those with slower discharge practices. In conclusion, it is possible for facilities to develop a practice of early discharge after lobectomy without increasing their rate of readmission.

The Evolution of Restraint in American Psychiatry

Author: Danilo Rojas-Velasquez

In psychiatry, restraint generally refers to direct methods such as mechanical restraints or the use of drugs. Despite psychiatrists’ best efforts to utilize restraint judiciously, many patients still view it as the field’s defining feature, especially on inpatient units with involuntary commitment, medications against will, and locked doors. This essay is an attempt to understand the pervasiveness of restraint in psychiatry. It uses changes in the practices of restraint to examine the growth of the field over time. To accomplish this, the paper identifies three distinct regimes of restraint: the moral treatment of the 1800s, associated with the asylum; the somatic treatment of the early 1900s, associated with psychosurgery; and the pharmaceutical treatment of the later 1900s, associated with pills. The essay analyzes primary sources drawn from the scientific and psychiatric literature of each period, in addition to marketing materials. It also examines the work of prominent figures associated with each regime, including Samuel Tuke, Clifford Beers, and Walter Freeman. The paper engages the work of a range of historians of psychiatry, including David Rothman, Michel Foucalt, Andrew Scull, and David Herzberg. Two major conclusions are drawn. First, restraint evolved from the physical form seen in madhouses to self-restraint first seen in the early asylum. The role of the psychiatrist followed this evolution, as the psychiatrist increasingly became the figure to help patients achieve self-restraint. Secondly, because psychiatrists became the judges of how much self-restraint is acceptable, they have come into conflict with society during periods of change, which contributes to the stigma and backlash against psychiatric practice.

functional assays. In particular, we demonstrated that our mutant cell clones secrete high levels of 2HG, and confirmed that the levels of this oncometabolite can be suppressed with small molecule inhibitors of mutant IDH1. We then performed a focused drug screen using select small molecule inhibitors of DNA repair, leveraging our observation that IDH1 mutant cells are more sensitive to radiation. We report that IDH1-R132h confers increased sensitivity to BMN-673, a PARP inhibitor known to preferentially kill cells with decreased homologous recombination (HR) functionality. We also demonstrate synergy between BMN-673 and the platinating agent, cisplatin, that is enriched by the IDH1-R132H mutation. Finally, preliminary gene expression analysis does not identify any significant decreased expression in a panel of DNA repair-related genes, suggesting that some alternative mechanism may be responsible for the drug sensitivity effect we see. Taken together, these findings suggest that IDH1 mutant tumors may be sensitive to PARP inhibition, representing a new treatment strategy for a devastating disease.

Growing evidence suggests that immunotherapy and radiation therapy can be synergistic in the treatment of cancer. We sought to determine the effect of the relative timing and type of immune checkpoint therapy on the response of melanoma brain metastases to treatment with stereotactic radiosurgery (SRS) by reviewing our institutional experience.

Patients with melanoma brain metastases were identified from an institutional database of patients treated with Gamma Knife SRS. SRS treatment plans, follow-up imaging, and details of immunotherapy treatment were reviewed. Immunotherapy and radiosurgery treatment to any single lesion was considered concurrent if SRS was administered within four weeks of immunotherapy. The primary outcome of interest was lesional response, as measured by changes in individual lesion volume during follow-up. The impact of timing and type of immunotherapy on lesional response was determined using the Wilcoxon rank sum test to compare median percent lesion volume change at 1.5 months, 3 months, and 6 months after SRS treatment, with significance determined by p=0.0167, per the Bonferroni correction for multiple comparisons. Secondary endpoints included lesional regrowth, distant brain recurrence, and overall survival.

75 melanoma patients with 566 brain metastases treated between 2007 and 2015 were included in this study. Concurrent use of immunotherapy and SRS resulted in significantly greater median percent reduction in lesion volume at 1.5 months (-63.1% vs -43.2%, p<0.0001), 3 months (-83.0% vs -52.8%, p<0.0001), and 6 months (-94.9% vs -66.2%, p<0.0001) compared to non-concurrent therapy. Median percent reduction in lesion volume was also significantly greater for anti-PD-1 than for anti-CTLA-4 at 1.5 months (-71.1% vs -48.2%, p<0.0001), 3 months (-89.3% vs -66.2%, p<0.0001), and 6 months (-95.1% vs -75.9%, p=0.0004). The median overall survival for the entire cohort was 18.5 months.

Our study shows that administration of immunotherapy within four weeks of SRS results in improved lesional response of melanoma brain metastases compared to treatment separated by greater than four weeks. Anti-PD-1 therapy also results in greater lesional response than anti-CTLA-4 after SRS.

Lauren E Provini, Michael J Corwin, Nicole L Geller, Timothy C Heeren, Rachel Y Moon, Denis V Rybin, Carrie K Shapiro-Mendoza, and Eve R Colson. Section of General Pediatrics, Department of Pediatrics, Yale University, School of Medicine, New Haven, CT.

Given that data assessing the heterogeneity of infant care practices among Hispanics are lacking, the objective of this study was to assess the association between maternal birth country and adherence to the American Academy of Pediatrics (AAP) safe sleep recommendations in a national sample of Hispanic mothers. We used a stratified, 2-stage, clustered design to obtain a nationally representative sample of mothers from 32 U.S. intrapartum hospitals. 907 completed follow-up surveys (administered 2-6 months postpartum) were received from mothers who self-identified as Hispanic/Latina, forming our sample, which we divided into 4 subpopulations by birth country (U.S., Mexico, Central/South America, and Caribbean). Prevalence estimates and aORs were determined for infant sleep position, location, breastfeeding, and maternal smoking. When compared with U.S.-born mothers, we found that: mothers born in the Caribbean (aOR 4.56, 95% CI 1.07-19.5) and Central/South America (aOR 2.68, 95% CI 1.38-5.22) were more likely to room share without bed sharing. Caribbean-born mothers were less likely to place infants to sleep supine (aOR 0.41, 95% CI 0.22-0.77). Mothers born in Mexico (aOR 1.67, 95% CI 1.03-2.72) and Central/South America (aOR 2.57, 95% CI 1.09-6.07) were more likely to exclusively breastfeed; Caribbean-born mothers (aOR 0.13, 95% CI 0.03-0.63) were less likely to do so. Foreign-born mothers were significantly less likely to smoke before and during pregnancy. In conclusion, among U.S. Hispanics, adherence to AAP safe sleep recommendations varies widely by maternal birth country. These data illustrate the importance of examining behavioral heterogeneity among ethnic groups and have potential relevance for developing targeted interventions for safe infant sleep.

dramatically, demanding the close integration of business and management with the

delivery of clinical care. In response, there has been a continuation of the trend towards

additional training for physicians through an MBA program that has been seen over the

last thirty years. However, some medical students have encountered some negative

perceptions voiced by senior physicians about MD/MBA training. As most MD/MBA

joint-degree candidates consider clinical careers, it is vital to understand the views of

residency program directors who hold the gates to graduate medical education.

Purpose: Therefore in this paper, we will investigate the following hypotheses:

Completing an MBA as a medical student will be perceived positively by residency

directors, and the global opinion of MD/MBA candidates has changed over the last

decade.

Methods: An electronic survey was sent to residency directors in most major

specialties across the United States to ascertain their opinions of MD/MBA residency

candidates. A Likert score was tabulated corresponding to the level of MBA-favorability

of each program. Statistical correlations were performed based on medical specialty,

demographics, geographical region, the experience of the program director with an MBA

curriculum, faculty with an MBA, or residents with an MBA. Data were compared with

a similar survey by Lyssy et al performed in 2006.

Results: 578 residency program directors responded to our survey, a response rate

of 22.2%. No statistically significant difference was found in the calculated Likert score

of MBA candidate favorability across the medical specialties. A statistically significant

difference in the proportion of program directors with interactions with faculty and

residents with an MBA was found among the medical specialties; however, no

statistically significant difference in the proportion of program directors who personally

hold an MBA was found. Program directors who had direct experience working with

residents with an MBA reported higher Likert positivity scores compared to those who

did not. Additionally, departments with a higher number of faculty with an MBA were

positively correlated with a greater number of residents with an MBA in that program.

Residency program director age was negatively correlated with the Likert MBA

candidate favorability score. Compared to the 2006 dataset, there were minimal changes

in the Likert-type question scores in 2016.

Conclusions: Residency directors across multiple specialties positivity regard

MD/MBA candidates and the candidates’ training for their residency programs.

Moreover, this regard has remained generally stable over the past decade.

At the main oncology unit in T&T, female breast cancer patients who met the NCCN criteria were recruited for this study. We conducted interviews inquiring about their personal breast cancer diagnosis, as well as any relevant family history. This was followed by the collection of saliva samples using Oragene kits, which were then analyzed by Color Genomics Inc. for 30 genes associated with hereditary cancers. Finalized results were returned to patients by genetic counselors from Color Genomics. In total, 58 patients who met NCCN guidelines were sequenced and results were returned. They showed that of 58 samples, 15 patients tested positive for deleterious HBOC germline mutations: 9 - BRCA1, 3 - BRCA2, 1 – CHEK2, 1- PALB2 and 1 – PTEN, with an overall prevalence rate of 25.8%. This prevalence rate is remarkable, given that HBOC mutations among U.S. women with breast cancer are found in only 5-10% of patients. These initial results clearly demonstrate the need to include genetic counseling and testing in the local oncology management in T&T, as the identification of HBOC mutations can influence treatment options, as well as help identify family members who are at high risk for cancer predisposition. Ultimately, this implementation could help alleviate the country’s high incidence and mortality rates with respect to breast cancer.

In 2012, Khayelitsha Hospital (KH) was opened with the Clinicom EMR installed. In 2013, KH received the ministerial award for clinical excellence as a result of the effective EMR rollout. Due to a high incidence of violence around Khayelitsha, and the need for effective trauma surveillance for efficient resource allocation, the authors sought to calculate the Emergency Center (EC) trauma rate for KH. Retrospective review of both paper charts and the Clinicom EMR from July 2012 to May 2013 was performed.

The Clinicom EMR yielded a mean monthly EC trauma visit of 280 patients, mean EC census of 3537 patients and a mean trauma rate of 8%. The latter was much lower than the nationwide estimate of 33%. For the month of July, 66 additional cases were found on the paper registry but not on the EMR. Furthermore for the months of January to May 2013, scanned copies of patient records were unavailable on the EMR online database.

The discrepancy between the paper records and the EMR suggest potential difficulties with the implementation of the Clinicom EMR that were overlooked by the award committee, and call into question the 8% trauma rate calculated. The results highlight the need for further evaluation of the functioning of the Clinicom EMR system to identify difficulties in implementation and correct them before the system is utilized in hospitals across South Africa.

It has been previously shown that application of a NgR1 decoy receptor (AA-NgR(310)ecto-Fc) increases sprouting below the site of the lesion in rats with spinal cord contusion injuries. Likewise, application of this same decoy receptor effectively disinhibited functional recovery as exemplified by the increase in percentage of weight-bearing rats treated with the decoy receptor. Noting the more ideal synthetic properties of a small molecule pharmaceutical, here we attempt to use a small molecule inhibitor of NgR1 to induce the in vitro regeneration of axons following scrape injury as well as in an in vivo model of mice with SCI. Additionally, noting the importance of LPA1 as shown through previous studies, we also attempt to utilize a small molecule inhibitor of LPA1 to promote axonal regeneration.

Our results show that inhibition of NgR1 with the small molecule inhibitor YU-NR-008 did not significantly improve axonal regeneration in vitro. Application of the NgR1 inhibitor YU-NR-008 alone showed a trend toward improved axonal regeneration, albeit insignificant (mean signal intensity for YU-NR-008 treated animals at 1.243 ± 0.128 vs. control 1.00 ± 0.00, p = 0.0787). Co-treatment with YU-NR-008 and Nogo-22 did not rescue Nogo-22-mediated inhibition of axonal regeneration (Nogo-22 0.771 ± 0.051 vs. Nogo-22 with YU-NR-008 at 0.801 ± 0.073). Additionally, functional recovery as measured by the Basso Mouse Scale (BMS) was not improved with the administration of YU-NR-008 following SCI for 2 or 4 weeks (D32 BMS scores were 4.643 ± 0.713 (SEM) for control vs. 3.550 ± 0.669 for animals treated with YU-NR-008 for 4 weeks). Likewise, administration of the LPA1 antagonist AM095 did not improve functional recovery following SCI (mean BMS at 54 days for AM095-treated animals was 3.182 ± 0.532 vs. 5.033 ± 0.448 for vehicle-treated animals). We conclude that the tested doses of YU-NR-008 and AM095 were ineffective in promoting recovery in a rodent model spinal cord injury. Additional studies will be needed to determine whether axonal growth was stimulated by these doses, or if drug doses failed to achieve the cellular target effect following spinal cord injury.

Celiac disease and inflammatory bowel disease (IBD) are immune mediated chronic gastrointestinal disorders that present with similar symptoms such as abdominal pain, diarrhea and growth failure. In adults, some studies have shown increased prevalence of celiac disease in patients with IBD while other studies have not. There is no published data looking at the prevalence of celiac disease in children with IBD. We aim to investigate the prevalence of celiac disease and anti-tissue transglutaminase (tTG) antibodies in children with IBD.

Methods

We performed a retrospective chart review of children with IBD followed routinely in our clinic over the last 4 years. These children were matched with a non-IBD control group of children presenting with gastrointestinal symptoms to GI clinic. All children were routinely screened for celiac disease with tTG-IgA and total IgA. Clinical, laboratory, and histological data were collected in all children. Abnormal tTG was defined as >30u/ml.

Results

The study population included 130 children with IBD and 258 in the control group. The mean age of children was 14.6 ± 3 years and 58% were males. The IBD group included 75 children with Crohn’s disease and 55 with ulcerative colitis. Abnormal tTG levels were found in 6 patients in the IBD group and 20 patients in the control group (4.6% vs. 7.8%, p = 0.24). To further evaluate the presence of celiac disease, tTG positive children were tested with endomysial autoantibodies (EMA) and duodenal histology, leading to celiac diagnosis in one patient with IBD and 12 patients in the control group. Celiac disease prevalence was lower among children with IBD compared with the control group (0.8% vs. 4.7%, p = 0.07). False positive rates were 3.9% and 3.3% for the IBD and the control groups, respectively.

Discussion

Children with IBD do not have higher prevalence of celiac disease compared to other children who present with gastrointestinal symptoms to a GI clinic. Prevalence of celiac disease in children with IBD (0.8%) is similar to that of reported prevalence (1%) in general population. Rates of false positive tTG antibodies are similar in children with IBD and the control group. Routine screening for celiac disease in children with IBD is not warranted.

population of patients at Yale-New Haven Hospital (YNHH) across a range of gestational

ages, and to evaluate the association between small EPV and low birthweight (BW). From

2009 to 2011, 366 patients at YNHH received ultrasound scans between 11+0 to 38+6

weeks gestational age (GA) to measure placental dimensions from 2009 to 2011. EPVs were

calculated using a previously validated convex-concave shell equation. An EPV vs GA best

fit curve was generated. The relationship between EPV and BW was analyzed. Subgroup

analyses were performed to evaluate differences between study participants who delivered at

YNHH, and those who did not. Analysis of EPV versus gestational age revealed a parabolic

curve with the following best fit equation: EPV = (0.372 GA – 0.00364GA2)3. One hundred

seventy four of the 366 women who underwent EPV measurement delivered at Yale-New

Haven Hospital (YNHH) and had their infants’ BWs recorded. The remaining patients

delivered at outlying hospitals, where BWs were not available to the investigators. However,

parabolic EPV GA curves generated from these two patient populations were superimposable. YNHH patients with an EPV in the bottom 50th percentile had 2.42 times the odds of having a newborn with a BW in the bottom 50th percentile (95% CI 1.27 – 4.68). Microscopic evaluation of two placentas corresponding to the smallest EPV outliers revealed significant placental pathology. We conclude that placental volume increases throughout gestation and follows a predictable parabolic curve. Very low EPV measurements are associated with low BWs. Therefore, EPV may be useful as a screen to identify women who are carrying fetuses who are at risk for intrauterine growth restriction.

Methods: We retrospectively reviewed all chart records for patients diagnosed with septic arthritis at the West Haven VA Medical Center between January 2006 and December 2015. Treatment outcomes were recorded at one year post-diagnosis.

Results: Sixty-two patients were diagnosed with native joint septic arthritis during the study period, 19 who were managed medically and 43 surgically. There was no significant difference in demographic variables, risk factors, clinical presentation, laboratory parameters, or duration of antibiotic therapy between the two groups. 81% of medically-managed patients and 82% of surgically-managed patients achieved a full recovery within 12 months. Overall mortality was 3.2%, and both patients who died were managed surgically. There were no significant predictors of poor treatment outcome aside from Black race.

Conclusion: Our findings suggest that medical management of septic arthritis with closed-needle aspiration is non-inferior to surgical management in terms of both morbidity and mortality. Given the small sample size, a prospective randomized control trial is needed to guide definitive recommendations on the best form of joint drainage

This prospective cohort pilot and feasibility study was conducted in the adult and pediatric EDs of an urban, academic, Level 1 trauma center. Twelve patients with concussions and twelve age- and gender-matched control patients presenting within 6 hours of an injury sustained during recreational activity were enrolled. ED blood specimens were banked for future proteomic analysis. Clinical outcomes were collected via online survey. Patient recruitment strategies were refined in three phases: (1) identification and notification by clinical staff, (2) email notification automatically-generated by the electronic health record (EHR), and (3) patient financial incentives provided at enrollment and upon completion of the symptom diary.

After Phase 1, the patient identification rate improved from 0% to 22% (p<0.001) with EHR-generated notifications. In Phase 3, the enrollment rate improved from 38% to 100% (p=0.01) with financial incentives.

This pilot and feasibility project is the first step to toward the goal of identifying new diagnostic and prognostic biomarkers as well as novel targets for therapy. EHR-based paging and financial incentives for participation increased subject identification and enrollment rate to inform and optimize the enrollment and recruitment strategies for the eventual larger study.

Ichthyoses are a highly heterogenous group of disorders, both genetically and clinically. Despite the identification of causative mutations in over 50 genes, clear genotype-phenotype correlations have been difficult to establish due to the rarity of the ichthyoses and because mutations in the same gene can have widely divergent phenotypes, even in individuals with the same disease-causing mutation.

In partnership with the Foundation for Ichthyosis and Related Skin Types (FIRST), we have served as a major national and international referral center and recruited over 800 kindreds with DOK, utilizing whole exome sequencing to identify novel, rare mutations that cause DOK, many of which represent phenotypic expansion.

We systematically evaluated genotypic and phenotypic data for 407 kindreds with ichthyosis, the largest cohort published to date. We identified 156 novel mutations and assessed phenotypic features of DOK with respect to genetic diagnosis to help refine our understanding of this heterogeneous group of disorders (Chapter 1). Such a systematic classification of DOK holds promise for the development of customized management plans, generation of targeted therapeutics, and improved understanding of prognostication based on genetic diagnosis.

In the process of performing a large scale genotype-phenotype characterization, we identified kindreds with rare clinical subtypes of ichthyosis, including Bathing Suit Ichthyosis (BSI), a rare disorder caused by mutations in the transglutaminase 1 gene (TGM1) and characterized by the restriction of scale to sites of relatively higher temperature such as the trunk, with cooler areas remaining unaffected (Chapter 2). We identified novel and recurrent mutations in sixteen subjects with BSI, the largest cohort published to date. Our findings expand the genotypic spectrum of BSI and the understanding of temperature-sensitivity of TGM1 mutations. Increased awareness of temperature-sensitive TGM1 genotypes should aid in genetic counseling, and provide insights into the pathophysiology of TGM1 ichthyoses, and potential therapeutic approaches.

In the course of unifying genetic and clinical data to advance the understanding of the different subtypes of ichthyosis, we realized the critical importance of tools to assess the clinical severity of ichthyosis. We designed a Visual Index for Ichthyosis Severity (VIIS) and tested the instrument for reliability and reproducibility using two different settings: one that utilized scoring of 60 test photographs by 10 dermatologists, and one with in-person evaluations on 85 subjects by 12 dermatologists at the Foundation for Ichthyosis and Related Skin Types (FIRST) conference (Chapter 3). The validation process revealed high reliability and reproducibility for both scale and erythema and indicated that the VIIS performs better in person than with photographs, an important consideration in design of clinical trials. This index provides a tool for clinical phenotyping and assessment of therapeutic response for many disorders of keratinization.

Pierre Martin M.Ed.(1), Lisa DelSignore M.D.(2), and Traci A. Wolbrink M.D. M.P.H.(2)

(1)Yale School of Medicine (2)From the Division of Critical Care Medicine, Department of Anesthesia, Perioperative and Pain Management, Boston Children’s Hospital and the Department of Anesthesia, Harvard Medical School, Boston, MA, USA. (Sponsored by JH, Department of Pediatrics, Yale School of Medicine)

Virtual patient simulation has been utilized to teach interviewing skills, often employing selection-based methods (e.g., multiple-choice lists and menu-based prompts) to simulate doctor-patient conversations. Users have evaluated these systems as inauthentic, which can diminish user immersion (influenced by control, realism, distraction and sensory factors) and, in turn, negatively affect skill acquisition, mastery and transfer.

Our objectives were to design and develop PBLCloud, a scenario-based and highly interactive platform that uses natural language processing to support a more realistic doctor-patient conversation and create an immersive clinical learning environment.

PBLCloud was developed utilizing an iterative design thinking process and its initial evaluation involved a mixed methods approach. We recruited a convenience sample of 11 participants: three (27%) fourth-year medical students from Harvard Medical School as well as two (18%) residents, four (36%) fellows and two (18%) attendings from Boston Children’s Hospital. There were two rounds of formative evaluation testing with eight participants in Round 1 and three participants in Round 2. Each participant completed a semi-structured think–aloud protocol exploring our pilot case, 10-item system usability scale (SUS) and 10-item open-ended questionnaire.

The chat-based functionality provides users with computer-generated context-specific responses during the historical encounter. Users have the opportunity to perform physical examinations, review incorporated multimedia, order and interpret diagnostic investigations, order therapeutic interventions that have appropriate effects on patient vitals and laboratory data, formulate and refine a differential diagnosis, receive just-in-time feedback regarding user-initiated actions and complete embedding learning exercises. 73% of participants strongly agreed that PBLCloud was useful (i.e., it is clinically-oriented, realistic, provides helpful feedback and is widely applicable) and 64% of participants strongly agreed that their experience with the system was enjoyable (i.e., it is relevant with an engaging interface). It was deemed to be more interactive and engaging than other simulators and 82% of participants were very interested in utilizing the system in the future. The average SUS score for Round 1 and 2 were 79.7 ± 12.0 and 82.5 ± 19.8 respectively. Areas of improvement were identified, in particular, the unsatisfactory response accuracy of the chat-based functionality.

Future work will include the investigation of various strategies to optimize the platform’s natural language processing algorithm as well as the formal evaluation of the system’s validity, reliability, level of induced user immersion and educational impact. We anticipate that PBLCloud will serve as a cost-effective and scalable approach for the instruction and assessment of clinical reasoning.

Brian S. Marcus, Mark R. Mercurio, Denise Esserman, Gary Kopf

Section of Cardiothoracic Surgery, Department of Surgery, Yale University, School of Medicine, New Haven, CT Objective: To evaluate the cognitive and affective factors which contribute to ethical evaluations by members of ethics committee (EC). Method: A total of 70 EC members at 4 large medical centers were given clinical vignettes and asked to make judgments. Each vignette had two versions and each EC member was randomly assigned one of the two versions to evaluate. The two versions differed only in affective content (likeability of agents) which had no bearing on the ethical issues. The EC members were then asked, using a 7-point Likert scale, to make an ethical judgment, and give an analysis as to what ethical principles they used in making their decision. Results: In two pairs of clinical ethical dilemmas, positive affect influenced the ethical evaluation in favor of the agent portrayed positively, and negative affect resulted in the opposite effect. In the clinical vignettes where there was no significant difference in any affective response to the agents of the evaluator, no difference in ethical decisions were seen between different versions of the vignette. Emotional content also had an effect on which ethical principles were used by EC members to explain their evaluations. Conclusion: Ethical evaluations by EC members are effected by non-ethically relevant emotive content of clinical scenarios. Emotional biases may play a significant role in the evaluation of ethical dilemmas, even though EC members may not be explicitly aware of them. Bioethical principles areused to rationalize these judgments after decisions have been made.

Methods: As part of a case-control study within the Kaiser Permanente Northern California system, pathologic and clinical records were obtained for RCC cases (2,152 white, 293 black) diagnosed from 1998 to 2008. Patient demographics, comorbidities, tumor characteristics, and treatment status were compared between whites and blacks. Overall survival and disease specific survival (DSS) were calculated by the Kaplan-Meier method. A Cox proportion hazards model was used to estimate the independent associations of race, comorbidity, and clinico-pathologic variables with DSS.

Results: Compared to whites, blacks were diagnosed at a younger age (median 62 vs. 66 years, p<0.001), were more likely to have papillary RCC (15% vs. 5.2%, p<0.001), and had similar rates of surgical treatment (78.8% vs. 77.9%, p=0.764). On multivariate analysis, advanced AJCC stage, lack of surgical treatment, larger tumor size, and higher grade were predictors of worse DSS. Race was not an independent predictor of survival.

Conclusions: Within a single healthcare system, we observed differences in characteristics of black and white patients with RCC; black patients had different comorbidities, were younger, and had decreased tumor stage. However, unlike other series, race was not an independent predictor of DSS, suggesting that survival differences in large registries may result from barriers to healthcare access and/or comorbidity rather than disease biology.

New models for data sharing can help clinicians and patients decide which interventions will best drive improved outcomes if they allow the exploration of the totality of available evidence. The controversy surrounding the safety of recombinant human bone morphogenetic protein-2 (rhBMP-2) provides a case example of putting the Yale Open Data Access (YODA) Project into practice. A review of the existing literature showed complications including heterotopic ossification, postoperative radiculitis, and endplate osteolysis associated with the use of rhBMP-2 in posterior spinal fusion techniques.

Given that two expert groups were provided with the same data and general aims, it was hypothesized that approaches and results of two independent systematic reviews and meta-analyses would be highly similar. An analysis of these reviews showed that they used different methodological approaches, with one center choosing to stratify by surgical approach and the other group combining data from all approaches. Results were broadly similar though there were some important differences. One center found a statistically significant, but small, benefit whereas the other center reported no advantage in comparing rhBMP-2 and control groups. In the analysis of harms, neither showed an increased cancer risk at 48 months, although one center reported a significant increase at 24 months. Conclusions reflected these differences in summary estimates of benefit balanced with small but potentially important risk of harm. Replication of systematic reviews are valuable as even expert groups looking at the same data may come to different conclusions.

Though efforts like data sharing can help clinicians better understand healthcare outcomes, measuring costs is a critical step in understanding and managing resource usage to improve value in healthcare. Time Driven Activity based Costing (TDABC) is a financial accounting methodology which can more accurately measure costs in orthopaedic surgery than commonly used hospital accounting methodologies. This methodology was put into practice in an academic medical center for primary total knee and hip arthroplasty patients of three surgeons. This accounting method requires the creation of process maps which detail the various steps of delivering total joint arthroplasty before combining them with financial data to derive an episode cost. This process itself can identify areas for process improvement. This process identified actionable areas of process improvement in weekend discharge, blood transfusion rates, anesthesia impacts on length of stay, patient waiting times, operating room turnover delays due to peripheral nerve block placement, and an opportunity for patient clinical and administrative optimization before surgery.

The implementation of TDABC methodology not only can lead to more accurate cost accounting in total joint replacement but uncover opportunities for operational improvement.

We analyzed exome sequences of unpaired tumor samples, collected prior to ACT chemotherapy, in 17 TNBC patients who exhibit complete pathologic response to neoadjuvant chemotherapy (pCR) and 15 patients who had extensive residual disease (RD).

In the process, we created one of the first analytical pipelines capable of performing comprehensive integrated analysis of somatic point mutations and structural variation in unpaired tumor exome samples. Validation on tumor-normal matched samples demonstrated >95% specificity for point mutations, LOH, and CNV calling compared to standard tumor-normal somatic analysis.

When applied to the TNBC cohort, our somatic mutation caller identified multiple damaging somatic mutations in genes linked to EMT. LOH analysis showed significantly greater LOH in pCR (Complete Response) than RD patients (Residual Disease) (p=6.5E-12). The five regions with greatest LOH difference between pCR and RD subgroups each contained a HRR gene locus. Overall high LOH burden was associated with the presence of TP53 point mutations (p=0.002).

By integrating data from all three methods, we found significantly more pCR patients with high mutation burden (including CNV and LOH) in homologous recombination repair genes than RD patients (83% vs 20%). With this metric, we can predict 83% of positive response and 80% of negative response based on our patients’ genomic profiles prior to chemotherapy initiation (OR=18.7, 95% CI= 3.2 to 110.3, p=0.0012).

This result offers a potential significant improvement in our ability to personalize therapy in TNBC and may facilitate development of targeted PARP inhibitor therapeutics.

Persistent activity across a delay period is considered the neuronal basis of working memory, which is generated by recurrent activity among a group of layer III PFC pyramidal cells, that excite each other via NMDA receptor synapses on dendritic spines. NMDAR requires a depolarized membrane to open; in some areas of the brain (e.g. visual cortex) this is done by AMPAR, but acetylcholine performs this function in layer III PFC, through actions at nicotinic a7 receptors [4], and possibly, through muscarinic M1 receptor (M1R) closing of M-type KCNQ potassium channels in the synaptic membrane.

Immunoelectron microscopy data has demonstrated the presence of both M1R and KCNQ in glutamate-like synapses in layer III of primate dlPFC. This study aimed to test the hypothesis that stimulation of the M1R would enhance working memory though closure of hyperpolarizing KCNQ channels. To examine the role of the M1-KCNQ signaling pathway in working memory, we tested the effect of agents that modulated M1R and the KCNQ channel on neuronal activity in dlPFC.

Single unit neuronal recordings were performed in non-human primates performing a working memory task. Drugs manipulating M1R and KCNQ were iontophoresed onto the recording site and the effect on firing was captured by a carbon fiber electrode.

We found that both M1R activation and KCNQ blockade enhanced working memory related activity in the monkey dlPFC (p<0.01) while M1R blockade (p<0.0001) and KCNQ channel opening (p<0.01) reduced neuronal activity in the dlPFC. Finally, the effect of the M1R was found to be influenced by the open state of the KCNQ channel. Reduced firing from M1R blockade in the dlPFC was reversed by closure of KCNQ channel (p<0.01).

The current study shows that both M1R and KCNQ channels regulate working memory circuitry at the neuronal level. The effects of M1R on neural activity in the working memory circuitry, is mediated by the open state of KCNQ channel. These findings suggest that M1R activation is a viable strategy to enhance working memory circuits and maybe be a potential target for agents developed to improve cognition. This strategy would have special relevance to cognitive dysfunction in schizophrenia, as patients have been found to have working memory deficits and downregulation of the M1R in working memory circuitry of the dlPFC [5].

Peripheral blood-derived monocytes from healthy donors (HD) (n=10) as well as CF subjects (n=10) were differentiated into macrophages using RPMI media with 50 ng/mL of MCSF for 10 days. Cells were treated with 100 ng/mL of LPS plus 20 ng/mL of INF-gamma to activate the M1 phenotype or 20 ng/mL of IL-4 to activate the M2 phenotype. Cells were treated for 24 hours in preparation for RNA isolation and 3 hours in preparation for protein isolation. RNA was isolated and reverse transcribed to cDNA in preparation for qualitative, reverse transcription polymerase chain reaction (qRT-PCR). qRT-PCR was performed using IL-6, IL-1B, and TNF-alpha primers as M1 markers and TGF-B1, MRC, PPAR-gamma as M2 markers. Expression levels were performed in seven experiments, each using a different set of HD and CF subjects, utilizing 8.0 x 10^5 to 1.0 x 10^6 cells per well, while three protein experiment were done using different sets of HD and CF subjects.

Stimulation of CF macrophages induces greater phosphorylation of proteins that regulate macrophage activation, as compared to HD macrophages. When CF macrophages are challenged with LPS/INF-gamma, they show greater phosphorylation of STAT1 protein as compared to HD macrophages. Similarly, when CF macrophages are challenged with IL-4, they exhibit greater phosphorylation of STAT6 protein, as compared to HD macrophages. Finally, CF macrophages show greater phosphorylation of AKT protein as compared to HD macrophages when stimulated with both M1 and M2 cytokines. As expected, HD macrophages (n=7, experiments) demonstrated appropriate plasticity and polarization with increased expression of M1 markers and decreased expression of M2 markers in response to LPS/INF-gamma. Furthermore, they demonstrated modestly increased expression of M2 markers and decreased expression of M1 markers in response to IL-4. In marked contrast, when stimulated with LPS/INF-gamma CF macrophages (n=7, experiments) demonstrated hyper-inflammation with dramatically increased expression levels of M1 markers as well as aberrant polarization as evidenced by increased expression levels of some M2 makers. Aberrant polarization was further characterized by increased expression of M1 markers in the presence of IL-4 in addition to increased expression of M2 markers.

These data suggest that lack of functioning CFTR in macrophages leads to the inability of macrophages to adequately respond to environmental cues and activate into appropriate phenotypes suggesting there is an intrinsic cellular defect. Moreover, the mechanism(s) that underlie these aberrant responses likely involve altered intracellular signal transduction which is currently under investigation.

Objective: To compare the efficacy of tranexamic acid and aminocaproic acid in decreasing blood loss and blood transfusion requirements during posterior spinal fusion for the treatment of adolescent idiopathic scoliosis.

Background: Due to the extent of the operation, posterior spinal fusion is associated with significant blood loss often requiring blood transfusions that increase the risk of morbidity and mortality. Antifibrinolytic medications, mainly tranexamic acid (TXA) and aminocaproic acid (Amicar), have been shown to reduce blood loss and blood transfusion requirements in studies on surgery for scoliosis. Our study compares the efficacy of using TXA and Amicar to using no anti-fibrinolytic in reducing blood loss and blood transfusion requirements.

Methods: A retrospective chart review was performed on all patients with idiopathic scoliosis undergoing exclusive posterior spinal fusion from 2008 to 2016 at one institution. Patients were put into three groups, a historical control group that was not given anti-fibrinolytics (67), a group given TXA (46), and a group given Amicar (21). There were no significant differences in age, gender, number of fused vertebrae, or Major Cobb angle between the three groups.

Results: The TXA group required significantly fewer average units of packed red blood cell (PRBC) transfusion (1.76 ± 1.25) than the control group (2.57 ± 1.41). The Amicar group (2.24 ± 1.04) did not demonstrate a significantly reduced blood transfusion requirement. There were no significant differences seen in intraoperative estimated blood loss across the three groups. Multiple regression analysis showed that TXA was significantly associated with a reduction in PRBC transfusion and that the number of vertebrae levels fused was significantly associated with an increase in blood transfusion. In contrast, Amicar did not demonstrate a statistically significant reduction in blood transfusion requirements.

Conclusion: We analyzed 134 patients who underwent posterior spinal fusion for adolescent idiopathic scoliosis to compare the effects of tranexamic acid (TXA) and aminocaproic acid (Amicar) to a control group given no anti-fibrinolytic therapy. TXA was found to significantly decrease packed red blood cell transfusion requirements while Amicar did not show a statistically significant change. Neither were associated with a decrease in intraoperative estimated blood loss.

Thomas Lazzarini, Crhistinne Gonçalves, Julio Croda, Albert Ko, Walter Benites, Liliane da Silva, Kristen McLean, Jason Andrews, Daniel Henrique Tsuha, and Robert Rohrbaugh. Department of Psychiatry, Yale University, School of Medicine, New Haven, CT.

Abstract: The purpose of this study was to conduct a retrospective cohort study as well as a qualitative study to better understand the current context of suicide among the indigenous population living on the reservations surrounding Dourados, Brazil. Critical questions included: What are the most important suicide risk factors? Which communities have the highest rates? What differences exist between higher-risk and lower-risk communities? The Brazilian National Mortality Database (SIM), the Special Indigenous Information System (SIASI), as well as the national census (IBGE) were utilized to estimate an overall suicide rate of 73·4 per 100,000 population per year in the reservation communities. The peak risk for males is during adolescence (15-19) and the peak risk for females is during late childhood (10-14). There was strong evidence for the clustering of suicides by time and geography as well as within extended families among this population. A comparison between two neighboring communities with a 5-fold difference in suicide rate demonstrated that the community with a higher suicide rate suffers from greater poverty and structural barriers to health care. Interventions must focus on children and youth living within the household of suicide victims as well as structural interventions to improve economic opportunities and access to healthcare in highly affected communities.

To accomplish this objective, patients with oral cavity cancers in the National Cancer Database (NCDB) were stratified into clinically lymph node-negative (cN0) and clinically lymph node-positive (cN1) cohorts to reflect the differing surgical management for these diseases. Univariate and multivariate analyses were performed to assess the relation between lymph node yield and overall survival, adjusting for other prognostic factors. Thresholds derived from the NCDB were validated in the Surveillance, Epidemiology, and End Results (SEER) database.

We found that increasing lymph node yield was linearly associated with improved survival up to 35 lymph nodes when controlling for other prognostic factors. In patients with cN0 cancers of the oral cavity from the NCDB, those who had <16 lymph nodes had significantly decreased survival. The proportion of positive lymph nodes was higher for patients who had 16 lymph nodes (27.2% vs 16.3% for<16 lymph nodes; p<0.001). This threshold was validated in 2,715 lymph node-negative cancers from SEER, with a mortality hazard ratio of 0.818 for ≥16 lymph nodes (95% confidence interval, 0.695-0.963; p=0.016). In patients with cN1 oral cavity cancers from the NCDB, groups with <26 lymph nodes had significantly decreased survival. This threshold was validated in 1,903 lymph node-positive cancers from SEER, with a mortality hazard ratio of 0.790 (95% confidence interval, 0.692-0.902; p<0.001). Academic centers, higher volume centers, and geographic location predicted higher lymph node yields.

In summary, more extensive neck dissection (≥16 lymph nodes in cN0, ≥26 lymph nodes in cN1) was associated with better survival. These findings are applicable as quality metrics are established in performance measurements in healthcare. Further evaluation of practice patterns in lymph node yield may represent an opportunity for improved quality of care.

Method: PubMed was searched for randomized controlled trials assessing the efficacy of benzodiazepines as short-term anxiolytics in pediatric patients. Twenty-one trials involving a total of 1,416 participants were included. A fixed effects model was used to examine the standardized mean difference of improvement in anxiety levels compared to control conditions. In stratified subgroup and meta-regression, the effect of the specific agent, dose, timing, and setting of benzodiazepine treatment was examined.

Results: A significant benefit was seen for benzodiazepines compared to control (standardized mean difference = 0.71 [95% confidence interval, 0.60-0.82], k = 24, z = 12.7, p<0.001). There was also funnel plot asymmetry in this meta-analysis, suggesting some evidence of publication bias. Moderator analyses found that when benzodiazepines were used in dental or non-operating room procedures, they were more effective than when they were used in operating room procedures (test for subgroup differences Q2 = 6.34, p=0.04). Tolerability analysis revealed there was no significant difference in the risk of developing irritability or behavioral changes between benzodiazepine and control groups.

Conclusions: Benzodiazepines are effective and well-tolerated when used as short-term anxiolytics in procedural settings for pediatric patients. Further research is needed to determine whether benzodiazepines are effective in pediatric anxiety disorders.

Retention in care is an essential component of meeting the UNAIDS “90-90-90” HIV treatment targets. In Khayelitsha township (population ~500,000) in Cape Town, South Africa, more than 50,000 patients have received ART since the inception of this public sector program in 2001. Disengagement from care remains an important challenge. We sought to determine the incidence of and risk factors associated with disengagement from care in 2013-4, and outcomes for those who disengaged.

Methods

We conducted a retrospective cohort study of all patients >= ten years of age who visited one of the 13 Khayelitsha ART clinics in 2013-14 regardless of the date they initiated ART. We described the cumulative incidence of first disengagement (>180 days) between 1 Jan 2013- 31 Dec 2014 by time on ART and time in the study using a competing risks model, which enabled us to extrapolate disengagement incidence up to ten years after ART initiation. We also described risk factors for disengagement based on a Cox proportional hazards model. We ascertained outcomes (death, return to care, hospital admission, other hospital contact, alive but not in care, no information) after disengagement until 30 June 2015 using province-wide health databases and the National Death Registry.

Results

Of 39,884 patients meeting our eligibility criteria, the median time on ART to 31 December 2014 was 33.6 months (IQR 12.4-63.2). Of the total study cohort, 592 (1.5%) died in the study period, 1,231 (3.1%) formally transferred out, 987 (2.5%) were “silent transfers” and visited another provincial clinic within 180 days, 9,005 (22.6%) disengaged, and 28,069 (70.4%) remained “in care.” Cumulative incidence of disengagement from care was 25.1% by five years on ART and 37.7% by ten years on ART estimated from time contributed in the study window in the competing risks model. Key factors associated with disengagement were age <30 years (10-20 years HR 1.38, 95% CI 1.24-1.54; 20-30 years HR 1.46, 95% CI 1.38-1.54; both relative to reference 30-40 years), male sex, pregnancy at ART start, and last CD4 count <350 cells/μl (CD4 <50 HR 3.34, 95% CI 2.92-3.83; CD4 50-200 HR 3.07, 95% CI 2.84-3.31; CD4 200-350 HR 2.03, 95% CI 1.91-2.15, all relative to reference CD4 >350 cells/μl); protective factors were ART club membership and baseline CD4 <350 cells/μl (CD4<50 HR 0.39, 95% CI 0.35-0.44; CD4 50-200 HR 0.46, 95% CI 0.43-0.50; CD4 200-350 HR 0.6, 95% CI 0.56-0.65, all relative to reference CD4 > 350 cells/μl) (Table 1). Of those who disengaged, the two most common outcomes by 30 June 2015 were return to ART care after 180 days (33%), and being alive but not in care in the Western Cape (25%). After disengagement, a total of 1,459 (16.2%) patients were hospitalized and 237 (2.6%) died.

Conclusions

One quarter of patients in Khayelitsha, one of the longest running ART programs in South Africa, disengaged from ART care at least once in a contemporary two-year period. One key aspect of the HIV care cascade is retention in care. While the majority of patients either subsequently returned to care or remained alive without hospitalization, a challenge to meeting retention targets, and the broader 90-90-90 HIV treatment targets, is developing, testing, and implementing program designs to target mobile populations and retain them in lifelong care. This should be guided by risk factors for disengagement as observed in this and other studies. Additionally, some focus should be placed on preventing misclassification of patients with regards to retention status by utilizing better patient tracking systems.

Materials and Methods: The websites of all NCI-designated Cancer Centers and all affiliated institutions were examined for publicly available data regarding head and neck cancer patients. For Yale data, IRB approval was obtained to use tumor registry data, as well as chart review, to create a comprehensive database for all new head and neck cancer patients at Yale in 2013 and 2014. The patients with oropharyngeal cancers were then isolated and all living patients were called to survey them about long term treatment effects, using a standardized survey. The data gathered was then analyzed using univariate and multivariate analysis.

Results: Only 6 institutions across the country had any publicly available data regarding head and neck patients, and only three of them had information beyond the number of patients seen. The database of head and neck cancer patients at Yale was created successfully, and compiled into outcomes books for each year that presented the relevant data. Analysis of oropharyngeal patients focused on HPV status, insurance type, academic vs. non-academic centers, and distance from radiation treatment site for patients treated with that modality. Many factors were found to be significant on univariate analysis. On multivariate analysis, it was found that HPV positive patients had better outcomes in various functional and patient reported outcomes. It was also found that private practice patients had improved outcomes compared to Medicare patients. Finally, it was also found that recurrence rates were higher for patients that lived over 15 miles away from their treatment site.

Conclusions: There is a paucity of publicly available data regarding head and neck cancer outcomes at NCI designated cancer centers around the country. At Yale, the data showed that standard metrics are in line with national outcomes. The institution can improve significantly in terms of various process metrics, most specifically in terms of having various ancillary staff work with patients who are diagnosed or treated for a head or neck cancer. Analysis of oropharyngeal patients demonstrated that patients with HPV negative cancers need closer monitoring for various functional and patient reported outcomes. It also demonstrated that patients on Medicare need monitoring for various other functional and patient reported metrics. Patients who live further from their treatment sites have higher recurrence rates, indicating that patients who have to travel further might be at higher risk for missing treatment or for receiving adequate follow-up.

Melissa A. Herrin and Sitaram M. Emani. Department of Cardiac Surgery, Boston Children’s Hospital, Harvard Medical School, Boston, MA. (Sponsored by James J. Yun, Department of Surgery, Yale University School of Medicine.)

Patients with borderline left heart, hypoplastic left heart syndrome variant (bHLHS) or unbalanced atrioventricular canal (uAVC), who undergo initial single ventricle palliation may be candidates for biventricular (BiV) conversion following left ventricle (LV) recruitment procedures. Our objective was to investigate the association between preoperative parameters and postoperative outcomes in a cohort of patients undergoing BiV conversion. We performed a retrospective review of patients who underwent BiV conversion to determine variables associated with clinical outcomes. Predictor variables included cardiac diagnosis, age and weight, LV dimension, LV end diastolic volume, LV mass, preoperative LV end diastolic pressure (LVEDP), and preoperative left atrial pressure. Primary outcome was a composite of death, heart transplant, or BiV reversal to SVP (reversal). Of 51 patients, 11 experienced primary outcome (22%). Patients with bHLHS were more likely to experience primary outcome than those with uAVC (30% vs. 6%, P = .03). Receiver operating characteristic analysis demonstrated that preoperative LVEDP had good predictive accuracy in classifying patients with and without the primary outcome (area under the curve = 0.757, 95% confidence interval: 0.594 - 0.919, P = .012). The Youden J-index indicated a cutoff of LVEDP ≥ 13 mmHg as optimal for predicting the primary outcome. Multivariable Cox regression demonstrated that LVEDP > 13 mmg was associated with primary outcome, independent of age, weight, gender and diagnosis (adjusted hazard ratio = 4.00, P = .037). Multivariable Cox regression analysis also demonstrated that elevated postoperative right ventricular pressure (>3/4 systolic blood pressure) was significantly associated with primary outcome (adjusted hazard ratio =21.75, P< .001) independent of age, weight, and diagnosis. Elevated preoperative LVEDP is a risk factor for suboptimal postoperative hemodynamics and adverse following BiV conversion after single ventricle palliation.

Kayleigh Herrick-Reynolds (Sponsored by Anna Reisman). Section of General Medicine, Department of Internal Medicine, Yale University, School of Medicine, New Haven, CT.

Medical students experience challenging situations during the clerkship year, but often lack opportunities to debrief. The near-peer reflective writing workshop at Yale School of Medicine (YSM) is a student-led session that is integrated into the didactic portion of the clerkships. We hypothesized that near-peers can create a safe environment for honest reflection, that this experience can create a rigorous perception of reflection, and that the mixed peer groups in these workshops can increase the sense of community.

Data was collected over one year from students in their clinical clerkships. Facilitators were trained near-peer volunteers who led students through a series of progressively more challenging writing prompts in a 1 hour and fifteen minute session. Open-ended evaluations were distributed after each session, and a final more detailed questionnaire was distributed at the end of the year. Facilitators also completed an end-of-year questionnaire. Open-ended answers were analyzed through an inductive, randomized, and iterative consensus process within a phenomenological framework.

Most students (62%) indicated that they would choose to attend the workshop even if they were not mandatory. Participants and facilitators alike agreed that they should become a standardized part of medical education. Facilitators and participants reported an increased sense of community from these workshops, and they provided a safe space for reflection. Students were surprised by how willing peers were to share their experiences, and they valued having protected time for reflection. The structure of the workshop helped facilitate recall and reflection. The workshops helped students process experiences, and encouraged them to focus more on empathy and emotions in clinical care. Students indicated that they would be more likely to share various challenging topics in a group led by a near-peer rather than a faculty member. Many indicated that this workshop would encourage them to increase reflective practice moving forward.

The YSM near-peer reflective writing workshop had an overwhelmingly positive reception by medical students, and has the potential to mitigate isolation and stress during the clerkship year. It may encourage medical students to change their perception of reflection, and continue to practice it throughout their careers.

Two main analyses are presented in this study. First, using publicly available information, the timing of generic drug approvals and the total number of generic manufacturers for all small-molecule drugs approved between 1984 and 2015 were characterized. Second, this study investigated the impact on drug prices and shortages of a specific FDA regulation, called the Unapproved Drugs Initiative.

The first analysis demonstrates that among 417 FDA-approved drugs, 210 were eligible for generic competition, and 77 (37%) had three or fewer generic drugs approved: 16 had three generic approvals, 9 had two, 16 had one, and 36 had zero. Among the 174 drugs with at least one generic approval, the median number of generic approvals was 7 (IQR, 4-12). Generic approvals were fewer among orphan-designated drugs when compared with non-orphan-designated drugs (18 of 33 [55%] vs. 156 of 177 [88%]; p<0.001).

The second analysis found that since 2006, 34 unapproved prescription drugs had been addressed by the Unapproved Drugs Initiative (UDI). Nearly 90% of those that went on to receive FDA approval were supported by literature reviews or bioequivalence studies, not new clinical trials. In addition, once targeted by the UDI, drugs experienced price and shortage increases of nearly 40% and 74%, respectively.

Overall, more than one-third of drugs approved after 1984 and without protection from patents have three or fewer generic competitors, making them vulnerable to price increases. By unintentionally reducing the number of manufacturers for specific drugs, the FDA’s Unapproved Drugs Initiative led to higher prices and more frequent and longer shortages, highlighting the importance of robust generic competition.

In conclusion, insufficient pharmaceutical competition has created an environment enabling price increases of old, off-patent generic drugs, such as Daraprim and Epipen. This study highlights that a substantial number of additional, similar drugs is vulnerable to such price increases for a variety of reasons. Future efforts to reform generic drug policy should seek to boost generic competition, more carefully regulate drug prices, and address brand-name pharmaceutical companies’ strategies to obstruct the ability of generic manufacturers to compete. In addition, physicians and patients should be bettered educated on the fact that a lack of generic competitors may mean that simply prescribing generic drugs will not make medications affordable for patients; alternative options may have to be explored. Such efforts are essential in ensuring continued patient access to affordable drugs.

Purpose: Little data exists to guide the appropriate use of invasive mediastinal staging in patients with clinically node-negative NSCLC staged by PET-CT. We examined a large cohort of patients with clinical stage I NSCLC determined by PET-CT to determine which patients benefit from invasive staging.

Materials/Methods: We identified consecutive clinical T1-2N0 NSCLC patients being evaluated for curative-intent therapy between 2011 and 2015. None had evidence of nodal disease by PET-CT; the endpoint was pathologic confirmation of occult N2 disease. Tumor size, location, histology, SUVmax, and radiographic appearance were evaluated as determinants of occult N2 disease. Two group comparisons of continuous variables were done with independent t-tests and categorical variables were compared with or Fisher’s exact test.

Results: In 284 patients with PET-CT-staged clinical T1-2N0 disease, the prevalence of occult N2 metastases was 7.0%. The negative predictive value of PET-CT was 92.9% and the negative predictive value of mediastinoscopy/EBUS was 96.3%.

T2 tumors were more likely to have occult N2 disease than T1 tumors (11.8% v 3.6% p=0.009). Pure solid tumors had greater involvement of N2 nodes than tumors with any ground glass component (12.6% v 1.5%, p=0.001). 17.5% of central tumor cases were found to have occult N2 metastases while 4.4% of patients with peripheral tumors (P<0.001). 33.3% of patients with solid central T2 tumors had occult N2 metastases whereas 3.6% of patients with peripheral T1 tumors with a ground glass component and 1.2% of patients with peripheral T1 solid tumors had N2 metastases.

Conclusions: Invasive mediastinal staging should be strongly encouraged in central tumors and solid T2 tumors because the risk of occult nodal involvement is greater than 10% in these cohorts. However, for patients with peripheral T1 tumors, the yield of invasive staging after a negative PET-CT is very low and invasive staging may not be warranted.

In a secondary study, a 15-item survey was administered to the current YNHH Neonatal-Perinatal fellows to evaluate the nature of the prenatal consult and the fellow’s recommendations for management of infants born in the gray zone. All five respondents recommend against resuscitation at 22 weeks, 80% recommend full resuscitation at 24 weeks and most prefer a case-by-case consideration at 23 weeks.

These findings stress the importance of frequent re-evaluation of patient choice for newborn management in borderline viable births as this may change with each additional day of pregnancy. Our study has important implications for the decision-making process between expectant parents, neonatologists, and obstetricians.

Background: Conventional amino-ester and amino-amide local anesthetics used to treat ocular pain from corneal injury and ophthalmic surgery, such as the commonly used proparacaine, require repeated administration and may delay corneal healing with long-term use. An ocular anesthetic formulation with extended effect and minimal toxicity is needed. Here, we studied the corneal anesthetic effect of two site 1 sodium channel blockers, tetrodotoxin (TTX) and saxitoxin (STX), individually or in combination with 2-adrenergic receptor agonists (dexmedetomidine and clonidine), and compared them with proparacaine. Finally, we characterized the biocompatibility of each formulation through in vitro cytotoxicity studies and studied the effect of test solutions on corneal healing.

Methods: Solutions of TTX dexmedetomidine, TTX clonidine, STX dexmedetomidine, dexmedetomidine or proparacaine were topically applied to the rat cornea. The duration of corneal anesthesia was measured by recording the blink response to probing of the cornea using a Cochet-Bonnet esthesiometer. Cytotoxicity from anesthetic solutions was measured in vitro using human corneal limbal epithelial cells. The effect on corneal healing was measured in an in vivo rat model after corneal debridement followed by repeated drug administration.

Results: Addition of dexmedetomidine to TTX or STX significantly prolonged the duration of corneal anesthesia beyond that of either drug alone, whereas clonidine did not. Topical co-administration of either TTX or STX with dexmedetomidine resulted in two to three times longer corneal anesthesia than did proparacaine. Site 1 sodium channel blocker and dexmedetomidine formulations were not cytotoxic and corneal healing was not delayed significantly by any of the tested formulations.

Conclusions: Topical co-administration of site 1 sodium channel blockers with dexmedetomidine results in prolonged duration corneal anesthesia without delaying corneal wound healing. Such formulations may be useful for the management of acute surgical and nonsurgical corneal pain.

Joyce Y Cheng, Fang-yong Li, Christine J Ko, Oscar R Colegio. Department of Dermatology, Yale School of Medicine, New Haven, CT

Solid organ transplant recipients (SOTR) have an approximately 100-fold increased risk of developing cutaneous squamous cell carcinoma (SCC). These SCCs may behave more aggressively than SCCs developing in non-immunosuppressed individuals. The purpose of this study was to gather more data regarding aggressive behavior of SCCs in immunosuppressed SOTRs compared with SCCs occurring in an immunocompetent control group. An 8-year retrospective observational cohort study was conducted to compare the demographics, care received by, and outcomes of 98 adult SOTR and immunocompetent patients with at least one histopathologically confirmed SCC. The two groups were statistically comparable with regards to demographics, patient care, follow-up time, and numbers of skin lesions they developed, though the SOTR group had significantly higher annual visit frequency (4 office visits per patient per year vs. 3, p = 0.025) and annual biopsy rates (6 biopsies per patient per year vs. 5, p = 0.039). In this setting, the SCCs developed by SOTRs did not appear to be significantly more aggressive than those in the immunocompetent control group. Our SOTRs did not develop significantly thicker tumors than the immunocompetent controls. One SOTR developed an SCC with perineural invasion, two SOTRs had regional metastasis, and one SOTR had nodal metastasis. An increased risk of carcinogenesis with increasing cumulative years of immunosuppression was surprisingly not observed among the SOTRs. SOTRs had a 90% greater risk of developing SCCs in the head and neck region compared with the immunocompetent group (RR=1.89, 95% CI 1.04-1.37). Taken together, our findings suggest that the drastically increased risk of SCCs in SOTRs and potential for aggressive behavior may be successfully managed to a level comparable to that in high-risk immunocompetent individuals with close adherence to current dermatologic surveillance recommendations for SOTRs and a marginally lower threshold for biopsy of suspicious lesions.

Twenty standardized teams (lead MD + bedside MD + two confederate clinical team members) were randomized to have a telepresent or an in-person leader. Telepresent leaders were connected via videoconference from a remote location and displayed on a screen at the bedside. All teams participated in a standardized, pre-programmed 20-minute simulated resuscitation with a scripted parent actor present. Simulations were video recorded and scored on teamwork and communication as well as clinical performance metrics using the validated STAT instrument. After each case, team members completed demographic, workload (NASA rTLX), and teamwork and communication (TeamMonitor) surveys. Post-simulation debriefings were scripted to collect qualitative data from participants regarding utility, effectiveness, and acceptability of telepresence.

There was no difference in STAT teamwork and communication scores (73 v 66; p=0.118), TeamMonitor scores (91 v 94; p=0.251), or teamwork and communication global rating scores (91 v 77; p=0.143). There was no difference in rTLX workload scores compared between team leaders (51 v 55; p=0.983) or between junior team members (44 v 59; p=0.123). Similarly, no difference was found in STAT clinical performance scores (72 v 64; p=0.168) or in time-to-defibrillation (238 sec v 253 sec; p=0.762).

Participating providers shared perspectives on the use of telepresence during resuscitation and expressed varying levels of comfort using the modality. Providers also highlighted strategies for the effective use of telepresence in the acute care setting, including enhanced verbal communication, role delineation, and mutual trust in clinical acumen of each provider involved.

Telepresence did not significantly impact teamwork and communication, workload, or clinical performance. Participating providers shared perspectives on the impacts of telepresence as well as strategies for effective use of telepresence in the acute care setting. Together, these data may inform future implementation of telepresence technology in emergency settings.

Jessica L. Buckley, Albert J. Sinusas, Mitchel R. Stacy. Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University School of Medicine, New Haven, CT.

Impaired lower extremity perfusion is a hallmark of peripheral arterial disease (PAD) and is particularly problematic in diabetic patients, who suffer from high rates of PAD, ulceration, and lower extremity amputation. The ability to non-invasively detect deficits in microvascular perfusion within vascular territories, or angiosomes, of the feet may provide information related to tissue viability and guide therapeutic interventions. In this study, we sought to apply single photon emission computed tomography (SPECT)/CT imaging to quantify volumetric microvascular perfusion within specific angiosomes containing non-healing foot ulcers in diabetic patients with critical limb ischemia (CLI). Additionally, we sought to assess the value of SPECT/CT perfusion imaging for predicting limb salvage in CLI patients undergoing lower extremity endovascular revascularization.

Forty-one diabetic patients (mean age, 66±12 yrs) with non-healing ulcers and nine healthy control subjects (mean age, 50±10 yrs) underwent SPECT/CT imaging of the feet following a resting injection of technetium-99m (99mTc)-tetrofosmin (dose, 550.6 ± 37 Mbq). CT images of diabetic feet were segmented into five angiosomes and used for quantifying relative radiotracer uptake, expressed as standardized uptake values (SUVs). SUVs were assessed for each CLI patient in the angiosome containing the non-healing ulcers, while average whole foot perfusion was assessed for healthy control subjects. Percent change in SPECT SUVs of ulcerated angiosomes was quantified following endovascular revascularization in patients, and 3-, 6-, and 12-month limb salvage outcomes were assessed.

SPECT/CT imaging allowed for visualization of perfusion deficits under resting conditions. 99mTc-tetrofosmin SPECT/CT imaging of angiosome foot perfusion demonstrated a significant difference in baseline perfusion values (SUVs) between diabetic patients with CLI and healthy control subjects (p = 0.02). Analysis of baseline SPECT/CT imaging and ankle-brachial index (ABI) measurements in CLI patients and healthy control subjects demonstrated a significant and positive relationship between SPECT/CT angiosome perfusion and ABI (p = 0.01; r = 0.41). Serial evaluation of relative changes in SPECT angiosome foot perfusion following revascularization revealed significant quantitative changes in perfusion after treatment, whereas ABI measurements did not demonstrate significant changes after revascularization. Changes in SPECT/CT-derived angiosome perfusion significantly differed between patients with and patients without amputation in the 3 (p = 0.01), 6 (p = 0.03), and 12 (p = 0.03) months following revascularization.

SPECT/CT imaging provides a useful non-invasive tool for evaluating microvascular perfusion within specific angiosomes of the foot under resting conditions. SPECT/CT imaging also allows for serial assessment of sensitive changes in angiosome microvascular perfusion following revascularization that are undetected by ABI. Perfusion imaging with SPECT/CT offers a novel quantitative imaging approach for assessing the efficacy of revascularization strategies targeted at restoring perfusion to non-healing wounds of the foot and may assist with predicting limb salvage outcomes in CLI patients undergoing revascularization. Future application of SPECT/CT perfusion imaging may provide additional value for detection and targeting of ischemic tissue for therapeutic interventions in the PAD patient population.

Part I is a single-center, retrospective analysis of a cohort of 144 post-cardiac arrest patients. Clinical findings, test results, and timing and rationale for WLST were abstracted from the medical record. Part II is a 28-question survey of physicians in the United States and abroad who manage post-cardiac arrest patients. A total of 895 respondents provided data on technique and the perceived value of each prognostic test. All data was analyzed using descriptive statistical methods.

WLST due to poor neurologic prognosis accounted for 61% of deaths in the single-center cohort. Results of both studies suggest that physicians used multiple modalities to assess prognosis, but often in a way inconsistent with evidence-based recommendations. Rationale for WLST referred generally to poor neurologic prognosis (e.g., “no chance of meaningful recovery”) rather than specific indicators. WLST in the absence of rigorous prognostic indicators undermines accurate neurologic prognostication in post-cardiac arrest patients and perpetuates a self-fulfilling prophecy of poor outcome.

Our derivation dataset consisted of 210 ischemic stroke patients receiving IV rt-PA from January 2009 until July 2013 at Yale New Haven Hospital. Our validation dataset included 303 patients who received IV rt-PA during the NINDS rt-PA trial. Predictive ability and goodness of fit were quantified by odds ratios (OR) and areas under the receiver operating characteristic curve (AUROC). Patient outcomes included sICH, brain swelling, 90-day severe outcome and 90-day mortality. Severe outcome was defined as 90-day modified Rankin Scale (mRS) scores ≥ 5, 90-day Barthel Index (BI) scores < 60 and 90-day Glasgow Outcome Scale (GOS) scores > 2.

Out of seventeen clinical parameters tested, three were independent predictors of sICH: prestroke mRS score (OR 1.54, P = 0.02), baseline National Institutes of Health Stroke Scale (NIHSS) score (OR 1.13, P = 0.002), and platelet count (OR 0.99, P = 0.04). We combined these three parameters to form the TURNP (Thrombolysis risk Using mRS, NIHSS and Platelets) score. For added simplicity, prestroke mRS score and baseline NIHSS score were also combined to form the TURN (Thrombolysis risk Using mRS and NIHSS) score, which predicted sICH without a significant drop in OR or AUROC. TURN predicted sICH with AUROC 0.74 (0.58 – 0.90) in the derivation dataset, and AUROC 0.65 (0.54 – 0.77) in the validation dataset. In the validation dataset, TURN predicted 24-hour brain swelling with AUROC 0.69 (0.63 - 0.75), 90-day mRS ≥ 5 with AUROC 0.83 (0.77, 0.89), 90-day BI < 60 with AUROC 0.81 (0.76 – 0.86), 90-day GOS > 2 with AUROC 0.81 (0.76 – 0.86) and 90-day mortality with AUROC 0.82 (0.76 – 0.88).

To improve the clinical utility of TURN, we developed and tested a mobile application Risk rtPA based on TURN for predicting 90-day outcome after rt-PA treatment. Risk rtPA returned predictions of severe outcome for a range of hypothetical patients with varying clinical characteristics, demonstrating broad applicability. This mobile application brings computationally simple prediction of post-thrombolysis risk to the bedside for real-time stroke prognostication.

Psoriasis is a chronic inflammatory disorder associated with both cardiovascular disease and mood disorders such as depression and anxiety. Multiple behavioral and physiological processes link depression, psoriasis and atherosclerosis, but inflammation is a major player implicated in the pathogenesis of all three. Psoriasis is associated with vascular inflammation, measured by 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT), and an increased risk of myocardial infarction. Vascular inflammation by FDG PET/CT predicts cardiovascular risk and outcomes. Studies have also shown that patients with psoriasis are more likely to suffer from comorbid depression and anxiety. However, whether these comorbidities may accelerate the development of cardiovascular disease in psoriasis is not well-characterized.

Hypothesis

Our primary hypothesis is that aortic vascular inflammation and coronary plaque burden will be increased in patients with psoriasis who have depression when compared to psoriasis patients who do not. In a follow-up study we hypothesize that metabolic activity in the anatomic location of the amygdala will correlate with increased vascular inflammation and decreased vascular function.

Methods

An ongoing psoriasis study cohort was chosen for the patient population. Patients who reported a history of depression (n=36) on survey were matched by age and sex to patients who reported no history of psychiatric illness (n=36). FDG PET/CT scans were used to assess vascular inflammation. From these scans, standardized uptake values were calculated by analyzing axial slices of the aorta. Target-to-background ratios were then calculated to standardize for background luminal FDG uptake. Coronary computed tomography angiography scans were analyzed in order to determine coronary plaque composition and to quantify plaque burden. Multivariate linear regression analyses were performed to understand the potential effect of psychiatric diagnoses on aortic vascular inflammation and coronary plaque burden. Traditional cardiovascular risk factors were adjusted for (standardized β reported). In a follow-up study, metabolic activity of the amygdala was quantified via analysis of FDG uptake in the anatomic location of the amygdala. These uptake values were divided by FDG uptake in the adjacent ipsilateral temporal lobe for standardization. Target-to-background ratios were then used to quantify amygdala activity. Vascular function was studied via aortic distensibility. Using phase contrast MRI scans throughout the duration of one cardiac cycle, axial slices of the aorta were analyzed to calculate aortic distensibility.

Results

Aortic vascular inflammation and coronary plaque burden were increased in psoriasis patients with comorbid depression as compared to those without. After adjustment for Framingham Risk Score, vascular inflammation (β=0.26, p=0.02), total plaque burden (β=0.17, p=0.03), and non-calcified plaque burden (β=0.17, p=0.03) associated with comorbid depression. In a subsequent study, patients with comorbid depression and/or anxiety had higher left amygdala activity. In unadjusted analyses vascular inflammation significantly associated with amygdala activity, although this relationship did not retain significance after adjustment for traditional cardiovascular risk. Aortic distensibility was significantly and inversely associated with left amygdala activity both in unadjusted analyses and after adjustment for traditional cardiovascular risk.

Conclusions

Patients with psoriasis who suffer from comorbid depression have greater burden of subclinical cardiovascular disease. Furthermore, vascular function is reduced in these patients and correlates with the degree of activity in the amygdala, a region of the brain that plays an important role in the modulation of stress. Targeted assessment of psychological stress and mood disorders in psoriasis patients may be warranted for further cardiovascular risk reduction in this high-risk population.

Response-to-treatment criteria are measures that radiologists use to quantify how well patients respond to therapy. Traditional response criteria are time-consuming to calculate and suffer from inter-operator variability. They are not uniformly performed despite an abundance of evidence demonstrating their value. The chief objection to their use is the time required for their manual calculation.

A fully automated method for response-to-treatment calculations would remove the time burden on radiologists and facilitate standardized reporting on HCC lesions. Successfull automation requires automatic liver and tumor segmentations and automatic metric calculation. Given the former (i.e. segmentations), the latter becomes straightforward.

The first project demonstrates a simple heuristic to fully automate the quantitative European Association for the Study of the Liver (qEASL) response criterion, assuming liver and tumor segmentations are provided. Our method automates parenchymal region-of- interest (ROI) selection and shows result equivalent to manual ROI selection achieved by the current qEASL standard.

The second project applies recent developments in deep learning for computer vision to the problem of automatic liver and tumor segmentation. We train a deep convolutional neural network classifier to detect liver and tumor margins in contrast-enhanced MR imaging.

These projects represent an end-to-end solution to tumor response criteria. They provide radiologists a straightforward method for response calculations and produce consistent, reproducible results.