Normal Workflow Research Articles

Implementation of the standardized, evidence-based three-stage suicide risk screening: Experience of a large Veterans Affairs medical center.

Suicide is a global public health concern and may be preventable with early identification. The suicide rate among US veterans is increasing. In response to the increase, Veterans Health Administration recommended a new standardized three-step, evidence-based suicide risk screening process across all Veterans Health Administration sites. The purpose of this project was to implement the new three-step suicide screening method and evaluate the rate of provider adherence. The implementation occurred in seven clinical sites in the Veterans Affairs Greater Los Angeles Health care System. Following initial implementation, two Plan-Do-Study-Act (PDSA) evaluated provider adherence to the screening processes. Staff members at each site received suicide prevention education. Staff members had the option of using an embedded template in the course of normal patient care workflow. Plan-Do-Study-Act 1 measured the early results. Staff members achieved a performance adherence rate of 18%, indicating that staff were less likely to proactively screen for risk of suicide. In PDSA-2, the mandatory use of screening replaced the optional use. Staff members achieved a 95% adherence rate after 3 months. Changing the workflow within the electronic health record from optional to mandatory utilization brought forth improvements in suicide prevention screening.

Perspectives and perceived barriers to pharmacist-prescribed tobacco cessation services in the community pharmacy setting

Objective(s)Tobacco use is a leading cause of preventable morbidity and mortality in the United States. Community pharmacists are in an advantageous position to increase patient accessibility to tobacco cessation medications and counseling. Pharmacists are permitted to prescribe or furnish tobacco cessation medications in 13 states with varying requirements and limitations. The primary objective of this study was to evaluate the perspectives and perceived barriers that pharmacy personnel have in providing pharmacist-prescribed tobacco cessation services in the community pharmacy setting. The secondary objectives were to evaluate current practices of the 5 A’s model and to identify additional training needed to provide the service. MethodsThis study was a descriptive cross-sectional electronic survey of pharmacy personnel from a large grocery pharmacy chain. The respondents were asked about their demographics, current practices of the 5 A’s model, perceived barriers and attitudes toward pharmacist-prescribed tobacco cessation services, and additional training needed to provide the service. ResultsThe survey received 79 responses from pharmacists. The results showed that 92.4% (73/79) of the respondents agreed that community pharmacists should provide tobacco cessation services. The biggest barrier identified was “lack of time during normal workflow to deliver tobacco cessation services” at 54.4% (43/79). Regarding the 5 A’s model, 74.7% (59/79) of the pharmacists responded “never” or “somewhat infrequently” to how often they practice the “Ask” step, with other steps in the 5 A’s model reported at similar frequencies. The top 3 additional trainings that the pharmacists identified were “strategies developing a follow-up plan with patients,” “incorporating service into workflow,” and “strategies providing counseling on tobacco cessation.” ConclusionEfforts should be made to give community pharmacy personnel more time and guidance to provide patient care services beyond traditional dispensing roles to provide tobacco cessation services.

COVID-19 risk in elective surgery during a second wave: a prospective cohort study.

The COVID-19 pandemic has greatly affected access to elective surgery, largely because of concerns for patients and healthcare workers. A return to normal surgery workflow depends on the prevalence and transmission of coronavirus in elective surgical patients. The aim of this study was to determine the prevalence of active SARS-coronavirus-2 infection during a second wave among patients admitted to hospital for elective surgery in Victoria. Prospective cohort study across eight hospitals in Victoria during July-August 2020 was conducted enrolling adults and children admitted to hospital for elective surgery or interventional procedure requiring general anaesthesia. Study outcomes included a positive polymerase chain reaction (PCR) test for SARS-CoV-2 in the preoperative period (primary outcome), and for those with a negative test preoperatively, the incidence of a positive PCR test for SARS-CoV-2 in the post-operative period. We enrolled 4965 elective adult and paediatric surgical patients from 15 July to 31 August 2020. Four patients screened negative on questionnaire but had a positive PCR test for coronavirus, resulting in a Bayesian estimated prevalence of 0.12% (95% probability interval 0-0.26%). There were no reports of healthcare worker infections linked to elective surgery during and up to 2 weeks after the study period. The prevalence of SARS-CoV-2 in asymptomatic elective surgical patients during a second wave was approximately 1 in 833. Given the very low likelihood of coronavirus transmission, and with existing current hospital capacity, recommencement of elective surgery should be considered. A coronavirus screening checklist should be mandated for surgical patients.

Successes and challenges of implementing a lung cancer screening program in federally qualified health centers: a qualitative analysis using the Consolidated Framework for Implementation Research.

In recent years, studies have shown that low-dose computed tomography (LDCT) is a safe and effective way to screen high-risk adults for lung cancer. Despite this, uptake remains low, especially in limited-resource settings. The American Cancer Society (ACS) partnered with two federally qualified health centers and accredited screening facilities on a 2 year pilot project to implement an LDCT screening program. Both sites attempted to develop a referral program and care coordination practices to move patients through the screening continuum and identify critical facilitators and barriers to implementation. Evaluators conducted key informant interviews (N = 46) with clinical and administrative staff, as well as regional ACS staff during annual site visits. The Consolidated Framework for Implementation Research guided our analysis of factors associated with effective implementation and improved screening outcomes. One study site established a sustainable lung screening program, while the other struggled to overcome significant implementation barriers. Increased time spent with patients, disruption to normal workflows, and Medicaid reimbursement policies presented challenges at both sites. Supportive, engaged leaders and knowledgeable champions who provided clear implementation guidance improved staff engagement and were able to train, guide, and motivate staff throughout the intervention. A slow, stepwise implementation process allowed one site’s project champions to pilot test new processes and resolve issues before scaling up. This pilot study provides critical insights into the necessary resources and steps for successful lung cancer screening program implementation in underserved settings. Future efforts can build upon these findings and identify and address possible facilitators and barriers to screening program implementation.

Blockchain-driven anomaly detection framework on edge intelligence

There are a large number of end devices in an IoT system, which may malfunction due to various reasons, such as being attacked. Anomaly detection of the devices and the whole IoT system normally rely on the analysis of the huge amount of log records generated by the end devices. How to protect the log records from being tampered with and realize the real-time anomaly detection is a challenging task which is still not addressed. Existing works on anomaly detection by the emerging and effective deep learning algorithms require the transfer of log data to cloud servers which incurs high communication overhead and long detection latency, and is subject to the risk of being tampered. In this paper, we propose a novel and efficient hierarchical framework for online anomaly detection in IoT systems atop Blockchain and smart contracts. At the device layer of the hierarchical framework, an efficient feature extractor is developed to preprocess the raw log data which greatly reduces the size of data to be transferred while keeps sufficient information for the anomaly detection model to use. At the cloud layer of the framework, deep learning models use the processed data from the device layer to build the detection model and output normal workflow patterns. In the edge layer of the framework, a permissioned blockchain is built and a series of smart contracts are developed which can guarantee data integrity and achieve automatic anomaly detection based on the model output from the cloud layer. Extensive experiments demonstrate that our framework can reduce the ledger size by 7.1% without detection accuracy reduction compared with traditional centralized solutions and the detection latency is only 0.47ms in our prototype. Our feature extractor can speed up by 3.6x–7.3x times on the execution time with almost the same CPU usage rate compared with state-of-the-art log parsers and encryption solutions, such as AES and RSA.

Mutational Patterns and Correlation to Chip-Related Mutations in Hematological Malignancies - a Study on Mutation Frequencies of 122 Genes in 28 Entities Including 3096 Cases

Background: Acquired somatic mutations are crucial for the development of the majority of cancers. In hematological malignancies, some molecular mutations are very specific for certain entities (e.g. BRAF in HCL, MYD88 in LPL), while others were detected in a variety of malignancies (e.g. mutations in TP53, TET2, DNMT3A, RUNX1). Moreover, mutations in genes related to CHIP (clonal haematopoiesis of indeterminate potential; ASXL1, TET2,DNMT3A) were detected in an age-related manner. Aim: (1) Analysis/comparison of mutation frequencies of 122 selected genes in 3096 cases with 28 different hematological malignancies for identification of "mutation-driven" entities. (2) Correlation of CHIP-related mutations with mutational landscapes. Methods: Whole-genome sequencing (WGS) was performed for all 3096 patients. For this, 151bp paired-end reads were generated on NovaSeq 6000 and HiSeqX machines (Illumina, San Diego, CA). The Illumina tumor/unmatched normal workflow was used for variant calling. All reported p-values are two-sided and were considered significant at p<0.05. Results: Entities with the highest numbers of mutations (median n=4) and thus potentially with the largest impact of mutations on pathogenesis comprised aCML (range: 1-7), CMML (1-6), MDS/MPN-U (2-5) and s-AML (2-8), whereas the lowest numbers (median n=0) were observed for CML (0-3), MGUS (0-2), MLN_eo (0-3), NK cell neoplasm (0-3) and PPBL (0-2). In the total cohort of 3096 cases, the most frequently mutated genes were TET2 (14%), ASXL1 (13%), TP53 (10%), SF3B1 (9%), DNMT3A (9%) and SRSF2 (9%). Entities with high frequencies of specific mutations (> 50%) comprised: aCML (ASXL1, 86%), BPDCN (TET2, 67%), BL (TP53, 60%), CMML (TET2, 67%; ASXL1, 58%), FL (KMT2D, 87% and CREBBP, 73%), HCL (BRAF, 100%), LPL (MYD88, 98%; CXCR4, 51%), MDS/MPN-U (ASXL1, 60%), MPN (JAK2, 68%), B-NHL (TP53, 50%) and T-NHL (STAT3, 52%). Mutations enriched in distinct entities included SETBP1 (26% in MDS/MPN overlaps), CSF3R (30% in MDS/MPN-U), STAT3 (only in T-NHL and NK cell neoplasm, 52% and 23%), NOTCH1 and PHF6 (T-ALL, 38% and 30%) and MYC and ID3 (almost exclusively in BL, 30% each). Genes predominantly mutated in myeloid neoplasms comprised e.g. SF3B1 (with the exception of CLL), JAK2, NPM1, RUNX1, IDH2, CEBPA, STAG2, NF1 and GATA2. By contrast, mutations in KMT2D, MYD88, ARID1A, ATM, CXCR4, BIRC3 and CD79B were detected almost exclusively in lymphoid malignancies. A broad distribution across entities was observed for mutations in TET2, ASXL1,DNMT3A, TP53, BCOR and ETV6. Thus, the first three, i.e. CHIP-related genes were also mutated with a high frequency in lymphoid neoplasms. In line with this, gene mutations found in the largest number of entities comprise DNMT3A (n=23 entities), TET2 (n=21), ASXL1, TP53, NRAS (n=19, respectively), KRAS and BCOR (n=17, respectively). Further, we compared the mutational patterns of cases with at least one CHIP-associated mutation (n=920 cases in the total cohort, "CHIP+") with cases without such mutations (n=2176, "CHIP-") to decipher CHIP-correlated mutation patterns. Significant differences with respect to accompanying mutations were mainly detected for myeloid neoplasms (MDS, mutations in n=12 genes significantly different in CHIP+ vs. CHIP- without CHIP genes themselves; AML, n=7; MPN, n=4; aCML, n=2; CMML, n=1) but also for MPAL (n=3), T-ALL (n=2), B-ALL, FL and LPL (n=1, respectively). Mutations in TP53 were found significantly enriched in CHIP- cases in 4 different entities, moreover mutations in KRAS, WT1 and SF3B1 were more abundant in CHIP- cases (in CMML, AML and aCML, respectively). By contrast, CHIP+ cases were characterized by high frequencies of mutations in RUNX1 (in n=4 entities), SRSF2, IDH2, NRAS (n=3) and EZH2 (n=2). Conclusions: (1) Certain mutations showed a broad distribution within or even across the myeloid/lymphoid lineage, including CHIP-related mutations frequently detected also in lymphoid malignancies. (2) The median numbers of mutations were low in entities that are defined by chromosomal fusions (CML, MLN_eo) or in entities that are regarded as "pre-malignant" (MGUS, PPBL), while especially MDS/MPN overlap cases seem to be mutation-driven (high number of mutations). (3) We deciphered different mutation patterns in CHIP+ (RUNX1, SRSF2, IDH2, NRAS, EZH2) and CHIP- (TP53, KRAS, WT1, SF3B1) cases across all entities, suggesting differences in pathophysiology. Disclosures No relevant conflicts of interest to declare.

Bedside Rounds in Intensive Care Units during the COVID-19 Pandemic and Beyond.

A survey-based pilot study was performed to examine the feasibility of videoconferencing to facilitate multidisciplinary rounds following the initiation of strict isolation and social distancing policies in a pediatric intensive care unit (PICU). The use of a mobile workstation was implemented as the central hub for rounding at the bedside by the attending physicians, while other members of the multidisciplinary and multispecialty team joined rounds from other locations with maintaining appropriate social distance. Fifty-eight staff members who participated in videoconferencing rounds completed the postimplementation survey. Eighty-eight per cent of staff agreed that the use of videoconferencing to facilitate rounds was an effective strategy to maintain social distancing between team members during the pandemic. Sixty-four percent of staff agreed that the use of videoconferencing improved participation of the PICU team and consultants by increasing access to rounds. Over 50% of staff agreed that the use of videoconferencing improved the efficiency of rounds and team productivity. Only 4% of staff responded that videoconferencing increased the duration of rounds and 37% responded that it decreased resident and team education. Fifty-five percent of staff agreed that videoconferencing was used to promote parental participation during this pandemic month. Videoconferencing was found to be a feasible solution to safely conduct multidisciplinary rounds while maintaining social distancing, and participants found it effective without interfering with normal workflow. Incorporating videoconferencing into traditional rounding practices may be advantageous following the pandemic to improve team and family access to rounds and workflow efficiency and rounding structure.

Establishing an Office-Based Framework for Resuming Otolaryngology Care in Academic Practice During the COVID-19 Pandemic.

The COVID-19 health crisis abruptly disrupted the practice of otolaryngology. This article aims to define the changes needed to operate an academic otolaryngology practice safely and efficiently from within the epicenter of the pandemic. We define the areas of normal patient workflow that have been affected by COVID-19, and we offer mitigation strategies with attention paid to the specific needs of subspecialties. The article includes data specific to the office practice metrics of the Weill Cornell Medicine Department of Otolaryngology-Head and Neck Surgery, as well as publically available data from New York Presbyterian Hospital system and the New York Times. Expert opinion. Through careful planning and execution, it is possible to reestablish safe otolaryngologic patient care during the COVID-19 pandemic. It will require a significant change from prior practice models for successful implementation. Additionally, telemedicine can be positively integrated into the treatment of otolaryngology diseases for new and established patients. The information conveyed in this review can be used as a guide by large and small otolaryngology groups to identify aspects of the patient visit that are "at risk" due to COVID-19, and it suggests sensible responses that can be made without a significant disruption to normal practice. The methods used to identify vulnerabilities with the patient visit process can be applied to future unforeseen crises, such as a resurgence of COVID-19 or a novel pandemic.

Laboratory practice in the face of Covid-19

In December 2019, the COVID-19 outbreak began in Wuhan and quickly spread over the world. Hospitals have had to make drastic changes to normal workflows and practices to handle the current pandemic. Laboratories face unique challenges in the management of the investigation of COVID-19. Not only must we consider the safe collection and delivery of samples, but we must also observe the latest guidelines in testing for the virus. We have introduced several new measures in our laboratory to accommodate the collection and testing of samples for COVID-19 from both within the hospital and external screening sites. These changes encompass the pre-analytical (sample collection, packaging, and delivery), analytical (evaluation, handling and preparation of samples) and post-analytical (result reporting within the hospital and to external bodies) aspects of both routine (biochemistry, haematology, transfusion and urine, stool and body fluid testing) and COVID-19 testing, to ensure the safe and efficient testing of any patient samples. In addition, more practical matters, such as laboratory staffing and continuing staff education, have also been changed to ensure the safety and well-being of laboratory staff. With the implementation of new rules and regulations, we seek to safeguard the health of all healthcare workers while streamlining the workflow for the large amount of testing required during this period. As the pandemic continues, new tests (e.g. COVID-19 serology and IL6 testing for prognosis or monitoring) are being requested, requiring even more changes and assay evaluation before implementation. All laboratories must be ready to adapt to these new challenges during the COVID-19 pandemic.

Opioids, Polypharmacy, and Drug Interactions: A Technological Paradigm Shift Is Needed to Ameliorate the Ongoing Opioid Epidemic.

Polypharmacy is a common phenomenon among adults using opioids, which may influence the frequency, severity, and complexity of drug–drug interactions (DDIs) experienced. Clinicians must be able to easily identify and resolve DDIs since opioid-related DDIs are common and can be life-threatening. Given that clinicians often rely on technological aids—such as clinical decision support systems (CDSS) and drug interaction software—to identify and resolve DDIs in patients with complex drug regimens, this narrative review provides an appraisal of the performance of existing technologies. Opioid-specific CDSS have several system- and content-related limitations that need to be overcome. Specifically, we found that these CDSS often analyze DDIs in a pairwise manner, do not account for relevant pharmacogenomic results, and do not integrate well with electronic health records. In the context of polypharmacy, existing systems may encourage inadvertent serious alert dismissal due to the generation of multiple incoherent alerts. Future technological systems should minimize alert fatigue, limit manual input, allow for simultaneous multidrug interaction assessments, incorporate pharmacogenomic data, conduct iterative risk simulations, and integrate seamlessly with normal workflow.

A Narrative Review of Strategies to Increase Patient Safety Event Reporting by Residents.

Because residents are often on the frontlines of patient care and are likely to witness adverse events firsthand, it is critical they report patient safety events. They may, however, be underreporting. We examined the current literature to identify strategies to increase patient safety event reporting by residents. We used CINAHL (EBSCO Information Services, Ipswich, MA), EMBASE (Elsevier, Amsterdam, the Netherlands), PsycINFO (APA Publishing, Washington, DC), and PubMed (National Center for Biotechnology Information, Bethesda, MD) databases. The search was limited to English-language articles published in peer-reviewed journals through March 2020. Key terms included "residents, trainees, fellows, interns, graduate medical education, house staff, event reporting, patient safety reporting, incident reporting, adverse event, and medical error." To organize findings, we adapted a published framework of strategies for encouraging self-protective behavior. We identified 68 articles that described strategies used to increase event reporting. The most sustainable interventions used a combination of 3 of the 5 strategies: behavior modeling, surveys and messaging, and required limited financial support. The survey creates awareness; the behavior modeling is critical for educational purposes, and the reminders help to reinforce the new behavior and embed it into routine patient care activities. We noted a dearth of studies involving trainees in root cause analysis following submission of event reports. The most successful sustainable interventions were those that combined strategies that minimized time for busy physicians, incorporated accessible event reporting in already existing medical records, and became part of a normal workflow in patient care.

Talking about the Cost of Care: A Critical Component of Shared Decision Making Patient and Provider Perspectives

Transparency about the costs that patients incur is an increasingly important factor in making decisions throughout the course of diagnosis and treatment. Both patients and providers regard honest, informed discussions about these costs as critical to providing person-centered care, but often encounter a range of barriers to initiating and maintaining these conversations. The Patient Advocate Foundation/National Patient Advocate Foundation with partial funding from the Robert Wood Johnson Foundation has conducted both quantitative and qualitative research with patients and providers designed to identify key issues related to cost of care discussions. These include components of financial toxicity for patients, both related to direct costs and to quality of life, as well as questions as to who should initiate these conversations, when they should take place, and what kind of information and training are needed to make these conversations meaningful. This article reports the results of this work with recommendations for providers about how to incorporate cost of care discussions into the normal clinical work flow and patient life flow.

MARIN: an open-source mobile augmented reality interactive neuronavigation system.

Neuronavigation systems making use of augmented reality (AR) have been the focus of much research in the last couple of decades. In recent years, there has been considerable interest in using mobile devices for AR in the operating room (OR). We propose a complete system that performs real-time AR video augmentation on a mobile device in the context of image-guided neurosurgery. MARIN (mobile augmented reality interactive neuronavigation system) improves upon the state of the art in terms of performance, allowing real-time augmentation, and interactivity by allowing users to interact with the displayed data. The system was tested in a user study with 17 subjects for qualitative and quantitative evaluation in the context of target localization and brought into the OR for preliminary feasibility tests, where qualitative feedback from surgeons was obtained. The results of the user study showed that MARIN performs significantly better in terms of both time ([Formula: see text]) and accuracy ([Formula: see text]) for the task of target localization in comparison with a traditional image-guided neurosurgery (IGNS) navigation system. Further, MARIN AR visualization was found to be more intuitive and allowed users to estimate target depth more easily. MARIN improves upon previously proposed mobile AR neuronavigation systems with its real-time performance, higher accuracy, full integration in the normal workflow and greater interactivity and customizability of the displayed information. The improvement in efficiency and usability over previous systems will facilitate bringing AR into the OR.

Talking About the Cost of Care: A Critical Component of Shared Decision Making: Patient and Provider Perspectives

IntroductionTransparency about the costs that patients incur is an increasingly important factor in making decisions throughout the course of diagnosis and treatment. Both patients and providers regard honest, informed discussions about these costs as critical to providing person-centered care, but often encounter a range of barriers to initiating and maintaining these conversations. MethodsThe Patient Advocate Foundation/National Patient Advocate Foundation with partial funding from the Robert Wood Johnson Foundation has conducted both quantitative and qualitative research with patients and providers designed to identify key issues related to cost of care discussions. ResultsKey issues identified include components of financial toxicity for patients, both related to direct costs and to quality of life, as well as questions as to who should initiate these conversations, when they should take place, and what kind of information and training are needed to make these conversations meaningful. ConclusionsThis article reports the results of this work with recommendations for providers about how to incorporate cost of care discussions into the normal clinical work flow and patient life flow.

Factors associated with workarounds in barcode-assisted medication administration in hospitals.

Aims and objectivesTo identify that workarounds (defined as “informal temporary practices for handling exceptions to normal procedures or workflow”) by nurses using information technology potentially compromise medication safety. Therefore, we aimed to identify potential risk factors associated with workarounds performed by nurses in Barcode‐assisted Medication Administration in hospitals.BackgroundMedication errors occur during the prescribing, distribution and administration of medication. Errors could harm patients and be a tragedy for both nurses and medical doctors involved. Interventions to prevent errors have been developed, including those based on information technology. To cope with shortcomings in information technology‐based interventions as Barcode‐assisted Medication Administration, nurses perform workarounds. Identification of workarounds in information technology is essential to implement better‐designed software and processes which fit the nurse workflow.DesignWe used the data from our previous prospective observational study, performed in four general hospitals in the Netherlands using Barcode techniques, to administer medication to inpatients.MethodsData were collected from 2014–2016. The disguised observation was used to gather information on potential risk factors and workarounds. The outcome was a medication administration with one or more workarounds. Logistic mixed models were used to determine the association between potential risk factors and workarounds. The STROBE checklist was used for reporting our data.ResultsWe included 5,793 medication administrations among 1,230 patients given by 272 nurses. In 3,633 (62.7%) of the administrations, one or more workarounds were observed. In the multivariate analysis, factors significantly associated with workarounds were the medication round at 02 p.m.–06 p.m. (adjusted odds ratio [OR]: 1.60, 95% CI: 1.05–2.45) and 06 p.m.–10 p.m. (adjusted OR: 3.60, 95% CI: 2.11–6.14) versus the morning shift 06 a.m.–10 a.m., the workdays Monday (adjusted OR: 2.59, 95% CI: 1.51–4.44), Wednesday (adjusted OR: 1.92, 95% CI: 1.2–3.07) and Saturday (adjusted OR: 2.24, 95% CI: 1.31–3.84) versus Sunday, the route of medication, nonoral (adjusted OR: 1.28, 95% CI: 1.05–1.57) versus the oral route of drug administration, the Anatomic Therapeutic Chemical classification‐coded medication “other” (consisting of the irregularly used Anatomic Therapeutic Chemical classes [D, G, H, L, P, V, Y, Z]) (adjusted OR: 1.49, 95% CI: 1.05–2.11) versus Anatomic Therapeutic Chemical class A (alimentary tract and metabolism), and the patient–nurse ratio ≥6–1 (adjusted OR: 5.61, 95% CI: 2.9–10.83) versus ≤5–1.ConclusionsWe identified several potential risk factors associated with workarounds performed by nurses that could be used to target future improvement efforts in Barcode‐assisted Medication Administration.Relevance to clinical practiceNurses administering medication in hospitals using Barcode‐assisted Medication Administration frequently perform workarounds, which may compromise medication safety. In particular, nurse workload and the patient–nurse ratio could be the focus for improvement measures as these are the most clearly modifiable factors identified in this study.

Quantification of Osteosarcoma Mineralization on Plain Radiographs - Novel Software Applications to Assess Response to Chemotherapy.

Osteosarcoma is the most common primary malignant bone tumor in children. Patient survival with osteosarcoma is heavily influenced by the response to chemotherapy, measured by tumor necrosis upon histological analysis. Unfortunately, response is not measurable until the time of surgery and therefore modifications to chemotherapy protocol are only made after several weeks of treatment and surgery. Osteosarcoma tumors often demonstrate increased mineralization following the onset of chemotherapy. Furthermore, it has been hypothesized that this mineralization-apparent on radiographs-may correlate with chemotherapy response, however, this has not been demonstrated with qualitative visual evaluation. The ability to non-invasively measure a patient's response to chemotherapy using plain radiographs, which is currently included in the normal clinical workflow, would guide the medical oncologists to tailor treatment for patients with osteosarcoma. After obtaining appropriate multi-center institutional review board approvals, we identified 31patients that possess a pair of pre-and post-chemotherapy radiograph along with the necrosis measure. The images were digitized scans of physical radiographs between 1999 and 2013. Software was designed to measure the signal intensities in the tumor, a region of the soft tissue, air, and healthy bone. The tumor signals were normalized based on the random combination of air, soft tissue or bone, by subtraction or division. The differences in tumor signal between pre-and post-image were plotted against the percent necrosis determined by histological analysis. Different combinations of the normalization methods were compared 2based on the slope, coefficient of determination (R2) and Pearson correlation coefficient (ρ).

Identification of a Specific Immunophenotype Associated with a Consistent Pattern of Genetic Mutations Including Srsf2 and Gene Expression Profile in MDS

Background: Myelodysplastic syndromes (MDS) comprise a heterogeneous group of diseases classified by comprehensive diagnostics. Identification of homogeneous subgroups is desirable to understand differences in clinical course and to develop targeted treatment approaches. We identified a specific immunophenotype in MDS and analyzed its genetic background. Aim: To assess the genetic background of MDS with specific CD11b/CD16 expression pattern in granulocytes and reduced CD45 expression in myeloid progenitor cells (MPC) by whole genome (WGS) and whole transcriptome sequencing (WTS). Methods: During routine flow cytometric evaluation for suspected MDS a specific immunophenotype was identified (Figure) which was found consistently associated with SRSF2 mutations. Normal granulocytic maturation starts with steep increase in CD11b expression paralleled by only slight increase of CD16 expression. This is followed by steep increase in CD16 expression paralleled by only slight further increase in CD11b expression (giraffe pattern). The immunophenotype newly identified in MDS patients is characterized by an increase of CD11b expression without CD16 expression followed by an increase in CD16 expression without further CD11b expression (rectangle pattern) and only dim CD45 expression of MPC. 32 such cases, all SRSF2mut (group 1) were selected to evaluate the molecular differences to 51 MDS SRSF2mut cases without this specific immunophenotype (group 2) by WGS and WTS. Variants were called with the Illumina tumor/unmatched normal workflow. A mixture of genomic DNA from multiple anonymous donors served as control sample. Variants with a global population frequency > 1% were removed and the final analysis was performed on protein-altering and splice-site variants only. For gene expression analysis, estimated gene counts were normalized and the resulting log2 counts per million were used as a proxy of gene expression in each sample. Unsupervised hierarchical clustering (HC) with Ward's method as clustering method and Euclidean distance as distance measure were used to segregate the patients according to their whole transcriptional profile. Analysis of differentially expressed (DE) genes was performed considering an adjusted p<0.05 and an absolute logFC>2 as significant. Results: HC resulted in a dendrogram clearly showing two distinct clusters of samples consistent with groups 1 and 2 as defined based on the immunophenotype. 116 genes were DE, including 13 transcription factors (e.g. MYCN, ZNF462, HOXA10), 4 FDA approved drug targets (MMP2, PRLR, AR, MME) and 6 potential drug targets (AASS, PXDN, CYP27A1, SLC16A2, UCHL1, GUCY1A3) as indicated by the human protein atlas (Uhlén et al, 2015). The identified gene signature was significantly enriched for cellular developmental process, cell adhesion and extracellular matrix organization (adjusted p<0.05 for each pathway). Analysis of the mutational profiles strikingly revealed that STAG2 mutations were exclusively found in group 1 (17/32, 53% vs. 0%, p<0.001). Hence, >50% of cases with the specific immunophenotype were characterized by co-mutations in SRSF2 and STAG2. STAG2 is a key member of the cohesion complex and its regulatory functions include chromosome segregation and formation of DNA loops that influence gene expression. Recently it was shown that the deletion of STAG2 severely impacts chromatin accessibility and cell fate decisions in hematopoiesis (Viny et al. 2019). Furthermore, group 1 was characterized by higher frequencies of mutations in ASXL1 (25/32, 78% vs. 9/51, 18%, p<0.001), NRAS (6/32, 19% vs. 3/51, 6%, p=0.07) and RUNX1 (11/32, 34% vs. 5/51, 10%, p=0.007). Of the 15 STAG2 unmutated cases from group 1, 6 (40%) had mutations in at least two genes out of ASXL1, NRAS and RUNX1. The same constellation was present in only 2/51 (4%) cases from group 2. Taken together, the immunophenotype of group 1, besides predicting SRSF2 mutations, is significantly associated with a specific mutation profile which is found in 23/32 (72%) such cases and which occurs infrequently in other SRSF2mut MDS cases (2/51, 4%, p<0.001). Conclusions: We here for the first time describe a specific immunophenotype which defines MDS cases with SRSF2 mutations and a consistent and specific mutational and gene expression profile. This comprehensive data may qualify for the definition of a new sub-entity which should be addressed in future research. Disclosures Kern: MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Walter:MLL Munich Leukemia Laboratory: Employment. Meggendorfer:MLL Munich Leukemia Laboratory: Employment. Baer:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

Challenging Blast Counts By Machine Learning Techniques and Genome Sequencing for Discriminating AML and MDS

MDS and AML are separated according to the percentage of bone marrow blasts. Although this is an arbitrary threshold, it leads to different thinking and therapeutic strategies. Following the WHO classification for myeloid neoplasms an increasing number of specific genetic aberrations are incorporated in the classification, like t(15;17), t(8;21), inv(16) or NPM1 and CEBPA mutations for AML. Some of these are classified as AML even irrespective of blast counts. Our aim was to stratify today´s AML and MDS patients by using genome sequencing data and a combination of machine learning techniques to identify the most prominent discriminative features, challenging the blast count as gold standard for discriminating AML from MDS. The analysis was based on a cohort of 1,292 patients (pts) morphologically diagnosed according to WHO classification: 591 AML and 701 MDS. Whole genome sequencing (WGS) was performed with 90x coverage for all samples to assess their mutational profiles. The Illumina tumor/unmatched normal workflow was used for variant calling. To remove the most frequent germline variants, each variant was queried against the gnomAD database, variants in non-coding regions and with global population frequencies >1% where excluded. The resulting dataset was filtered to exclude genes with a mutation frequency <1% resulting in a variant list assigned to 2,918 different genes. Additionally, the 10 most abundant cytogenetic aberrations in AML and MDS were used to train the model. The dataset was randomly divided into a training (90%) and validation (10%) set ensuring that all morphology based AML and MDS phenotypes according to WHO were present in both sets. Subsequently we applied LASSO regression to identify the features that optimize the classification accuracy of AML versus MDS. 500 models were built with 10-fold cross-validation to stratify the patients with an accuracy range from 66.7% - 95.2%. All the models differed slightly in their composition of selected features and, hence, the models with the highest accuracy (top 5%) were chosen, the selected features assessed and features that occurred in more than 50% of the models were kept to train a Naïve Bayes classifier. Using this final model to stratify patients of the validation cohort we achieved an accuracy of 83.7%. The model consisted of only 74 genetic markers, covering 8 cytogenetic aberrations and 66 affected genes. 26 of those genes belong to the COSMIC cancer gene list and of these 21 are well known markers included in myeloid screening panels. 40 genes and their variants (including rare polymorphisms, as well as variants or somatic mutations) have not been described in association with myeloid neoplasms so far. The model assigned 35/132 (27%) AML cases and 10/140 (7%) MDS cases to their respective counterparts. The median bone marrow blast count was 72% for concordant AML cases, 4% for concordant MDS cases, 44% for the falsely assigned AML cases and 8% for falsely assigned MDS patients, indicating that the blast count was lower in patients assigned to MDS instead of AML. However, the classification was based on the sample genetic background only and, hence, we grouped the model features based on relational self organizing maps (SOM) to identify entity-specific co-occurrence networks (7 AML and 4 MDS clusters, Figure 1). In AML the dominant cluster features were: normal karyotype with NPM1 mutation, t(8;21), t(15;17), inv(16), WT1 mutation, KRAS mutation, and TP53 mutation with co-occurring del(7) and del(5q). In MDS the 4 clusters were described by: del(5q), SF3B1 mutation, normal karyotype, and del(20q). All the misclassified cases showed a molecular profile that could be clearly associated with the ascertained entity-defining molecular features, explaining the divergent group assignment. Therefore, the molecular profiles indicate a considerable genetic similarity of the two diseases independent of the blast count. Considering the possible synergistic effects of co-occurring variants, the higher density of the interaction network in AML compared to MDS is noteworthy. In conclusion, our study shows that a depth of information is hidden in the exome of myeloid neoplasms that can be uncovered by machine learning techniques and allows a genetic-based classification of AML and MDS. These techniques not only challenge the morphological classification, but also suggest that blast counts might not be the best parameter for treatment decision. Disclosures Meggendorfer: MLL Munich Leukemia Laboratory: Employment. Walter:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

2137. Impact of Accelerate Pheno™ Rapid Blood Culture Detection System with Real-time Notification vs. Standard Antibiotic Stewardship on Clinical Outcomes in Bacteremic Patients

BackgroundAccelerate Pheno™ blood culture detection system (AXDX) provides identification (ID) and antimicrobial susceptibility testing (AST) within 8 hours of growth in blood culture. We previously reported length of stay (LOS), time to optimal therapy (TTOT), and antibiotic days of therapy (DOT) decrease following AXDX implementation alongside an active antimicrobial stewardship program (ASP). It is unclear whether real-time notification (RTN) of results further improves these variables.MethodsA single-center, quasi-experimental before/after study of adult bacteremic inpatients was performed after implementation of AXDX. A 2017 historical cohort was compared with two 2018 intervention cohorts. Intervention-1: AXDX performed 24/7 with results reviewed by providers or ASP as part of their normal workflow. Intervention 2: AXDX performed 24/7 with RTN to ASP 7 days per week 9a-5p and overnight results called to ASP at 9a. Interventions 1 and 2 were utilized on an alternating weekly basis during the study (February 2018–September 2018). Historical ID/AST were performed using VITEK® MS and VITEK®2. Exclusion criteria included polymicrobial or off-panel isolates, prior positive culture, and patients not admitted at the time of AST. Clinical outcomes were compared with Wilcoxon rank-sum and χ 2 analysis.Results540 (83%) of 650 positive cultures performed on AXDX had on-panel organisms. 308 (57%) of these cultures and 188 (77%) of 244 reviewed historical cultures met inclusion criteria. Baseline illness severity and identified pathogens were similar between cohorts. Clinical outcomes and antimicrobial DOT are reported in Tables 1 and 2.ConclusionFollowing our implementation of AXDX, clinical outcomes including LOS, TTOT, total DOT, BGN DOT, and frequency of achieving optimal therapy were significantly improved compared with a historical cohort. Addition of RTN for AXDX results in the setting of an already active ASP did not further improve these metrics. However, compared with historical arm, AXDX with RTN did significantly impact specific subsets of antibiotic use while AXDX alone did not. This may be due to earlier vancomycin de-escalation. These results support the benefit of integration of AXDX into healthcare systems with an active ASP even without the resources to include real-time notification. Disclosures All authors: No reported disclosures.

Printing with light: daylight–fluorescent inks for large-gamut multi-color printing

With printing technologies continuously reaching ever higher degrees of performance and quality, the need for novelty and impact and also keeps increasing. Specialty inks have always played an important role here, albeit not without challenges. Often the use of such materials involved dedicated solutions that deal with these inks outside of the constraints of normal pipelines and workflows, which constrains their application and results in limited use. This is so since specialty inks, such as fluorescents, behave differently to traditional dye or pigment-based ones. So much so that most applications use specialty inks as spot colors, explicitly determining (by means of a separate layer in the input) where they are to be used for given content. For example, for materials such as fluorescents or quantum dots, the possibility of presenting more light at a given wavelength than is incident at that wavelength, breaks some of the basic assumptions of current processes, from how they are measured to how they can be used in building color separations all the way to how color management deals with them. In this paper we describe first experiments of using fluorescent inks that are activated by visible – instead of the more customary UV – radiation, showing performance of spectrophotometer measurement with a dedicated model to handle the specific properties of these inks, as well as results of the impact such inks can have on extending color gamuts that go significantly beyond current printing gamuts and therefore also pose new challenges.