Machine Learning-Based Genome-Wide Salivary DNA Methylation Analysis for Identification of Noninvasive Biomarkers in Oral Cancer Diagnosis.

This study aims to examine the feasibility of ML-assisted salivary-liquid-biopsy platforms using genome-wide methylation analysis at the base-pair and regional resolution for delineating oral squamous cell carcinoma (OSCC) and oral potentially malignant disorders (OPMDs). A nested cohort of patients with OSCC and OPMDs was randomly selected from among patients with oral mucosal diseases. Saliva samples were collected, and DNA extracted from cell pellets was processed for reduced-representation bisulfite sequencing. Reads with a minimum of 10× coverage were used to identify differentially methylated CpG sites (DMCs) and 100 bp regions (DMRs). The performance of eight ML models and three feature-selection methods (ANOVA, MRMR, and LASSO) were then compared to determine the optimal biomarker models based on DMCs and DMRs. A total of 1745 DMCs and 105 DMRs were identified for detecting OSCC. The proportion of hypomethylated and hypermethylated DMCs was similar (51% vs. 49%), while most DMRs were hypermethylated (62.9%). Furthermore, more DMRs than DMCs were annotated to promoter regions (36% vs. 16%) and more DMCs than DMRs were annotated to intergenic regions (50% vs. 36%). Of all the ML models compared, the linear SVM model based on 11 optimal DMRs selected by LASSO had a perfect AUC, recall, specificity, and calibration (1.00) for OSCC detection. Overall, genome-wide DNA methylation techniques can be applied directly to saliva samples for biomarker discovery and ML-based platforms may be useful in stratifying OSCC during disease screening and monitoring.

Mock clinical testing in the validation of fluid-phase biomarkers for head and neck carcinoma diagnosis: Scoping review.

This review sought to determine the range and nature of prospective-sampling and blinding methods for validating nonviral biofluid markers diagnostic of head and neck carcinomas. Electronic database searching was conducted to identify studies published in English from January 1, 2009 to August 1, 2020. Sixteen studies from 17 articles published between 2011 and 2020 were included in this review. We found that about 3 out of 100 studies utilized at least one of the mock testing approaches for biomarker validation. Protein, mRNA, and metabolomic markers also represented the only groups whose validation has been attempted using these methods. Furthermore, studies that utilized both methods were found to have lower bias concerns on the quality assessment of diagnostic accuracy studies (QUADAS-2) tool. Overall, there is a need to include these protocols in research endeavours verifying diagnostic biomarkers for head and neck carcinomas following the preliminary establishment of their classification accuracy.

Engaging Frontline Leaders and Staff in Real-Time Improvement.

BACKGROUND: The relationship of staff satisfaction and engagement to organizational success, along with the integral influence of frontline managers on this dimension, is well established in health care and other industries. To specifically address staff engagement, Virginia Mason Medical Center, an integrated, single-hospital health system, developed an approach that involved leaders, through the daily use of standard work for leaders, as well as staff, through a Lean-inspired staff idea system. METHODS: Kaizen Promotion Office (KPO) staff members established three guiding principles: (1) Staff engagement begins with leader engagement; (2) Integrate daily improve- ment (kaizen) as a habitual way of life not as an add-on; and (3) Create an environment in which staff feel psycho- logically safe and valued. Two design elements--Standard Work for Leaders (SWL) and Everyday Lean Ideas (ELIs) were implemented. RESULTS: For the emergency department (ED), an early adopter of the staff engagement work, the challenge was to apply the guiding principles to improve staff engagement while improving quality and patient and staff satisfaction, even as patient volumes were increasing. Daily huddles for the KPO staff members and weekly leader rounds are used to elicit staff ideas and foster ELIs in real time. Overall progress to date has been tracked in terms of staff satisfaction surveys, voluntary staff turnover, adoption of SWL, and testing and implementation of staff ideas. For example, voluntary turnover of ED staff decreased from 14.6% in 2011 to 7.5% in 2012, and 2.0% in 2013. Organizationwide, at least 800 staff ideas are in motion at any given time, with finished ones posted in an idea supermarket website. CONCLUSION: A leadership and staff engagement approach that focuses on SWL and on capturing staff ideas for daily problem solving and improvement can contribute to organization success and improve the quality of health care delivery.

Using Lean to Advance Quality Improvement Research.

INTRODUCTION: Quality improvement research skills are not commonplace among quality improvement practitioners, and research on the effectiveness of quality improvement has not always kept pace with improvement innovation. However, the Lean tools applied to quality improvement should be equally relevant to the advancement of quality improvement research. METHODS: We applied the Lean methods to develop a simplified quality improvement publication pathway enabling a small research methodology group to increase quality improvement research throughout the institution. The key innovations of the pathway are horizontal integration of the quality improvement research methods group across the institution, implementation of a Lean quality improvement research pathway, and application of a just-in-time quality improvement research toolkit. RESULTS: This work provides a road map and tools for the acceleration of quality improvement research. At our institution, the Lean quality improvement research approach was associated with statistically significant increases in the number (annual mean increase from 3.0 to 8.5, p = .03) and breadth of published quality improvement research articles, and in the number of quality improvement research projects currently in process. DISCUSSION: Application of Lean methods to the quality improvement research process can aid in increasing publication of quality improvement articles from across the institution.

Implementation of a custom alert to prevent medication-timing errors associated with computerized prescriber order entry.

PURPOSE: The implementation of a custom alert to prevent medication-timing errors associated with the use of a computerized prescriber order-entry (CPOE) system is described. METHODS: In early 2013, Virginia Mason Medical Center began work to modify the CPOE system to make it impossible for a medication-timing error to occur. A visual control, a custom alert that forced prescribers to "self-inspect with pause" (a soft-stop override) before signing an order that may result in a medication-timing error, was developed. The prescriber could choose to modify the order to skip the next dose or change the start date or time of the order, cancel the order, or continue without altering the order. The custom alert was designed to fire only for medications that are administered once or twice daily. The effectiveness of the intervention was evaluated with interrupted time series before-and-after analysis of medication-timing errors. Outcomes measured included the proportion of errors that were corrected by the prescriber before signing the order and the number of medication-timing errors that reached patients. RESULTS: The proportion of orders where a prescriber modified the order as a result of the alert increased from 12% before the intervention to 29% postintervention (p > 0.001), and this percentage continued to increase at five months postintervention (p < 0.001). The percentage of medication-timing errors that reached the patient decreased by 50% in the postintervention period. CONCLUSION: Medication-timing errors associated with the use of a CPOE system were corrected through implementation of a customized alert.

Using lean "automation with a human touch" to improve medication safety: a step closer to the "perfect dose".

BACKGROUND: Virginia Mason Medical Center (Seattle) employed the Lean concept of Jidoka (automation with a human touch) to plan for and deploy bar code medication administration (BCMA) to hospitalized patients. METHODS: Integrating BCMA technology into the nursing work flow with minimal disruption was accomplished using three steps ofJidoka: (1) assigning work to humans and machines on the basis of their differing abilities, (2) adapting machines to the human work flow, and (3) monitoring the human-machine interaction. Effectiveness of BCMA to both reinforce safe administration practices and reduce medication errors was measured using the Collaborative Alliance for Nursing Outcomes (CALNOC) Medication Administration Accuracy Quality Study methodology. Trained nurses observed a total of 16,149 medication doses for 3,617 patients in a three-year period. RESULTS: Following BCMA implementation, the number of safe practice violations decreased from 54.8 violations/100 doses (January 2010-September 2011) to 29.0 violations/100 doses (October 2011-December 2012), resulting in an absolute risk reduction of 25.8 violations/100 doses (95% confidence interval [CI]: 23.7, 27.9, p < .001). The number of medication errors decreased from 5.9 errors/100 doses at baseline to 3.0 errors/100 doses after BCMA implementation (absolute risk reduction: 2.9 errors/100 doses [95% CI: 2.2, 3.6,p < .001]). The number of unsafe administration practices (estimate, -5.481; standard error 1.133; p < .001; 95% CI: -7.702, -3.260) also decreased. CONCLUSION: As more hospitals respond to health information technology meaningful use incentives, thoughtful, methodical, and well-managed approaches to technology deployment are crucial. This work illustrates how Jidoka offers opportunities for a smooth transition to new technology.

Why Lean doesn't work for everyone.

Popularisation of Lean in healthcare has led to emphasis on Lean quality improvement tools in isolation, with inconsistent results. We argue that delivery of safer, more efficient, and higher quality-patient focused care requires organisational transformation of which the Lean toolkit is only one component. To successfully facilitate system transformation toward higher quality care at lower cost, Lean tools must be part of a comprehensive management system, within a supportive institutional culture, and with committed leadership.

Opening the ICU doors.

Family member presence may contribute to the healing of hospitalized patients, but may also be in conflict with the perceived needs of delivering intensive care. We detail our experience with "opening the doors" of the intensive care unit (ICU), allowing family members to be present and participate in the care of loved ones without restriction. "Opening the doors" challenged the traditions, legacy and sense of professional entitlement that were a part of ICU culture and generated considerable initial resistance among nurses and physicians. We describe our "opening the doors" transformation to more patient- and family-centered care in four steps: (1) enlist support of administrative and local leaders; (2) create a collective aim; (3) test on a small scale, and (4) scale up after initial successes. Preparing ICU staff so that they are comfortable with more "on stage" time (i.e., greater family presence) was critical to our success. "Opening the doors" now serves as a guiding vision to organizing the ICU's work.

Using lean to improve medication administration safety: in search of the "perfect dose".

BACKGROUND: At Virginia Mason Medical Center (Seattle), the Collaborative Alliance for Nursing Outcomes (CALNOC) Medication Administration Accuracy Quality Study was used in combination with Lean quality improvement efforts to address medication administration safety. METHODS: Lean interventions were targeted at improving the medication room layout, applying visual controls, and implementing nursing standard work. The interventions were designed to prevent medication administration errors through improving six safe practices: (1) comparing medication with medication administration record, (2) labeling medication, (3) checking two forms of patient identification, (4) explaining medication to patient, (5) charting medication immediately, and (6) protecting the process from distractions/interruptions. RESULTS: Trained nurse auditors observed 9,244 doses for 2,139 patients. Following the intervention, the number of safe-practice violations decreased from 83 violations/100 doses at baseline (January 2010-March 2010) to 42 violations/100 doses at final follow-up (July 2011-September 2011), resulting in an absolute risk reduction of 42 violations/100 doses (95% confidence interval [CI]: 35-48), p < .001). The number of medication administration errors decreased from 10.3 errors/100 doses at baseline to 2.8 errors/100 doses at final follow-up (absolute risk reduction: 7 violations/100 doses [95% CI: 5-10, p < .001]). The "perfect dose" score, reflecting compliance with all six safe practices and absence of any of the eight medication administration errors, improved from 37 in compliance/100 doses at baseline to 68 in compliance/100 doses at the final follow-up. CONCLUSION: Lean process improvements coupled with direct observation can contribute to substantial decreases in errors in nursing medication administration.