Graph Neural Network Modelling as a potentially effective Method for predicting and analyzing Procedures based on Patient Diagnoses

BackgroundCurrently, the healthcare sector strives to increase the quality of patient management and improve the economic performance of healthcare providers. The data contained in electronic health records (EHRs) offer the potential to discover relevant patterns that aim to relate diseases and therapies, and thus discover patterns that could help identify empirical medical guidelines that reflect best practices in the healthcare system. Based on this pattern identification, it is then possible to implement recommendation systems based on the idea that a higher volume of procedures is associated with high-quality models.MethodsAlthough there are several applications that use machine learning methods to identify these patterns, this identification is still a challenge, in part because these methods often ignore the basic structure of the population, considering the similarity of diagnoses and patient typology. To this end, we have developed graph methods that aim to cluster similar patients. In such models, patients are linked when the same or similar patterns can be observed for these patients, a concept that enables the construction of a network-like structure. This structure can then be analyzed with Graph Neural Networks (GNN) to identify relevant labels, in this case the appropriate medical procedures.ResultsWe report the construction of a patient Graph structure based on basic patient’s information like age and gender as well as the diagnoses and trained GNNs models to identify the corresponding patient’s therapies using a synthetic patient database. We compared our GNN models against different baseline models (using the SCIKIT-learn library of python) and compared the performance of the different model methods. We have found that GNNs are superior, with an average improvement of the f1 score of 6.48% respect to the baseline models. In addition, the GNNs are useful for performing additional clustering analyses that allow specific identification of specific therapeutic clusters related to a particular combination of diagnoses.ConclusionsWe found that GNNs are a promising way to model the distribution of diagnoses in a patient population and thus better model how similar patients can be identified based on the combination of morbidities and comorbidities. Nevertheless, network building is still challenging and prone to prejudice, as it depends on how ICD distribution affects the patient network embedding space. This network setup requires not only a high quality of the underlying diagnostic ecosystem, but also a good understanding of how to identify related patients by disease. For this reason, additional work is needed to improve and better standardize patient embedding in graph structures for future investigations and applications of services based on this technology, and therefore is not yet an interventional study.

Society of Vascular and Interventional Neurology Standards and Parameters for Guideline Development and Publication

BACKGROUND As much of the scope of neurointerventional practice falls outside data covered by existing randomized clinical trials, and, as a result, may have failed to enter into existing guidelines, an evidence‐based framework for guideline and standards development is needed. We establish an evidence‐based framework to guide all subsequent guidelines and brief practice updates produced by the Society of Vascular and Interventional Neurology (SVIN). METHODS The SVIN formed the Guidelines and Practice Parameters committee to develop the structure and procedures for guidelines and brief practice updates. RESULTS In this article, the Guidelines and Practice Parameters committee has outlined the process by which the guidelines will be created and approved by the SVIN. Additionally, the Guidelines and Practice Parameters committee has adopted the American College of Cardiology/American Heart Association framework of Class of Recommendation and Level of Evidence. A unique, additional separation of the Expert Opinion endorsement category has been developed when high‐quality evidence does not exist at the time of the publication. CONCLUSIONS The SVIN has developed an evidence‐based framework for all guideline statements and brief practice updates. The SVIN guidelines and brief practice updates will guide clinicians in the field of interventional neurology to improve and standardize patient care.

Facilitators and Barriers to Clinical Pathway Uptake and Utilization among Primary Care Providers in Saskatchewan – A Qualitative Study

Background: Clinical Pathways (CPWs) are multidisciplinary, evidence based, complex interventions designed to standardize patient care. In Saskatchewan, development, implementation, and evaluation of seven provincial CPWs (Hip & Knee, Spine, Pelvic Floor, Prostate Assessment, Fertility Care, Lower Extremity Wound Care, and Acute Stroke) present significant challenges, leading to lower uptake and utilization. This study aimed to identify facilitators and barriers to CPW uptake and utilization by Family Physicians in Saskatchewan. Methods: A qualitative interpretive approach was used consisting of eight one-on-one key informant (KI) interviews and five focus groups (FG) in identifying the facilitators and barriers to CPWs. KIs had been involved in the design and implementation of CPWs. FGs were held with 30 Family Physicians in two urban and two rural Saskatchewan cities. All interviews were audio recorded and transcribed. Inductive, thematic analysis of the interviews based on the Theoretical Domain Framework (TDF) for behavioral changes was used to identify facilitators and barriers to CPW uptake and utilization by Family Physicians.Results: KI interviews informed the FG interview guide. From 5 FG discussions, 51 themes emerged and were mapped under 14 TDF domains. Family Physicians notably emphasized the barriers for utilizing CPWs. Major barriers were: system-level (knowledge & communication, social/professional identity, Family Physician engagement and education); objective clarification (goals, belief about consequences of implementing CPW) and technical and resource related (administrative, access to local specialists, enforcement and incentives). The most prominent barrier was lack of systematic CPW promotion and inconsistencies in communication between: organization to practitioner; organization to organization; and practitioner to practitioner. Facilitators were recognized to mitigate barriers and ranged from need for optimized and integrated IT services (i.e. Electronic Medical Records), to optimism towards CPW usage and patient outcomes. Conclusions: Informed by data from KIs and FGs, this exploratory study identified specific improvements required to promote uptake of CPWs based on perceived facilitators and barriers. Recommendations are provided to enhance uptake among FPs. These initial findings will inform the creation of a theory-based, province -wide survey instrument to further evaluate CPWs.

Navigating a Path to Equity in Cancer Care: The Role of Patient Navigation

Notable barriers exist in the delivery of equitable care for all patients with cancers. Social determinants of health at distal, intermediate, and proximal levels impact cancer care. Patient navigation is a patient-centered intervention that functions across these overlapping determinants to increase access to cancer services throughout the cancer care continuum. There is a need to standardize patient navigation training while remaining responsive to local contexts of care and a need to implement patient navigation programs with a health equity lens to address cancer care inequities.

Feasibility of Fitness Tracker Usage to Assess Activity Level and Toxicities in Patients With Colorectal Cancer

PURPOSE Performance status (PS) is a subjective assessment of patients' overall health. Quantification of physical activity using a wearable tracker (Fitbit Charge [FC]) may provide an objective measure of patient's overall PS and treatment tolerance. MATERIALS AND METHODS Patients with colorectal cancer were prospectively enrolled into two cohorts (medical and surgical) and asked to wear FC for 4 days at baseline (start of new chemotherapy [± 4 weeks] or prior to curative resection) and follow-up (4 weeks [± 2 weeks] after initial assessment in medical and postoperative discharge in surgical cohort). Primary end point was feasibility, defined as 75% of patients wearing FC for at least 12 hours/d, 3 of 4 assigned days. Mean steps per day (SPD) were correlated with toxicities of interest (postoperative complication or ≥ grade 3 toxicity). A cutoff of 5,000 SPD was selected to compare outcomes. RESULTS Eighty patients were accrued over 3 years with 55% males and a median age of 59.5 years. Feasibility end point was met with 68 patients (85%) wearing FC more than predefined duration and majority (91%) finding its use acceptable. The mean SPD count for patients with PS 0 was 6,313, and for those with PS 1, it was 2,925 (122 and 54 active minutes, respectively) ( P = .0003). Occurrence of toxicity of interest was lower among patients with SPD > 5,000 (7 of 33, 21%) compared with those with SPD < 5,000 (14 of 43, 32%), although not significant ( P = .31). CONCLUSION Assessment of physical activity with FC is feasible in patients with colorectal cancer and well-accepted. SPD may serve as an adjunct to PS assessment and a possible tool to help predict toxicities, regardless of type of therapy. Future studies incorporating FC can standardize patient assessment and help identify vulnerable population.

MO030PATIENT SAFETY IN A LARGE MULTINATIONAL RENAL SERVICES PROVIDER NETWORK

Background and Aims Patient safety is considered of paramount importance under any qualified provision of care, but results from routine tracking of incidents have scarcely been reported, even when that may negatively impact survival. To analyze all types of incidents in a multinational renal service provider network from Jan.1st to Sept. 30th, 2019. Method For the last 10 years, our institution has tracked all incidents under a structured process program, as well as, educated our staff in the importance of proactively reporting and analyzing incidents in a quarterly basis at the clinic, by country and globally. Incidents are categorized in 4 different types: A-Patient related; B-Staff and visitors; C-Products and D-Equipment. Different incident codes are assigned to each type. Results A total of 68.399 incidents (2.7 incidents/patient/year) have been reported during Q1-Q3 2019 (higher than in 2018: 2.2). This means an increase of 20% in the total number of reported incidents. Total incidents/1000 treatments was 17.1 (12.1 patient-related incidents). Reporting follows a heterogeneous pattern among countries, being lowest in Argentina and highest in the UK. Top 5 reported incidents were as follows: Codes A15 (voluntarily shortened treatment) and A14 (Patient did not show up), both related to patient adherence to treatment, accounted for 36% of total incidents, vascular access (VA) complications (A4) for 10.5%, change of dialyzer and/or blood lines due to clotting (A2) for 8.6% and recurrent minor monitor malfunction (D1) for 6.7% of incidents. Codes related with unexpected death or cardiorespiratory arrest are not present among the total global top 10 incidents. Conclusion Detailed tracking of incidents and comparison between countries have potential to increase quality of care. Room for improvement recently made the Corporate Medical Office to launch new strategies on VA management, anticoagulation and patient compliance, among others. This large series may help other institutions to better monitor and standardize patient safety.

Developing the Synergy Model Patient Acuity Tool for Admission, Discharge, and Transfer

Background The Admission Discharge Transfer–Synergy Model Acuity Tool (ADT-SMAT) was developed to quantify patient intervention intensity and patient response variability and to capture nurses’ critical thinking. The tool is based on the American Association of Critical-Care Nurses Synergy Model for Patient Care. Objective To determine whether the ADT-SMAT is reliable and valid for predicting the level of care for admission, discharge, and transfer of critically ill patients. Methods Reliability was examined by using interrater reliability, intraclass coefficient, and effect size analyses to evaluate physiological variables and total calculated ADT-SMAT score in 246 patients. Content validity was determined in consultation with critical care nurses, and construct validity was examined by assessing the correlation between ADT-SMAT scores and other convergent and divergent constructs. Results The ADT-SMAT showed strong reliability for measuring the physiological variables and total score, with an intraclass coefficient of 0.930. The value of Cohen d determining the effect size for each element of the ADT-SMAT was less than 0.20 for every element, indicating that substantial differences in scoring did not occur. The validity of the ADT-SMAT requires additional testing. Conclusions This is the first study attempting to correlate Synergy Model patient characteristics and acuity while integrating nurses’ critical decision-making process. With further testing, the ADT-SMAT could be a valuable tool to quantify and standardize patient characteristics in determining the appropriate level of care associated with admission, discharge, and transfer decisions.

Developing an emergency medicine handoff tool: an electronic Delphi approach

Background Handoffs at the end of clinical shifts occur with high frequencies in the emergency department setting and they pose an increased risk to patients. There is a need to standardize handoff practices. This study aimed to use an electronic Delphi method to identify the core elements essential for an emergency department physician to physician handoff and propose a framework for implementation. Methods An electronic Delphi-style study with a national panel of board-certified emergency physicians in Saudi Arabia. The panel was conducted over four rounds. The first to identify elements relevant to the end of shift handoff and categorize them into domains, while the remaining three to score and debate individual elements. Results Twenty-five board-certified emergency physicians from various cities and practice settings were enrolled. All panelists completed the entire Delphi process. Thirty-two elements were identified and classified into 4 domains. The top five rated handoff elements were patient identification, chief complaint history, clinical stability, working diagnosis, and consulting services involved. Panel scores showed convergence as rounds progressed and the final list of elements had a high-reliability score (Cronbach’s alpha 0.93). Conclusions This study yielded an itemized and ranked list of elements that are easy to implement and could be used to standardize patient handoffs by emergency physicians. While this study was conducted on an emergency medicine panel, the methods used may be adapted to develop standardized handoff frameworks that serve different disciplines or practice settings.