The Evaluation of a Clinical Decision Support Tool Using Natural Language Processing to Screen Hospitalized Adults for Unhealthy Substance Use: Protocol for a Quasi-Experimental Design.

BACKGROUND: Automated and data-driven methods for screening using natural language processing (NLP) and machine learning may replace resource-intensive manual approaches in the usual care of patients hospitalized with conditions related to unhealthy substance use. The rigorous evaluation of tools that use artificial intelligence (AI) is necessary to demonstrate effectiveness before system-wide implementation. An NLP tool to use routinely collected data in the electronic health record was previously validated for diagnostic accuracy in a retrospective study for screening unhealthy substance use. Our next step is a noninferiority design incorporated into a research protocol for clinical implementation with prospective evaluation of clinical effectiveness in a large health system. OBJECTIVE: This study aims to provide a study protocol to evaluate health outcomes and the costs and benefits of an AI-driven automated screener compared to manual human screening for unhealthy substance use. METHODS: A pre-post design is proposed to evaluate 12 months of manual screening followed by 12 months of automated screening across surgical and medical wards at a single medical center. The preintervention period consists of usual care with manual screening by nurses and social workers and referrals to a multidisciplinary Substance Use Intervention Team (SUIT). Facilitated by a NLP pipeline in the postintervention period, clinical notes from the first 24 hours of hospitalization will be processed and scored by a machine learning model, and the SUIT will be similarly alerted to patients who flagged positive for substance misuse. Flowsheets within the electronic health record have been updated to capture rates of interventions for the primary outcome (brief intervention/motivational interviewing, medication-assisted treatment, naloxone dispensing, and referral to outpatient care). Effectiveness in terms of patient outcomes will be determined by noninferior rates of interventions (primary outcome), as well as rates of readmission within 6 months, average time to consult, and discharge rates against medical advice (secondary outcomes) in the postintervention period by a SUIT compared to the preintervention period. A separate analysis will be performed to assess the costs and benefits to the health system by using automated screening. Changes from the pre- to postintervention period will be assessed in covariate-adjusted generalized linear mixed-effects models. RESULTS: The study will begin in September 2022. Monthly data monitoring and Data Safety Monitoring Board reporting are scheduled every 6 months throughout the study period. We anticipate reporting final results by June 2025. CONCLUSIONS: The use of augmented intelligence for clinical decision support is growing with an increasing number of AI tools. We provide a research protocol for prospective evaluation of an automated NLP system for screening unhealthy substance use using a noninferiority design to demonstrate comprehensive screening that may be as effective as manual screening but less costly via automated solutions. TRIAL REGISTRATION: ClinicalTrials.gov NCT03833804; https://clinicaltrials.gov/ct2/show/NCT03833804. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/42971.

Patients' and clinicians' perceptions of oral anticoagulants in atrial fibrillation: a systematic narrative review and meta-analysis.

BACKGROUND: Atrial fibrillation (AF) increases the risk of developing a stroke by 20%. AF related strokes are associated with greater morbidity. Historically, warfarin was the anticoagulant of choice for stroke prevention in patients with AF but lately patients are being switched or started on direct oral anticoagulants (DOACs). DOACs are promoted as safer alternatives to warfarin and it is expected that they will be associated with fewer challenges both for patients and healthcare professionals. This systematic narrative review aimed to explore perspectives of patients and professionals on medicines optimisation of oral anticoagulation with vitamin K antagonists and DOACs in atrial fibrillation. METHODS: Prospero registration CRD42018091591. Systematic searches undertaken of research studies (qualitative and quantitative), published February 2018 to November 2020 from several databases (Web of Science, Scopus, Medline Via Ovid, CINHAL via Ebsco, and PubMED via NCBI) following PRISMA methodology. Data were organised using Covidence software. Two reviewers independently assessed the quality of the included studies and synthesized the findings (thematic analysis approach). RESULTS: Thirty-four studies were included. Studies were critically appraised using established critical appraisal tools (Qualsyst) and a risk of bias was assigned. Clinicians considered old age and the associated complexities such as co-morbidities and the increased potential for bleeding as potential barriers to optimising anticoagulation. Whereas patients' health and medication beliefs influenced adherence. Notably, structured patient support was important in enhancing safety and effective anticoagulation. For both patients and clinicians, confidence and experience of safe anticoagulation was influenced by the presence of co-morbidities,  poor knowledge and understanding of AF and the purpose of anticoagulation. CONCLUSION: Age, complex multimorbidity and polypharmacy influence prescribing, with DOACs being perceived to be safer than warfarin. This systematic narrative review suggests that interventions are needed to support patient self-management. There are residual anxieties associated with long term anticoagulation in the context of complexities. TRIAL REGISTRATION: Not applicable.

Dialysis, COVID-19, Poverty, and Race in Greater Chicago: An Ecological Analysis.

RATIONALE & OBJECTIVE: Persons with end-stage kidney disease receiving in-center maintenance hemodialysis may be at high risk for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) exposure and severe outcomes with coronavirus disease 2019 (COVID-19). The objective of this study was to examine the correlation of SARS-CoV-2 positivity rate per capita and COVID-19-associated deaths with number of dialysis stations and demographics of residents within zip codes in Cook County, IL. STUDY DESIGN: Ecological analysis. SETTING & PARTICIPANTS: Data for SARS-CoV-2 test results and COVID-19-associated deaths during January 21 to June 15, 2020, among the 5,232,412 residents living within the 163 zip codes in Cook County, IL, were merged with demographic and income data from the US Census Bureau. The total number of positive test results in this population was 84,353 and total number of deaths was 4,007. ASSESSMENTS: Number of dialysis stations and stations per capita within a zip code were calculated. SARS-CoV-2-positive test results per capita were calculated as number of positive test results divided by the zip code population. COVID-19-associated deaths per capita were calculated as COVID-19 deaths among residents for a given zip code divided by the zip code population. ANALYTIC APPROACH: Spearman rank correlation coefficients were calculated to examine the correlation of SARS-CoV-2-positive tests per capita and COVID-19-associated deaths per capita with dialysis stations, demographics, and household poverty. To account for multiple testing, statistical significance was considered as P < 0.005. RESULTS: Among the 163 Cook County zip codes, there were 2,501 dialysis stations. Positive test results per capita were significantly associated with number of dialysis stations (r = 0.25; 95% CI, 0.19 to 0.29; P < 0.005) but not with dialysis stations per capita (r = 0.02; 95% CI, -0.03 to 0.08; P = 0.7). Positive test results per capita also correlated significantly with number of households living in poverty (r = 0.57; 95% CI, 0.53-0.6; P < 0.005) and percentage of residents reporting Black race (r = 0.28; 95% CI, 0.23-0.33; P < 0.005) and Hispanic ethnicity (r = 0.68; 95% CI, 0.65-0.7; P < 0.001;). COVID-19-associated deaths per capita correlated significantly with the percentage of residents reporting Black race (r = 0.24; 95% CI, 0.19-0.29; P < 0.005) and with percentage of households living in poverty (r = 0.34; 95% CI, 0.29-0.38; P < 0.005). The association between the number of COVID-19-associated deaths per capita and total number of dialysis stations (r = 0.20; 95% CI, 0.14-0.25; P = 0.01) did not achieve a priori significance, whereas the association with dialysis stations per capita (r = 0.12; 95% CI, 0.07-0.17; P = 0.01) was not significant. LIMITATIONS: Analysis is at the zip code level and not at the person level. CONCLUSIONS: The number of dialysis stations within a zip code correlates with the SARS-CoV-2 positivity rate per capita in Cook County, IL, and this correlation may be driven by population density and the demographics of the residents. These findings highlight the high risk of SARS-CoV-2 exposure for patients with end-stage kidney disease living in poor urban areas.