Karim Keshavjee

ORCID: 0000-0003-1317-7035
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Electronic Health Records Systems
  • Mobile Health and mHealth Applications
  • Artificial Intelligence in Healthcare
  • Diabetes, Cardiovascular Risks, and Lipoproteins
  • Machine Learning in Healthcare
  • Diabetes Management and Research
  • Clinical practice guidelines implementation
  • Diabetes Management and Education
  • Digital Mental Health Interventions
  • Data Quality and Management
  • Healthcare Systems and Technology
  • Heart Failure Treatment and Management
  • Telemedicine and Telehealth Implementation
  • Primary Care and Health Outcomes
  • Ethics in Clinical Research
  • Medication Adherence and Compliance
  • Chronic Disease Management Strategies
  • Healthcare Policy and Management
  • Big Data and Business Intelligence
  • Health Policy Implementation Science
  • Diabetes Treatment and Management
  • Pharmaceutical Practices and Patient Outcomes
  • Health Literacy and Information Accessibility
  • Health Systems, Economic Evaluations, Quality of Life
  • Business Process Modeling and Analysis

University of Toronto
2015-2025

Public Health Ontario
2018-2025

Toronto Metropolitan University
2016-2025

Institute for Work & Health
2024-2025

Institut d'Histoire et de Philosophie des Sciences et des Techniques
2025

Social Research Association
2024

Institute of Health Services and Policy Research
2024

Canada Health Infoway
2003-2023

Concordia University
2018-2023

University of Ottawa
2023

Diabetes Mellitus is one of the major health challenges all over world. The prevalence diabetes increasing at a fast pace, deteriorating human, economic and social fabric. Prevention prediction mellitus increasingly gaining interest in healthcare community. Although several clinical decision support systems have been proposed that incorporate data mining techniques for course progression. These conventional are typically based either just on single classifier or plain combination thereof....

10.1016/j.procs.2016.04.016 article EN Procedia Computer Science 2016-01-01

Diabetes Mellitus is an increasingly prevalent chronic disease characterized by the body's inability to metabolize glucose. The objective of this study was build effective predictive model with high sensitivity and selectivity better identify Canadian patients at risk having based on patient demographic data laboratory results during their visits medical facilities.Using most recent records 13,309 aged between 18 90 years, along information (age, sex, fasting blood glucose, body mass index,...

10.1186/s12902-019-0436-6 article EN cc-by BMC Endocrine Disorders 2019-10-15

<h3>ABSTRACT</h3> Recent studies indicate that U.S. consumers9 feminist attitudes can influence positively their evaluations of sexual images women in advertisements. This study tested the generalizability finding Korea—a country with a much different cultural background. Employing Feminist Perspectives Scale, which provides more complex view attitudes, results reaffirm positive effect general but also show consumers holding conservative and liberal perspectives evaluated favorably—and...

10.1503/cmaj.081272 article EN cc-by-nc-nd Canadian Medical Association Journal 2009-07-06

The development of a pan-Canadian network primary care research networks for studying issues in has been the vision Canadian researchers many years. With opportunity funding from Public Health Agency Canada and support College Family Physicians Canada, we have planned developed project to assess feasibility family medicine practices that exclusively use electronic medical records. Primary Care Sentinel Surveillance Network will collect longitudinal data across epidemiology management 5...

10.3122/jabfm.2009.04.090081 article EN The Journal of the American Board of Family Medicine 2009-07-01

This article proposes a Clinical Adoption Framework for making sense of health information system (HIS) success in Canada. It extends Canada Health Infoway's Benefits Evaluation with contextual factors that influence HIS adoption by clinicians, which include people, organization, implementation, and the macro environment. Our hypothesis is successful clinical an requires explicit recognition, strategies actions address described framework. Validation this framework stakeholders literature...

10.12927/hcq.2011.22157 article EN Healthcare Quarterly 2011-01-27

The objective of this inductive research was to investigate: 1) the relationship between diabetes mellitus and individual risk factors metabolic syndrome (MetS), in a non-conservative setting; 2) prediction future onset using relevant MetS; 3) investigate relative performance machine learning methods when data sampling techniques are used generate balanced training sets. dataset contains 667 907 records for period ranging from 2003 2013. Quantifying contribution MetS development setting...

10.1109/access.2018.2884249 article EN cc-by-nc-nd IEEE Access 2018-12-21

Prevention and diagnosis of NAFLD is an ongoing area interest in the healthcare community. Screening complicated by fact that accuracy noninvasive testing lacks specificity sensitivity to make stage diagnosis. Currently no non-invasive ATP III criteria based prediction method available diagnose risk. Firstly, objective this research develop machine learning order identify individuals at increased risk developing using factors clinical updated 2005 for Metabolic Syndrome (MetS). Secondly,...

10.1038/s41598-018-20166-x article EN cc-by Scientific Reports 2018-01-26

Health information technology implementations frequently fail despite extensive research on success factors over the past three decades. This paper introduces Playing-to-Win Digital Strategy Canvas, an adaptation of Martin and Lafley’s framework, tailored for healthcare. The canvas integrates business strategy principles with evidence-based insights to address unique challenges in digital health implementation. Key elements include prioritizing high-risk populations, co-designing solutions...

10.3233/shti250027 article EN cc-by-nc Studies in health technology and informatics 2025-02-18

Clinical researchers use prognostic modeling techniques to identify a-prior patient health status and characterize progression patterns. It is highly desirable predict future condition especially implement preventive intervention strategies in pre-diabetic individuals. Hidden Markov Model (HMM) its variants are a class of models that provide predictions concerning by exploiting sequences clinical measurements obtained from longitudinal sample patients. Despite the advantages using these for...

10.1109/access.2020.2968608 article EN cc-by IEEE Access 2020-01-01

Abstract Stratifying individuals at risk for developing diabetes could enable targeted delivery of interventional programs to those highest risk, while avoiding the effort and costs prevention treatment in low risk. The objective this study was explore potential role a Hidden Markov Model (HMM), machine learning technique, validating performance Framingham Diabetes Risk Scoring (FDRSM), well-respected prognostic model. Can HMM predict 8-year an individual effectively? To our knowledge, no...

10.1038/s41598-019-49563-6 article EN cc-by Scientific Reports 2019-09-24

Abstract Depression is disproportionately prevalent among individuals with diabetes compared to the general populace, underscoring critical need for predictive mechanisms that can facilitate timely interventions and support. This study explores use of machine learning forecast depression in those at risk or diagnosed diabetes, leveraging extensive primary care data from Canadian Primary Care Sentinel Surveillance Network. Six models including Logistic Regression, Random Forest, AdaBoost,...

10.1101/2024.02.03.24302303 preprint EN cc-by-nc-nd medRxiv (Cold Spring Harbor Laboratory) 2024-02-07

<h3>Background</h3>Computerized decision support systems (CDSSs) linked with electronic medical records (EMRs) are promoted as an effective means of improving patient care. However, very few high-quality studies set in routine, community-based clinical care, and no consistent evidence effect on outcomes has been found.<h3>Methods</h3>A randomized controlled trial among EMR-using primary care practices Ontario, Canada. Patients 55 years or older previous vascular events, diabetes mellitus,...

10.1001/archinternmed.2011.471 article EN Archives of Internal Medicine 2011-10-24

This study evaluates mobile apps using a theory-based evaluation framework to discover their applicability for patients at risk of gestational diabetes. assessed how well the existing on market meet information and tracking needs with diabetes evaluated feasibility integrate these into patient care. A search was conducted in Apple iTunes Google Play store that contained keywords related following concepts nutrition: diet, tracking, diabetes, pregnancy. Evaluation criteria were developed...

10.1177/1460458219896639 article EN cc-by-nc Health Informatics Journal 2020-01-08

Prediabetes is a critical health condition characterized by elevated blood glucose levels that fall below the threshold for Type 2 diabetes (T2D) diagnosis. Accurate identification of prediabetes essential to forestall progression T2D among at-risk individuals. This study aims pinpoint most effective machine learning (ML) model prediction and elucidate key biological variables distinguishing individuals with prediabetes. Utilizing data from Canadian Primary Care Sentinel Surveillance Network...

10.1101/2024.02.03.24302301 preprint EN cc-by-nc-nd medRxiv (Cold Spring Harbor Laboratory) 2024-02-05

Abstract Type 2 Diabetes Mellitus (T2DM) is a chronic metabolic disorder with increasing population incidence. However, T2DM takes years to develop, allowing onset prediction and prevention be clinically effective treatment strategy. In this study we propose assess novel approach diabetes which integrates specialized extension of the random forest algorithm known as survival (RSF). Rather than predicting binary outcome, machine learning model incorporates analysis methodology predict time...

10.1101/2024.02.03.24302304 preprint EN cc-by-nc-nd medRxiv (Cold Spring Harbor Laboratory) 2024-02-07

Despite the growing availability of artificial intelligence models for predicting type 2 diabetes, there is still a lack personalized approaches to quantify minimum viable changes in biomarkers that may help reduce individual risk developing disease. The aim this article develop new method, based on counterfactual explanations, generate recommendations one-year diabetes. Ten routinely collected extracted from Electronic Medical Records 2791 patients at low and high diabetes were analyzed....

10.1371/journal.pone.0272825 article EN cc-by PLoS ONE 2022-11-17

Diabetes mellitus type 2 is increasingly being called a modern preventable pandemic, as even with excellent available treatments, the rate of complications diabetes rapidly increasing. Predicting and identifying it in its early stages could make easier to prevent, allowing enough time implement therapies before gets out control. Leveraging longitudinal electronic medical record (EMR) data deep learning has great potential for prediction. This paper examines predictive competency models...

10.1371/journal.pdig.0000354 article EN cc-by PLOS Digital Health 2023-10-25
Coming Soon ...