- 60
- 122 624
Population Data BC
เข้าร่วมเมื่อ 3 ก.พ. 2015
Population Data BC (PopData) is a multi-university, data and education resource facilitating interdisciplinary research on the determinants of human health, well-being and development.
We offer researchers access to one of the world’s largest collections of health care, health services and population health data, and a comprehensive education and training service on how to best use those data.
Linkage of data across sectors, such as health, education, early childhood development, workplace and the environment, facilitates advances in understanding the complex interplay of influences on human health, well-being and development.
The videos in this channel describe what we do and how researchers have used linked data to answer questions about the health and well-being of British Columbians. There are also instructional videos from our Education and Training Unit and presentations from conferences and other events.
We offer researchers access to one of the world’s largest collections of health care, health services and population health data, and a comprehensive education and training service on how to best use those data.
Linkage of data across sectors, such as health, education, early childhood development, workplace and the environment, facilitates advances in understanding the complex interplay of influences on human health, well-being and development.
The videos in this channel describe what we do and how researchers have used linked data to answer questions about the health and well-being of British Columbians. There are also instructional videos from our Education and Training Unit and presentations from conferences and other events.
Review of new health core tables for Discharge Abstracts Database and NACRS data sets
Review of new health core tables for Discharge Abstracts Database and NACRS data sets
มุมมอง: 57
วีดีโอ
Review of new health core tables for Medical Services Plan data set
มุมมอง 343 หลายเดือนก่อน
Review of new health core tables for Medical Services Plan data set
Identification algorithms and administrative health data
มุมมอง 5102 ปีที่แล้ว
This webinar focuses on identification algorithms and related considerations within the context of epidemiological research leveraging linked multi-source administrative health data in Canada.
Closing the Loop: From system-based data to evidence influenced policy and practice
มุมมอง 1952 ปีที่แล้ว
The Manitoba Centre for Health Policy (MCHP) has a long and successful history of leveraging the whole-population administrative data to generate evidence to answer real-world policy questions. This webinar will provide an overview of the Manitoba Population Research Data Repository at the Manitoba Centre for Health Policy, the methodologies MCHP uses to generate evidence, and the partnership-b...
Regression Discontinuity Design
มุมมอง 1.5K2 ปีที่แล้ว
Social policy and clinical care are filled with thresholds-people receive or don’t receive interventions or treatments based on factors such as income, test results, year of birth, and others. These thresholds are a potential source of quasi-randomization for strong observational studies. Regression discontinuity designs are a quasi-experimental research design that utilize these thresholds to ...
Record Linkage Methodology Under Fellegi-Sunter Paradigm, with Extensions
มุมมอง 2.1K2 ปีที่แล้ว
In 1969, Ivan Fellegi and Alan Sunter formalized a strategy for conducting probabilistic record linkage that had been developed previously. Included in this formalization was the demonstration that the scoring method used with this is optimal under certain assumptions. While other record linkage methods have been developed (including Bayesian-based ones) for large-scale linkages the Fellegi-Sun...
Using administrative data to support Ontario's COVID-19 response
มุมมอง 1662 ปีที่แล้ว
Throughout the COVID-19 pandemic, real-time data and analytics have been important to both understanding the virus’ impact on populations and health systems and informing the public health response. Using existing research infrastructure and public health expertise, ICES was well-positioned to provide both rapid, ad hoc reports and routine reporting and analytics to policymakers and Public Heal...
Splink: a software package for probabilistic record linkage and deduplication at scale
มุมมอง 15K2 ปีที่แล้ว
In this seminar, we will introduce Splink, a software package developed for probabilistic record linkage at scale. This is free software provides a toolkit for record linkage of datasets of tens or even hundreds of millions of records, guiding the user through the various stages of linkage
Sensing pedestrian flows for assessment of non-pharma policy interventions during COVID-19 pandemic
มุมมอง 4362 ปีที่แล้ว
Reducing the number of social contacts within a population has shown to be an effective measure to reduce the transmission of the SARS-CoV-2 virus. Some prominent non-pharmaceutical interventions (NPIs) have been, e.g., stay-at-home orders or closing schools/businesses. To assess changes in spatial mobility of the German population, sensor data of pedestrian flows in 49 metropolitan areas at 10...
Survival Analysis, Censoring and Time Scales
มุมมอง 2.7K2 ปีที่แล้ว
Survival analysis or time-to-event analyses have been popularized for making predictions about future events based on some exposure in the past. The methods are familiar to epidemiologists but also to actuarial scientists / insurance companies to estimate risk. This presentation will give a quick overview of Cox proportional hazard models, covering the assumption of independence of censoring, a...
Introduction to Statistical Disease Cluster Detection with Health Administrative Data
มุมมอง 8092 ปีที่แล้ว
Disease processes have patterns and when disease varies in space, spatial patterns are produced. Identifying and quantifying patterns of disease in geographic areas may lead to a better understanding of why a disease occurs. This webinar will focus on some statistical cluster detection tests that can identify geographic areas with a higher prevalence of disease than expected.
Linking Education and Hospital Data in England
มุมมอง 1752 ปีที่แล้ว
The educational and health records for nearly all (>95%) school children in England have for the first time, been successfully linked to create an anonymised, securely held database for research. This ECHILD (Education and Child Health Insights from Linked Data) database will enable a significant improvement in the scale and depth of research into the relationships between health, education and...
Planting the S.E.E.D.S of Indigenous Population Health Data Linkage
มุมมอง 1613 ปีที่แล้ว
The increasing accessibility of data through digitization and linkage has resulted in Indigenous and allied individuals, scholars, practitioners, and data users recognizing a need to advance ways that assert Indigenous sovereignty and governance within data environments. Advances are being talked about around the world for how Indigenous data is collected, used, stored, shared, linked, and anal...
Introduction to The Data Innovation Program for Academic Researchers
มุมมอง 2733 ปีที่แล้ว
The Government of British Columbia, Ministry of Citizens' Services, in partnership with Population Data BC, recently launched the Data Innovation Program for academic researchers, allowing access to cross-sector data from multiple provincial ministries and organizations for the first time. The Data Innovation Program (DI Program) is a data integration and analytics program for government analys...
Air Pollution, housing & respiratory tract Infections in Children: National birth Cohort study
มุมมอง 1693 ปีที่แล้ว
In this talk we will present the PICNIC study. The aim of PICNIC is to quantify the contribution of in-utero, infant and childhood exposure to ambient air pollution and adverse housing conditions that are associated with the risk of RTI admissions in children less than 5 years old. We will use administrative data from national birth cohorts of all children born in England between 2005 and 2014 ...
Creation of the first national linked colorectal cancer dataset in Scotland: lessons learned
มุมมอง 2533 ปีที่แล้ว
Creation of the first national linked colorectal cancer dataset in Scotland: lessons learned
Use of Causal Diagrams in Variable Selection for Causal Observational Studies
มุมมอง 1.6K3 ปีที่แล้ว
Use of Causal Diagrams in Variable Selection for Causal Observational Studies
Development of a prognostic prediction model to estimate the risk of multiple chronic diseases
มุมมอง 1K3 ปีที่แล้ว
Development of a prognostic prediction model to estimate the risk of multiple chronic diseases
Enabling Insight: Tools for Exploration and Data Quality Assessment of Administrative Data Files
มุมมอง 3483 ปีที่แล้ว
Enabling Insight: Tools for Exploration and Data Quality Assessment of Administrative Data Files
Developing a Data Integrated COVID-19 Tracking System for Decision-Making and Public Use
มุมมอง 2613 ปีที่แล้ว
Developing a Data Integrated COVID-19 Tracking System for Decision-Making and Public Use
Interrupted Time Series and its applications
มุมมอง 10K3 ปีที่แล้ว
Interrupted Time Series and its applications
Unlocking the Potential of Electronic Health Records for Health Research
มุมมอง 6994 ปีที่แล้ว
Unlocking the Potential of Electronic Health Records for Health Research
Why the Public Needs a Say in How Patient Data are Used for Covid 19 Responses
มุมมอง 454 ปีที่แล้ว
Why the Public Needs a Say in How Patient Data are Used for Covid 19 Responses
Estimating the Clinical & Economic Burden Using Prediction & Simulation Modeling: COPD in Ontario
มุมมอง 7384 ปีที่แล้ว
Estimating the Clinical & Economic Burden Using Prediction & Simulation Modeling: COPD in Ontario
Achieving quality primary care EMR data: a description of the CPCSSN data in Alberta
มุมมอง 1174 ปีที่แล้ว
Achieving quality primary care EMR data: a description of the CPCSSN data in Alberta
Intro to Multistate Modeling Approaches for Analyzing Population-wide Health Administrative Data
มุมมอง 9K4 ปีที่แล้ว
Intro to Multistate Modeling Approaches for Analyzing Population-wide Health Administrative Data
Introduction to Causal Inference: Propensity Score Analysis in Healthcare Data
มุมมอง 2.3K4 ปีที่แล้ว
Introduction to Causal Inference: Propensity Score Analysis in Healthcare Data
Holy s***. I have been thinking about this problem for years and at least conceptually this seems like the solution to my problem. Very excited to try this out. Thank you!
Absolutely one of the best talks on youtube for survival analysis with competing risks. Cringed with a yikes at the interruption for questions at 55:40 when he is mid-sentence and mid-concept trying to get to cause-specific vs subdistribution etc.
Excellent presentation, thank you Prof. Sutradhar.
thanks alot
Great talk, tyvm!
Excellent presentation. Thank you for taking the time to explain all of this.
Hello, thank you for this great repo. But can you please tell me how can we get deduplicated enity for eg in fake_1000 dataset, let say I trained a model but now I want to use this to get deduplicated enityt/ cluster Id with most common values for each field. Please help me
ImportError: cannot import name 'exact_match' from 'splink.comparison_library' (C:\Python311\Lib\site-packages\splink\comparison_library.py) i am getting this issue how to resolve it?
Excellent presentation! Thanks so much for sharing!
Thanks for the presentation. We are trying to begin using Splink. Can you please let me know how to reach you the get some practice files.
the practice files are in the git repo
@@AshleyMillsTube Thank you so much. I will review them. Much appreciated.
an excellent lecture that saved me during epidemiology class. Thank you !
Amazed by how simple she made quantile regression to understand. Thanks!
any comparison to graph algorithms on the same use case?
Excellent, one of the best lectures ever on survival analysis and competing risks. Thanks Peter.
Bohnenmafia! #neverforget
Das war schön damals :)
Thank you. Very insightful
Thank you so much! great lecture
When using the msm procedure in R is censored modeled as a state, with time set to zero or is it not included as a state? Thanks
Excellent speech - Thank you. Which procedure in SAS are you using?
The SAS procedure is ‘proc phreg’.
@@populationdatabc1184 I know the proc phreg very well, but have do you model a Multistate model in proc phreg? Do you have an exampel or a paper descriping it? Thanks 🙂
Hi! Quick Question: what linker method corresponds to the chart displayed after training the u parameter (25:16)?
Hiya - it's `linker.match_weights_chart()` - documentation here: moj-analytical-services.github.io/splink/linkerqa.html#splink.linker.Linker.match_weights_chart
Great video for understanding Survival Analysis!
woow very telling graph 24:15
Thank you! Very well explained.
I will be forever greatful to Dr Okosodo , because I was to that herpes virus got no cure not until get in contact with Dr Okosodo , he is indeed a great traditional healer and really good at what he does and he is also kind hearted Words can't express my gratitude just know I am greatful to you and will always be loyal to you anytime any day ...... Pm for information
That is a very good piece of work in explaning quantile regression. Blessings!
je peux avoir le script s'il vous plait?
high quality, thank you!
Very well explained. Thanks a lot.
Is it possible to get the slideshow in pdf form? Also, to whom do we write if we have questions?
Hi Anne. Please email ann.greenwood@popdata.bc.ca regarding your query
Thanks for posting this, is the presentation available somewhere?
excellent talk!
Very excellent speech. Thank you.
Outstanding presentation - thank you for posting!
Is it possible to get the r script?
Thanks so much. Excellent explanation, clear and understandable. Blessings
We were recently asked the following question from a past webinar participant: Question: "When we use IPTW as different time points to get the estimates, how will we pool them together to sum up a final hazard ratio?" Answer: This is done using a weighted model approach. For code and an example using Cox refer to academic.oup.com/aje/article/180/2/160/2739160
very nice presentation. I have a question about the competing risk in multi-state model. what kind of method is used in multi-state model, cause-specific or sub-distribution hazard model? Or another method?
Multistate models are a framework by which to examine event transitions over time. Survival models are simply a 2-state unidirectional multistate model. Competing risks models are also multistate models. If the question is about regression methods under a multistate framework, then both cause-specific and sub-distribution models may be implemented (similar to other multistate models, such as competing risks models).
Well explained. Thanks
Thank you! really useful ad helpful
je peux avoir le script (R) pour tester sur les données de mon pays??
Thank you very much. very useful.
This is one of the best lectures on Survival Analysis.
ppo
Really helpful, thanks!
Nice presentation ! However I have some questions. First, I don't see the motivation behind disease identification, isn't the disease already identified by the physician and stored somewhere in the database ? Also, when you refer to data "linkage", do you imply the creation of a knowledge graph / graph database ( linked data ), or do you simply mean the data combination of these data sources ? Thanks :)
Hi, MrDonald911. Thanks for your interest in this presentation! Frank Lee and Adam D'Souza have provided the following response to your questions. Re: The motivation: Any disease diagnosis is ultimately always determined by a physician. However, it's not always easily available in a structured form. Many diagnoses ARE stored in the administrative health database (the DAD) in a structured form, but not all of them. The structured DAD records are generated by human medical coders who manually read clinical documentation and turn it into structured coded data. But not every disease that a patient may have will be coded, for various reasons (incomplete clinical documentation, diagnosis occurred in primary care rather than acute care, etc.). Most relevant to our talk, some conditions that were explicitly documented in the free-text clinical notes may not have been deemed clinically significant to that particular admission (e.g. patient is hypertensive, but was admitted for surgery on a broken leg), and may therefore not be coded, given the time constraints placed on the coders. These are the 'missing' diagnoses in the structured data that we are hoping to capture via NLP applied directly to the free-text documents. For additional details please refer to the following articles: • Barriers to data quality resulting from the process of coding health information to administrative data: a qualitative study- link.springer.com/article/10.1186/s12913-017-2697-y • Coder perspectives on physician-related barriers to producing high-quality administrative data: a qualitative study - cmajopen.ca/content/5/3/E617.abstract Re: Data linkage: We were referring to the combination of different data sources (in our case, clinical data from the EMR, administrative data from the DAD, and registry data from APPROACH). For additional details please refer to the following article: CREATE: A New Data Resource to Support Cardiac Precision Health - www.cjcopen.ca/article/S2589-790X(20)30239-0/fulltext
The video starts at 9:00
how can we calculate sample size for computing risk regression model is it differ from survival technique or different? 2. can we apply competing risk with prediction model in asingle study?
The following is a reply from Dr. Peter Austin. 1. For a cause-specific hazard model, you can use methods that you would for a conventional Cox model. I haven’t seen examples of power calculations for subdistribution hazard models. 2. Can we apply competing risk with prediction model in a single study? I’m not entirely clear on what is meant by this question. One can develop and internally validate a model in a single study. However, one would want to subsequently validate the model in an external sample.
Thanks for this presentation. In the context of bank credit risk: If the primary event is time to loan default and say the the client closes and repays the loan account prior to loan maturity, i.e. loan is repaid and the account is closed (this event happened prior to loan default) and default is thus no longer possible. Is the account closure event a competing risk?
Yes, because default and prepayment are terminal states.
Really well explained! Thank you! Has helped me a lot!
I love this! Thank you so much