MMI 409 - Biostatistics
Evidence based medicine stems from the measurement of outcomes. Analysis of those outcomes allows providers to talk with patients about the likelihood of benefit and harm based on how much they are similar to the population being studied. This analysis is used in the development of decision tools, manual or electronic, that guide providers with every decision they make.
As informaticists we are not expected to be able to perform complex summaries or analysis of research-derived data designed to address IT, clinical, research, and administrative needs within health care. But we will be expected to have some understanding of whether the approach to this work is appropriate, helpful and provides results consistent with the approach. This is ideally done by every scientist as they review the methods used by other researchers within their field. Perhaps more challenging than the mathematics is the process of formulating the questions to be asked. The proper question is at the heart of research-based solutions. That will give the answers we seek to medical informatics issues or problems within a health care context.
Learning the basic tools of statistical analysis - the population and test characteristics, margins of error, confidence intervals, and types of error - are all part of understanding this essential field of study. Learning how to use software to complete the analysis makes the process simpler, but can readily be misleading without an understanding of what to expect in the results. It is this core concept, reasonableness, that we will always need to apply to statistics and what we consider evidence based medicine.
As a part of my interest in developing Clinical Decision Support tools, I plan to help educate users about the benefits and limits of statistical analysis, and the ways in which the results of tools should be considered. In most cases the results will offer a set of diagnosis from which the provider can choose based on their clinical sense - something that we have yet to come close to reproducing in a computer.
Syllabus Midterm Exam Final Exam MMI Overview Coursework Projects Home
*
As informaticists we are not expected to be able to perform complex summaries or analysis of research-derived data designed to address IT, clinical, research, and administrative needs within health care. But we will be expected to have some understanding of whether the approach to this work is appropriate, helpful and provides results consistent with the approach. This is ideally done by every scientist as they review the methods used by other researchers within their field. Perhaps more challenging than the mathematics is the process of formulating the questions to be asked. The proper question is at the heart of research-based solutions. That will give the answers we seek to medical informatics issues or problems within a health care context.
Learning the basic tools of statistical analysis - the population and test characteristics, margins of error, confidence intervals, and types of error - are all part of understanding this essential field of study. Learning how to use software to complete the analysis makes the process simpler, but can readily be misleading without an understanding of what to expect in the results. It is this core concept, reasonableness, that we will always need to apply to statistics and what we consider evidence based medicine.
As a part of my interest in developing Clinical Decision Support tools, I plan to help educate users about the benefits and limits of statistical analysis, and the ways in which the results of tools should be considered. In most cases the results will offer a set of diagnosis from which the provider can choose based on their clinical sense - something that we have yet to come close to reproducing in a computer.
Syllabus Midterm Exam Final Exam MMI Overview Coursework Projects Home
*