Center for Clinical Health Policy Research
Jump to Main Content

Disease Modeling

 
Disease ModelingDisease models are mathematical representations of clinical conditions, intended to summarize what is known about the disease epidemiology, prevention and treatment, and can therfore be of great benefit to clinicians, researchers, professional societies, manufacturers, and policy makers.  They can provide a foundation for research planning, cost-effectiveness analysis, clinical trial analysis, policy making, and education.
Disease modeling is a major component of our Center's efforts to synthesize the best evidence and promote informed clinical and clinical policy decision making.  In addition technical innovation, extensive  validation, and careful documentation, our modeling efforts are distinguished by attention to the needs of end-users.
 
 

 

Decision Analysis on Screening Cervical Cancer 

Dates: 6/2007-5/2008 
 
PI: Shalini Kulasingam, MPH, PhD
 
Project Type: Decision Modeling
 
Sponsor/Funder: Agency for Healthcare Research Quality (AHRQ) and the Centers for Disease Control and Prevention (CDC)
Aims/objectives:
OBJECTIVE: The overall goal of the proposed research will be to conduct a decision analysis to address gaps in the existing knowledge base that are needed to promote the effective implementation of cervical cancer screening in primary care practice. This research will complement work being conducted by the United States Preventive Services Task Force (USPSTF) and the Oregon Evidence Practice Center
 
SPECIFIC AIMS: The aims of this decision analysis are as follows:
 
1) Compare the overall effectiveness (defined below) and cost-effectiveness of four screening strategies for cervical cancer, under assumptions of fixed age for beginning and ending screening and test interval
  1. HPV first, then Pap (liquid-based and/or conventional)
  2. HPV with concurrent Pap (liquid-based and/or conventional)
  3. Pap first, HPV if ASC-US compared to Pap only (with repeat Pap for ASC-US) (liquid-based and/or conventional)
  4. Pap Only  (liquid-based or conventional)
2) To provide estimates of the impact of different strategies for following women with different combinations of test results on effectiveness and cost-effectiveness based on the initial screening strategy
 
3) To provide estimates of the impact of different ages for beginning and ending screening on effectiveness and cost-effectiveness based on the initial screening strategy
 
4) To provide estimates of the impact of different screening intervals on effectiveness and cost-effectiveness based on the initial screening strategy
 
We anticipate that the results of these aims will assist the USPSTF in developing recommendations for cervical cancer screening practice.
 

ACME Model

PI: David B. Matchar, MD, FACP, FAHA

Anticoagulation Management Event/Cost Model

ACME provides estimates of event rates and cost for various strategies to improve anticoagulation quality, for example, usual physician care, anticoagulation service care, and patient self-testing. Inputs can be modified to reflect local experience (eg, distributions of INRs for different management strategies, costs for products and services, and so on.) This Excel model is available for download, to be run locally.
 

AFAM

PI: David B. Matchar, MD, FACP, FAHA

Atrial Fibrillation Antithrombosis Module

This discrete semi-Markov model is programmed in C++. Two web-based applications are available on this link: a "natural history" simulation and a cost-effectiveness model of stoke prevention strategies for patients with atrial fibrillation. Note that model inputs are relatively old (from the early 1990s) and are currently undergoing revision.

Agreement Information
 

Alzheimer's Disease Model

PI: David B. Matchar, MD, FACP, FAHA

The Alzheimer's Disease Model was developed for the Center for Medicare and Medicaid Services (CMS) to assist in the evaluation of various strategies for testing patients at risk for Alzheimer's disease. This model is programmed in DATA. Individuals interested in more details can download the report to CMS and can contact: david.matchar@duke.edu
 

CDC State Stroke Policy Model

Dates: 10/2004-9/2005
 
PI: David B. Matchar, MD, FACP, FAHA

Revise the DSPM to incorporate state level information and to generate state-level estimates of the burden of disease and the impact of new public health initiatives. Though this can be applied to a wide variety of programs and regions, we will initially focus on improvement of hypertension control in the state of Mississippi. This project will be done in collaboration with the Centers for Disease Control, Cardiovascular Health Branch, the Mississippi Health Policy Center, and Dr. George Howard, a leading stroke epidemiologist at University of Alabama, Birmingham. Specific Aims: 1) Develop a state-level version of the Duke Stroke Policy model based on state-level data regarding demographics, the distribution of risk factors, and local patterns of resource use; 2) Create a standard report that responds to the information needs of specific policy makers, for example, legislators and their staffers involved in oversight of health insurance.
 

Judging the Value of New Therapies for the Care of Patients with CKD: Developing and Applying a Comprehensive Model-based Approach 

Dates: 11/2004-5/2006
 
PI: David Matchar, MD, FACP, FAHA
To develop an advanced chronic kidney disease pharmacoeconomic model as a publicly available repository of best epidemiological evidence, which can be used to evaluate the development and outcomes of advance chronic kidney disease.
 
 

Cost and Effectiveness of Guideline-based Cancer Pain Management

PI: Amy Abernethy, MD

In a tailored cost-effectiveness analysis, we studied differences in cancer pain management between guideline-based care (GBC), oncologist-based care (OBC), and usual care (UC). GBC, though modestly more expensive, led to improved cancer pain control as compared to OBC and UC. We created an evidence-based decision analytical model to provide insight into the burden of cancer pain, and to illustrate the potential impact of following cancer pain guidelines.This Excel model, designed to assist clinicians in weighing the effectiveness and costs of different cancer pain management strategies, is available for download at:
 

Stroke Policy Model (SPM)

PI: David Matchar, MD, FACP, FAHA and Gregory P. Samsa, PhD
Related:
Stroke Policy Model (SPM)

This discrete semi-Markov model is programmed in several versions including C++ and SAS-IML. This is the parent from which several applications are derived including the Atrial Fibrillation (AFAM), Antithrombosis Module, the CDC State Stroke Policy Model, and the Anticoagulation Management Event/Cost Model (ACME). We also provide screen shots to illustrate how other applications can be generated from this parent model.

SPM Background
SPM Calculation Engine
Printer Friendly PageSend this Story to a Friend
© 2008, Duke University Health System   DHTS Web Services DHTS Web Services