Site Index:
More
Information:
In This Section:
News and Info:
|
|
The following is an
annotated list of my experiences with Markov models.
I wrote this in 1970 with
Will Gersh and Eugene Dong at Stanford, just as I was beginning to
study engineering and mathematics. It was my introduction to this
modeling technique. According to a PubMed search it is the 7th
published Markov model in medicine. It was used in at least one
commercially successful automated ICU monitoring system. However I
am sure it has been supplanted by better algorithms by now.
(to top)
 |
Gersh W, Eddy
DM, and Dong E. "Cardiac Arrhythmia Classification: A Heart-Beat
Interval-Markov Chain Approach." Computers and Biomedical
Research 1970; 4:385-392.
|
I studied Markov models
with Dick Smallwood and Ron Howard in the Department of
Engineering Economic Systems at Stanford. They were both superb
practitioners and teachers. Dick was my mentor. Ron "wrote the
book" on Markov models.
(to top)
The Blue Cross Blue
Shield Association (BCBSA) asked me to help them design a cancer
screening benefit program. This was the first time that an
indemnity insurer decided to cover a preventive activity. This is
the project that stimulated me to build a Markov model of
sequential, or intermittent screening. Based on my analyses, the
association created policies to cover screening for mammography,
Pap smears (three year) and colon cancer (three to five-year).
(to top)
|
Eddy DM.
"Rationale for the Cancer Screening Benefit Program Screening
Policies." Blue Cross Association and the National Cancer
Institute, 1977 |
This is described in the top 10 projects.
The problem is described there. The solution required developing
many new methods. They are described in detail in the book
"Screening for Cancer". They include a new method for time-varying
and state-varying models, and the automatic calculation of
transition probabilities "on the fly", as functions of the
system's past history.
|
Eddy DM. "A
Methodology for Evaluating Breast Cancer Screening Programs" in
the Proceedings of the International Conference on
Cybernetics and Society, November, 1976 |
|
Eddy DM. Screening for Cancer:
Theory, Analysis and Design. Englewood Cliffs NJ,
Prentice-Hall, Inc 1980 (book) |
|
A paper in the
mathematical journal Operations Research, which I can't
find. |
|
Eddy DM. "A Mathematical Model for Timing of Repeated Medical
Tests." Medical Decision Making 1983; 3:45-62. (Reprinted
in Scandinavian Journal of Clinical and Laboratory
Investigation, Supplement 171, 1984.)
(to top) |
This is described in the
top 10 projects. Even though I did
not have to develop any new mathematics for this project, beyond
what was already in the
theory of intermittent inspections, this project is very
important because it represents the first time a mathematical
model was used to design a national guideline. The model was
scoured for two years by very skeptical physicians and managers.
The original paper was
|
Eddy DM. "Guidelines for the Cancer-Related Checkup.
Recommendations and Rationale." Ca-A Cancer Journal for
Clinicians 1980; 30:193-240. |
At least 20 other publications are related to
this work. Some of the highlights (in reverse order) are
(to top)
|
Eddy
DM. "Screening for Breast Cancer." Annals of Internal
Medicine 1989; 111:389-399
Reprinted in Common Screening Tests, DM Eddy (Ed)
Philadelphia, American College of Physicians, 1991. Reprinted in
Selected
Readings in General Surgery
1990 |
|
Eddy DM. "Screening for Lung Cancer." Annals of Internal
Medicine 1989; 111:232-237. Reprinted in Common Screening
Tests, DM Eddy (Ed) Philadelphia, American College of
Physicians, 1991 |
|
Eddy DM, Hasselblad V, McGivney W, Hendee W. "The Value of
Mammography Screening in Women under Age 50." Journal of the
American Medical Association 1988; 259:1512-1519 |
|
Eddy DM, Nugent FW, Eddy JF, Coller J. Gilbertsen V, Gottlieb
LS, Rice R, Sherlock P. Winawer S. "Screening for Colorectal
Cancer in a High-risk Population: Results of a Mathematical
Model. "Gastroenterology. 1987 March; 92(3):682-92. |
|
Eddy DM. "Technology Assessment: The Role of Mathematical
Modeling." Section in Assessing Medical Technologies,
Institute of Medicine, Washington DC, National Academy Press,
1985 |
|
Eddy DM. "Finding Cancer in Asymptomatic People." Cancer
1983; 51:2440-2445. Biases. New York, Cambridge
University Press, 1982 |
|
Eddy DM, Shwartz M. "The Application of Mathematical Models in
Screening." Chapter in D Schottenfeld, J Fraumeni (Eds)
Cancer Epidemiology and Prevention. WB Saunders
Company, 1982 |
|
McNeil B,
Eddy DM. "The Costs and Effects of Screening for Cancer Among
Asbestos-Exposed Workers." Journal of Chronic Diseases
1982; 35:351-385 |
|
Eddy DM. "The Economics of Cancer Prevention and Detection:
Getting More for Less. "Cancer. 1981 Mar 1;47(5 Suppl):1200-9. |
|
Eddy DM. "Detection of Lung Cancer." JAMA.1981 Aug 7;246
(6):621-2 |
|
Eddy DM. "Appropriateness of Cervical Cancer Screening."
Gynecologic Oncology 1981 Oct 12: S168-S187 |
|
Eddy DM. "The Economics of Cancer Prevention and Detection."
Cancer 1981; 47:1200-1209 |
|
Eddy DM.
"Evaluation of a Screening Program in Oncology." Bulletin
Cancer (Paris) 1980; 67:515-523 |
|
Eddy DM.
"Computer Models and the Evaluation of Colon Cancer Screening
Programs." Chapter in S Winawer, D Schottenfeld and P Sherlock (Eds)
Colorectal Cancer: Prevention, Epidemiology and Screening.
New York, Raven Press, 1980
(to top) |
IN the years after the ACS report I was called on to help
develop a case for coverage of cancer screening by Medicare. (By
statute, Medicare does not cover preventive services. The decision
to cover cancer screening required a congressional mandate. I was
asked to do the cost-effectiveness analyses. Much of this work,
which involved Markov models, was commissioned by Congress through
the Office of Technology Assessment.
|
Eddy DM. "Screening for Colon
Cancer: A Technology Assessment." Background Paper #2: Case
Studies of Medical Technologies in The Implications of
Cost-Effectiveness Analysis of Medical Technology. Congress
of the United States Office of Technology Assessment, 1981 |
|
Eddy DM. "Testimony on
Technology Assessment to the Committee on Labor and Human
Resources." U.S. Senate, 1984 |
|
Eddy DM. "The Effectiveness,
Cost, and Cost-Effectiveness of Breast Cancer Screening with
Annual Mammography and Breast Physical Examination for Women
over 65." Committee on Ways and Means, Subcommittee on Health,
U.S. House of Representatives, 1986 |
|
Eddy DM. "Breast Cancer
Screening for Women over Age 65." U.S. Congressional Office of
Technology Assessment for submission to the U.S. Congress,
Senate Committee on Labor and Human Resources, as part of the
report "Breast Cancer Screening for Medicare Beneficiaries.
Effectiveness and Costs to Medicare and Medical Resources
Required." Washington DC, U.S. Congressional Office of
Technology Assessment, November 1987.
(to top) |
This is described in the
top 10 projects
|
Eddy DM. "Setting Priorities for Cancer Control Programs."
Journal of the National Cancer Institute 1986; 76:187-199 |
|
Eddy DM. "The Use of
Quantitative Methods in Planning National Cancer Control
Programs." Bulletin of the World Health Organization
64:683-693, 1986 |
|
Stjernswärd J, Stanley K, Eddy DM, et al. " National Cancer
Control Programs and Setting Priorities." Cancer Detection
and Prevention 1986; 9:113-124 |
|
Eddy DM. "Secondary Prevention of Cancer: An Overview."
Bulletin of the World Health Organization 1986; 64:421-429
(to top) |
This is described in the
top 10 projects
|
Eddy DM. "A Computer-Based Model for the Design of Cancer
Control Strategies." National Cancer Institute Monographs
1986; 2:75-82 |
|
Eddy DM. "CAN*TROL: A Computer Model for Designing National
Cancer Control Strategies.."Bull Cancer.
1987;74(3):323-32 |
|
Levin DL, Gail MH, Kessler LG, Eddy DM. "A Model for Projecting
Cancer Incidence and Mortality in the Presence of Prevention,
Screening, and Treatment Programs." National Cancer Institute
Monographs 1986; 2:83-93
(to top) |
The publication
of a paper by the International Agency for Research on Cancer (IARC)
gave me an opportunity to conduct an independent validation of the
screening model. The results I published in 1980 predicted what
they found 7 years later using data from about a million women, to
within 3% (relative). This is the only independent, prospective
validation cited in the book on Cost-Effectiveness in Health and
Medicine (Gold et al editors).
(to top)
|
Eddy DM. "The Frequency of Cervical Cancer Screening. Comparison
of a Mathematical Model with Empirical Data." Cancer
1987; 60:1117-1122 |
In 1990 I became a Senior Advisor to Kaiser Permanente Southern
California. There I began to work with practicing clinicians and
administrators on information systems, guidelines, assessment of
new technologies and medical necessity decisions, and priority
setting. These decisions involve a much higher level of clinical
and administrative detail, and much more realism than the national
and international-level work I had been doing. I found that the
Markov structure was not well suited to these
"rubber-meets-the-road" types of problems. As a consequence, I
began to think of other ways to apply quantitative methods. The
result was the Archimedes project. It is entirely different than a
Markov model, driven by different types of problems. For more
information on Archimedes,
click here. (If it is password protected,
contact me).
(to top)
-
Biography • Top 10 papers • Top 10 projects • Top disappointments • Markov models
|
|
