1. The Structure of Medical
Data
• Medicine is remarkable for its failure to
develop a standarized vocabulary and
nomenclature
• Issues of data retrieval and analysis are
confounded by discrepancies between
observers and data analysts
• Imprecission and a lack of standarized
vocabulary are problematic when we wish to
aggregate data recorded by multiple health
professionals
• EHR’s, their encoding must be able to
presume a specific meaning for the terms
2. Coding systems
Because of the needs to know about
health trends for populations and to
recognize epidemics in the early
stages, there are various health
reporting requirements for hospitals
and physicians.
Also reporting discharge
diagnosis, procedures performed on
pacients
The codes used must be well defined
3. Coding systems
Coding systems have limitations when
are applied in more general clinical
settings
Researchers have worked to develop
a unified medical language system
(UMLS), a common structure that ties
together the various vocabularies the
have been created
4. The Data Knowledge
Spectrum
A central focus in medical informatics
is the information base that constitutes
the substance of medicine
Three terms are frequently used to
describe the content of computed
based systems: data, information and
knowledge
A database is a collection of individual
observations without summarizing
analysis
5. Strategies of Medical Data
Selection and Use
All medical databases are basically
incomplete because they reflect the
selective collection of data by the
health care personnel
The challenge is to ask only questions
that are necessary, to perform only the
examinations that are required and to
record only pertinent data
6. The Hypothetico - Deductive
Approach
Studies of medical decision makers have
shown that strategies for data collection
and interpretation may be imbedded in
an interative process known as
hypothetico-deductive approach
The central idea is one of
seuqential, staged data
collection, followed by data interpretation
and the generation of
hypotheses, leading to hipothesis-
directed selection of the next most
appropiate data to be collected.
7. The Hypothetico - Deductive
Approach
Physicians refers to the set of active
hypothesis as the differential
diagnosis
Physicians have developed safety
measures to avoid missing important
issues. They are focused in four
categories:
◦ Past medical history
◦ Family history
◦ Social history
◦ Review of systems
8. The Hypothetico - Deductive
Approach
After finishing with the physical
examination the list of hipotheses may
be narrowed down sufficiently that the
physician can undertake specific
treatment
It often is necessary to gather aditional
data
The response of the patient to
treatment is itself a datum point that
may affect the hypotheses about a
patient’s illness
9. The relationship between data
and hypotheses
Observation evokes a hypothesis
Sensitivity: the likehood that a given
datum will be observed in a patient
with a given disease or condition
Specificity. An observation is highly
specific for a disease if it is generally
not seen in patients who do not have
that disease
10. The relationship between data
and hypotheses
The prevalence of a disease is simply
a measure of the frequency with which
the disease occurs in the population of
interest
The predictive value of a test is simply
the post-test probability that a disease
is present based on the results of a
test
11. Methods for selecting questions
and comparing tests
Given a set of current
hypotheses, how does the physician
decide what additional data should be
collected?
12. The Computer and Collection of
Medical Data
The need for data entry by physicians
has posed a problem for medical
computing systems since the earliest
days of the field
In some applications is possible for
data to be enterred automatically into
the computer
The use of touch-
screens, mouse, PDA’s can help to
reduce the resistance to computer use