U.S. patent application number 13/556706 was filed with the patent office on 2012-11-15 for searching an electronic medical record.
This patent application is currently assigned to CERNER INNOVATION, INC.. Invention is credited to CHRISTOPHER S. FINN, MARGARET CUSHING KOLM, DAVID P. MCCALLIE, JR..
Application Number | 20120290328 13/556706 |
Document ID | / |
Family ID | 42319690 |
Filed Date | 2012-11-15 |
United States Patent
Application |
20120290328 |
Kind Code |
A1 |
MCCALLIE, JR.; DAVID P. ; et
al. |
November 15, 2012 |
SEARCHING AN ELECTRONIC MEDICAL RECORD
Abstract
A method, system, and medium are provided for searching an
electronic medical record. Search results are returned in response
to a search query. The search query may be one or more designated
medical concepts. The search results may be displayed according to
a ranking that determines which search results are likely to be the
most responsive to a query submitted by a particular clinician
based on matching the most important clinical concepts in each
document to the most important clinical concepts in the search
query.
Inventors: |
MCCALLIE, JR.; DAVID P.;
(STILWELL, KS) ; FINN; CHRISTOPHER S.; (LIBERTY,
MO) ; KOLM; MARGARET CUSHING; (KANSAS CITY,
MO) |
Assignee: |
CERNER INNOVATION, INC.
OVERLAND PARK
KS
|
Family ID: |
42319690 |
Appl. No.: |
13/556706 |
Filed: |
July 24, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12351288 |
Jan 9, 2009 |
8239216 |
|
|
13556706 |
|
|
|
|
Current U.S.
Class: |
705/3 |
Current CPC
Class: |
G16H 50/70 20180101;
G16H 40/67 20180101; G06F 16/3331 20190101; G16H 10/60
20180101 |
Class at
Publication: |
705/3 |
International
Class: |
G06Q 50/24 20120101
G06Q050/24 |
Claims
1. One or more non-transitory computer-storage media having
computer-executable instructions embodied thereon for performing a
method of finding information in an electronic medical record, the
method comprising: receiving a search query from a user to search
the electronic medical record that is associated with a patient and
that includes a plurality of electronic documents that describe a
medical history for the patient; identifying one or more components
of the electronic medical record that contain text with one or more
words that matches the search query; determining, for each
particular clinical concept recited in the one or more components,
a patient-subject status that indicates whether the patient is an
object of a particular clinical concept recited in a particular
component; determining a query-responsiveness score using the
patent-subject status for each of the one or more components that
match the search query, wherein the query-responsiveness score
indicates how responsive an individual component is to the search
query; presenting search results that communicate information
describing each of the one or more components displayed ordered
according to the query-responsiveness score assigned to each of the
one or more components.
2. The media of claim 1, wherein the method further includes
displaying one or more filter options that allows the user to
filter the search results, wherein the one or more filter options
are based on one or more of a clinical-concept in an individual
search result, an aggregation of related clinical concepts, a
clinical facility associated with the individual search result, and
a document class associated with the individual search result.
3. The media of claim 1, wherein the method further includes
receiving a selection of a search mode, wherein the search mode is
a medical synonym match, wherein the one or more components of the
electronic medical record matches the search query when at least a
medical synonym of the one or more words in the search query is
found within the one or more components of the electronic medical
record, and wherein the medical synonyms match search mode includes
matching two or more words or phrases that have a similar medical
meaning.
4. The media of claim 1, wherein the method further includes
receiving a selection of a search mode, wherein the search mode is
set to match related concepts, and wherein the one or more
components of the electronic medical record matches the search
query when a clinical concept conveyed by the search query matches
at least one related clinical concept recited in the one or more
components of the electronic medical record, and wherein the
clinical concept describes any aspect of a person's health
condition.
5. The media of claim 4, wherein determining the
query-responsiveness score of each of the one or more components
further includes: determining a truth status for said each
particular clinical concept recited in the one or more components,
wherein the truth status indicates whether said each particular
clinical concept was expressed positively, negatively, ambiguously,
or unknown; determining a clinical-usage context for each
particular clinical concept recited in the one or more components,
wherein the clinical-usage context describes how the clinical
concept was used in a component; determining a document-importance
factor for each particular clinical concept recited in the one or
more components, wherein the document-importance factor measures a
relevance of the particular clinical concept to a main subject of a
particular document by analyzing other clinical concepts that are
used in the particular document with the particular clinical
concept; determining a specificity factor for each particular
clinical concept recited in the one or more components based on a
degree of narrowness for a scope of said each particular clinical
concept; and wherein the query-responsiveness score for the
individual component is increased when the patient-subject status
for one or more of the clinical concepts identified within the
component is affirmative, when the truth status for one or more of
the clinical concepts identified within the component equals
positive, when the clinical-usage context for one or more of the
clinical concepts identified within the component directly relates
to the patient, when the document-importance factor is high, and
when the specificity factor for one or more of the clinical
concepts identified within the component is high.
6. The media of claim 1, wherein the method further includes:
expanding one or more primary clinical concepts recited in the
search query to additional related clinical concepts, thereby
generating a plurality of expanded clinical concepts to match with
a particular one of the one or more components of the electronic
medical record; assigning a boost factor to each combination of an
expanded clinical concept and the particular matching component of
the electronic medical record, wherein the boost factor is higher
when the expanded clinical concept is close to the one or more
primary clinical concepts on a clinical-concept ontology and lower
when the expanded clinical concept is remote from the one or more
primary clinical concepts; and wherein the boost factor is used to
calculate the query-responsiveness score.
7. The media of claim 1, wherein the method further includes
increasing the query-responsiveness score when a role of the user
submitting the search query is in the same category as the role of
a person that created a matching component.
8. One or more non-transitory computer-storage media having
computer-executable instructions embodied thereon for performing a
method of searching an electronic medical record for selected
clinical concepts, the method comprising; receiving a search query
from a user to search the electronic medical record that is
associated with a patient and that includes a plurality of
electronic documents that describe a medical history for the
patient; identifying one or more components of the electronic
medical record that contain text with one or more words that
matches the search query; determining, for each clinical concept
recited in the one or more components, a clinical-usage context
that describes how the clinical concept was used in a component;
determining a query-responsiveness score using the clinical-usage
context for each of the one or more components that match the
search query, wherein the query-responsiveness score indicates how
responsive an individual component is to the search query;
presenting search results that communicate information describing
each of the one or more components displayed ordered according to
the query-responsiveness score assigned to each of the one or more
components.
9. The media of claim 8, wherein the method further includes:
determining a truth status for said each particular clinical
concept recited in the one or more components, wherein the truth
status indicates whether said each particular clinical concept was
expressed positively, negatively, ambiguously, or unknown; wherein
the query-responsiveness score is based on the clinical-usage
context and the truth status for said each of the one or more uses
of the clinical concept.
10. The media of claim 9, wherein the method further includes:
determining, for each particular clinical concept recited in the
one or more components, a patient-subject status that indicates
whether the patient is an object of a particular clinical concept
recited in a particular component; and wherein the
query-responsiveness score is based on the clinical-usage context,
the truth status, and the patient-subject status for said each of
the one or more uses of the clinical concept.
11. The media of claim 10, wherein the method further includes:
determining a document-importance factor for each particular
clinical concept recited in the one or more components, wherein the
document-importance factor measures a relevance of the particular
clinical concept to a main subject of a particular document by
analyzing other clinical concepts that are used in the particular
document with the particular clinical concept; determining a
specificity factor for each particular clinical concept recited in
the one or more components based on a degree of narrowness for a
scope of said each particular clinical concept; and wherein the
query-responsiveness score is based on the clinical-usage context,
the truth status, the patient-subject status, the
document-importance factor, and the specificity factor for the
particular use.
12. The media of claim 8, wherein the query-responsiveness score is
based on a boost factor, wherein the boost factor includes a
role-boost factor that is based on a role of the user that
submitted the search query and a role of a person associated with a
use of the clinical concept in the electronic medical record, and
wherein the role-boost factor is increased when the role of the
user is in the same category as the role of the person associated
with a document within the electronic medical record that includes
the use of the clinical concept from the search query.
13. The media of claim 8, wherein the query-responsiveness score is
based on at least one boost factor, wherein the at least one boost
factor includes a clinical-facility-boost factor, wherein the
clinical-facility-boost factor is based on a clinical facility
associated with the user that submitted the search query and the
clinical facility associated with a use of the clinical concept in
the electronic medical record, and wherein the
clinical-facility-boost factor is increased for the use of the
clinical concept in the electronic medical record when the use of
the clinical concept in the electronic medical record is in a
portion of the electronic medical record that is associated with
the clinical facility from which the user submitted the search
query.
14. The media of claim 8, wherein the method further includes:
expanding a clinical concept recited in the search query into
additional related clinical concepts, thereby generating a
plurality of expanded clinical concepts to match to each of the
uses of the plurality of expanded clinical concepts within
documents in the electronic medical record; assigning a
closeness-boost factor to each combination of an expanded-clinical
concept and a matching use of the expanded-clinical concept within
a document in the electronic medical record, wherein the
closeness-boost factor is higher when the expanded-clinical concept
is close to the clinical concept on a clinical-concept ontology and
lower when the expanded-clinical concept is remote from the
clinical concept; and wherein the query-responsiveness score is
also based on the closeness-boost factor.
15. One or more non-transitory computer-storage media having
computer-executable instructions embodied thereon for performing a
method of preparing an electronic medical record for electronic
searching, the method comprising: receiving a search query from a
user to search the electronic medical record that is associated
with a patient and that includes a plurality of electronic
documents that describe a medical history for the patient;
identifying one or more components of the electronic medical record
that contain text with one or more words that matches the search
query; determining, for each particular clinical concept recited in
the one or more components, a specificity factor for the particular
use of the clinical concept based on a degree of narrowness for a
scope of the particular use of the clinical concept; determining a
query-responsiveness score using the specificity factor for each of
the one or more components that match the search query, wherein the
query-responsiveness score indicates how responsive an individual
component is to the search query; presenting search results that
communicate information describing each of the one or more
components displayed ordered according to the query-responsiveness
score assigned to each of the one or more components.
16. The media of claim 15, wherein the method further includes: for
each of said clinical concepts, determining a document-importance
factor for each of said clinical concepts, wherein the
document-importance factor measures a relevance of a particular
clinical concept to a main subject of a particular document by
analyzing other clinical concepts that are used in the particular
document with the particular clinical concept wherein the
query-responsiveness score for each use of the clinical concept in
the electronic medical record is also based on the
document-importance factor for each use of the clinical
concept.
17. The media of claim 15, wherein the method further comprises,
for each of the clinical concepts, determining a clinical-usage
context for each use of the clinical concept in the electronic
medical record, wherein the clinical-usage context describes how
the clinical concept was used in the electronic medical record,
wherein the clinical-usage context includes one or more of a
presenting complaints section, a patient history, a family history,
a review of systems section, a physical exam record, a
prescription, an order, a lab result, a vital sign, a diagnosis,
and a procedure, and wherein the query-responsiveness score for
each use of the clinical concept in the electronic medical record
is also based on the clinical-usage context for each use of the
clinical concept.
18. The method of claim 17, wherein the query-responsiveness score
for an individual use of a clinical-concept is increased when the
clinical-usage context for the individual use of the clinical
concept directly relates to the patient.
19. The media of claim 15, wherein the method further includes
increasing the query-responsiveness score when a role of the user
submitting the search query is in the same category as the role of
a person that created a matching component.
20. The media of claim 19, wherein the query-responsiveness score
is based on a boost factor calculated based on one or more of: a
hierarchical closeness of an expanded-clinical concept, wherein the
boost factor is increased if the expanded-clinical concept is
closely related to one or more clinical concepts within the search
query based on a clinical-concept ontology; a date of the matching
document; a document class for the matching document; a primary
focus of the matching document; and a clinical facility associated
with the matching document.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 12/351,288, filed Jan. 9, 2009, entitled
"Searching an Electronic Medical Record."
BACKGROUND
[0002] Clinical facilities (e.g., hospital, therapy center,
practice group) or other managers of medical information may
maintain a patient's medical record in an electronic database. An
individual patient's medical record is called an electronic medical
record or personal health record. The electronic medical record may
include electronic documents and database entries. Over time, a
patient's electronic medical record may contain a lot of
information. The information may be reached by browsing through the
electronic medical record and opening documents to look for
information. The large amount of information makes it difficult to
find desired information.
SUMMARY
[0003] Embodiments of the invention are defined by the claims
below, not this Summary. A high-level overview of various aspects
of the invention are provided here for that reason, to provide an
overview of the disclosure, and to introduce a selection of
concepts that are further described below in the Detailed
Description. This Summary is not intended to identify key features
or essential features of the claimed subject matter, nor is it
intended to be used as an aid in isolation to determine the scope
of the claimed subject matter.
[0004] In a first aspect, one or more computer-storage media having
computer-useable instructions embodied thereon for performing a
method of finding information in an electronic medical record are
provided. The method includes receiving a search query from a user
to search the electronic medical record, wherein the electronic
medical record is associated with a patient. The electronic medical
record includes a plurality of electronic documents that describe a
medical history for the patient and is stored on the
computer-storage media. The method also includes identifying one or
more components of the electronic medical record that contain text
that matches the search query, wherein each of the one or more
components is a section of text with in the electronic medical
record that includes one or more words. The method also includes
determining a query-responsiveness score for each of the one or
more components that match the search query. The
query-responsiveness score indicates how responsive an individual
component is to the search query. The method further includes
presenting search results that communicate information describing
each of the one or more components. The search results are
displayed ordered according to the query-responsiveness score
assigned to each of the one or more components.
[0005] In a further aspect, one or more computer-storage media
having computer-useable instructions embodied thereon for
performing a method of searching an electronic medical record for a
selected clinical concept are provided. The method includes
receiving a search query that includes a clinical concept from a
user. The clinical concept is an aspect related to a person's
health. The method also includes identifying one or more uses of
the clinical concept in the electronic medical record, wherein the
electronic medical record is an electronic description of a medical
history for a patient and is stored on the computer-storage media.
The method further includes determining a query-responsiveness
score for each document within the electronic medical record in
which the clinical concept is used. The query-responsiveness score
describes how important a particular document is likely to be to
the user. The query-responsiveness score is determined based on a
clinical-importance score for each of the one or more uses of the
clinical concept and at least one boost factor that is based on the
search query and the user. The method also includes presenting
search results based on the query-responsiveness score associated
with said each document within the electronic medical record in
which the clinical concept is used.
[0006] In a further aspect, one or more computer-storage media
having computer-useable instructions embodied thereon for
performing a method of preparing an electronic medical record for
electronic searching are provided. The method includes receiving
the electronic medical record that includes information describing
at least a portion of a medical history associated with a patient.
The method also includes identifying clinical concepts within the
electronic medical record, wherein the clinical concept is an
aspect related to a person's health. The method also includes, for
each of the clinical concepts, determining a patient-subject status
for each use of the clinical concept in the electronic medical
record, wherein the patient-subject status indicates whether the
patient is a subject of a particular use of the clinical concept.
The method further includes, for each of the clinical concepts,
determining a truth status for each use of the clinical concept in
the electronic medical record, wherein the truth status indicates
whether the clinical concept was expressed positively, negatively,
ambiguously, or unknown. The method further includes, for each of
the clinical concepts, determining a clinical-usage context for
each use of the clinical concept in the electronic medical record,
wherein the clinical-usage context describes how the clinical
concept was used in the electronic medical record. The method also
includes, for each of the clinical concepts, determining a
specificity factor for each use of the clinical concept in the
electronic medical record based on the degree of specificity,
precision or narrowness of scope of the concept, as derived from
the concept's position in a clinical ontology, or other reference
information. The method further includes assigning a
clinical-importance score to each use of the clinical concept in
the electronic medical record based on a the patient-subject
status, the truth status, the clinical-usage context, and the
specificity factor and storing the clinical-importance score
associated with each use of the clinical concept in each document
within the electronic medical record in a data store.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0007] Illustrative embodiments of the present invention are
described in detail below with reference to the attached drawing
figures, wherein:
[0008] FIG. 1 is a block diagram depicting an exemplary computing
environment suitable for use in implementing embodiments of the
present invention;
[0009] FIG. 2 is a block diagram depicting an exemplary computing
architecture suitable for searching an electronic medical record
for components that are responsive to a search query, in accordance
with an embodiment of the present invention;
[0010] FIG. 3 is a flow diagram showing a method of preparing
electronic medical records for electronic searching, in accordance
with an embodiment of the present invention;
[0011] FIG. 4 is a flow diagram showing a method of finding
information in electronic medical record, in accordance with an
embodiment of the present invention;
[0012] FIG. 5 is a flow diagram showing a method of searching an
electronic medical record for a selected clinical concept, in
accordance with an embodiment of the present invention;
[0013] FIG. 6 is an illustrative screen display of an interface to
receive a search query, in accordance with an exemplary embodiment
of the present invention;
[0014] FIG. 7 is an illustrative screen display of search results
that are responsive to a submitted clinical concept, in accordance
with an exemplary embodiment of the present invention;
[0015] FIG. 8 illustrates the identification of clinical concepts
in a document that is part of the electronic medical record, in
accordance with an exemplary embodiment of the present invention;
and
[0016] FIG. 9 illustrates indexing clinical concepts and related
information in an index, in accordance with an exemplary embodiment
of the present invention.
DETAILED DESCRIPTION
[0017] Embodiments of the present invention allow a user to search
for information in an electronic medical record ("EMR"). An EMR is
a collection of information describing the medical history of a
patient. The EMR may be managed by a variety of sources including a
clinical facility, such as a hospital, and the patient. In one
embodiment, the EMR is personal health record. The EMR for a single
patient may contain combinations of database entries and electronic
documents that are related to the patient's medical history. The
database entries may be created by filling out an electronic form
presented in a user interface. The documents and database entries
may include encoded data that describes a portion of a patient's
medical history. For example, a diagnosis for diabetes may be
codified as "D234539A293" and recorded in a document or database
entries. EMRs for groups of patients may be collected in a single
data store.
[0018] Embodiments of the present invention allow a user to submit
a search query through an interface and return search results that
are responsive to the search query. As will be described in more
detail subsequently, the search results may be ordered according to
a query-responsiveness score so that the most important matching
components of the EMR can be quickly located at the top of the
result list. Search results may present components of the EMR at
any level of granularity. A component of the EMR may be any text
within the EMR including a document within the EMR or a section of
text within a document (e.g., a paragraph, document section) in the
EMR, or any structured and/or codified element of information
contained within the EMR. For example, components may be documents
within the EMR, a word within the EMR, a sentence within the EMR, a
single use of a clinical concept within the EMR, or a component of
a document within the EMR. Different embodiments evaluate
components of the EMR and present search results based on those
components at different levels of granularity. The analysis used to
determine the responsiveness of a search result, which is described
hereafter, may be performed at whatever level of granularity the
search results are presented. Throughout this disclosure, the level
of granularity will most commonly be described as a document in the
EMR or a component of the EMR, but embodiments of the present
invention are not intended to be limited to these descriptions. As
will be pointed out subsequently, some of the factors used to
determine the responsiveness of a search result may not be used
when the search results are at a very low level of granularity,
such as a single word or a single use of a clinical concept. In one
embodiment, the search results are matched based on clinical
concepts in the query and clinical concepts in components of the
EMR. A clinical concept is an aspect related to a person's health.
A clinical concept describes any aspect of a person's health
condition, or any object, action, attribute or idea that is related
to a health condition. Examples include: diseases; symptoms;
clinical observations and findings; diagnostic tests; diagnostic or
therapeutic procedures; organisms, substances, devices or products
related to health conditions; anatomic structures including
genomic; phenotypic expression; behavior, family and social context
related to health conditions; risk factors and outcomes; facilities
and care providers. For example, heart disease and a heart attack
are examples of clinical concepts.
[0019] The search results are ordered according to a
query-responsiveness score that is calculated for each matching
search result (e.g. document, document portion). In one embodiment,
the query-responsiveness score is calculated by combining a set of
clinical-importance scores with a set of boost factors. The
clinical-importance score measures how important the clinical
concepts used within each specific component of the EMR are, apart
from a query. In one embodiment, the clinical-importance score is
generated for each use of a clinical concept within the EMR in
advance and stored in an index.
[0020] There are at least two categories of boost factors that may
be combined with the clinical-importance score to calculate the
query-responsiveness score. Document-boost factors measure the
responsiveness of a search-result (e.g., a document, a component of
a document, a component of the EMR) without considering the query.
Query-boost factors measure the document's responsiveness to the
query using information related to the query, such as the role of
the user submitting the query. The boost factors are assigned to
produce a good fit between the query and the potential search
result. One or more boost factors may be combined with the set of
clinical-importance scores to generate a query-responsiveness score
for the component of the EMR.
[0021] Having briefly described embodiments of the present
invention, an exemplary operating environment suitable for use in
implementing embodiments of the present invention is described
below. Some of the wording and form of description is done to meet
applicable statutory requirements. Although the terms "step" and/or
"block" or "module" etc. might be used herein to connote different
components of methods or systems employed, the terms should not be
interpreted as implying any particular order among or between
various steps herein disclosed unless and except when the order of
individual steps is explicitly described.
[0022] Referring to the drawings in general, and initially to FIG.
1 in particular, an exemplary computing system environment, for
instance, a medical information computing system, on which
embodiments of the present invention may be implemented is
illustrated and designated generally as reference numeral 20. It
will be understood and appreciated by those of ordinary skill in
the art that the illustrated medical information computing system
environment 20 is merely an example of one suitable computing
environment and is not intended to suggest any limitation as to the
scope of use or functionality of the invention. Neither should the
medical information computing system environment 20 be interpreted
as having any dependency or requirement relating to any single
component or combination of components illustrated therein.
[0023] The present invention may be operational with numerous other
general purpose or special purpose computing system environments or
configurations. Examples of well-known computing systems,
environments, and/or configurations that may be suitable for use
with the present invention include, by way of example only,
personal computers, server computers, hand-held or laptop devices,
multiprocessor systems, microprocessor-based systems, set top
boxes, programmable consumer electronics, cellular telephones,
network PCs, minicomputers, mainframe computers, distributed
computing environments that include any of the above-mentioned
systems or devices, and the like.
[0024] The present invention may be described in the general
context of computer-executable instructions, such as program
modules, being executed by a computer. Generally, program modules
include, but are not limited to, routines, programs, objects,
components, and data structures that perform particular tasks or
implement particular abstract data types. The present invention may
also be practiced in distributed computing environments where tasks
are performed by remote processing devices that are linked through
a communications network. In a distributed computing environment,
program modules may be located in local and/or remote
computer-storage media including, by way of example only, memory
storage devices.
[0025] With continued reference to FIG. 1, the exemplary medical
information computing system environment 20 includes a general
purpose computing device in the form of a control server 22.
Components of the control server 22 may include, without
limitation, a processing unit, internal system memory, and a
suitable system bus for coupling various system components,
including database cluster 24, with the control server 22. The
system bus may be any of several types of bus structures, including
a memory bus or memory controller, a peripheral bus, and a local
bus, using any of a variety of bus architectures. By way of
example, and not limitation, such architectures include Industry
Standard Architecture (ISA) bus, Micro Channel Architecture (MCA)
bus, Enhanced ISA (EISA) bus, Video Electronic Standards
Association (VESA) local bus, and Peripheral Component Interconnect
(PCI) bus.
[0026] The control server 22 typically includes therein, or has
access to, a variety of computer-readable media, for instance,
database cluster 24. Computer-readable media can be any available
media that may be accessed by control server 22, and includes
volatile and nonvolatile media, as well as removable and
non-removable media. By way of example, and not limitation,
computer-readable media may include computer-storage media and
communication media. Computer-storage media may include, without
limitation, volatile and nonvolatile media, as well as removable
and non-removable media implemented in any method or technology for
storage of information, such as computer readable instructions,
data structures, program modules, or other data. In this regard,
computer-storage media may include, but is not limited to, RAM,
ROM, EEPROM, flash memory or other memory technology, CD-ROM,
digital versatile disks (DVDs) or other optical disk storage,
magnetic cassettes, magnetic tape, magnetic disk storage, or other
magnetic storage device, or any other medium which can be used to
store the desired information and which may be accessed by the
control server 22. Communication media typically embodies computer
readable instructions, data structures, program modules, or other
data in a modulated data signal, such as a carrier wave or other
transport mechanism, and may include any information delivery
media. As used herein, the term "modulated data signal" refers to a
signal that has one or more of its attributes set or changed in
such a manner as to encode information in the signal. By way of
example, and not limitation, communication media includes wired
media such as a wired network or direct-wired connection, and
wireless media such as acoustic, RF, infrared, and other wireless
media. Combinations of any of the above also may be included within
the scope of computer-readable media.
[0027] The computer-storage media discussed above and illustrated
in FIG. 1, including database cluster 24, provide storage of
computer readable instructions, data structures, program modules,
and other data for the control server 22.
[0028] The control server 22 may operate in a computer network 26
using logical connections to one or more remote computers 28.
Remote computers 28 may be located at a variety of locations in a
medical or research environment, for example, but not limited to,
clinical laboratories (e.g., molecular diagnostic laboratories),
hospitals and other inpatient settings, veterinary environments,
ambulatory settings, medical billing and financial offices,
hospital administration settings, home health care environments,
and clinicians' offices and the clinician's home or the patient's
own home or over the Internet. Clinicians may include, but are not
limited to, a treating physician or physicians, specialists such as
surgeons, radiologists, cardiologists, and oncologists, emergency
medical technicians, physicians' assistants, nurse practitioners,
nurses, nurses' aides, pharmacists, dieticians, microbiologists,
laboratory experts, laboratory technologists, genetic counselors,
researchers, veterinarians, students, and the like. The remote
computers 28 may also be physically located in non-traditional
medical care environments so that the entire health care community
may be capable of integration on the network. The remote computers
28 may be personal computers, servers, routers, network PCs, peer
devices, other common network nodes, or the like, and may include
some or all of the elements described above in relation to the
control server 22. The devices can be personal digital assistants
or other like devices.
[0029] Exemplary computer networks 26 may include, without
limitation, local area networks (LANs) and/or wide area networks
(WANs). Such networking environments are commonplace in offices,
enterprise-wide computer networks, intranets, and the Internet.
When utilized in a WAN networking environment, the control server
22 may include a modem or other means for establishing
communications over the WAN, such as the Internet. In a networked
environment, program modules or portions thereof may be stored in
the control server 22, in the database cluster 24, or on any of the
remote computers 28. For example, and not by way of limitation,
various application programs may reside on the memory associated
with any one or more of the remote computers 28. It will be
appreciated by those of ordinary skill in the art that the network
connections shown are exemplary and other means of establishing a
communications link between the computers (e.g., control server 22
and remote computers 28) may be utilized.
[0030] In operation, a user may enter commands and information into
the control server 22 or convey the commands and information to the
control server 22 via one or more of the remote computers 28
through input devices, such as a keyboard, a pointing device
(commonly referred to as a mouse), a trackball, or a touch pad.
Other input devices may include, without limitation, microphones,
satellite dishes, scanners, or the like. Commands and information
may also be sent directly from a remote healthcare device to the
control server 22. In addition to a monitor, the control server 22
and/or remote computers 28 may include other peripheral output
devices, such as speakers and a printer.
[0031] Many other internal components of the control server 22 and
the remote computers 28 are not shown because such components and
their interconnection are well known. Accordingly, additional
details concerning the internal construction of the control server
22 and the remote computers 28 are not further disclosed
herein.
[0032] Although methods and systems of embodiments of the present
invention are described as being implemented in a WINDOWS or LINUX
operating system, operating in conjunction with an Internet-based
delivery system, one of ordinary skill in the art will recognize
that the described methods and systems can be implemented in any
system supporting the search of electronic medical records. As
contemplated by the language above, the methods and systems of
embodiments of the present invention may also be implemented on a
stand-alone desktop, personal computer, cellular phone, smart
phone, PDA, or any other computing device used in a healthcare
environment or any of a number of other locations.
[0033] Turning now to FIG. 2, a block diagram is illustrated that
shows an exemplary computing-system architecture 200 suitable for
searching an electronic medical record for components that are
responsive to a search query, in accordance with an embodiment of
the present invention. The computing-system architecture 200 shown
in FIG. 2 is merely an example of one suitable computing system and
is not intended to suggest any limitation as to the scope of the
use or functionality of the present invention. The computing system
architecture 200 should not be interpreted as having any dependency
or requirement related to any single component/module or
combination of component/modules illustrated therein.
[0034] The components in computing-system architecture 200 work
together to analyze and index an EMR and analyze a search query to
provide a user with the search results that are most responsive the
user's search query. Computing-system architecture 200 includes
clinical-concept component 210, clinical-usage component 215,
frequency component 220, specificity-factor component 225, truth
component 230, patient-subject component 235, clinical-importance
component 240, document-scoring component 245, search-engine
component 250, search-interface component 255, query-expansion
component 260, query-scoring component 265, EMR data store 270,
search-index data store 275, and clinical-concept data store 280.
Computing-system architecture 200 may operate in a single computing
device, such as control server 22 shown in FIG. 1. In the
alternative, computing-system architecture 200 may operate in a
distributed computing environment that includes multiple computing
devices coupled with one another via one or more networks. Such
networks may include, without limitation, one or more local area
networks (LANs) and/or one or more wide area networks (WANs). Such
network environments are commonplace in offices, enterprise/wide
computer networks, intranets, and the Internet. Accordingly, the
network, or combination of networks, is not further described
herein.
[0035] Clinical-concept component 210 may identify clinical
concepts within a search query and components of an EMR. The
clinical-concept component 210 may interact with the
clinical-concept data store 280 in order to identify clinical
components. The clinical-concept data store 280 contains at least
one clinical-concept nomenclature and may contain several. A
clinical-concept nomenclature is a vocabulary of medical
terminology. A vocabulary contains terms and expressions related to
a specific domain. Clinical nomenclatures may describe conditions,
symptoms, diagnostic procedures, organisms, substances, treatments
and other clinically relevant concepts. The data store 280 may also
contain local, non-clinical concepts such as facilities, locations,
care providers. A clinical-concept ontology may contain information
about the attributes and relations of the terms in a
clinical-concept nomenclature. At the simplest level, the
clinical-concept ontology organizes terms around clinical concepts,
and relates "synonyms" or terms that describe the same concept. For
example, the terms "heart attack" and "myocardial infarction"
describe the same clinical concept, and can be related as synonyms.
The ontology can also contain more complex relationships.
Specificity is expressed as a hierarchy of "is-a" relationships.
For example, the concept described as "left main coronary artery
disease" can be related to the more general concept of "heart
disease," of which it "is a" specific form. Additional relations
depend on the nature of the clinical-concept ontology. They can
include relations that link disease with similar causes, a disease
to a symptom that it causes, a medication to the disease it treats,
and many more. In one embodiment, SNOMED CT (Systematized
Nomenclature of Medicine-Clinical Terms) is used as the
clinical-concept ontology. SNOMED CT is a systematically organized,
computer-processable collection of medical terminology.
[0036] Clinical-concept component 210 may identify each of the
clinical concepts within an electronic medical record. The clinical
concepts may be identified by performing a textual analysis on the
EMR and matching phrases to clinical concepts within a
clinical-concept ontology. Clinical concepts may also be identified
by mapping codified data in the EMR to the clinical-concept
ontology. For example, the codified data describing a diagnosis may
be mapped to the clinical concept describing the diagnosis. Once
identified, the clinical concepts within the EMR may be stored in
an index used for searching the electronic medical record.
Additional information describing the use and context of the
clinical concept within the EMR may also be indexed. The additional
information may be provided by other components. The function of
those components is described subsequently. In one embodiment, each
use of a clinical concept is indexed according to the document or
component of the EMR in which the clinical concept is used.
[0037] The clinical-concept component 210 may also identify
clinical concepts within the search query using a similar
methodology. The clinical concepts identified in the search query
may be communicated to the search-engine component 250.
Clinical-concept component 210 may expand a specific
clinical-concept found in the search query by traversing the
clinical-concept ontology to find more general, more specific,
and/or sibling clinical concepts. The expanded set of clinical
concepts may be used to expand the search criteria and/or to expand
the information indexed that describes a particular document or
component in the EMR. For example, the primary clinical concept
identified in the search query and expanded clinical concepts may
be used to identify matching clinical concepts in an index.
Additionally, expanded clinical concepts may be indexed along with
the primary clinical concept used in the documents or components in
the EMR.
[0038] Clinical-usage component 215 performs part of an analysis
that is used to determine the importance of a particular use of a
clinical concept within an EMR. The clinical-usage component 215
ascertains the context or role in which a clinical concept is used
the EMR. The particular use of a clinical concept may be assigned a
pre-defined clinical-context category. The clinical-usage context
may be determined by ascertaining in what part of a document the
clinical concept is used, or how the clinical concept is used in
certain sentences. Examples of clinical-usage contexts include: the
presenting complaint, patient history, family history, review of
systems, physical exam, prescription, order, lab result, vital
sign, diagnosis, procedure, and others. This list of clinical-usage
contexts is not meant to be exhaustive.
[0039] The clinical-usage component 215 may perform natural
language processing and grammatical analysis to determine the
clinical-usage context. Document metadata may also be analyzed to
determine what role a portion of a document or a sentence in a
document plays in the electronic medical record. Other markers and
headings included within the EMR or within documents that are
within the EMR may be used to help identify the clinical-usage
context. The clinical-usage context may be stored in an index, such
as the index in search-index data store 275, along with the use of
the associated clinical concept. The clinical-usage context may be
used as a factor in assigning a clinical-importance score to a
particular use of a clinical concept. The determination of a
clinical-importance score is described in more detail
subsequently.
[0040] The frequency component 220 analyzes the frequency with
which a clinical concept occurs within an electronic medical
record. The frequency component 220 may assign a frequency score to
each use of a clinical concept. In one embodiment, the frequency
score is the result of a calculation. For example, the number of
uses of a particular clinical concept in an EMR divided by the
number of uses of all clinical concepts in the EMR. Clinical
concepts that occur alone in a document may be less important than
clinical concepts that occur multiple times in a document. In other
words, multiple appearances of the same or related clinical
concepts within a document may indicate that the clinical concept
is a main subject of the document. A clinical concept that occurs
rarely across multiple electronic medical records may have more
importance than common clinical concepts. In one embodiment, two
frequencies scores are calculated. The first frequency score is
based on occurrences of the clinical concept within a document in
the EMR and the second score is based on the number of occurrences
of the clinical concept within the entire EMR or across multiple
EMRs. The actual frequency score or a frequency rank may be stored
in an index with the associated the use of the clinical concept.
The frequency score may be used to determine a clinical-importance
score for the use of the clinical concept.
[0041] The specificity component 225 determines the specificity
factor of a clinical concept based on the degree of specificity,
precision or narrowness of scope of the concept, as derived from
the concept's position in a clinical ontology, or other reference
information. The specificity factor represents the precision, or
degree of detail of a clinical concept. For example, "heart
disease" is a general, or a fairly non-precise concept, whereas
"right coronary artery occlusion" (a form of heart disease) is a
precise concept. The specificity factor is a powerful tool for
ranking and evaluation because precise clinical concepts tend to be
more clinically interesting and significant, and also because the
documents that describe concepts in very precise language are
usually the documents that contain the most clinically interesting
information about that concept. The specificity factor may be
derived from a combination of the concept relations in a
clinical-concept ontology, and additional content and algorithms.
The specificity factor may be expressed as a numeric level within a
hierarchy. Alternatively, the specificity factor may be expressed
as a category. For example, the most general clinical concepts may
be designated with a specificity factor of "1" or as "low"
specificity. More precise concepts could be grouped as medium or
high. The specificity factor may be indexed in association with
each use of the clinical concept. The specificity factor may be
used as an additional factor to determine a clinical-importance
score for the clinical concept.
[0042] The truth component 230 performs a grammatical analysis to
assign a truth status to a use of a clinical concept. The truth
status indicates whether the use of the clinical concept is
positive, negative, ambiguous, or unknown. For example, the phrase
"the patient complained of chest pain" would create a positive
truth status for to the clinical concept "chest pain." On the other
hand, the phrase "the patient denies chest pain" would warrant a
negative truth status for "chest pain" because the patient said he
had not had chest pain. The truth status may be stored in an index
in association with the particular use of the clinical concept. The
truth status may be used to determine an importance score for a
particular use of the clinical concept. A negative truth status may
make a use of a clinical concept less important. A positive truth
status may increase the importance score calculated for the
particular use of the clinical concept.
[0043] The patient-subject component 235 determines the
patient-subject status of a use of a clinical concept. The
patient-subject status indicates whether the patient is the subject
being described by the clinical concept. For example, a medical
history indicating the patient's father died of cancer would not
have the patient as the subject of the clinical concept "cancer"
and the patient-subject status would be false. On the other hand,
the patient is the subject of the clinical concept "chest pain" in
a reference to the patient complaining of chest pain in a
presenting complaint. In this case, the patient-subject status
would be true. The clinical subject may be stored in an index in
association with the particular use of the clinical concept. The
patient-subject status may be used as an additional factor to
determine a clinical-importance score for the use of the clinical
concept.
[0044] The clinical-importance component 240 assigns a
clinical-importance score to each use of a clinical concept within
the electronic medical record. A use of a clinical concept is a
single occurrence of the clinical concept within the EMR. A single
clinical concept may be used multiple times within a single
document within the EMR as well as within other documents in the
EMR. Each use of a clinical concept may be assigned a
clinical-importance score. The purpose of the clinical-importance
score is to quantify how important a particular use of a clinical
concept is likely to be to a person searching for components of the
EMR in which the clinical concept is used. The more important the
use, the more likely the document containing the use is interesting
to the person submitting the search query. However, the
clinical-importance score is calculated based on the use of a
clinical concept within the EMR and without reference to any
specific search query.
[0045] The clinical-importance score may be calculated based on
several factors. Different embodiments of the present invention
combine different factors to calculate each clinical-importance
score. It may not be necessary to use every factor explained herein
to calculate the clinical-importance score. Further, different
weights may be given to different factors when calculating the
clinical-importance score. Embodiments of the present invention are
not limited to the specific examples given. In one embodiment, the
clinical-importance score is calculated based on the clinical-usage
context, the truth status, the patient-subject status, the
specificity factor, and the frequency associated with the
particular use of the clinical concept in this document compared to
other matching documents in the patient's EMR. In another
embodiment, the clinical-importance score is calculated based on
the truth status, the patient-subject status, and the
clinical-usage context. In another embodiment, the
clinical-importance score is calculated based only on the
clinical-usage context.
[0046] As stated previously, different weighting may be given to
the different factors used to calculate the clinical-importance
score. Regardless of the weight given to a factor in a particular
embodiment, in general, the factors may increase or decrease the
ultimate score as explained subsequently. When the patient is the
subject of the use of the clinical concept, the importance score
may be increased. A negative truth status for a use of a clinical
concept may lower the importance score. The importance score may be
raised if the clinical-usage context is in a category that is
authoritative and related to the patient. For example, if the
clinical-usage context is the presenting complaint or the patient
history, the importance score could be raised more than if the
clinical-usage context is a patient's family history, since the
family history is less directly related to the patient. Different
categories of clinical-usage context may be given different values
to plug into the calculation of the clinical-importance score. In
general, the more closely the clinical concept is related to the
patient and the more authoritative, the higher the importance score
will be. A relatively high frequency of occurrence in a document
within the EMR compared to other documents may increase the
importance score because the clinical concept is more likely to be
the subject of the document, and thus, more important than a
clinical concept that may be mentioned tangentially or that is used
indiscriminately in many documents.
[0047] Document-analysis component 245 gathers attributes
describing components of the electronic medical record. In one
embodiment, the components of the EMR are analyzed at the document
level of granularity. A component of the EMR may contain multiple
uses of one or more clinical concepts. The document or component
attributes may be stored in the index and used to calculate
document-boost factors that increase a responsiveness rank of a
matching component within the search results.
[0048] In addition to collecting document attributes, the
document-analysis component 245 may also assigns each document one
or more primary-focus domains and one or more generic-document
domains. For example, the primary-focus domain may be "physician"
for a particular document authored by a physician. Other examples
of primary-focus domains include "nurse-focused" and
"social-worker-focused." The primary-focus domain and the
generic-domain may be used for calculating boost factors. The
generic-document domains may include a designation that the
document is a clinical document, lab result, vital sign, problem,
order, or other clinical event. The primary-focus domain and
generic-documents domains may be stored in an index.
[0049] Many different attributes and collections of attributes may
be used to calculate document-boost factors. Document-boost factors
include a source-boost factor, a class-boost factor, and a
document-type-boost factor. A source-boost factor may be calculated
based on the source of a document. The source-boost factor may be
increased or decreased according to the importance of the source of
the document. For example, the source-boost factor may be increased
if the document is a primary source, such as a discrete lab result
in contrast to a secondary source, such an end-of-the-day summary
that repeats the lab result. In addition to the source-boost
factor, a class-boost factor may be used. The class-boost factor is
based on the importance of a document as determined by the
importance of the class of which the document is a part. The
class-boost factor may be increased if the document class is a
"clinician-authored document." In contrast, the class-boost factor
may be lowered if the document class is "a procedure note," or is
authored by a non-clinician. In addition to the class-boost factor
and source-boost factor, the document type may be considered when
calculating a document-type-boost factor. For example, the
document-type-boost factor may be increased if the document is a
discharge summary or transfer summary. Both the discharge summary
and transfer summary tend to be more important because they contain
authoritative summaries of a patient's progress. Similarly, an
admission note may be more important because it may define the
cause of a new episode. Daily progress notes, medical student
notes, chart abstractor notes, and other notes may be less
important and would tend to decrease the boost factor.
[0050] The domains may also be incorporated into document-boost
factors or query-boost factors that are used as part of the
calculation of the query-responsiveness score. For example, the
domains may be used in combination with information related to the
query to calculate a role-boost factor. The role-boost factor is
based on matching a search query submitted by a particular category
of clinician (e.g., nurse, doctor, social worker) to a document
associated with the same category of clinician. For example, if the
search query is submitted by a nurse, then the document's
query-responsiveness score may be increased, through a boost
factor, if the primary-focus domain is nurse. Similarly, if a
search query is entered by a social worker, then the
query-responsiveness score may be increased if the primary focus of
the document is social worker.
[0051] Search-engine component 250 receives a search query and
retrieves documents or components of documents from the electronic
medical record that are responsive to the search query. A document
or component of a document is responsive to the search query when a
portion of the document or component of a document matches the
search query or an expansion of the search query. Search-engine
component 250 may use the search index in search-index data store
275 to find search results. The search results may be presented in
an order intended to present the most responsive results first. The
most responsive results may be determined using a
query-responsiveness score (to be discussed subsequently). The
search-engine component 250 may present search results that
describe component of an EMR, a document in an EMR, or a component
of a document in the EMR. A component of a document may be one or
more words within the document.
[0052] Search-interface component 255 presents a graphical user
interface to a user for receiving search criteria and presenting
search results. An example of a graphical user interface is shown
in FIGS. 6 and 7. These figures will be described in more detail
subsequently. In addition, the search-interface component 255 may
also present filter options.
[0053] Query-expansion component 260 expands the search query to
create a plurality of expanded search terms. The plurality of
expanded search terms are used by the search-engine component 250
to find and rank additional search results. The degree of expansion
performed by the query-expansion component 260 may differ depending
on the search mode selected. For example, in a text only search
mode, the query-expansion component 260 may be limited to only
stemming each word within the search query. Stemming a word reduces
the word to its root. For example, the stem of "dogs" is "dog," and
the stem of "changing" is "change." Query-expansion component 260
may, in conjunction with other components, expand terms in the
search query. For example, the plurality of expanded search terms
may include medical synonyms of words submitted in the search
query. In addition, clinical concepts identified within the search
query may be expanded to include sibling, child, and/or related
clinical concepts from the clinical-concept ontology. The
query-expansion component 260 may transmit the search terms and
expanded search terms to the search-engine component 250.
[0054] Query-scoring component 265 assigns a query-responsiveness
score to each search result returned in response to a search query.
As described previously, a search result may be any component of
the EMR, including a document or a part of a document. For the sake
of simplicity, the calculation of a query-responsiveness score will
be described using a document as a search result. However, a
query-responsiveness score may be calculated for any portion of the
EMR for which the necessary information is available. The search
results may be received from the search-engine component 250
[0055] The query-responsiveness score for an individual document
matching the query may be calculated by combining the
clinical-importance scores assigned to each use of a clinical
concept in the individual document with one or more document-boost
factors and query-boost factors. The query-scoring component 265
may calculate one or more boost factors including the
closeness-boost factor.
[0056] A closeness-boost factor may be calculated for each matching
combination of clinical concepts when the primary clinical concept
in the query is expanded. A matching combination of clinical
concepts includes a clinical concept related to the query and a use
of the same clinical concept in the particular document. The
clinical concept related to the query may be the primary clinical
concept and/or expanded clinical concepts. The closeness-boost
factor may be used to increase the query-responsiveness score for
combinations including the primary clinical concept and expanded
clinical concepts that are close to the primary clinical concept.
Since a user's query may be exploded into a set of synonyms, child,
and related concepts, in one embodiment, the closeness-boost
factors are assigned to combinations with the primary clinical
concept and to combinations with the expanded clinical concepts
such that combinations with the primary clinical concept have the
most importance, and combinations with the more distant `related`
concepts have less importance. In general, the closer on the
clinical-concept ontology the expanded clinical concept is to the
primary (e.g. original) clinical concept, the higher the
closeness-boost factor. In contrast, the further away on the
clinical-concept ontology that the expanded clinical concept is,
the lower the boost factor. The closeness-boost factors are
combined with the clinical-importance scores to determine the
query-responsiveness score of the document or component. For
example, a document including a clinical concept with a high
clinical-importance score matching to an expanded clinical concept
that is a sibling of (i.e., close to) the primary clinical-concept
would receive a higher query-responsiveness score than when the
same clinical concept is combined with an expanded clinical concept
several layers away from the primary clinical concept. A boost
factor based on the closeness of the clinical concept from the
query is only used when the query is expanded. Other boost factors
may be used to calculate into the query-responsiveness score when
the query is not expanded or in combination with the
closeness-boost factor given combinations of the primary-clinical
concept and closely related expanded-clinical concepts.
[0057] In addition to the closeness-boost factor, additional boost
factors may be used based on other document and/or query
attributes. For example, a role-boost factor based on the role to
the person submitting the query and the role of the person
associated with a component of the EMR may be used. The role-boost
factor may be calculated by the query-scoring component 265. The
role-boost factor may be increased for a document with a
primary-focus domain of "nurse" when the query is submitted by a
nurse. Similarly, the class-boost factor used. As described
previously, the class-boost is based on the importance of a
document-class. For example, the class-boost factor may be higher
if the document class is "physician authored" and lower if the
document class is "medical-student authored." In another
embodiment, a boost factor is time weighted. The
time-weighted-boost factor increases for documents that are more
recent. In yet another embodiment, the clinical-facility-boost
factor increases if the document was generated in the same clinical
facility as the one from which the query is submitted. Additional
boost factors may be calculated based on one of more of the
previously described attributes. All of the previously described
boost factors may be combined, but not all boost factors need be
used to calculate a query-responsiveness score. A
query-responsiveness score may be calculated using just one, or
none, of the previously described boost factors.
[0058] The Filter component 267 allows the user to filter the
search results according to suggested or submitted criteria. For
example, the filter component may provide an interface allows the
user to filter search results according to date, document class,
clinical facility, and the document's primary focus. In one
embodiment, the filter component 267 suggests a filter criteria for
the user to select along with an indication of how many of the
search results match the filtered criteria. For example, the
interface could indicate that 20 search results are in the document
class "physician authored." In one embodiment, the search results
may be filtered by clinical concepts found within the search
results. Related clinical concepts may be aggregated into a general
filter option that would present search results that include any of
the related clinical concepts. The filter options could be
presented with the clinical concepts having the highest aggregation
of clinical-importance scores. These filter examples are not meant
to be exhaustive, other filters based on factors store in the index
are within the scope of this disclosure.
[0059] The Electronic medical record data store 270 contains the
electronic medical records for one or more patients. An EMR is a
collection of information describing the medical history of a
patient. In addition, the EMR data store 270 may include electronic
medical records from one or more clinical facilities. The EMR data
store 270 may be accessed by other components within
computing-system architecture 200.
[0060] Search-index data store 275 includes a search index which
stores the words and the clinical concepts extracted from the
patient's documents as described above. The search index may be
isolated on a per patient or per clinical facility basis. In other
embodiments, multiple patients and even multiple EMRs may be
searched concurrently.
[0061] Clinical concept data store 280 contains one or more
clinical-concept ontologies as described previously. In one
embodiment, the clinical-concept ontology is based on a combination
of SNOMED CT (to represent clinical conditions, symptoms, therapy,
organisms, etc.) and RxNorm (to represent medications). Embodiments
of the present invention are not limited to using SNOMED CT. Other
hierarchies of medical terminology may be used.
[0062] Turning now to FIG. 3, a method 300 of preparing electronic
medical records for electronic searching is provided, in accordance
with an embodiment of the present invention. At step 310, an
electronic medical record that includes information describing at
least a portion of a medical history associated with a patient is
received. As described previously, an electronic medical record
includes information that describes a patient's medical conditions,
treatments received, and other medical information for the patient.
The EMR may be recorded by clinicians, the patient, or others.
Electronic medical records may be generated by a single clinical
facility or by multiple clinical facilities. The electronic medical
record may be stored in a data store that is accessible to a single
clinical facility or to multiple clinical facilities.
[0063] In one embodiment, any information in the electronic medical
record that is not already formatted as an displayable document is
converted into a new displayable document. The electronic medical
record may contain structured and/or codified data including
database entries. In one embodiment of the present invention, the
database entries are converted into electronic documents. The
converted electronic documents include entries from the database
fields and descriptions of the database fields. Thus, once
non-document portions are formatted, an electronic medical record
may consist of a plurality of electronic documents that are
displayable by a search engine as a search result. In one
embodiment, the newly created electronic documents may be used as
component of the EMR to associate with boost factors.
[0064] At step 320, clinical concepts in the electronic medical
record are identified. A clinical concept describes any aspect of a
person's health condition, or any object, action, attribute or idea
that is related to a health condition. Examples include: diseases;
symptoms; clinical observations and findings; diagnostic tests;
diagnostic or therapeutic procedures; organisms, substances,
devices or products related to health conditions; anatomic
structures including genomic; phenotypic expression; behavior,
family and social context related to health conditions; risk
factors and outcomes; facilities and care providers. For example,
heart disease and a heart attack are examples of clinical
concepts.
[0065] The identification of clinical concepts in a document that
is part of an electronic medical record is illustrated in FIG. 8.
For the sake of clarity, only a few of the clinical concepts and
significant words are shown. The electronic document 800 in FIG. 8
is a summary of a patient's visit to a doctor's office. The
electronic document 800 includes a document type 810. The document
type 810 field is a description of what category the electronic
document is in. In this case, the document type 810 is an office
visit. The document date 812 is Oct. 4, 2005. The document title
814 is "Mr. X., pepperoni pizza with hot peppers." The document
author 816 is "T. Jones, M.D." The document source 818 is "River
Heights Medical Association." These document attributes may be
associated with each word or use of a clinical concept within the
document. These document attributes may also be included as
document metadata. Factors, such as the clinical-usage context may
be determined using these factors.
[0066] The words in electronic document 800 that are in blocks may
be clinical concepts. The blocked words include "peppers" 820,
"heartburn" 822, "peppers" 824, "heartburn" 825, "heart attack"
826, "G.I. ulcer" 828, "abdominal hernia repair" 830, and "heart
disease" 832. These words may be selected based on analysis that
looks for medical terms, keywords and other words/terms of
interest.
[0067] Returning now to FIG. 3, at step 330, a patient-subject
status for each use of a clinical concept of the electronic medical
record is determined. The patient-subject status indicates whether
the patient is the subject of the clinical concept. For example,
the patient is the object of the sentence "he has never been
diagnosed with a G.I. ulcer" in electronic document 800. The
patient-subject status may be determined using a grammatical
analysis of text in which the clinical concept occurs. Each
instance of a clinical concept may be associated with a different
patient-subject status. The patient-subject status may be recorded
as true\false. True indicates that the patient is the subject of
the clinical concept. In another embodiment, the patient-subject
status may be recorded with the name of the subject associated with
the clinical concept. For example, the patient-subject status may
be recorded as "patient's father."
[0068] At step 340, a truth status of the particular clinical
concept is determined. The truth status indicates whether the
clinical concept was expressed positively, negatively, ambiguously,
or unknown. For example, referring to electronic document 800 "the
worst heart burn I ever had, so bad I thought I was having a heart
attack," refers positively to heartburn and ambiguously to heart
attack. In this example, the truth status could be recorded as
positive for both heartburn and ambiguous for heart attack.
[0069] At step 350, a clinical-usage context of each clinical
concept is determined. The clinical-usage context describes how the
clinical concept was used in electronic medical record. The
clinical-usage context may be determined based on the section of
the electronic document in which the use of the clinical concepts
occurs or by analysis of the grammar of the sentence in which the
concept is used. The clinical-usage context may also be determined
based on the type of document in which the use of the clinical
concept occurs. For example, "heartburn" 825 in electronic document
800 is in the clinical-usage context of the presenting complaint.
"Abdominal hernia repair" 830 in electronic document 800 is in the
clinical-usage context of a patient medical history. Each clinical
concept may be categorized into one or more of a predefined group
of clinical-context categories. Examples of clinical-usage contexts
include a presenting complaint section, a patient history, a family
history, a review of systems section, a physical exam record, a
prescription, order, a lab result, a vital sign, a diagnosis, and a
procedure record.
[0070] At step 360, a specificity factor is determined for the
particular clinical concept based on the degree of specificity,
precision or narrowness of scope of the concept, as derived from
the concept's position in a clinical ontology, or other reference
information. As described previously, the clinical-concept ontology
describes relationships between clinical concepts. The specificity
factor could be recorded as a group category, such as high medium,
or low. The specificity factor could also be recorded as a level in
the clinical-concept ontology.
[0071] In one embodiment, a document-importance factor is
determined for the particular use of the clinical concept. The
document importance is the relevance of the particular clinical
concept to the main subject of the document. This is determined by
clustering the clinical concepts in the document into clinical
categories. In the case of diseases and symptoms, the cluster of
related concepts would correspond to body-system-condition
categories. The frequency of references to each category is
evaluated to determine the "subject(s)" or areas of focus of the
document. Weighting is also applied based on the clinical usage of
the concepts. For example, consider two documents that each contain
one use of the concept "heart disease."One document also contains
many concepts related to heart disease, such as the diagnostic test
"echocardiogram," the medication therapy "statin," and the clinical
finding "S-T segment depression." In addition, the therapeutic
intervention "cardiac catheterization" is contained in the
significant clinical usage of "assessment and plan." In the second
document, "heart disease" is also mentioned, but without any
closely related concepts. The "document importance" of the concept
"heart disease" is high in the first instance, low in the
second.
[0072] At step 370, a clinical-importance score is computed and
assigned to each clinical concept. The clinical-importance score
may be determined based on the patient-subject status of the
particular clinical concept, the truth status of the particular
clinical concept, the clinical-usage context of the particular
clinical concept, and the specificity factor of the particular
clinical concept. The various factors may be combined and given
different weights to arrive at the clinical-important score. In
other embodiments of the present invention, additional factors are
used to calculate the clinical-important score. For example, as
described in FIG. 2, the frequency of a particular clinical concept
within the electronic medical record or a plurality of electronic
medical records may be used.
[0073] At step 380, the clinical-importance score associated with
each clinical concept is stored in the document index. The index
may store a clinical-importance score for each use of a clinical
concept in association with the document in which each clinical
concept is used. In another embodiment, the index may store a
clinical-importance score for each use of a clinical concept in
association with a component of the EMR in which each clinical
concept is used. The storage of clinical-important scores in a
searchable index is illustrated by FIG. 9. The index 900 in FIG. 9
includes text 910 from electronic document 800 shown in FIG. 8. The
first row of text 910 includes "heartburn" 822, "heartburn" 825,
"heart attack" 826, "heart attack" 828, "abdominal hernia repair"
830, and "G.I. ulcer" 832. Words such as "pepper" 820 are not
included in the concept index because they are not associated with
clinical concepts, thought they may be stored in an ordinary "word"
index so that they can help refine searches for subjects that
aren't clinical concepts. The next row lists the clinical concept
912 with which the word is associated. For example the word
"heartburn" 822 is associated with clinical concept "heartburn"
913. Similarly "heartburn" 825 is associated with "heartburn" 914.
Both "heartburn" 913 and "heartburn" 914 have the same SNOMED CT
code. As described previously, the SNOMED CT is a hierarchy of
medical terminology. Embodiments of the present invention are not
limited to using the SNOMED CT medical hierarchy. Because each use
of a clinical concept within the electronic medical record may be
indexed, the same clinical concept may be indexed several times.
The scores given to each use of a clinical concept may be
different.
[0074] Continuing with FIG. 9, clinical concepts 912 include
"myocardial infarction" 915, "myocardial infarction" 916, "repair
of hernia of abdominal wall" 917, and "gastrointestinal ulcer" 918.
Each clinical concept includes a series of attributes that are
determined as described previously. The truth status 920,
clinical-usage context 930, patient-subject status 940, clinical
concept type 950, specificity factor 960, and document importance
970 are all recorded. Finally, the clinical-importance score 980
for each clinical concept is recorded. In the embodiment shown in
FIG. 9, the clinical-importance score is summarized as a category
of high, medium, or low. In another embodiment, the
clinical-importance score could be recorded as a numerical value
that is the result of a calculation.
[0075] Turning now to FIG. 4, a method 400 of finding information
in electronic medical record is shown, in accordance with an
embodiment of the present invention. At step 410, a search query is
received from a clinician. An interface to receive a search query
is illustrated in FIG. 6. The search interface 600 includes a
patient-banner bar 610 to display identification information for
the patient whose electronic medical record is being searched. The
tab section 620 allows a clinician to navigate to various functions
of the search interface 600. Several tabs are includes in the tab
section 620. The tab section 620 includes a today tab 621, a
documents search tab 622, a lab\vitals search tab 623, a graphing
tab 624, a spark lines tab 625, and an IntelliStrip tab 626. The
documents search tab 622 is shown as selected in FIG. 6. The search
interface 600 includes a search input area 630. In an embodiment of
the present invention, a list of suggested clinical concepts 632 is
presented in response to entering a partial search query 631. Each
suggested clinical concept includes the name of the clinical
concept and the SNOMED CT identification number. The clinical
concept "heart disease" 633 is paired with ID number 634. The
clinical concept "heart disease due to ionizing radiation" 635 is
paired with ID number 636. The clinical concept "heart disease due
to radiation" 637 is paired with ID number 638. The clinical
concept "heart disease during pregnancy" 639 is paired with ID
number 640. The clinical concept "heart disease excluded" 641 is
paired with ID number 642. The clinical concept "heart disease in
mother complicating pregnancy, childbirth and/or puerperium" is
paired with ID number 644. The start of each clinical concept
begins with "heart di." the list of suggested clinical concepts 632
will show fewer clinical concepts as additional letters are typed
into search input area 630.
[0076] Returning to FIG. 4, at step 420, one or more components of
the electronic medical record that match the search query are
identified. A component is a section of text including one or more
words/terms. Embodiments of the present invention allow a user to
specify the type of match required to return a search result. For
example the user may select a text-only match. With a text-only
match, stems of words in the search query must match stems of words
within the electronic medical record to generate a search result.
In another embodiment, the query terms are expanded to include
medical synonyms of the query terms. The medical synonyms are used
as additional query terms to match components of an electronic
medical record to generate a search result. For example, "heart
attack" and "myocardial infarction" may be medical synonyms. A
search result would be identified if either "heart attack" or
"myocardial infarction" were found in the electronic medical
record. In another embodiment, the user may request matches based
on "child" clinical concepts. At query time, the clinical concept
identified in the search query would be expanded to include
"children" of the base concept as defined by the associated
clinical-hierarchy. These child concepts would also be used to
match clinical concepts identified within the electronic medical
record to produce a search result. In another embodiment, the user
may search based on related concepts. With the related concepts
choice, the original query term is expanded to include additional
related clinical concepts which may be matched to documents in the
patient's electronic medical record. The text-only matching option
should return the least results and the related-concepts option
should return the most results.
[0077] FIG. 6 illustrates an interface 600 for the user to select
the matching criteria. The matching interface includes a text-only
mode 652, a clinical concept mode 654, and related concepts mode
656. In one embodiment, the search results are dynamically updated
as the user toggles between different matching criteria.
[0078] Returning to FIG. 4, at step 430, a query-responsiveness
score is determined for each of the one or more components that
match the search query. The query-responsiveness score indicates
how responsive the individual component is to the search query. As
described previously, the query-responsiveness score may be
calculated based on a combination of the clinical-importance score
of the matching clinical concepts with the one or more components
and the associated boost factors, as described above. The
calculation of a clinical-importance score and the various factors
that may be considered when calculating the clinical-importance
score have been described previously. Similarly, determining the
boost factors has also been described previously.
[0079] At step 440, search results that communicate information
describing each of the one or more matching components are
displayed. As described previously, the matching components may be
documents, portions of documents, or other portions of the EMR. The
search results are displayed ordered according to the
query-responsiveness score assigned to each of the one or more
components.
[0080] FIG. 7 illustrates the display of search results that are
responsive to a submitted clinical concept. Result interface 700
includes a search result display area 720. The search results
display area 720 includes two search results, both of which are
documents. Though not shown in FIG. 7, the search results may be
components of a document. Each search result is displayed with an
expansion column 726. Selecting the expansion button shows other
documents that are similar to search result. Each search result
also has a rank 728, a query-responsiveness score 730, an document
age 732, a date 734, clinical-context information 736, a document
title 738, the clinician that authored or authenticated the search
result 740, and text summarizing the search result 742. The first
search result is displayed first because it has the highest
query-responsiveness score, which in this case is 1.00 746. The
highest ranked search result is designated as number 1 744 in the
list. The document age is 3 yrs 748, the date of the document is
Oct. 4, 2007 750, the clinical-usage context of the document 752 is
"Cardio Assoc. of Olathe/MD Consult." The title of the document 754
is "Cardiac Cath Evaluation." The clinician is Tom Clark 756. The
text excerpts 722 for the first search results includes use of the
clinical concept heart disease 633. The search results are based on
matching criteria of related concepts 656. Clinical concepts
related to heart disease are outlined in the text excerpts. The
outlined text includes "heart disease" 758, "heart disease" 759,
"heart disease" 760, "coronary artery disease" 761, and "occluded
right coronary artery" 762.
[0081] The second search result is designated as search result
number two 768, and has a query-responsiveness score 769 of 0.540.
The document age is 3 yrs 770, and the date is Oct. 4, 2007. The
clinical-usage context is "Beacon Health/MD Consult" 772 and the
title of the document is "Card consult: exertional angina" 773. The
clinician responsible for the second search result is Jordan Jones
774. The text excerpts 724 show outlined text that matches the
search query 733. The outlined text includes "heart disease" 775,
"coronary artery disease" 776, "angina" 777, "angina" 778,
"coronary artery disease" 779, "heart disease" 780, and "sinus BR"
781.
[0082] The result interface 700 also includes a filter interface
785. The filter interface 785 allows the clinician to filter by
year 786, encounter locations 787, or document class 788. The
number of search results that match a particular filter criteria
are displayed in parentheses adjacent to the suggested filter
criteria. For example, 17 search results are available for the year
2007. The result interface 700 also includes query reset button 710
that allows the user to submit a new query. A
sort-documents-by-date button 712 is also included in the search
result interface 700.
[0083] Turning now to FIG. 5, a method 500 of searching an
electronic medical record for a selected clinical concept is
provided, in accordance with an embodiment of the present
invention. At step 510, a search query that includes a clinical
concept is received from a user. As described previously, the
clinical concept describes any aspect of a person's health
condition, or any object, action, attribute or idea that is related
to a health condition. Examples include: diseases; symptoms;
clinical observations and findings; diagnostic tests; diagnostic or
therapeutic procedures; organisms, substances, devices or products
related to health conditions; anatomic structures including
genomic; phenotypic expression; behavior, family and social context
related to health conditions; risk factors and outcomes; facilities
and care providers. For example, heart disease and a heart attack
are examples of clinical concepts. The search query may be received
through an interface, such as search interface 600 described above
with reference to FIG. 6. The clinical concept may be entered
directly into the search interface word-for-word. In another
embodiment, the clinician selects the clinical concept from a list
of clinical concepts. In another embodiment, clinical concepts are
derived from natural language text submitted into the search
interface. The derivation of clinical concepts from natural
language text may utilize a medical synonym dictionary. Any and all
such variations, and any combination thereof, are contemplated to
be within the scope of embodiments of the present invention.
[0084] At step 520, one or more uses of the clinical concept from
the search query are identified in the electronic medical record.
The electronic medical record is an electronic description of a
medical history for a patient and is stored on one or more
computer-readable media. In one embodiment, the clinical concepts
within a medical record had previously been extracted and indexed.
In this case, the index is searched for clinical concepts that
match the clinical concept within the search query. As described
previously the clinical concept from the search query could be an
expanded-clinical concept.
[0085] At step 530, a query-responsiveness score is determined for
each document within the electronic medical record that uses the
clinical concept contained within the user's query. The
query-responsiveness score describes how important a particular
document or component is likely to be to the clinician. As
described previously, the query-responsiveness score may be
determined based on the combination of clinical-importance scores
for each of the one more uses of the clinical concept and a set of
boost factors calculated based on the expansions of the search
query submitted by the clinician. The factors, such as
clinical-usage context, specificity factor, frequency, truth
status, and others that may be used to calculate the
clinical-importance score have been described previously.
Similarly, the factors utilized to calculate the boost factors have
also been described previously. For example, the closeness-boost
factor may be increased when the clinician that submitted the
search query is in the same category of role as the clinician
associated with the use of the clinical concept.
[0086] At step 540, search results are displayed based on the
query-responsiveness score associated with each of the one or more
uses of the clinical concept. The search result with the highest
query-responsiveness score may be displayed first. Search results
associated with lower query-responsiveness scores may be displayed
subsequently.
[0087] As can be seen, embodiments of the present invention allow a
user to submit a search query through an interface and return
search results from an electronic medical record that are
responsive to the search query. The search results by default will
be ordered according to a query-responsiveness score so that the
clinician will find the most important document at the top of the
list.
[0088] It will be understood that certain features and
subcombinations are of utility and may be employed without
reference to other features and subcombinations and are
contemplated within the scope of the claims. Not all steps listed
in the various figures need be carried out in the specific order
described.
* * * * *