Frequently Asked Questions
Background
What is BigMouth?
BigMouth is a centralized
oral health data repository derived from electronic health records (EHRs) at multiple
dental schools participating in the Consortium of Oral Health Research and
Informatics (COHRI). With demographic, medical history and dental data on over 3 million
million patients, this oral health database is available for research and the
advancement of evidence-based dentistry.
Who are the participating institutions?
As of December 2020, we
have ten institutions who have contributed data to BigMouth. That includes Harvard School of Dental Medicine, University of California San Francisco School of Dentistry, University of Michigan School of Dentistry, UTHealth
School of Dentistry at Houston, Tufts University School of Dental Medicine, University of Colorado School of Dental Medicine, Loma Linda University School of Dentistry, University of Buffalo School of Dental Medicine, The University of Iowa College of Dentistry, The University of Minnesota School of Dentistry and
University of Pittsburgh School of Dental Medicine). We are expecting more institutions
to participate in the coming years.
Data
What data do I have access to?
Upon login, a user is
presented with two folders. The folders are:
a.
axiUm
- Contains data from the user's own local axiUm
Instance.
b.
COHRI - Contains data from all the
participating sites. This folder is normally used when a user is interested in
querying data from all COHRI members. The folder is designed in such a way that
the user is able to get a summary from all sites at the same time keeping the
source of data anonymous.
What does axiUm folder contain?
The
axiUm node contains the (user's) site specific
terminologies. Users can query for data available in axiUm
at their own site. The number displayed adjacent to each folder/leaf node
represents the number of patients.
'
Demographics:
contains age, gender, and race/ethnicity. User can query age by value with the
following operators: <, >, ≥, ≤, =, between.
'
Diagnosis:
contains Dental Diagnostic System (DDS) terminologies, formerly known as the EZCodes diagnostic terminology. User can query a diagnosis
with or without date range.
'
Forms:
contains medical, dental history of patients, and caries risk assessment. In
addition, each site may have other specific forms. User can query a form question
with or without date range.
'
Insurance:
contains the names of insurance companies that insure patients.
'
Odontogram: contains
existing materials, existing conditions on a specified tooth or surface and the
total number of missing teeth.
'
Perio:
contains clinical periodontal parameters which could be queried by examination
types (e.g. initial examination, re-evaluation examination).
'
Medication:
contains prescription and patient's current medication information and can be queried with or without
date range.
'
Practice:
contains details about the different practices in a site.
'
Procedures:
contains Dental Procedure Codes (CDT) and Current Procedural Terminology (CPT)
procedures which could be queried with or without date range.
'
Providers:
contains different dental health providers such as dental student, dentist,
resident etc.
How is the age calculated?
Age
is calculated from the date of patient's last recorded observation (procedure,
diagnosis etc.).
What does the COHRI folder contain?
This
folder provides an integrated terminology system that allows users to query for
data from all schools that contribute data to BigMouth.
'
Demographics:
contains age, gender, and race/ethnicity from all schools.
'
Diagnosis:
contains DDS terminologies from Harvard, UCSF, and UTH.
'
Forms:
contains medical and dental history from all schools. COHRI terminology
contains caries risk assessment from all schools except UPITT.
'
Odontogram: contains
existing materials, existing conditions on a specified tooth or surface and the
total number of missing teeth from all schools.
'
Perio:
contains clinical periodontal parameters from all schools.
'
Procedures:
contains CDT and CPT procedures from all schools.
'
Providers:
contains different dental health providers from all schools.
Where is the Race/ethnicity standard derived from?
Race/ethnicity
standard derived based on NIH guidelines.
Query
How can I get data for my research from BigMouth?
The BigMouth interface
provides query tools to user to get the number of patients having conditions or
diseases of interest such as number of patients with caries or periodontal
diseases, etc...
There are two levels of
access:
Level 1: Users who have
access to BigMouth can run queries themselves and use the count for their
research.
Level 2: If the user
wants to get detailed data for each individual patient, a3\ copy of the IRB
approval document will need to be provided along with the data request
template. The template will facilitate us to extract the requested data which
will be presented in an Excel sheet with requested variables in columns and
individual patients in rows.
In addition to it, the
users are also required to sign a data access request form. If the user is
interested in accessing patient data from outside the school to which the user
is affiliated, in addition to the IRB approval document from the user's school,
the project will need to be approved by the COHRI project review committee. The
following documents contain all the required information:
a.
Clinical Research Committee Checklist
b.
COHRI Clinical Research Committee Proposal
Guidelines
Please contact the
BigMouth team for details of the form.
How can I make a 2x2 table using BigMouth? - For example, I am interested in a diagnosis A and a treatment B. I would like to see how many patients with that diagnosis A had that treatment, and how many did not. - I'd also like to see how many without that diagnosis A had that treatment B how many did not.
This can be achieved
using the exclude/include option in each column. Here are few screen shots:
|
Yes Diagnosis, Yes
treatment |
No Diagnosis, Yes
treatment |
|
|
|
|
Yes Diagnosis, No
treatment |
No Diagnosis, No
treatment |
|
|
|
How can I make a temporal query?
There
are five basic steps in defining a temporal query in the Query
Tool view:
- Change Temporal
Constraint to Define sequence of Events.
- Define Population in
which events occur (optional step).
- Define Events
- Define order of events (temporal relationships)
- Run the query
Step 1: Change Temporal Constraint
The first step is to change the Temporal
Constraint to Define sequence of Events.
Step 2: Define Population in which events occur
Once you have changed the Temporal
constraint to be Define sequence of Events a new Page
selection box will appear below the Temporal Constraint section. The
default page will be Population in which events occur. It is on
this page you will define your population requirements.
Step 3: Define Events
The events are the first component of a
temporal query. There are no restrictions on the number of events you wish to
define. The only requirement is that you have to define at least two events.
To define the events simply click on the Page
selection box and select Event 1 from the drop-down
list.
The groups and constraints for the events work
in the same manner as they did for a traditional i2b2 query. Simply drag the
items you want to include in Event 1 to the appropriate groups.
Once you have added your items to the groups
you can click on the Page selection box and select Event
2 from the drop-down list. This will change the page to display the
groups for Event 2.
If you need to add a third event you can click
on the New Event button located next to the Page
selection box.
Step 4: Define Order of Events (Temporal
Relationship)
As stated earlier the second component of a
temporal query is the relationship between the events (temporal relationship).
In the i2b2 Web Client this is done on the Define order of events page,
which is accessed by clicking on the Page selection box and
selecting Define order of events from the drop-down list.
The page will display as follows:
Step 5: Run Query
In the Web Client running a query works the
same regardless of whether or not it is a Temporal Query or a traditional i2b2
query. By clicking on the Run Query button, the i2b2 client will
send the request to the i2b2 server which will run the query as defined.
How do I query for a specific date range?
Users
can specify the date range of the observation as shown below
Why do the patient counts not add up in the navigation tree?
The
reason is because one patient may have more than one specific condition. For
example, total number of patients of the category 'Abnormalities of teeth' in
the picture below is 984. The number of patients of the subcategories of the
category 'Abnormalities of teeth' adds up to 1041. This is because one or more
patient may be in both subcategories such as 'Cementum Defect' and 'Dentin
Defect'.
