GE Centricity for Business (GECB) is the UT Physicians billing system. From Centricity we pull Demographics and Billing data, which we push into our i2b2 instance. Unlike Allscripts, Centricity contains information on patients seen in both outpatient and inpatient settings. Tables of commonly requested information are presented. Data are updated on a weekly basis through a set of extracts provided by the clinical IT Staff. Data are loaded into the CDW nightly.
Type of Query | Number of Records |
---|---|
Patients from Centricity | 5,501,776 |
Centricity contains demographic information on several million patients. However, the Clinical Data Warehouse has measures which reduce the total counts:
Thus, the final counts loaded into our i2b2 instance will be lower than the total records in Centricity.
Type of Query | Number of Records |
---|---|
Total Patients with administrative data | 3,126,610 |
Total Patients with demographic data | 5,352,828 |
Clinic | Total Patient Encounters |
---|---|
ALDINE | 387,582 - 1.61% |
ER LBJ | 1,563,348 - 6.51% |
ER MHH | 1,791,600 - 7.46% |
IP LYNDON B JOHNSON | 1,380,239 - 5.75% |
IP MHH | 4,566,126 - 19.0% |
IP MHH CHILDRENS | 569,590 - 2.37% |
OP LBJ CLINIC | 479,173 - 1.99% |
OP LYNDON B JOHNSON | 1,678,414 - 6.99% |
OP MHH | 1,165,494 - 4.85% |
UTP HPB CLINIC | 2,482,859 - 10.3% |
Race and Ethnicity are combined as one variable. Gender and marital status are reported as recorded in Centricity.
Race | Count |
---|---|
AFRICAN AMERICAN | 810,594 |
ALL OTHER | 189,164 |
AMERICAN INDIAN,ESKMO,OR ALEUT | 17,198 |
ASIAN OR PACIFIC ISLANDER | 53,361 |
CAUCASIAN | 1,577,209 |
Declined | 5,391 |
HISPANIC OR LATINO | 130,559 |
LATIN AMERICAN | 484,492 |
NOT ANSWERED | 120,387 |
Null | 105,050 |
OTHER | 1,555,114 |
Unknown | 418,198 |
Status | Count |
---|---|
CODE | 163,409 |
COHABITATING | 146 |
DESCRIPTION | 352,470 |
DIVORCED | 139,737 |
LEGALLY SEPARATED | 2,967 |
LIFE PARTNER | 1,368 |
MARRIED | 1,219,173 |
Null | 663,120 |
NUMERIC CODE | 16,987 |
SEPARATED | 18,526 |
SINGLE | 1,558,768 |
UNKNOWN | 1,226,344 |
WIDOWED | 103,702 |
Gender | Count |
---|---|
F | 2,912,521 |
I | 850 |
M | 2,432,317 |
Null | 103,586 |
U | 17,443 |
Invoices contain billing diagnosis and the billing diagnosis is what is provided in this dataset. Since it is a billing diagnosis, it may or may not represent clinical reality. For example if you search for patients with breast cancer (ICD9 174.*) among our outpatient clinic patients, we estimate that you will only identify around half of the actual patients that have, or have had, breast cancer. With endometrial cancer (ICD9 182.*), we estimate that the number of patients you identify will be roughly one quarter of the patients that have, or have had, endometrial cancer.
These estimates are based on a manual review we performed on a subset of the records. See Oncology research using electronic medical record data for more information.
The i2b2 terminology for diagnosis was built using the UMLS (Unified Medical Language System). A small portion of the diagnosis codes cannot be mapped to ICD10 or ICD9 from the UMLS, leaving them unsearchable.
Diagnosis Code | Total Records |
---|---|
250.00 Diabetes mellitus without mention of compli | 503,036 - 0.74% |
272.4 Other and unspecified hyperlipidemia | 481,243 - 0.71% |
401.1 Benign essential hypertension | 319,215 - 0.47% |
401.9 Unspecified essential hypertension | 980,127 - 1.45% |
786.50 Chest pain, unspecified | 462,479 - 0.68% |
E11.9 Type 2 diabetes mellitus without complicatio | 465,232 - 0.69% |
E78.5 Hyperlipidemia, unspecified | 423,320 - 0.62% |
I10 Essential (primary) hypertension | 1,265,095 - 1.87% |
V20.2 Routine infant or child health check | 382,142 - 0.56% |
Z23 Encounter for immunization | 640,847 - 0.95% |
CPT Procedure Code | Total Records |
---|---|
ELECTROCARDIOGRAM REPORT | 1,247,105 - 2.88% |
EMERGENCY DEPT VISIT | 879,269 - 2.03% |
OFFICE/OUTPATIENT VISIT EST | 4,899,482 - 11.3% |
OFFICE/OUTPATIENT VISIT NEW | 1,123,235 - 2.59% |
POSTOP FOLLOW-UP VISIT | 971,732 - 2.24% |
Procedures come from Centricity Transactions. These are billing transactions, as such, only the codes that the physicians bill for will appear in the procedures data. A large percentage of these procedures are coded as office visit, not necessarily a procedure.
The CPT Terminology is built from the UMLS as well. See the NIH reference page for more information.
8% of the procedure codes cannot be mapped to the CPT from the UMLS leaving them unsearchable.