Collecting Cancer Registry Data Through Linking Hospital Electronic Medical Records

Wednesday, June 12, 2013: 2:00 PM
104 (Pasadena Convention Center)
Youjie Huang , Florida Department of Health, Tallahassee, FL
Jason Feldman , Florida Department of Health, Tallahassee, FL
Monnique Hernandez , Florida Cancer Data System, Miami, FL
Jill Mackinnon , Florida Cancer Data System, Miami, FL
BACKGROUND:  Cancer registries collect cancer reports through manual reviewing and abstracting medical records. Many hospitals have implemented electronic medical record (EMR) in Florida. Collecting information on cancer diagnosis and treatment through linking hospital EMR may improve timeliness, completeness and quality of cancer registry data. The purpose of this study is to access feasibility and value of collecting cancer patient data through linking state cancer registry with hospital EMR system.

METHODS:  The Florida Department of Health and the Florida Cancer Data System (FCDS) established a partnership with a Florida hospital system to develop methods for hospital EMR extraction and transmission. Patients were identifies based on ICD-9-CM cancer diagnosis codes. The patient EMR submitted to FCDS data consist of detailed treatment, medication, and discharge reports. Once transmitted to FCDS, EMR were matched with registry data. Hospital pathology data were processed to identify records relevant to cancer. The study was funded by Agency for Healthcare Research and Quality (AHRQ), grant # R24 H2019658-01.

RESULTS: 11,506 patients diagnosed with or treated for breast cancer between 2007 and 2010 at a Florida hospital network comprised of 9 facilities whose EMR were submitted to FCDS and linked with cancer registry data. Patient demographics, socio-economic status, and co-morbidities were investigated in relationship to completion of treatment. Records represented patient data for each episode of care, including detailed treatment information for chemotherapy and radiation dosage, surgery, and pathology test results. EMR provided additional information to the cancer registry data: 5% of registry surgery information, 1% of radiation records, and 7% of chemotherapy records were updated using EMR data. EMR provided unique detail information on cancer treatments that cancer registry does not routinely collect.

 CONCLUSIONS:  Collecting cancer data through linking hospital EMR data is feasible when technical expertise is invested. Hospital EMR includes a range of important data elements not included in routine cancer reporting which are useful for analyzing treatment effectiveness. Data processing methodologies are limited to the capabilities of a given hospital EMR system due to lack of data collection and coding standards.