Florida Agency for Healthcare Administration logo

Adding Clinical Data to Statewide Administrative Data Pilot Project

The Agency for Health Care Administration (AHCA), Florida Center for Health Information and Policy Analysis, was awarded a contract from the Agency for Healthcare Research and Quality (AHRQ) that runs through September 2009 for a pilot project to study new ways to approach hospital quality measures. The pilot project funding is provided to add clinical laboratory data to the hospital administrative data already collected by the Agency under statutory authority. By adding clinical data to administrative data, AHRQ hopes to develop better predictors for hospital quality indicators. Florida is one of only three states that have received these grants.

Research has found that through the addition of present on admission to diagnoses it was possible to create risk adjustment models that have discriminatory power for rating hospital performance on selected AHRQ Inpatient Quality Indicators and Patient Safety Indicators.

One purpose of the pilot project is to demonstrate and evaluate the process required to 1) standardize laboratory data into a common nomenclature; 2) merge clinical data with hospital administrative data; 3) complete a statistical analysis of the merged dataset; 4) assess the added value of using clinical data to evaluate the quality of patient care within hospitals; 5) and describe all findings in a Final Report.

The Agency will work with 3M Health Information Systems, Inc (3M HIS) and the pilot hospitals to map their laboratory values to standardized Logical Observation Identifiers Names and Codes (LOINC) terminology and to evaluate the extent to which the 3M risk-adjustment model can be used in public hospital performance comparisons and can be made more accurate with the availability of the clinical data. During this contract (see Process Flowchart page 3):

The Agency will

• Develop a data sharing agreement with hospitals to support acquisition of clinical data for linkage to existing Agency information.
• Load the clinical lab data received from hospitals into an Oracle database in the Agency that holds the existing administrative inpatient data collected from the hospitals. The clinical and administrative datasets will then be joined into a single file for each hospital, combining the clinical with the administrative data.
• Send the combined file to 3M HIS using secure File Transfer Protocol (FTP).
• Provide participating hospitals with summary information and reports.
• Complete a final report that details the processes of normalizing laboratory terminology, linking clinical and administrative datasets and assessing the added value in the use of clinical data to determine the quality of patient care within the hospitals in the pilot project.

3M HIS will:

• Provide technical assistance to hospital quality and technical staff in creating a map of standardized values based on the coding sheet of their laboratory data terminology.
• Work with each hospital to standardize its laboratory data terminology and values to Logical Observation Identifiers Names and Codes (LOINC) standards. Currently 3M uses LOINC version 2.22, but will be using the new LOINC version released by Regenstrief after July 2008.
• Work with each hospital to verify accuracy of the final normalized map of laboratory values and then send each hospital its LOINC map.
• Group the merged clinical and administrative data into All Patient Refined Diagnosis Related Groups (APR DRGs) for analysis. 3M HIS will identify the individual clinical data elements, if any, that demonstrate an ability to improve the prediction of hospital-level quality measures.
• Provide a summary of its findings to the Agency.

Hospitals will:

• Attend a meeting with the Agency, other hospitals and subcontractors in May 2008.
• Send their laboratory data catalogue to 3M Terminology Consulting Services (3M TCS) to initiate the LOINC mapping (see 3M data dictionary template). The data collected for the pilot project, as shown in Table 1, will consist of specific clinical laboratory data elements and a set of demographic indicators that will be used to link the clinical data with the administrative data.
• Extract three quarters of laboratory data, from April 1, 2007 – December 31, 2007, on all patients, for all laboratory tests conducted.
• Apply the LOINC mapping to convert their unique laboratory values to LOINC standardized values and terminology.
• Conduct quality assurance to ensure data mapping is correct.
• Upload the standardized laboratory dataset, using standard messaging format, such as in a Comma Separated Value (CSV) to the secure FTP site at the Agency.

For more information, visit the website at: http://ahca.myflorida.com/search/index.shtml?cx=012521305296577922040%3Apfndlojhd-4&cof=FORID%3A11&ie=UTF-8&q=adding+clinical+data+project&siteurl=ahca.myflorida.com%2Findex.shtml#943.

Table 1: Clinical and Demographic Data Elements for the Pilot Project