To examine how Florida trauma centers bill their patients, the Tampa Bay Times relied on billing data collected by state health officials from hospitals across the state.
The databases, maintained by the Agency for Health Care Administration, include diagnosis, treatment and hospital charge information for every patient seen in an emergency room or admitted to a Florida hospital.
The database does not identify patients by name or address.
The Times analyzed more than seven years of data going back as far as 2006, the first year the state required hospitals to break out fees for trauma response.
To conduct its analysis, the Times sought additional information that is not contained in the state databases.
For example, the state data does not explicitly identify trauma patients. To identify them, researchers typically use the medical diagnostic codes recorded in the data. The Times consulted with experts and used a widely accepted formula that flags trauma patients based on injuries.
Among the experts was Etienne Pracht, a University of South Florida health care economist who is one of the state's leading experts in analyzing injury severity and trauma trends using the state's data.
The Times focused on patients with diagnostic codes widely considered trauma cases, including brain and spinal cord injuries, burns, broken skulls and complex limb fractures. We further limited our analysis to only those patients flagged in the data as emergencies or trauma cases.
In addition, the Times included any patient who was charged a trauma response fee.
To interpret diagnosis and treatment codes recorded in the data, the Times relied on information from Alkaline Software Inc., a company that publishes medical databases. It provides free information about the codes online at www.icd10data.com.
Finally, the Times relied on USF researchers, including Pracht, to provide a severity rating for each individual patient for the most recent year and a half of data. Pracht's methodology, widely used by researchers, allowed the Times to compare trauma centers based on the level of injuries sustained by its patients.
Ultimately, the Times analysis focused on 306,525 patients who met the criteria for trauma and were seen at trauma centers from 2006 to June 2013, the most recent data available.
The analysis included patients from 26 hospitals that operated trauma centers in that time. It did not include the two children's hospitals receiving trauma patients. One trauma center, Orange Park Medical Center, only operated from November 2011 to February 2013.
The Times analysis likely underestimates the number of patients who were treated at a trauma center and then released the same day. That's because one of the two databases state officials use to track hospital charges has limitations.
The database — for emergency department patients who don't get admitted to the hospital — does not track the status of patients upon arrival, which was needed in our methodology to identify true trauma patients.
Letitia Stein And Alexandra Zayas, Times staff writers