Healthcare Cost and Utilization Project

The Healthcare Cost and Utilization Project (HCUP) was developed by the Agency for Healthcare Research and Quality (AHRQ), in partnership with entities that maintain statewide hospital administrative databases, to provide multistate, population-based data on both insured and uninsured patients in a uniform format.

Participating states contribute their statewide hospital administrative databases, which contain information from discharge summaries, and HCUP makes the data available for public use. When a patient is discharged from a hospital or a hospital-affiliated facility, an abstract is created that summarizes the administrative information related to the hospitalization. Within individual states, the discharge summaries are incorporated into databases by an agency of the state government, a hospital association, or another organization designated to collect this information. The individual statewide databases contain similar information; however, data completeness and composition vary somewhat from state to state. The Center for Organization and Delivery Studies within AHRQ edits the state databases, applies a uniform coding system, and incorporates the uniformly coded data into the HCUP databases. The data available through HCUP increase as more states participate and as new databases are developed.

Five HCUP databases have been formed, including both inpatient and outpatient administrative data. The statewide files share common data elements, including primary and secondary diagnoses and procedures, admission and discharge status, patient demographics [Page 466](age, gender, median income for zip code; race/ethnicity is available for some states), expected payment source, total charges, and length of stay. Some states also include identifiers that enable linkage to other databases, including the AHA Annual Survey of Hospitals, and Medicare public release data.

The central database is the State Inpatient Database (SID), which is composed of annual, state-specific files, beginning with 1990. The 2005 SID file, with 39 states participating, includes about 90% of all discharges from community hospitals in the United States. Under the definition currently used, records from short-term general hospitals and some specialty hospitals are included in the SID; federal hospitals (Veterans Administration, military, and Indian Health Service hospitals), psychiatric hospitals, alcohol/chemical dependence treatment facilities, and hospitals within prisons are excluded.

Two HCUP databases have been developed based on samples drawn from the SID: (1) The Nationwide Inpatient Sample (NIS) is designed to approximate a20% sample of all U.S. community hospitals. The annual database includes all the discharge data from the sampled hospitals (about 1,000 hospitals in 2003). (2) The Kids’ Inpatient Database (KID) includes 10% of uncomplicated births and 80% of all other pediatric and adolescent hospitalizations. The NIS and KID samples are both drawn from a sampling frame stratified by number of beds, teaching status, ownership, rural/urban location, and region.

In addition to the inpatient databases (SID, NIS, and KID), HCUP has developed two outpatient databases: (1) The State Ambulatory Surgery Database (SASD) contains data from hospital-affiliated ambulatory surgery sites; data from some states include freestanding sites as well. These data can be linked to records in the SID. (2) The State Emergency Department Database (SEDD) contains data from hospitalaffiliated emergency department visits that do not result in hospital admission.

The HCUP Web site provides access to reports and summary analyses and provides software that allows users to query the databases. The databases can be purchased by researchers. Software tools have been developed by AHRQ that can be used on the HCUP databases and with other administrative databases; they can be downloaded without charge. These tools include (1) AHRQ Quality Indicators, three modules for measuring different aspects of the quality of inpatient care, including avoidable hospitalizations and iatrogenic events; (2) Clinical Classifications Software, which aggregates the codes from the International Classification of Diseases (ICD-9-CM) into a smaller number of clinically meaningful categories; and (3) Co-Morbidity Software, which identifies coexisting conditions, using ICD-9-CM diagnosis codes in hospital discharge records.

...