National Survey of Ambulatory Surgery Data Collection and Processing
Sample selection and abstraction of information from medical records were performed at facilities as was the case in the 1994-1996 NSAS. Facility staff did the sampling in about 40 percent of facilities that participated in the 2006 survey, and facility staff abstracted the data in about 30 percent of the participating facilities. In the remaining facilities, the work was performed by personnel of the U.S. Census Bureau acting on behalf of NCHS.
Many of the items on the abstract form were the same in 2006 as in the 1994-1996 NSAS. Included were items relating to the personal characteristics of the patients such as age, sex, race, and ethnicity. Also, there were administrative items such as date of procedure, disposition and expected sources of payment, and medical information about patients including diagnoses and procedures, which were coded according to the International Classification of Diseases, 9th Revision, Clinical Modification. In the 2006 NSAS, data were also collected on symptoms present during or after surgery, and follow-up of patients within 24 hours after they left the ambulatory surgery facility. Additional questions were asked about the facilities in the 2006 NSAS.
Data processing and medical coding were performed by Constella Group Inc., Durham, North Carolina. As in 1994-1996, a quality control program was conducted on the coding and entering of data from abstracts to electronic form. Approximately 10 percent of the abstracts were independently recoded by an NSAS coder, with discrepancies resolved by a chief coder. The overall error rate for the 2006 NSAS was 0.3 percent for diagnosis coding and keying, 0.2 percent for procedure coding and keying, and 0.3 percent for demographic coding and keying.
Also as in 1994-1996, editing and estimation of the data were completed at NCHS by the Technical Services Branch.
For additional information about the NSAS 2006 data collection and processing, see the 2006 NSAS public-use data file documentation [PDF - 375 KB].
Data collection for the 1994-1996 National Survey of Ambulatory Surgery (NSAS) consisted entirely of manual data collection. Sample selection and abstraction of information from medical records were performed at the facilities. Facility staff did the sampling in about 44 percent of the facilities that participated in the 1996 survey. Facility staff abstracted the data in approximately two-thirds of the participating facilities. In the remaining facilities, the work was performed by personnel of the U.S. Bureau of the Census on behalf of NCHS.
The 1994-1996 medical abstract form contains items relating to the personal characteristics of the patients such as age, sex, race, and ethnicity. There are administrative items such as date of procedure, disposition and expected sources of payment. The medical information about patients includes up to seven diagnoses and six procedures, which are coded according to the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM).
All medical coding and keying operations were subject to quality control procedures. These procedures utilized a two-way 10 percent independent verification procedure in which abstracts were independently recoded by a second coder, with discrepancies resolved by the chief coder. Error rates were calculated for nonmedical and medical information based on discrepancies in codes between the independent and the original coder. When error rates exceeded set standards for either nonmedical or medical coding, the medical or nonmedical information was recoded for the entire batch of abstracts and again subjected to quality control procedures.
After medical coding and keying were completed, extensive computer editing was conducted to assure that all responses were accurate, consistent, logical and complete. When necessary, records were reviewed manually to resolve inconsistencies. Missing age and sex data were imputed using a hot deck procedure. A detailed description of the NSAS data collection procedures is included in "The Plan and Operation of the National Survey of Ambulatory Surgery," [PDF - 3.9 MB] Vital and Health Statistics Series 1, Number 37.