STANDARD ANALYTIC FILES

SAFs contain final action claims data in which all adjustments have been resolved. When an error is discovered on a claim, a duplicate claim is submitted indicating that the prior claim was an error. A subsequent claim containing the corrected information may then be submitted. In SAFs, the impact of the error claims and their corrections are netted out.

The SAFs group includes eight files:

• Inpatient 100% SAF
• Outpatient 100% SAF
• HHA 100% SAF
• Hospice 100% SAF
• SNF 100% SAF
• Clinical Laboratory 100% SAF
• DME 100% SAF
• 5% Sample Beneficiary SAF

Because the information about file creation, file maintenance, file structure and usage, data structure and usage, methods of access, and migration of data is virtually identical for all SAFs, that information is presented once, followed by descriptions of features unique to each SAF.

File Creation

When analyzing NCH data, researchers must create decision rules and write algorithms to account for claim adjustments. The final action data SAFs were initially created to eliminate the need for this step. Final action data relieve users of the need to account for adjustments and provide a uniform file for analysis purposes.

The SAFs are constructed from weekly data submissions to the NCH 100% Nearline File. Final action claims data are obtained by processing NCH claims through a series of algorithms designed to match original claims with adjustment claims and resolve all adjustments. Files containing the current year's submissions are available beginning with data submitted from January to June of that year. Annual files are created in July for the prior calendar year (based on claim through date); this 18-month window captures approximately 98 percent of the claims for services provided in a given year. Availability of the SAFs is discussed in Data Structure and Usage.

File Maintenance

The SAFs are updated every quarter through June of the following year when the annual files are finalized. Current year's data are created six months into the year; then it is updated quarterly until finalized at 18 months. Once the annual files are finalized, they are not updated to include late arriving claims. Annual files are approximately 98% complete. Because the data for the SAFs come from the NCH 100% Nearline File, they have been subjected to the NCH/CWFMQAS.

File Structure and Usage

Tthe SAFs contain the same variables as the NCH 100% Nearline File records from which they are derived. The files are sorted by Beneficiary Claim Account Number and NCH Category Equatable BIC.

The SAFs can be used to support policy development decisions such as coverage determination and rate setting, quality of care measurement through development of geographic and beneficiary/provider profiles, monitoring of contractor activities, and program research. They are a valuable resource for studies requiring utilization data because adjustments have already been taken into account.

Data Structure and Usage

Each SAF is a different subset of NCH 100% Nearline File data. Thus, the information about record structure, the file sample, and the number of records varies. This information is presented in the individual SAF discussions.

All institutional and non-institutional SAFs (100% & 5%) are available from 1991 onward. Data are stored on tape at the HDC.

Methods of Access

DSAF is a menu-driven system for retrieving data from a variety of claims, utilization, and enrollment files including SAFs. Using DSAF, the 100% SAFs (inpatient, outpatient, HHA, hospice, SNF, and clinical laboratory) can be subset to obtain records from 1, 5, 20, 60, or 100 percent samples of beneficiaries. In addition, the 5% Sample Beneficiary SAF can be subset to obtain a one percent sample of beneficiaries. DSAF also allows beneficiary cohorts to be defined based on the values of a select group of data elements: beneficiary demographics (e.g., age, sex, and race); record types (e.g, inpatient, outpatient, HHA); data elements unique to the file; and service year.

The SAFs are also available as Beneficiary Encrypted Files (BEFs), with patient, physician, and other identifiers encrypted. The BEFs are described in the Beneficiary Encrypted Files Data chapter.

Migration of Data

Because the SAFs were first created in 1993, they have not been affected by the changing CMS (previously HCFA) data processing environment.

List of Data Elements

Data element lists have not been included for the SAFs. However, the data element lists for the corresponding NCH 100% Nearline File record types, which are similar to the SAFs, will be included in the future on the CMS web site at www.cms.hhs.gov.

Inpatient 100% Standard Analytic File (SAF)

The Inpatient 100% SAF contains final action inpatient claims data for all Medicare beneficiaries. It has the same structure as the NCH 100% Nearline inpatient/SNF record. However, this file contains only inpatient hospital data; it does not contain SNF data. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 Inpatient 100% SAF contains approximately 12 million records.

Outpatient 100% Standard Analytic File (SAF)

The Outpatient 100% SAF contains final action outpatient claims data for all Medicare beneficiaries. It is available from 1986 to 1988 in one format, and from 1989 onward in another format and has the same structure as the NCH 100% Nearline File outpatient record and are described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 Outpatient 100% SAF contains approximately 105 million records.

Home Health Agency (HHA) 100% Standard Analytic File (SAF)

The HHA 100% SAF contains final action HHA claims data for all Medicare beneficiaries. It has the same structure as the NCH 100% Nearline HHA record. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 HHA 100% SAF contains approximately 12 million records.

Skilled Nursing Facility (SNF) 100% Standard Analytic File (SAF)

The SNF 100% SAF contains final action SNF claims data for all Medicare beneficiaries. It has the same structure as the NCH 100% Nearline inpatient/SNF record. However, this file contains only SNF data; it does not contain inpatient hospital data. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 SNF 100% SAF contains approximately 3.3 million records.

Hospice 100% Standard Analytic File (SAF)

The Hospice 100% SAF contains final action hospice claims data for all Medicare beneficiaries. It has the same structure as the NCH 100% Nearline hospice record. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 Hospice 100% SAF contains approximately 1 million records. This file is the only Medicare claims file that provides hospice data alone

Clinical Laboratory 100% Standard Analytic File (SAF)

The Clinical Laboratory 100% SAF contains final action claims data for all clinical laboratory services reported on physician/supplier claims. Clinical laboratory services are identified by a predefined laboratory file (containing all lab HCPCS).

This file has the same record structure as the NCH 100% Nearline physician/supplier record. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File. The 1998 Clinical Laboratory 100% SAF contains approximately 129 million records. This file is the only Medicare claims file that provides clinical laboratory data alone.

Durable Medical Equipment (DME) 100% Standard Analytic File (SAF)

Beginning October 1, 1993, CMS changed the way it handled the reporting of DME claims. The establishment of DMERCs regionalized the processing for most of these claims. The transition period for this new procedure occurred between October 1, 1993 and June 30, 1994. During the phase-in, DME suppliers submitted claims for payment using the old method and the new. As a result, DME claims for 1993 and 1994 could be reported as either a Record Identification Code (RIC) "O" or a RIC "M." The 1998 file contains approximately 41 million RIC "M" and 2 million RIC "O" records. The 1994 DME file (RIC "M") contains claims for the last quarter of 1993 and includes all 1994 DME claims submitted to DMERCs. There will continue to be some DME claims submitted through local carriers as RIC "O."

5% Sample Beneficiary Standard Analytic File (SAF)

The 5% Sample Beneficiary SAF contains all final action claims data for a five percent sample of Medicare beneficiaries. Beneficiaries with a HIC number with the digits 05, 20, 45, 70, or 95 in the eighth and ninth positions are included in the sample. It has the same structure as the NCH 100% Nearline File. The record structure is described in Data Structure and Usage for the NCH 100% Nearline File.

The 5% Sample Beneficiary SAF includes the following component files: Inpatient, Outpatient, HHA, Hospice, SNF, Physician/Supplier, which includes pre-October 1993 DME data, and, effective March 1997, DME RIC "M" data from October 1993 onward. Prior to March 1997 these files did not include RIC "M" DME data. These 5% files are also available individually. The 1998 5% Sample Beneficiary SAF contains approximately 39 million records. The 5% Sample Beneficiary SAF is the only SAF that contains physician/supplier data. It replaces the BMAD Beneficiary File for accessing physician/supplier claims data. Institutional data are available in this file beginning in 1989; physician/supplier data are available from 1991 onward.