National Hospital Discharge Survey NHDS and National Survey

National Hospital Discharge Survey (NHDS) and National Survey of Ambulatory Surgery (NSAS) 2006 Data Users Conference July 10, 2006 3: 30 pm-5: 00 pm Session #15

Session Overview NHDS: Overview Carol De. Frances, Ph. D Analytic Issues Karen Lees, MPH Examples of Research Jean Kozak, Ph. D Accessing Data Marni Hall, Ph. D New Directions Bob Pokras, MA

Overview of the National Hospital Discharge Survey (NHDS) Carol De. Frances, Ph. D

NHDS Survey Years § Conducted annually 1965 -present • Latest data available: 2004 • 2005 will be available this Winter

NHDS Survey Design § Scope and coverage: • Short stay, non-Federal hospitals • ALOS < 30 days • General and children’s general hospitals

NHDS Sampling Plan § Three stage design: • Geographic units • Hospital • Discharge

NHDS Sample Size § Hospitals: • § About 500 hospitals sampled per year Discharges: • Over 300, 000 sampled per year

NHDS Data Collection § Manual hospitals - 56% § Automated hospitals - 44% § Overall response rate for the 2004 NHDS – 92%

NHDS Manual Data Collection § NCHS – Statistical Design § Census Bureau – Field Work § Contractor – Coding and Data Entry

NHDS Automated Data Collection § Electronic files obtained from: • States • Commercial firms • Individual hospitals

NHDS Data Processing § NCHS: • Editing • Estimation

NHDS Estimation § Weight: • Inverse of the probability of selection • Adjustments for non-response • Population weighting ratio adjustment

Variables on NHDS Public Use Data Files

Patient Data § Age § Sex § Race § Expected source of payment § Discharge status § Marital status

Hospital Characteristics § Geographic region § Bed size § Ownership

Medical Data § International Classification of Diseases, 9 th Revision, Clinical Modification (ICD-9 -CM) § Diagnoses (up to 7) § Procedures (up to 4)

Additional Variables § Days of care § Month of discharge § Diagnoses Related Group (DRG) § Weight

New Variables for NHDS § Available beginning in the 2001 NHDS: • Source of Admission • Type of Admission

New Variables for the NHDS § Beginning to collect with the 2007 NHDS: • Admitting Diagnosis • Present on Admission checkbox for all seven diagnoses collected

Analytic Issues Karen Lees, MPH

Topics § Utilization measures § Medical coding system § Statistical issues

NHDS provides data on hospitalizations not people

Measures Include § Discharges § Days of care § Average length of stay § Diagnoses § Surgeries/procedures

Discharges § Include deaths § Include transfers to other hospitals or long-term care facilities § Do not usually include newborn infants

Days of Care § Total number of days discharged patients spend in the hospital § All stays are counted as at least 1 day § The admission day is counted, but not the discharge day § The number of days divided by the number of discharges is the average length of stay

Diagnoses § Disease, injury or other reason for hospitalization § Coded according to US adaptations of the International Classification of Diseases (ICD-9 -CM)

Diagnoses § Principal diagnosis: chiefly responsible for hospitalization § First-listed diagnosis: principal if specified, otherwise one listed first

Diagnoses § All-listed: total number of times diagnoses appears on record § Any-listed: discharges with diagnosis in any position on record

Hospital discharges with fractures, 2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Surgery / Procedures § Surgical (appendectomy) § Diagnostic (spinal tap) § Therapeutic (chemotherapy) procedures § Coded according to US adaptations of the International Classification of Diseases

NHDS provides data on inpatient procedures not total procedures

Versions of the International Classification of Diseases § 8 th revision used 1970 -78 § 9 th revision used 1979 -present § 10 th revision for use in future

ICD 8 th Revision § Some codes different than in 9 th Revision § Did not use E-codes § Made modifications in coding to accommodate available data

ICD 9 th Revision § Addenda added annually since 1986 § Codes added, deleted, expanded, and revised § Lists of changes available in annual summary reports, file documentation

Weights § Must use weighted data to obtain unbiased national estimates § Each record has a weight § Sum the weights of the records

Reliability § To be reliable, estimates must be based on at least 30 records § And have a relative standard error of less than 30 percent § Estimates based on 30 -59 records should be used with caution

Standard Errors § Some standard errors are in Advance Data summaries § Generalized error curves are in the Series 13 Annual Summaries and data documentation § Use SUDAAN for specific standard errors - need access to confidential data

Examples of Research Jean Kozak, Ph. D.

Percent distribution of discharges and days of care by age, 2004 Days of care Discharges 85 and over 75 -84 Age in years 65 -74 55 -64 45 -54 35 -44 25 -34 15 -24 5 -14 0 -4 20 10 Percent distribution Source: CDC/NCHS, National Hospital Discharge Survey (NHDS) 20

Average length of hospital stay by age, 1980 -2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Percent of hospital inpatients transferred to long-term care facilities, by age, 1990 -2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Hospitalization rates for major diagnostic categories by sex, 2004 Males Females Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Percent distribution of patients hospitalized for heart disease, 2003 All other diagnoses 87% Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Hospitalization rate for cancer, 1990 -2004 1990 63. 0 2004 41. 2 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Number of discharges with HIV diagnoses by age, 1995, 2000 and 2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Hospitalization rates for depression for boys and girls 5 -19 years of age, 1990 -92 and 2002 -04 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Hospitalization rates for inpatients with Clostridium difficile by age, 1996 -2003 Source: Mc. Donald LC, et al. Emerg Infect Dis. 2006; 12(3): 410 -5.

Hospitalization rates for avoidable conditions by race, patients under 65 years of age, 1990 -2004 Black patients White patients Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Average length of stay for selected diagnostic categories, 2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Average length of stay for hip fractures, 1990 -2004 12. 8 days 1990 6. 4 days 2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Length of hospitalization for childbirth, 1980 -2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Rate of cardiac catheterizations by sex, age, and geographic region, 2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Coronary artery bypass grafts (CABGs) and percutaneous coronary interventions (PCIs), 1980 -2004 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Percent distribution of white and black discharges with coronary angioplasty according to stent use, 2003 30. 5% White patients 22. 0% Black patients Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Hospitalization rates for obese patients with gastric bypass, 1998 -2003 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Rate of cesarean delivery, 1970 -2004 29. 9 1970 5. 5 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Rate of vaginal birth after cesarean (VBAC), 1990 -2004 1990 20. 4 2004 10. 9 Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Rate of forceps and vacuum extraction, 1990 -2004 Vacuum Forceps Source: CDC/NCHS, National Hospital Discharge Survey (NHDS)

Data users § Hospitals § Universities and medical schools § Professional organizations, such as AMA, AHA, WHO § Government agencies, such as NIH, CMS, CDC, IHS § Medical research organizations, pharmaceutical and medical supply manufacturers, insurance companies § News media § Other users such as publishing houses, market research groups, free lance writers

Accessing Data from the National Hospital Discharge Survey Marni Hall, Ph. D.

How can you get the data? § In publications, including annual reports § From data tabulations on selected topics § Using public-use data files (micro-data) § Calling NCHS to request tabulations

On the NHDS/NSAS homepage www. cdc. gov/nchs/about/major/hdasd/nhds. htm § You’ll find: • A Description and Methodology of NHDS (and NSAS) • Downloadable Publications from NCHS • List of other publications using NHDS or NSAS data • Public Use Files – micro data

How to get to the NHDS/NSAS webpage from the NCHS homepage www. cdc. gov/nchs § Under Surveys & Data Collection Systems on the far left of the screen § Select NHCS which stands for National Health Care Survey § When a pop-up list of surveys appears, select National Hospital Discharge Survey

Annual Publications using NHDS § Advance Data reports • § 2004 available on the website Annual Summary with Detailed Diagnosis and Procedure Data • 2003 available on the website • 2004 available later this summer

How to download publications § Publications, data tables, and data file documentation are in Adobe Acrobat PDF format § Require use of the free Adobe Acrobat Reader software, available for download at: www. adobe. com

Public-Use Files Available on the Web § Data files and documentation available for free from the NHDS webpage § NHDS: 1996 through 2004 • NSAS: 1994, 1995, 1996* *2006 NSAS data will be available early in 2008

Obtaining and using data files § Data files are in ASCII format and must be downloaded via FTP server. § Downloadable public-use data files are “zipped” for a speedier download § “Unzip” these files with free data extraction software such as: • Win. Zip (www. winzip. com) • PKunzip (www. pkware. com)

Analyzing Public Use Data § Files require the use of standard statistical software packages, such as SAS, SPSS, Stata, etc.

Multi-year data files § Two separate multi-year files on CD-ROM are available for free from NCHS: 1970 -1978 data years (ICD-8 coding) 1979 -2004 data years (ICD-9 -CM coding) § DRG’s are not included on the multi-year files. § For NSAS a 1994 -1996 data file is available.

Where to find ICD-9 -CM codes for Diagnoses or Procedures www. cdc. gov/nchs/icd 9. htm § Full-text ICD-9 -CM documents are RTF (Rich Text Format) files and can be handled with any word processing package § Addenda and conversion tables are PDF documents § Very lengthy document

Publications and Information Products http: //www. cdc. gov/nchs/products. htm • Lists electronic products • Additional Information about obtaining reports, data, and other products

Other sources of tabulated NHDS Data • FASTATS A to Z • Aging – Trends in Aging Database provides information on national trends and key variables that depict the health status of older Americans compiled from a number of surveys – mostly on persons 50 years old and over • Healthy People • Health, United States

Restricted Data in NHDS http: //www. cdc. gov/nchs/r&d/rdc. htm § HCSB maintains confidential information in files which are not publicly disseminated § Restricted data pose disclosure risks to survey respondents § These data are available to researchers through the NCHS Research Data Center (RDC)

Through the RDC Researchers Can Use: § Confidential files for NHDS and NSAS variance estimation § NHDS and NSAS analytic files that have been linked with outside data sources

Confidential Variables Available only on Restricted Files § ZIP Codes for residence of patient and for hospitals § State/County FIPS CODE for both patient and hospital § AHA ID for hospital § Design Variables needed to run SUDAAN NOTE: Identifiers such as patient name, address, and SSN are not collected in the NHDS

NHDS Can be Linked with Other Files, for example § American Hospital Association File (AHA) which has data on hospital characteristics § Area Resource File (ARF) which has county level data (e. g. SES variables and health service availability) § Linkage is with contextual NOT personal / demographic information

NHDS or NSAS Questions? § § By telephone: • Hospital Care Statistics Branch: 301 -458 -4321 • NCHS Information Dissemination Staff (toll free) -866 -441 -NCHS [1 -866 -441 -6247] Send an email to: NHDS@cdc. gov 1

Future Activities and Directions Robert Pokras, MA

Future Activities and Directions § Redesigning the NHDS § Refielding the NSAS

Context for Redesigning the NHDS § Data elements limited to UB 92 (UB 04) § The value of primary data collection to meeting current and future policy and research needs § To take a fresh look at current and future policy and research issues and associated data needs and gaps § Meetings and interviews § Working group meeting in March, 2006

Priority Issues § Cost of care/ use of resources • § including efficiency/ waste Quality of care/ safety • including disparities § Care delivered in the hospital § Surveillance and public health • § including surge capacity Globalization • including outsourcing

Highest Rated Options for Redesign § Coordinate with AHRQ data collection (HCUP) § Add resource use/ cost/ billings/ payments § Add clinical depth • § Improve patient demographics • § e. g. , medications, tests e. g. , ethnicity Link to health-related outcomes • e. g. , death index

Clinical Depth Core Data Set (All Sampled Cases) § Height § Weight § Administered drugs

Clinical Depth Diagnostic/ Procedure/ Demographic Module (Subsample of Cases) § Lab values § Pathology results

Resource Use/ Cost/ Efficiency § Billings -- charges § Medicare program • Cost-to-charge ratios • DRG hospital-specific payments

Improved Patient Demographics § Race § Ethnicity

Link to Health-Related Outcomes § Link with the National Death Index to look at 30 day mortality

Additional Aspects of the Redesign § Link mother and newborn records § Collect data from ED and observation status (for sampled inpatients) § Collect data from last previous admission • Length of stay • Principal diagnosis

Next Steps § Summer 2006: • Final conceptual framework • Strategy and data elements § 2006 -2007: Feasibility test in 9 hospitals § 2007: Contract for field test § 2008: Field test § 2010: New survey fielded

Refielding the National Survey of Ambulatory Surgery (NSAS) § First fielded in 1994 -1996 § Back in the field for 2006 § Data collect ends in March 2007 § Public Use files will be available in early 2008

Breast lumpectomy procedures Total ? . . Ambulatory Inpatient Sources: CDC/NCHS, National Survey of Ambulatory Surgery (NSAS) and National Hospital Discharge Survey (NHDS)

Any questions?