Understanding and Using NAMCS and NHAMCS Data Tools

Understanding and Using NAMCS and NHAMCS Data Tools and Basic Programming Techniques Donald Cherry Ambulatory and Hospital Care Statistics Branch Division of Health Care Statistics 1

Overview n n n Some important features of NAMCS & NHAMCS File structure SETS Exercises using SAS Proc Surveyfreq/Proc Surveymeans, SUDAAN, STATA u Downloading data & creating a SAS dataset u Simple frequencies with/without standard errors u Creating a new variable-Asthma u Visit rates for asthma-male/female u Total number of digestive write-in procedures u Time spent with physician Considerations Summary 2

NAMCS and NHAMCS n National Ambulatory Medical Care Survey (NAMCS) u Visits to nonfederal, office-based physicians u CHC’s sampled beginning in 2006 n National Hospital Ambulatory Medical Care Survey (NHAMCS) u Visits to hospital outpatient and emergency departments 3

NAMCS Sample Design n n n Three stage design u 112 PSUs u Physician practices within PSUs u Patient visits within practices One-week reporting period About 30 visits per doctor are typically sampled For 2006— 3, 350 doctors sampled 104 CHC’s sampled & physician visits included in sample Total visits 29, 392 4

Scope of the NAMCS Basic unit of sampling is the physician-patient visit n In scope visits: u Must occur in physician’s office u Must be for medical purposes u Administrative visits not sampled u House calls, emails, phone calls not sampled n 5

Scope of the NAMCS (cont. ) n n Physicians must be: u Classified by AMA or AOA as primarily engaged in office-based patient care u nonfederally employed u not in anesthesiology, radiology, or pathology u 64 percent unweighted response rate in 2006 CHC’s are Federally Qualified or “look alike” 6

NHAMCS Sample Design Multistage probability design n First stage sample of 112 PSUs n Hospitals within PSUs n Clinics within OPDs, Emergency Service Area (ESA) within EDs n Patient visits within clinics, ESAs n 4 -week reporting period n 382 hospitals sampled in 2006; 35, 849 ED visits and 35, 105 OPD visits n 7

Scope of the NHAMCS Basic unit of sampling is patient visit n Emergency and outpatient departments of noninstitutional general and shortstay hospitals n Not Federal, military, or Veterans Administration facilities n Located in 50 states and D. C. n 8

Sample Weight Each NAMCS record contains a single weight, which we call Patient Visit Weight n Same is true for OPD records and ED records n This weight is used for both visits and drug/procedure mentions n 9

Data Items n n Patient characteristics u Age, sex, race, ethnicity Visit characteristics u Source of payment, continuity of care, reason for visit, diagnosis, treatment Provider characteristics u Physician specialty, hospital ownership… MULTUM drug characteristics added in 2006 10

Coding Systems Used Reason for Visit Classification (NCHS) n ICD-9 -CM for diagnoses, causes of injury and procedures n Drug Classification System-MULTUM n 11

File Structure n Download data and layout from website http: //www. cdc. gov/nchs/about/major/ahcd 1. htm n Flat ASCII files for each setting and year: NAMCS: 1973 -2006 NHAMCS: 1992 -2006 n STATA files on Web: NAMCS: 2003 -2005 NHAMCS: 2003 -2005 12

Creating a usable STATA dataset n Two options: 1) Use the self-extracting file in STATA folder to open a complete dataset for the 2003 -2005 NAMCS, NHAMCS-ED, & NHAMCS-OPD 2) Use the DO file (*. do) and the dictionary file (*. dct) along with the flat data file (*. exe) to create a dataset 3) Stat. Transfer 13

Organizational structure-NAMCS data Provider provider info practice info geographic info Write-in scope procedure 1 Visit patient & visit info treatment & outcome info medications Medications 1 -8 Primary reason for Visit Primary diagnosis Write-in scope procedure 2 Other test/service 1 Other test/service 2 Surgical procedure 1 Other Reason for Visit MULTUM Categories Other diagnosis Surgical procedure 2 Non-surgical procedure 1 Other Reason for Visit Other diagnosis Non-surgical procedure 14 2

SETS-Statistical Export and Tabulation System 15

Hands-on Exercises STATA Users n Double-click: My ComputerLocal Disk C: DUC_08 n Open STATA n In the command window type: u Set mem 1000 m u Set matsize 5000 n Under the “File” icon-doubleclick namcs 05. dta n Under “New Do File Editor”double-click: STATA exercises. do SAS/SUDAAN Users n Double-click: My ComputerLocal Disk C: DUC_08 n Double-click: Final Exercises 16

Visit rate estimates Female population=800 Calculation* New variable Phycode Sex Patwt (Patwt/Pop)*10 0 Sexwt 1401 1 100 (100/800)*100 12. 5 1820 1 300 (300/800)*100 37. 5 1001 1 50 (50/800)*100 6. 25 500 1 120 (120/800)*100 15 Sample size=4 Visits=570 71. 25 visits per 100 persons *Note: Rate=est/pop=Σ patwt/pop=1/pop*Σ patwt. 17

Calculating Total Number of Write-in Procedures Record Proc 1 Proc 2 Proc 3 Proc 4 Proc 5 Proc 6 Proc 7 Proc 8 Totproc 1 1911 0000 0000 1 2 2186 0000 0000 2 3 5490 0000 0000 1 4 0000 0000 0 5 8192 0000 0000 8200 2 Note: 0000=No procedure recorded. 18

Data Considerations 19

NAMCS vs. NHAMCS n Consider what types of settings are best for a particular analysis u Persons of color are more likely to visit OPD’s and ED’s than physician offices u Persons in some age groups make disproportionately larger shares of visits to ED’s than offices and OPD’s 20

Procedures Program Categorical Variables Continuous Variables SAS PROC SURVEYFREQ PROC SURVEYMEANS STATA SVY: TAB SVY: MEAN SUDAAN PROC CROSSTAB PROC DESCRIPT 21

How Good are the Estimates? Depends … In general, OPD estimates tend to be somewhat less reliable than NAMCS and ED. n Since 1999, our Advance Data reports include standard errors in every table so it is easy to compute confidence intervals around the estimates. n 22

RSE improves incrementally with the number of years combined n RSE = SE/x n RSE for percent of visits by persons less than 21 years of age with diabetes ü ü ü 1999 RSE =. 08/. 18 =. 44 (44%) 1998 & 1999 RSE =. 06/. 18 =. 33 (33%) 1998, 1999, & 2000 RSE =. 05/. 21 =. 24 (24%) 23

Some User Considerations n n n NAMCS/NHAMCS sample visits, not patients No estimates of incidence or prevalence No state-level estimates May capture different types of care for solo vs. group practice physicians Data is only as good as what is documented in the medical record 24

If nothing else, remember…The Public Use Data File Documentation is YOUR FRIEND! n Each booklet includes: u A description of the survey u Record format u Marginal data (summaries) u Various definitions u Reason for Visit classification codes u Medication & generic names u Therapeutic classes 25