Linking Example: Merge data from the 2001 MEPS office-based Medical Provider Visits file with data from the 2001 MEPS full-year file. Aggregate event-level data to the person level.

1 The SAS System 13:11 Sunday, November 21, 2004

NOTE: SAS (r) Proprietary Software Release 8.2 (TS2M0)

Licensed to AGENCY FOR HEALTHCARE RESEARCH & QUALITY, Site 0040776001.

NOTE: This session is executing on the WIN_PRO platform.

1 options ls=120 ps=79 nodate;

2 ods noproctitle;

3 /************************************************************************************\

4 Program: c:\meps\prog\Example_L3.sas

6 Description: This example shows how to:

7 (1) Aggregate event records to the person level

8 (2) Make 1 annual variable from 3 round variables

9 (3) Make a categorical variable from a continuous variable

10 (4) Use SUDAAN to calculate standard errors

12 Input Files: c:\meps\data\h59g.sas7bdat (2001 Office-Based Visits)

13 c:\meps\data\h60.sas7bdat (2001 Full-Year File)

14 \************************************************************************************/

16 libname hc 'c:\meps\data';

NOTE: Libref HC was successfully assigned as follows:

Engine: V8

Physical Name: c:\meps\data

18 title 'AHRQ MEPS DATA USERS WORKSHOP (LINKING) -- NOV/DEC 2004';

19 title2 'Link 2001 Household File and 2001 Events File';

21 proc format;

22 value inscov 1='1 Any Private' 2='2 Public Only' 3='3 Uninsured';

NOTE: Format INSCOV has been output.

23 value insured 1='1 Insured' 2='2 Uninsured';

NOTE: Format INSURED has been output.

24 value agecat

25 1='1. 0-3'

26 2='2. 4-7'

27 3='3. 8-11'

28 4='4. 12-15'

29 5='5. 16-17'

30 6='6. 18+';

NOTE: Format AGECAT has been output.

31 value genckup 1='1 General Checkup' 2='2 No General Checkup';

NOTE: Format GENCKUP has been output.

32 run;

34 /* 2001 Office-based medical provider visits */

35 /* Identify persons with a visit for a general check-up & their expenditures */

37 title3 "# Persons with a General Checkup in a Provider's Office";

38 data h59g; set hc.h59g(keep=dupersid vstctgry obxp01x obsf01x); by dupersid;

39 if first.dupersid then do;

40 genckup=.;

41 ambtotpd=0;

42 ambfampd=0;

43 end;

44 retain genckup ambtotpd ambfampd;

45 if vstctgry = 1 then do;

46 genckup = 1;

47 ambtotpd + obxp01x;

48 ambfampd + obsf01x;

49 end;

50 if last.dupersid;

51 label

52 genckup ='Had Office-Based General Checkup'

53 ambtotpd='Total Amount Paid'

54 ambfampd='Amount Paid by Family';

55 keep dupersid genckup ambtotpd ambfampd;

56 run;

NOTE: There were 147490 observations read from the data set HC.H59G.

NOTE: The data set WORK.H59G has 23286 observations and 4 variables.

57 proc freq; tables genckup/missing; run;

2 The SAS System

NOTE: There were 23286 observations read from the data set WORK.H59G.

NOTE: The PROCEDURE FREQ printed page 1.

59 title3 'Variables from Full-Year File (Persons w/ Positive Weight)';

60 data h60; set hc.h60(keep=dupersid perwt01f varstr01 varpsu01 age31x--age53x inscov01);

61 by dupersid;

62 /* subset to positive weight persons */

63 if perwt01f > 0;

64 /* define AGE as last nonmissing age in 2001 */

65 if age53x ge 0 then age=age53x;

66 else if age42x ge 0 then age=age42x;

67 else if age31x ge 0 then age=age31x;

68 /* make age category variable */

69 agecat = (age ge 0) + (age gt 3) + (age gt 7) + (age gt 11) + (age gt 15) + (age gt 17);

70 /* make insurance status variable */

71 if inscov01>2 then insured=2; else insured=1;

72 label insured='Had Health Insurance in 2001';

73 run;

NOTE: There were 33556 observations read from the data set HC.H60.

NOTE: The data set WORK.H60 has 32122 observations and 11 variables.

74 proc freq; tables

75 insured*inscov01

76 agecat/list missing;

77 format insured insured. inscov01 inscov. agecat agecat.;

78 run;

NOTE: There were 32122 observations read from the data set WORK.H60.

NOTE: The PROCEDURE FREQ printed page 2.

79 proc means nmiss min max maxdec=0; class agecat; var age; format agecat agecat.; run;

NOTE: There were 32122 observations read from the data set WORK.H60.

NOTE: The PROCEDURE MEANS printed page 3.

81 title3 'Link Person-Level File from Events File with Full-Year Person File';

82 data pers; merge

83 h59g

84 h60(drop=age31x--age53x inscov01 in=a); by dupersid; if a;

85 if genckup = . then genckup=2;

86 run;

NOTE: There were 23286 observations read from the data set WORK.H59G.

NOTE: There were 32122 observations read from the data set WORK.H60.

NOTE: The data set WORK.PERS has 32122 observations and 10 variables.

87 proc freq; tables genckup; format genckup genckup.; run;

NOTE: There were 32122 observations read from the data set WORK.PERS.

NOTE: The PROCEDURE FREQ printed page 4.

89 /* sort for SUDAAN */

90 proc sort; by varstr01 varpsu01; run;

NOTE: There were 32122 observations read from the data set WORK.PERS.

NOTE: The data set WORK.PERS has 32122 observations and 10 variables.

92 title3 'Persons Age 18+';

93 proc crosstab data=pers filetype=sas design=wr notot norow;

94 subpopn agecat=6;

95 nest varstr01 varpsu01/missunit;

96 weight perwt01f;

97 subgroup genckup insured;

98 levels 2 2;

99 tables genckup genckup*insured;

100 print nsum colper secol / style=box wsumfmt=f15.0 nodate notime;

101 rformat genckup genckup.;

102 rformat insured insured.;

103 run;

3 The SAS System

Opened SAS data file PERS for reading.

NOTE: There were 32122 observations read from the data set WORK.PERS.

NOTE: The PROCEDURE CROSSTAB printed pages 5-7.

104

105 title3 'Persons Age 18+ with a General Checkup';

106 proc means data=pers(where=(agecat=6 & genckup=1)) n mean maxdec=2;

107 class insured;

108 format insured insured.;

109 var ambtotpd ambfampd;

110 weight perwt01f;

111 run;

NOTE: There were 9369 observations read from the data set WORK.PERS.

WHERE (agecat=6) and (genckup=1);

NOTE: The PROCEDURE MEANS printed page 8.

NOTE: SAS Institute Inc., SAS Campus Drive, Cary, NC USA 27513-2414

---------------------------------------------------------------------------------

----------------------------------------------------------------------------------