Information Technology Reference
In-Depth Information
Table 9. Correlation of antibiotic and hemodialysis
Table of hemod by antibiotic
hemod
antibiotic
Total
Frequency
Row Pct
Col Pct
0
1
0
220171
97.55
97.99
5535
2.45
98.19
225706
1
4513
97.79
2.01
102
2.21
1.81
4615
Total
224684
5637
230321
Table 10. Correlation of transfusion and venus catheterization
Table of vcath by transf
vcath
transf
Total
Frequency
Row Pct
Col Pct
0
1
0
213265
96.10
96.90
8649
3.90
84.53
221914
1
6824
81.17
3.10
1583
18.83
15.47
8407
Total
220089
10232
230321
if (rxmatch('9904',prcomb)> 0 ) then transf= 1 ;
if (rxmatch('3893',prcomb)> 0 ) then vcath= 1 ;
if (rxmatch('9671',prcomb)> 0 ) then ventl96= 1 ;
if (rxmatch('3324',prcomb)> 0 ) then biopbronc= 1 ;
if (rxmatch('9672',prcomb)> 0 ) then ventm96= 1 ;
if (rxmatch('3491',prcomb)> 0 ) then thorac= 1 ;
if (rxmatch('3995',prcomb)> 0 ) then hemod= 1 ;
if (rxmatch('9604',prcomb)> 0 ) then endotrach= 1 ;
if (rxmatch('9921',prcomb)> 0 ) then antibiotic= 1 ;
if (rxmatch('9394',prcomb)> 0 ) then respmed= 1 ;
if (rxmatch('8872',prcomb)> 0 ) then ultrasound= 1 ;
run;
Note that for the first table analysis in Table 9, 2% of patients with COPD receiving hemodialysis
also receive an antibiotic. Similarly, for Table 10, 18% of patients with COPD receiving a venous cath-
eter also receive a transfusion. The issue of multicollinearity is considerable and cannot just be ignored
when defining these indicator functions.
 
Search WWH ::




Custom Search