Information Technology Reference
In-Depth Information
Table 7. Disease staging: resource demand level
Table of class by DS_RD_Level
class
DS_RD_Level(Disease Staging: Resource Demand Level)
Total
Frequency
Row Pct
Col Pct
1
2
3
4
5
1
0
0.00
0.00
4
7.14
0.00
31
55.36
0.00
21
37.50
0.00
0
0.00
0.00
56
2
318258
49.05
79.50
143313
22.09
9.54
157849
24.33
4.55
28625
4.41
2.14
744
0.11
0.29
648789
3
75289
7.92
18.81
520571
54.78
34.66
291351
30.66
8.40
60766
6.39
4.54
2308
0.24
0.90
950285
4
5925
0.68
1.48
366778
42.27
24.42
399557
46.05
11.52
90580
10.44
6.77
4791
0.55
1.86
867631
5
678
0.08
0.17
218745
27.30
14.56
458848
57.27
13.22
114716
14.32
8.58
8245
1.03
3.20
801232
6
117
0.02
0.03
123952
16.66
8.25
475322
63.89
13.70
132058
17.75
9.88
12550
1.69
4.88
743999
7
18
0.00
0.00
67788
10.41
4.51
435412
66.86
12.55
132405
20.33
9.90
15589
2.39
6.06
651212
8
8
0.00
0.00
36216
6.26
2.41
387727
67.07
11.18
133492
23.09
9.98
20651
3.57
8.03
578094
9
5
0.00
0.00
18810
3.48
1.25
345482
63.99
9.96
145324
26.92
10.87
30251
5.60
11.76
539872
10
9
0.00
0.00
5785
0.49
0.39
517993
43.71
14.93
499057
42.11
37.33
162182
13.69
63.03
1185026
Total
400307
1501962
3469572
1337044
257311
6966196
the first five. The most critical codes should be listed first, so this truncation is reasonable.
But then, how do we determine whether patients with truncated codes are more severe compared
to patients who only have 5 codes to begin with? First, we look at the difference in primary diagnosis
between the two groups: less than 4 codes and 5 codes or more. Next, we use text mining to examine
the subgroup of patients with 5 or more codes, and compare the text clusters to patients who have 4 or
fewer codes. First, we look at the most frequent codes for both groups to determine if patients with 5
or more codes have more severe problems. To simplify the analyses, we use a random sample of both
groups. Table 10 gives the frequency list for the two groups.
At first glance, the two sets of codes look quite different. Yet, once the codes related to childbirth
are eliminated, 401.9 is first in both groups followed by 428.0. In other words, it appears that if we treat
childbirth, labor, and delivery as very different from the remaining reasons to be hospitalized, the codes
Search WWH ::
Custom Search