Statistical Methods - Text Mining Techniques for Healthcare Provider Quality Determination - page 137

Table 8. Mortality (all causes) by hospital

Table of DSHOSPID by DIED

Hospital Code DIED Total

Frequency
Row Pct
Col Pct

0 1

1 2795

97.12
9.21

83
2.88
12.56

2878

2 1460

96.95
4.81

46
3.05
6.96

1506

3 884

97.46
2.91

23
2.54
3.48

907

4 7652

97.76
25.21

175
2.24
26.48

7827

5 2369

97.89
7.80

51
2.11
7.72

2420

6 5237

96.84
17.25

171
3.16
25.87

5408

7 1476

98.07
4.86

29
1.93
4.39

1505

8 938

100.00
3.09

0
0.00
0.00

938

9 5370

98.62
17.69

75
1.38
11.35

5445

10 2172

99.63
7.16

8
0.37
1.21

2180

Total 30353 661 31014

mortality rate ranges from a low of 0 to a high of 3.16. We want to know if the hospital with zero deaths
has patients that are as severe as the hospital with 3.16% deaths.
Table 9 gives the proportion of patients with septicemia by hospital. Note that hospital #8 with zero
deaths also has zero patients with septicemia. Hospital #6 with the highest death rate has almost 7%
patients with septicemia, which is the highest of the ten hospitals. This hospital should probably be
investigated to determine whether this high rate of septicemia is a result of nosocomial infection, or
whether patients enter the hospital with it. Table 10 gives the rate of pneumonia. Does this hospital take
in sicker patients compared to the other nine hospitals?
Hospital #6 again has the highest rate of pneumonia to go with the highest death rate; hospital #8
has the lowest rate of pneumonia; in fact, it is the only hospital with a rate of less than 1%. Table 11