Mather et al reported a model for predicting 30-day readmission for a patient with pneumonia. This can help to identify a patient who may benefit from more aggressive management. The authors are from Hartford Hospital in Connecticut.
Patient selection: hospitalization for pneumonia
Parameters:
(1) sex
(2) number of previous admissions
(3) chronic lung disease
(4) cancer
(5) median income
(6) history of anxiety or depression
(7) hematocrit
Parameter
Finding
Points
sex
female
0
male
0.463
number of previous admissions
<= 2
0
>= 3
0.608
chronic lung disease
no
0
yes
0.489
cancer
no
0
yes
0.78
median income
< $43,000
0.599
>= $43,000
0
history of anxiety or depression
no
0
yes
0.481
hematocrit
>= 30%
0
< 30%
0.62
where:
• The cutoff for income needs to be adjusted with time.
total score =
= SUM(points for all 7 parameters)
X =
= SUM(points for all 7 parameters) - 2.976
probability of readmission =
= 1 / (1+ EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.71.
To read more or access our algorithms and calculators, please log in or register.