Russo et al reported a score for predicting hospital length of stay after anterior cervical discectomy and fusion (ACDF). The score was derived using basic statistics and machine learning. The authors are from Quinnipiac University and Thomas Jefferson University.
APSS is an abbreviation for ACDF Predictive Scoring System.
Parameters:
(1) ASA score
(2) age in years
(3) COPD
(4) diabetes mellitus
(5) body mass index (BMI)
(6) hypothyroidism
(7) sex
(8) hypertension
(9) depression
(10) 2-level ACDF
Parameter
Finding
Points Logistic
Points Lasso
ASA score
8 * (ASA)
7 * (ASA)
age in years
< 65 years
0
0
>= 65 years
10
7
COPD
no
0
0
yes
7
4
diabetes mellitus
no
0
0
yes
4
1
BMI
< 35 kg per sq m
0
0
>= 35 kg per sq m
3
0
hypothyroidism
no
0
0
yes
7
4
sex
male
0
0
female
3
0
hypertension
no
0
0
yes
5
3
depression
no
0
0
yes
5
3
2-level ACDF
no
0
0
yes
2
0
total score =
= SUM(points for all of the parameters)
Interpretation:
• minimum score: 0
• maximum score for logistic regression model: 78 (assuming surgery with ASA 5 unlikely)
• maximum score for Lasso model: 50 (assuming surgery with ASA 5 unlikely)
value of X =
= (0.146 * (Lasso score)) - 3.101
probability of a length of stay >= 2 days =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.68.
To read more or access our algorithms and calculators, please log in or register.