Blanc et al reported a score for predicting the risk of a potentially avoidable hospital readmission for a general internal medicine patient. This can help to identify a patient who may benefit from more aggressive management after discharge. The authors are from multiple institutions in Switzerland.
Patient selection: general internal medicine
Outcome: potentially avoidable hospital readmission
Parameters:
(1) admission in the 6 months previous tothe current admission
(2) length of stay in days
(3) anemia
(4) heart failure
(5) hypertension
(6) acute myocardial infarction
(7) chronic ischemic heart disease
(8) diabetes with organ failure
(9) cancer
(10) opioid therapy
(11) serum potassium in mmol/L
Parameter |
Finding |
Points |
admission previous |
no |
0 |
|
yes |
8 |
length of stay |
<= 4 days |
0 |
|
> 4 days |
3 |
anemia |
no |
0 |
|
yes |
2 |
heart failure |
no |
0 |
|
yes |
4 |
hypertension |
no |
0 |
|
yes |
3 |
acute myocardial infarction |
no |
0 |
|
yes |
-4 |
chronic ischemic heart disease |
no |
0 |
|
yes |
5 |
diabetes with organ damage |
no |
0 |
|
yes |
9 |
cancer |
no |
0 |
|
without metastases |
4 |
|
with metastases |
6 |
opioid therapy |
no |
0 |
|
yes |
3 |
serum potassium |
<= 5.5 mmol/L |
0 |
|
> 5.5 mmol/L (hyperkalemia) |
4 |
where:
• The scoring for cancer is based on maximum score of 47 reported in the paper.
total score =
= SUM(points for all of the parameters)
Interpretation:
• minimum score: -4
• maximum score: 47
Total Score |
Risk Group |
Risk 30-Day Readmission |
< 3 |
low |
2.4-2.6% |
3 to 10 |
medium |
5.2-5.7% |
> 10 |
high |
12.9-13.1% |
Performance:
• The area under the ROC curve is 0.69.