Derivation and Validation of a Scoring System to Predict LAAC
Readmissions
A simple scoring system was devised based on non-routine disposition and
the presence of chronic pulmonary disease, peripheral vascular disease,
chronic kidney disease/end stage renal disease and anemia. Non-routine
disposition, chronic pulmonary disease and renal disease were assigned 2
points each, while peripheral vascular disease and anemia were given 1
point each (Figure 4). In the derivation cohort, the scoring system had
a stepwise discriminatory ability in predicting readmission. For
example, patients with a score of 1 had a readmission rate of 4.2%,
while a score of 5 or more had a readmission rate of 23.1% (Figure 5).
The C-statistic for the scoring system for the derivation cohort was
0.6787.
Figure 4. Multivariable model of 30-day readmissions in
patients undergoing LAAC procedures with scoring system integrated.
Parentheses indicate 95% confidence intervals.
Figure 5. Readmission scoring model correlating likelihood of
readmission within 30 days with numerical score using the derivation
cohort.
This scoring system was applied to the 2014-2015 NRD using the ICD-9
codes for the risk factors noted above. The scoring system maintained
predictive ability and is represented in Figure 6. The C-statistic in
the validation cohort was 0.633.
Figure 6. Readmission scoring model correlating likelihood of
readmission within 30 days with numerical score using the validation
cohort.