Predicting Post-operative Readmissions in Pediatric Urology by Using Three Validated Comorbidity Indices
Sources of Funding: K08-DK100534 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Introduction
There is a lack of validated surgical risk assessment scores in pediatric urology to predict mortality and complications. Indices such as Charlson Comorbidity (CCI) and Van Walraven (VWI) were developed using adult cohorts, and the Rhee Index was established for pediatric general surgery patients. To assess their applicability in pediatric urology, we compared how well the three comorbidity indices predict post-operative readmissions to the emergency room and inpatient unit after urological procedures.
Methods
We analyzed the State Inpatient Databases (SID) from 2007 to 2010. The SID is specifically designed to track ER and inpatient visits. We included pediatric patients (< 18 y) who underwent the following urological procedures: ureteroneocystostomy, ureteroureterostomy, radical/partial nephrectomy, pyeloplasty, appendicovesicostomy, enterocystoplasty, vesicostomy, and bladder neck sling. 30-day ER and inpatient readmissions were extracted. Comorbidity scores were calculated using each index. We used descriptive analysis to describe the patient cohort. To compare the performance on predicting post-operative readmissions, receiver operating characteristics (ROC) were constructed for each index.
Results
We identified a total of 6,752 patients. The median age was 4 years; males accounted for 43.9% of the cohort, and 52.4% were privately insured. 7.4% had at least one inpatient readmission, and 8.1% had at least one ER admission. The CCI had the best predictability for 30-day inpatient readmissions (AUC=0.63) than VWI (AUC=0.54) and Rhee Index (AUC=0.56); p<0.0001. All three indices performed similarly poor in predicting 30-day ER admissions: CCI (AUC=0.52), VWI (AUC=0.51), and Rhee Index (AUC=0.50); p=0.5.
Conclusions
The Charlson Index was significantly better at predicting inpatient readmissions than Van Walraven or Rhee Index, but the three scores were equally poor in predicting post-surgical ER admissions. The three indices were designed to predict mortality and, thus, performed significantly less well in predicting readmissions. Our result supports that a new risk index needs to be developed to better predict post-operative readmissions in pediatric urology patients.
Funding
K08-DK100534 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Steven Wolf
J. Todd Purves
John S. Wiener
Jonathan C. Routh