Abstract MP31-09: A predictive model to help identify factors associated with submitting 24-hour urine collections (APL - 2018)

A predictive model to help identify factors associated with submitting 24-hour urine collections

View Poster

INTRODUCTION

A 24-hour urine is the standard for metabolic work up to help prevent recurrent nephrolithiasis in high-risk patient populations. Our objective was to identify how demographic factors, high-risk medical characteristics for stone formation, and differences in patient management influence post-operative follow-up in regards to workup for stone prevention.

METHODS

A retrospective chart review of patients treated for urolithiasis between August 2014 and March 2017 was performed. Patient demographics, medical characteristics, stone factors, type of treatment, and compliance data were compiled into our dataset. Patients who were requested to obtain a 24-hour urine collection were included. Bivariate analysis using Fishers exact test was used to identify patient variables that made 24-hour urine collection more likely. Variables that were statistically significant on bivariate analysis were then used to formulate a predictive model associated with 24-hour urine collection.

RESULTS

Of the 303 patients who met inclusion criteria, 184 (60.7%) submitted an initial 24-hour urine collection. On bivariate analysis patients older than 50 were more likely to submit a 24-hour urine collection (71.4% vs 51.5%; p <0.001), patients with a metabolic predisposition for stones were more likely to submit a 24-hour urine collection (70.6% vs 53.1%; p =0.003), and patients who did not have surgery were more likely to submit a 24-hour urine collection (97.7% vs 53.5%; p <0.001). Our three-variable prediction model found that not undergoing surgery was a strong predictor of 24-hour urine collection regardless of age or metabolic predisposition for nephrolithiasis. The patient least likely to collect a 24-hour urine was age 50 or younger, who had surgery, and did not have a metabolic predisposition for nephrolithiasis.

CONCLUSION

Our predictive model demonstrated a very high probability of 24-hour urine collection submission for patients who did not have surgery for their kidney stones vs those patients who did have surgery. We suspect that patients perceive surgery as a more definitive treatment for kidney stones when compared to conservative management, highlighting the importance of patient education regarding the natural history of nephrolithiasis and role of metabolic management for disease prevention.

Funding: none