This article examines the reliability of using administrative billing codes to identify pediatric urinary tract infection hospitalizations. Urinary tract infections are among the most common causes of hospitalization for children, and there is substantial variation in the treatment and outcomes in pediatric urinary tract infection hospitalizations. Administrative billing codes are frequently used in health services research, but it is not known whether administrative billing codes accurately capture pediatric urinary tract infections.
The authors analyzed health records for 833 children who underwent urinalysis or urine culture in one of five participating pediatric hospitals. They conducted a retrospective study by analyzing administrative billing codes against laboratory test results and diagnoses from patient medical records.
- Of the 833 children, 265 were found to have a true urinary tract infection based on laboratory results and 447 were found to have a urinary tract infection based on provider confirmation.
- The positive predictive value of administrative billing records for urinary tract infections was 50 percent with use of laboratory testing and 85 percent with provider confirmation.
- The positive predictive value of administrative billing was higher with the subgroup of patients with a principle diagnosis of urinary tract infection.
This research suggests that administrative billing codes can be used to identify patients with urinary tract infections, but that the billing codes are most accurate when the principle discharge diagnosis is a urinary tract infection.