A retrospective claims database analysis to assess patterns of interstitial cystitis diagnosis

ABSTRACT Objective: Interstitial cystitis (IC) is often misdiagnosed as one of several other conditions manifesting similar symptoms. This analysis assesses the potential extent of IC misdiagnosis while considering concomitant conditions in a managed care population and identifies predictors of IC diagnosis. Research design and methods: Administrative insurance claims data covering 1.7 million lives (1999–2003) were analyzed. Insurance enrollees with ≥ 1 IC diagnosis (ICD-9-CM of 595.1x) were identified as IC patients. A random sample of non-IC controls was selected using a 10:1 matching ratio. Six‐month incidence rates of ‘commonly misdiagnosed conditions’, (overactive bladder, urinary tract infection, chronic pelvic pain, endometriosis, prostatitis) were compared before and after patients’ initial IC diagnosis and the reduction in incidence rate of commonly misdiagnosed conditions was used as a suggestive measure of the extent of IC misdiagnosis. The Kaplan–Meier method was used to assess the extent that commonly misdiagnosed conditions were predictors of subsequent IC. A Cox Proportional Hazards regression model (that adjusts for patient demographics, concomitant and misdiagnosed conditions) was used to estimate the hazard ratio (HR) of these conditions. Similar analyses were performed for the ‘commonly concomitant conditions’ (fibromyalgia, irritable bowel syndrome, vulvodynia). Results: There were 992 IC patients and 9920 controls identified. The reduced incidence of commonly misdiagnosed conditions after initial IC diagnosis suggests that the misdiagnosis rate could be as high as 38% within the 6‐month period before initial IC diagnosis. Conclusions: Diagnoses of commonly misdiagnosed conditions are significant predictors of future IC diagnosis. When overlooked, potential misdiagnosis of IC can lead to underestimation of the true prevalence of IC. Similarly, diagnoses of commonly concomitant conditions are significant predictors of future IC diagnosis. These initial findings based on claims data suggest hypotheses for further investigation with clinical data. These results suggest more consideration of IC as a diagnosis is warranted, especially when certain diagnoses are repeatedly made and the resulting treatments do not alleviate the patient's symptoms.