Time to biologic treatment and to total colectomy associated with diagnostic delay in ulcerative colitis: Analysis of an Italian regional administrative healthcare database

Background: Diagnostic delay (DD) in Ulcerative colitis (UC) is mainly documented in studies based on disease registries, and may lead to disease complications and increased rate of surgery. Objectives: To quantify potential DD in a real-world cohort of UC patients, and to assess whether DD is associated with adverse clinical outcomes. Methods: This is a retrospective cohort study. Data was retrieved from administrative healthcare databases of Tuscany, an Italian region. We included patients with a first record of a diagnosis of UC or a first record of dispensation of oral mesalazine from 6/1/2011 to 6/30/2016 (index date, ID). Patients < 18 years old at ID or with lookback period < 5 years or follow-up period < 3 years were excluded. Patients with less than 2.4 mg/day average dose of mesalazine, calculated in the first year after ID, were excluded from the cohort. Patients were classified as exposed to DD when at least 1 Emergency Department admission or hospitalization for gastrointestinal causes were recorded in the look-back period, excluding the 6 months preceding ID. Patients without any record of interest in the look-back period or with a record of interest in the 6 months preceding ID were defined as not-exposed to DD. We investigated time from ID to the date of first prescription of a biologic drug (BIO) and to total colectomy (TCOL). Hazard ratio (HZ) was estimated using Cox models adjusted for age and gender. Results: The overall cohort included 3,951 patients. Females were 1,655 (41.9%). Mean age was 53.5 (standard deviation: 16.7). Patients exposed to DD were 1,640 (41.5%) with a median delay time of 534 days, and those not-exposed were 2,311 (58.5%). Out of the whole cohort, 286 patients had the first prescription of BIO and 54 had a TCOL with an average time of 2.86 and 2.98 years, respectively. Among patients non-exposed to DD, cases of BIO and TCOL were 160 (6.9%) and 30 (1.3%), and among exposed ones were 126 (7.7%) and 24 (1.5%), respectively. The HZ was 1.13 (confidence interval, CI, 95%: 0.89-1.43) for BIO, and 1.17 (CI 95% 0.68-2.00) for TCOL. Conclusions: A non-significant risk of biologic therapy and surgery is reported for UC patients with suspected DD compared with UC patients timely diagnosed. Further studies using other databases are needed to investigate UC patients' outcomes related to DD