MP24-02: Global variation in cancer detection rates in patients referred to secondary care with haematuria: do some over investigate? Results from the IDENTIFY collaborative study.
Friday, May 15, 2020
7:00 AM – 9:00 AM
Kevin Gallagher, Arighno Das, Sinan Khadhouri, Kenneth MacKenzie, Taimur Shah, Chuanyu Gao, Eleanor Zimmermann, Eric Edison, Matthew Jefferies, Arjun Nambiar, Matthew Nielsen, Joshua Meeks, John McGrath, Veeru Kasivisvanathan, The BURST research collaborative The IDENTIFY study group
Introduction: The aim of this study was to determine the prevalence of urinary tract cancer (UTC) in secondary care. Secondary aims were to explore reasons for differences in prevalence between countries.
Methods: IDENTIFY was a prospective international cohort study. Patients referred to secondary care with symptoms of urinary tract cancer (no previous history of UTC) were included. Data on presenting symptoms, investigations and diagnoses were recorded in an online database.
Results: Data on 10873 patients from 110 hospitals in 25 countries were included. 7155 (65.5%) had visible haematuria (VH), 3155 (28.9%) had non-visible haematuria (NVH), 614 (5.6%) were referred for a reason other than haematuria (NH). The rate of urinary tract cancer detected in secondary care varied between countries from 4.8% to 72% (median 26%, IQR 11.9-42.0). Using a mixed model with age, haematuria type, gender, smoking status and imaging protocol as fixed effects and country and hospital as random effects the adjusted prevalence of urinary tract cancer in this population was 26.2% (95% CI 18.6-35.5).
Country was associated with cancer detection rate independent of age, the type of haematuria investigated, gender and smoking (p<0.001). Higher “healthcare access and quality index” for a country was significantly correlated with lower cancer detection rates (r=0.51 p=0.008). The bladder cancer detection rate in a country and the proportion of cases found to have “high risk” bladder tumours in that country was not significantly correlated (r=0.32, p=0.13)
The proportion of a country's cohort of patients investigated with VH, the mean age of patients and the proportion of patients who smoked were all associated with higher rates of cancer detection in that country.
The incidence of bladder cancer in the population of each country in 2018 (according to Globocan) did not significantly correlate with the bladder cancer detection rate in the IDENTIFY study cohort from that country (r=0.24, p=0.25).
Conclusions: The IDENTIFY study demonstrates that a wide variation in the prevalence of urinary tract cancer detected in secondary care exists between different countries for patients referred with symptoms suggestive of urinary tract cancer. This appears to be due to differences in patient selection for investigation. These data support the case for a individualized risk-based approach to suspected urinary tract cancer investigation. Source of
Funding: Charitable grants from: The Urology Foundation, The Rosetrees Trust and Action Bladder Cancer UK.