Development and validation of a haematuria cancer risk score to identify patients at risk of harbouring cancer

• DOI: 10.1111/joim.12868
• Publication Date: 2018-12-06T19:33:58Z
• Authors: W. S. Tan; A. Ahmad; A. Feber; H. Mostafid; J. Cresswell; C. D. Fankhauser; S. Waisbrod; T. Hermanns; P. Sasieni; J. D. Kelly
• Keywords: Adult; Male; Adolescent; detection; 610 Medicine & health; Risk Assessment; Sensitivity and Specificity; nomogram; Young Adult; 03 medical and health sciences; 0302 clinical medicine; Risk Factors; 616; Humans; Prospective Studies; Aged; Hematuria; Aged, 80 and over; predict; bladder cancer, detection, haematuria, nomogram, predict, urinary tract cancer; Smoking; Reproducibility of Results; Original Articles; Middle Aged; haematuria; urinary tract cancer; 3. Good health; 10062 Urological Clinic; Urinary Bladder Neoplasms; 2724 Internal Medicine; bladder cancer; Female

Abstract Background A lack of consensus exists amongst national guidelines regarding who should be investigated for haematuria. Type haematuria and age‐specific thresholds are frequently used to guide referral the investigation Objectives To develop externally validate cancer risk score (HCRS) improve patient selection Methods Development cohort comprise 3539 prospectively recruited patients at 40 UK hospitals (DETECT 1; ClinicalTrials.gov: NCT02676180) validation 656 Swiss patients. All were aged >18 years referred hospital evaluation visible nonvisible Sensitivity specificity HCRS in derived from a cut‐off identified discovery cohort. Results Patient age, gender, type smoking history HCRS. achieves good discrimination (AUC 0.835; 95% CI: 0.789–0.880) calibration (calibration slope = 1.215) with no significant overfitting ( P 0.151). The detected 11.4% n 8) more cancers which would missed by National Institute Health Clinical Excellence guidelines. American Urological Association identify all 12.6% compared 30.5% achieved upper tract have been identified. Conclusion offers discriminatory accuracy is superior existing simplicity model facilitate adoption physician decision‐making.