In a case-note review of 120 women and 227 men presenting with muscle-invasive bladder tumours in 1998, survival was worse for women in 3 years of follow-up, with the greatest difference, of 19.9%, at 6 months. For more deprived women, 6-month survival was 52.3%, and 32 (37.2%) presented with advanced disease, compared with 73.5%, and three (8.8%) for less deprived women. Bladder cancer is the only common cancer for which women have worse prognosis than men as reported in the UK, continental Europe and North America (Micheli et al, 1998;Coleman et al, 1999;Ries et al, 1999). Studies based on routine registry data from the Netherlands and the US have reported later stage at diagnosis for women than for men and worse stage-specific survival (Fleshner et al, 1996; Mungan et al, 2000a, b). These findings are unexplained. We report the results of a case-note review of patients with newly diagnosed muscle-invasive bladder cancers to explore the reasons for differences in survival by sex.
MATERIALS AND METHODSPatients were eligible if they were diagnosed with a muscleinvasive bladder carcinoma in 1998, were residents of Greater Manchester or Lancashire (GML) and had no previous history of bladder cancer.All patients in GML diagnosed with a new bladder tumour in 1998 were identified from the database held at the North Western Cancer Registry, which receives information from all hospitals and pathology laboratories in GML, and death details from the Office of National Statistics (ONS). Pathology reports received by the Registry were reviewed for evidence of muscle-invasive disease (MID). Case notes were requested if pathology confirmed or suggested MID, or information was incomplete. In order to identify patients with a diagnosis of MID based on clinical or radiological, rather than histological, findings, the case notes of patients not in the above categories but who either underwent radiotherapy in 1998 or who died between 1st January 1998 and 31st December 2001 were also requested.Patients were included in the study if they had MID confirmed on histology, or they fulfilled one of the following conditions: residual mass present following resection of a bladder tumour; bladder mass on scan or ultrasound with bladder cancer considered the most likely diagnosis; or a large necrotic bladder tumour on cystoscopy which was not investigated further.The dates of death were obtained from the Registry, and deaths were classified into those due to bladder cancer or other causes, using death certificates and case notes. Stage was determined using pathology reports and case notes, with clinical staging used for all analyses. Initially, tumours were staged as T2, T3, T4a or T4b, but for analyses were classified into advanced (T4b and/or M1) or nonadvanced. Details of histology and grade were obtained from pathology reports.Material deprivation was measured using the Townsend score of the enumeration district (ED) in which the patient lived, determined from Registry postcode details. Scores were classified into five groups based on q...