Early diagnosis, place of diagnosis and persistent differences at 10 yearsin breast cancer survival. Hospitals and breast clinic cases prognosis

In the Province of Florence (central Italy), the Center for Cancer Preventi on (CSPO) runs a breast clinic for early diagnosis of breast cancer at whic h women can present as self-referrals or be referred by a general practitio ner. About 55% of breast cancers incident in the area were diagnosed at CSP O. This paper evaluates the differences observed in 10-year observed and re lative survival between hospitals and breast clinic cases. A population-bas ed data set of 1182 invasive cases diagnosed in the years 1985 and 1986 (Tu scany Cancer Registry) was analyzed. Cox models were used to evaluate the e ffects of place of diagnosis on observed survival. In comparison to hospita l cases, breast clinic cases showed a significantly higher frequency of loc alized cancer (55% vs 37%) and a better relative survival, persistent at 10 years of follow-up (69% vs 58%). The better survival of breast clinic case s persisted after age adjustment (hospitals cases; RR = 0.71, 95% CI: 0.60- 0.84) and after age + diffusion of disease adjustment (RR = 0.79, 95% CI: 0 .66-0.93). For regionally diffused cancers, breast clinic cases showed a 10 -year survival significantly better than hospital cases. In conclusion, eve n if some bias (i.e. lead-time and length bias) could affect the comparison , breast clinic cases showed a better survival persistent until 10 years af ter diagnosis. An earlier stage distribution explained, at least partly, th e different prognosis. (C) 1999 Lippincott Williams & Wilkins.