Relative and attributable risk for cervical cancer: a comparative study in the United States and Italy.

The attributable risk for invasive cervical cancer in the US and Italian populations has been estimated in relation to main 'aetiological' factors (number of sexual partners, age at first intercourse, parity, oral contraceptive use and smoking) and history of Pap smear using data from two case-control studies conducted in the US (466 cases and 788 controls) and Italy (528 cases and 456 controls). The risk of cervical cancer increased in both studies with multiple sexual partners, decreasing age at first intercourse, higher parity, oral contraceptive use and smoking. Levels of exposure to various risk factors were markedly different in the two countries (ie number of sexual partners, frequency of oral contraceptive use and smoking were greater in the US). Multiple Pap smears and a short interval since last Pap smear strongly reduced risk of cervical cancer in both populations, although screening was much more widespread in the US study population, with only 9% of controls reporting no previous smear versus 38% of the Italian control series. The combined population attributable risk for the five 'aetiological' risk factors was slightly greater in the US study (76%) than in the Italian one (69%), chiefly because of a higher prevalence of exposure to sexual factors in US study women. A substantially larger proportion of Italian cases were due in part to deficiency in screening (46% in US and 84% in Italy). Thus, further inclusion of the effect of screening programmes (number of Pap smears and time since last Pap) led to an overall proportion of cases attributable to the examined risk factors of 87% in the US and 95% in Italy.