Oral contraceptives and cervical cancer risk in Costa Rica. Detection bias or causal association?

To examine the relationship between cervical cancer and oral contraceptive (OC) use, we analyzed data from a population-based, case-control study in Costa Rica. Women aged 25 to 58 years in whom cervical cancer was diagnosed and reported to the National Tumor Registry were examined as two separate case groups: invasive cervical cancer and carcinoma in situ (CIS). Controls were women aged 25 to 58 years identified through a national survey. Women who had used OCs had no increased risk of invasive cervical cancer compared with women who had never used OCs (relative risk, 0.8; 95% confidence interval, 0.5 to 1.3). Women who had used OCs had an increased risk of CIS compared with those who had never used OCs (relative risk, 1.6; 95% confidence interval, 1.2 to 2.2). However, further analyses indicated that this increased risk was confined to those who had recently used OCs. Also, the risk of CIS was not elevated in subgroups in which a history of cervical smears was not strongly linked to OC use. The elevated risk of CIS among OC users may therefore reflect a bias caused by enhanced detection of disease rather than a causal association.