Strategies for the implementation of cervical and breast cancer screening of women by primary care physicians

Objective. While effective screening tests for the prevention and early det ection of cervical and breast cancers exist, poor screening rates are evide nt. The aim of this paper was to determine the most effective strategies fo r the implementation of breast and cervical cancer screening delivered to w omen. Methods. An in-depth search of the literature using Medline and the Cochran e Library was carried out between the years 1966 and 2000. Randomized contr olled studies addressing the delivery of both breast and cervical screening were retained for the purposes of this review. Absolute difference (AD) in screening was defined as screening rates in the intervention arm - screeni ng rates in the control arm. Number needed to intervene (NNI) is a new term developed for the purpose of this paper and refers to the number of physic ians or physician-patient pairs that must be exposed to the intervention be fore one screening test is performed. NNI is defined as 1/AD. Results. Strategies for the implementation of screening tests are divided i nto three categories: physician-only based, physician and patient based, an d patient-only based. Conclusions. Physician-based strategies, especially manual and computer-gen erated reminders, appear to be the most effective approach in the implement ation of breast and cervical cancer delivery to women. Absolute gains in sc reening rates were as high as 40% with an NNI of 2.5 physicians; therefore, approximately 3 physicians need to be exposed to a reminder notice before 1 physician actually orders the screening tests. (C) 2001 Academic Press.