Screening programmes for cervical cancer have been credited with reducing the incidence of and mortality from cervical cancer. The main components of these screening programmes are: (i) their level of organisation; (ii) the age at which women begin screening; (iii) the age at which women discontinue screening; (iv) the interval between repeat screens; (v) the frequency at which the programmes provide screening; and (vi) the response to an abnormal screening test. However, not all screening programmes are equally efficient and differences in programme components can result in big differences in their cost effectiveness. Studies that employ cost-effectiveness analysis (CEA) to examine the efficiency of different programme components can inform the development of cost-effective programmes. This article presents findings of an international review of cost-effectiveness studies of cervical cancer screening. These studies consistently find that certain types of programmes are more cost effective than others. Programmes that are centrally organised and implemented by the public sector are reported to be more cost effective than those that use public funds for screening at other medical visits (convenience screening), or those that provide guidelines for healthcare professionals and the public to promote spontaneous discretionary screening. There is also substantial agreement about the cost effectiveness of other programme components. When multiple screenings are possible, studies report that they should generally begin at age 25 to 35 years and end at age 65 to 70 years, although it is important that older women have 3 normal Papanicolaou (Pap) smears before the discontinuation of screening. The interval for repeat screens that is reported to provide the best balance between cost and life-years saved is between 3 and 5 years. However, when a choice must be made between screening more women fewer times, or screening fewer women more times, most studies indicate that it is more cost effective to prioritize resources to obtain at least one screening for each woman. The screening of previously unscreened and high-risk populations has been shown to be especially cost effective. Despite this agreement, many studies report that models of the cost effectiveness of screening for cervical cancer are sensitive to a number of parameters. Changes in the attendance rate of the programme, the quality of the Pap smear, and the cost of the Pap smear can markedly change the cost effectiveness of a screening programme. Finally, this review discusses different perspectives of social choice analysis (e.g. CEA and cost-benefit analysis), when the objective is to prevent cervical cancer and the options are to screen, detect and treat, to reduce behavioural risk factors, and/or to pursue promising biological research.
