Article Text
Abstract
Objectives When we make a recommendation, the balance of benefits and harms should always be considered. Although various strategies have been proposed to assess the balance of benefits and harms, there is no definitive method. MCDA has been used in health technology assessment, and it can be used for making recommendations. We used multiple criteria decision analysis (MCDA) to make recommendations for cervical cancer screening guidelines.
Methods Nine experts of different backgrounds were asked to be members of a committee involved in guideline development for cervical cancer screening. They included gynecologists, epidemiologists, a public health researcher, a methodologist, a nurse, and a patient. The GRADE approach was used to make recommendations based on the following four perspectives: benefits, harms, preference/value, and resource/cost. Before the committee meeting, an evidence report and questionnaire were sent to the members. An analytic hierarchy process, which was of the MCDA methods, was used, and the priory was assigned based on their values to cytology, co-testing (Pap cytology and HPV testing), HPV testing taken by the gynecologists, and self-sampling for HPV testing. Two models were used for assessment: the first included benefits and harms, and the second for the four perspectives. Based on the results, the scores were calculated, and the total score rankings were compared between the models.
Results In the first model, cytology had the highest score, followed by HPV testing taken by the gynecologist, co-testing, and self-sampling for HPV testing. In the second model, cytology had the highest score, followed by self-sampling for HPV testing and HPV testing taken by the gynecologists. Self-sampling for HPV testing scored higher from the perspective including preferences and resources. Based on the results, both cytology and HPV testing alone was strongly recommended, and co-testing was conditionally recommended.
Conclusion To develop the guidelines for cervical cancer screening, a systematic review was conducted, and evidence for HPV testing alone and for co-testing (Pap cytology and HPV testing) for cervical cancer screening was confirmed. We used multiple criteria decision analysis (MCDA) to make recommendations for cervical cancer screening guidelines. If there is clear evidence for several strategies for a specific intervention, MCDA can be used to determine the priority of recommendations.