The Use of the Prediction Model for the Duration of Contraceptive Use in Anticipation of Drop Out in the New Normal

The COVID-19 epidemic affects to the implementation of the Family Planning program. The implementation of procedures to prevent the spread of COVID-19 has prevented intensive family planning service activities and causing significant decrease in the number of active family planning participants in Indonesia. This study proposes the implementation of a prediction system as a solution to anticipating family planning dropouts when pandemic. The model is built through data mining classification techniques with historical data sources on the use of contraceptive tools/methods from the 2017 Indonesian Demographic and Health Survey. With accuracy level of 0.851, the system predicts duration of use of contraceptive methods in 3 alternative outputs, namely 1-3 months, 4-12 months and more than one year. The information can be used to anticipate the potential drop out of family planning participants. Program initiatives can be focused on family planning participants who are predicted to have low continuity of use of contraceptives. The results of predictions in regional units, can be used to map areas based on the level of continuity in the use of contraceptives. Program initiatives can be right on target, focusing on areas with low levels of continued use of contraceptives