Exploring the importance of method information index on early discontinuation of intrauterine contraceptive device in India using classification and regression tree analysis of survival data

Rajaram Subramanian Potty, Karnataka Health Promotion Trust
Jayachandran A A, Avenir Health (Track20 India Project)
AG Khan, Government First Grade College & Post Graduate Centre

In India, intrauterine device (IUD) is one of the most significant modern contraception due to its effectiveness and programmatic importance attached. However, high discontinuation of IUD use is a major issue though it is a priority method. We attempted to examine early IUD discontinuation in India using NFHS-4 calendar data. We extracted all the episodes and included only the IUD use that started a month after the start month of the calendar and three months prior to the survey date for the analysis. We used lifetable, cox proportional hazard model and classification and regression trees for survival data to examine the IUD discontinuation rates by 12 and 24 months. The hazard of discontinuation within 12 (RHR:2.96) and 24 (RHR: 2.91) months was relatively highest for women from southern India who did not receive any method information. Twelve-month discontinuation was lowest (RHR:0.15) among north Indian women who received method information with first episode having family planning exposure of average or below and no surviving sons. Twenty-four-month discontinuation rate was lowest (RHR:0.22) among women from other than west and south who received method information with first episode having ideal number of children less than two and wanted to have another child.

Keywords: Family planning and contraception, Life event calendar analysis, Event history analysis, Policy

See paper.

  Presented in Session 65. Factors Inhibiting Continuity in Contraceptive Use