Using Propensity Score Bootstrapping on Determining the Model of the HIV/AIDS Patients’ Assistance Mahdalena1,*, Mahpolah1, Rajiani Ismi2 1The Ministry of Health Polytechnic Banjarmasin, Indonesia 2STIA & Port Management Barunawati Surabaya, Indonesia *Corresponding Author: Mahdalena Head of Research and Community Service, The Ministry of Health Polytechnic Banjarmasin, Indonesia, Email: lenaf4dl1@gmail.com
Online published on 21 February, 2019. Abstract Background The combination of Propensity Score Stratification (PSS) with a bootstrap method will get an estimate that has better accuracy because the covariates will be more balanced. Propensity score Bootstrapping is used to estimate the response to mentoring HIV/AIDS patients in RSUD dr. H. Moch. Ansari Saleh Banjarmasin South Kalimantan. Method The research method used is non-reactive based on secondary data. The study sample was HIV/AIDS patients who had undergone treatment for more than one year at RSUD dr. H. Moch. Ansari Saleh Banjarmasin, South Kalimantan. The response variable is y)y) CD4 (Cluster of Differentiation) level, while predictor variables (Xj) knowledge, attitude, self-concept, family support, and treatment in the form of antiretroviral therapy. Results From the test, the significance of parameters bootstrap shows that at the significance level α = 5% indicating all covariates are significant, namely the covariate of knowledge, attitude, self-concept, family support. Estimation of the PSS bootstrap sample is best in the two strata group with a standard error of 0.0127 and a treatment effect of 0.4175. Conclusion The ATE (Average Treatment Effect) value in the univariate response using a proportion test with z test statistics showed that antiretroviral therapy and assistance affected Cluster of Differentiation (CD4) levels. Top Keywords Propensity Score Stratification, Bootstrap, Antiretroviral Therapy. Top |