25/11/2021 - 14:00 - 15:10
COC 72 - DESEMPENHO DE INSTRUMENTOS DE MENSURAÇÃO DE DIFERENTES DESFECHOS RELACIONADOS AO PARTO, INFÂNCIA E ADOLESCÊNCIA |
32604 - IDENTIFICATION OF HIGH-RISK GROUPS FOR NON-VACCINATION USING DECISION TREES
THIAGO MELO SANTOS - UNIVERSIDADE FEDERAL DE PELOTAS; CENTRO INTERNACIONAL DE EQUIDADE EM SAÚDE, BIANCA DE OLIVEIRA CATA PRETA - UNIVERSIDADE FEDERAL DE PELOTAS; CENTRO INTERNACIONAL DE EQUIDADE EM SAÚDE, ALUÍSIO JARDIM DORNELLAS DE BARROS - UNIVERSIDADE FEDERAL DE PELOTAS; CENTRO INTERNACIONAL DE EQUIDADE EM SAÚDE
Objective
Non-vaccinated children are a particularly vulnerable and understudied group. Our objective was to identify subgroups of children with high prevalence of zero dose (neither BCG, polio, DPT nor measles vaccines received) using machine learning implementations of decision trees.
Methods
We developed Classification and Regression Tree (CART) models using data from DHS surveys of India 2015 and Chad 2014 in order to identify risk groups of zero dose among children aged 12 to 59 months. 25 explanatory variables were investigated.
Results
India had a zero dose prevalence (ZDP) of 8%. Children born outside of a health facility were at high risk of zero dose (ZDP: 16%), and among those, children whose mother did not receive tetanus immunisation were at an even higher risk (ZDP: 46%).
For Chad (ZDP: 19%), administrative region was selected as the first split variable with 11 subnational units considered high-risk areas (ZDP: 40%). For those regions, and similar to India, children whose mother did not receive any dose of tetanus vaccine were also considered the highest risk subgroup (ZDP: 47%).
No further split was created for children born in health facilities in India (ZDP: 6%) nor for the other regions in Chad (ZDP: 9%).
Conclusions
Two trees were created with only two splits each. Subgroups with zero dose prevalence higher than 40% were identified.
|
