COIOSDeep risk insights through multi-disciplinary research across the globe
people walking on street during daytime
Item confirms Preprint · not peer-reviewed Added

Machine learning predicted HPV vaccine hesitancy no better than logistic regression

Among 1,156 parents in Buea Health District, Cameroon, logistic regression, random forest and gradient boosting identified hesitant parents equally well, with area under the curve between 0.841 and 0.870.

A secondary analysis of a survey of 1,156 parents of children aged 9 to 18 in Buea Health District, Cameroon, modelled HPV vaccine hesitancy. Logistic regression, random forest and gradient boosting performed equally well on the held-out fifth of the survey, area under the curve 0.841 to 0.870. Insufficient vaccine information, perceived unsafety and distrust in the Ministry of Health predicted hesitancy more strongly than sociodemographic disadvantage.

Why it is interesting: Added algorithmic complexity bought no accuracy over standard logistic regression on tabular survey data.

Source
medRxiv, 20 September 2026
Paper
Predicting parental Human papillomavirus vaccine hesitancy: development and internal validation of machine learning models
Authors
Alabi IO, Bekalo DB, Kartsonaki C, Agbor VN
DOI
10.64898/2026.09.18.26363393
Type
Preprint
Design
Secondary analysis of a cross-sectional survey of 1,156 parents in Buea Health District, Cameroon; 80/20 split, internal validation only
Verdict
Confirms prior evidence
Driver
AI in medicine

Other drivers in Data, models and systems