Predictive Modeling of Non-Communicable Diseases Using Social Determinants of Health as Features: A Review of Existing Approaches


Peatiphat Bhoothookngoen & Nattapong Sanchan

    Abstract:

    The paper presents a comprehensive review of the current state of predictive models for non-communicable diseases (NCDs) prevalence, specifically focusing on utilising social determinants of health (SDHs) as features for model training. The review's search strategy employed a thorough screening process to select sixteen studies for inclusion. These studies used supervised, unsupervised, and other algorithms to forecast NCDs' burden; the most frequently applied attributes were age, gender, Fasting Blood Sugar (FBS), physical inactivity, obesity, and smoking. The evaluation methods for the models included a range of metrics, such as Percent Accuracy, Receiver Operating Characteristic (ROC), and Hamming loss. The review concludes that predictive models have the potential to forecast NCD prevalence accurately and highlights the need for further research that focuses on incorporating SDH-related factors as features for model training.

    Keywords: Non-communicable diseases, social determinants of health, prediction, machine learning

    References:

    Download this paper: [PDF]
    Dowload bibtex entry: [BibTex] 
    @article{2023_Bhoothookngoen_Sanchan,
    place={Bangkok, Thailand}, 
    title={Predictive Modeling of Non-Communicable Diseases Using Social Determinants of Health as Features: A Review of Existing Approaches}, 
    volume={19}, 
    url={https://ph02.tci-thaijo.org/index.php/sej/article/view/249846}, 
    number={1}, 
    journal={Srinakharinwirot University Engineering Journal}, 
    author={Bhoothookngoen, Peatiphat and Sanchan , Nattapong}, 
    year={2023}, 
    month={Dec.}, 
    pages={79–88}
      }
    Rich text bibliography entry (for copy & paste into a word processor):
    Bhoothookngoen, P., & Sanchan , N. (2023). Predictive Modeling of Non-Communicable Diseases Using Social Determinants of Health as Features: A Review of Existing Approaches. Srinakharinwirot University Engineering Journal, 19(1), 79-88.