Abstract:
Introduction: Achieving and maintaining optimal glycemic targets can be difficult because of several factors that make uncontrolled diabetes a public health problem. The objective of this study was to propose a predictive score for prolonged poor glycemic control in the type 2 diabetes cohort. Methodology: This was a secondary data analysis of a crosssectional study. The dependent variable was prolonged poor glycemic control. The modality of the variable with the lowest adjusted OR in the model was assigned a point. The points of the other modalities were weighted proportionally to this variable. Logistic regression was performed and tested by a ROC curve.
Results: 270 patients were included in the study. In multivariate analysis, low educational level (OR=8.34, CI95% [1.97-35.22]); family support for diabetes management (OR=0.65, CI95% [0.45-0.94]); abdominal obesity (OR=2.27, CI 95% [1.08-4.77]); a history of hospitalization (OR=7.39, CI95% [2.97-18.39]); poor adherence to treatment (OR=2.97, CI 95% [1.42-6.18]); and microangiopathy (OR=5.05, CI 95% [2.36-10.81]) were factors independently associated with prolonged poor diabetes control. A score greater than or equal to 45 was found in this study. The sensitivity and the specificity in our study were respectively 78.89% and of 84.51% with a good performance (AUC= 0.87).
Conclusion: The Predictive score is made up of a triad of patient, family and caregiver factors. All of these
components are modifiable factors.
Keywords: Predictive Score, Prolonged Poor Glycemic Control, Type 2 Diabetes, Ouagadougou

Predictive Score, Prolonged Poor Glycemic Control, Type 2 Diabetes