Possibilities of logistic regression analysis in building a prognostic model of personal risk assessment of the development of osteopenic syndrome in children with juvenile idiopathic arthritis

Abstract

Osteopenic syndrome (OS) is one of the possible complications of the course of juvenile idiopathic arthritis (JIA) in children. The issue of development and clinical implementation of an individual methodology for the definition of prognostic criteria for objective quantitative assessment of the risk of developing osteopenia in children with JIA is relevant and requires further clarification. The logistic regression, which is used to predict and estimate the probability of an observation belonging to one or another gradation of a qualitative characteristic, can be useful for creating effective predictive models.

Aim - to evaluate the prognostic significance and informativeness of some clinical indicators with the selection of the most optimal potential factors within the framework of the development of a mathematical equation for calculating the personal probability of osteopenia in a patient with JIA and the construction of an adequate, convenient for use in practical medicine prognostic model.

Materials and methods. The results of clinical, anamnestic and laboratory-instrumental data of 50 children with JIA (average age - 13.0 (11.0; 16.0) years) were analyzed. Laboratory research methods, in addition to general clinical ones, included the determination of 25 hydroxyvitamin D level, parathyroid hormone, osteocalcin, bone alkaline phosphatase (ostease), the marker of bone resorption β-Cross Laps in blood serum. The bone mineral density was assessed using ultrasound densitometry.

Results. Two prognostic models of OS risk in patients with JIA were created based on logistic regression analysis taking into account the most informative predictors. The models have high-quality operational characteristics in terms of sensitivity, specificity, and diagnostic (prognostic) effectiveness.

Conclusions. The developed prognostic models can be used in clinical pediatrics for personal assessment of the degree of risk of developing osteopenic syndrome in children with JIA, selection of high-risk groups and prevention of possible complications.

The research was carried out in accordance with the principles of the Helsinki Declaration. The study protocol was approved by the Local Ethics Committee of participating institution. The informed consent of the patient was obtained for conducting the studies.

No conflict of interest was declared by the authors.