Prognostic model of skin cancer risk assessment

Abstract

OBJECTIVE: Introduction: Early detection of people at risk of skin cancer will reduce the incidence of disease, lower the cost of health technologies and decrease anxiety level in patients. The aim of the work is to create a prognostic model for identifying people at increased risk of skin cancer development. PATIENTS AND METHODS: Material and methods: We used the results of our previous research on identifying risk factors in patients with actinic keratosis (AK), squamous cell carcinoma in situ (SCCis) and cutaneous squamous cell carcinoma (cSCC), who were under dynamic observation at the State Institution of Science "Research and Practical Centre of Preventive and Clinical Medicine" State Administrative Department (hereinafter SIS) in 2014-2017. RESULTS: Results: The prognostic model is valid, AUC = 0.97 (95% CI 0.96 - 0.99) showing a significant association of the risk of skin cancer development with the following factors: patient's age, sunburns, using skin sunscreens, exposure to the sun in recent times, exposure to radiological materials, drug administration (antiarrhythmic drugs, antihypertensive medications, hormonal contraceptives, antibiotics), burdened family history (melanoma, squamous cell cancer). Model sensitivity was 95.1% (95% CI 91.6% - 97.4%), specificity - 88.5% (95% CI 84.6% - 91.8%). CONCLUSION: Conclusions: The developed and analysed mathematical risk prediction system made it possible to identify 11 factors which are significantly associated with risk of skin cancer development. The prognostic model might be offered for specialists in taking decision at the stage of primary and secondary prevention of skin cancer.