THE RISK PREDICTION OF THE FIRST TOE INGROWN NAIL DEVELOPMENT IN POPULATION AT THE OUTPATIENT STAGE

One of the most common reasons of patients’ referrals to surgeons at outpatient stage is the development of ingrown toenail (hereinafter – IT), predominantly affecting the first toe

The etiological component of IT is represented by pathogenic fungi and opportunistic microflora of the skin of the feet and toes, and in the pathogenesis of the disease, it is possible to distinguish groups of risk factors (hereinafter – RF) that produce and also contribute to the development of IT, such as feet hyperhidrosis and periungual roller injuries

The purpose of the study: elaboration of a methodology for predicting the development of an ingrown nail of the first tie in patients at the outpatient stage, followed by the creation of an eponymous software product.

By design, this study is a retrospective analysis with statistical processing and expert evaluation of medical data. We analyzed 4185 outpatient medical records of patients who received medical aid at the outpatient Center of the Samara State Medical University Clinics from September 2020 to January 2023 with various nosological forms of diseases belonging to the group of surgical infections. Outpatient records of 132 patients were selected and analyzed from the array of medical documents. The selection criterion for the study with further analysis and expert evaluation of the data was the presence of IT of the first toe in the patient.

At the next stage, based on the results of a retrospective analysis of outpatient medical records, RF were identified, contributing to the development of IT of the first toe in the patient (Table 1).

At the next stage, the identified RF were analyzed using the expert assessment technique. At the same time, a gradation and determination of the significance of each of the RF related to both productive and predisposing ones for the development of an IT of the first toe were carried out, as well as assignment of a point value to each of them.

The expert assessment was carried out by 18 field-oriented specialists – doctors of the surgical chairs of the Samara State Medical University, who have clinical experience in providing medical care to patients with an IT. The method of direct independent personal assessment was chosen, which, on the part of medical experts, was carried out by means of anonymous questionnaires. The experts were asked, considering their own clinical experience, to assign each RF a point value from 0 to 10, depending on its significance and influence on the occurrence of an IT of the first toe. Then the data of the expert assessment were compared with the data of a retrospective analysis of outpatient medical records of patients and subjected to mathematical analysis.

According to the clinical data, presented in the array of medical records, as well as considering the scores of various expert assessments, the total scores were calculated for each patient on 18 scales. Thus, for each patient in the data set, 18 new variables were created, corresponding to the scores collected on the assessment scales of medical experts.

When several medical experts participate in an anonymous survey, discrepancies in their assessments of the significance of certain RF are inevitable, but the magnitude of this discrepancy is important. A group assessment can be considered sufficiently reliable only if there is good agreement between the answers of individual specialists. To assess the significance of this parameter, Kendall's multiple rank correlation coefficient (W) was calculated. As a result of the calculations, the Kendall coefficient turned out to be 0.83 (p<0.001). Cronbach's alpha coefficient was used to determine the internal consistency of characteristics describing an object. The calculations showed that the value of Cronbach's alpha was 0.98. These values of the Kendall and Cronbach coefficients characterize the consistency of the assessments of expert doctors as rather high.

The third stage of the study consisted in assessing the quality of recognition of the possible development of an IT of the first toe in patients at the outpatient stage using the identified RF. To this end, we used ROC-assay, built characteristic ROC-curves with an estimate of the area under the graph and a proposal for possible threshold values of points on a scale, above which the risk of IT developing was considered increased with the corresponding sensitivity and specificity. The use of ROC-assay made it possible to assess the professional skills and qualifications of medical experts.

Figure 1 - Areas under the graphs of ROC-curves (AUC) according to expert assessments of RF

Note: DE – doctor-expert, “mustache” – the limits of the 95% confidence interval

DOI:10.23670/IRJ.2024.139.146.2

As a result, the areas under the ROC-curves (Figure 1) for all doctor-experts statistically significant exceeded the value of 0.5. This value corresponds to the area under the diagonal on the graph, and a random guess. It can also be seen from the graph that the assessments of all doctor-experts have approximately equal discriminatory characteristics, since their “mustache” – 95% confidence intervals, overlap.

The quality of recognition (sensitivity, specificity) of the risk of IT of the first toe developing in patients at the outpatient stage according to the scale of each of the doctor-experts is presented in Table 2.

At the final stage, we built a summary table of the predictive significance of RF by superimposing them on the results of the analysis of outpatient medical records of patients (Table 3).

The Pearson’s criteria for assessing the strength of the relationship between the RF and the outcome showed that strong and relatively strong factors include producing (trauma to the nail plate), from predisposing – fungal nail lesions, coupled with their deformation, as well as congenital deformities of feet, and from additional RF - the presence of diabetes mellitus and adiposis, as well as hyperhidrosis of the feet.

The generated table of outcomes (Table 4), in which the observed indicators according to the analysis of 132 medical records are opposed to those predicted on the basis of the calculated model, demonstrates a high quality of prognosis, namely 92.4%, in terms of the risk of IT of the first toe developing in patients at the outpatient stage.

Based on the developed forecasting methodology, the authors team created a computer software product – “Determination of the Risk of Primary Development of an Ingrown Nail of the First Toe”. For this application, a certificate of state registration of the computer program No. 201617934 of 07.15.2020 was received

The quality of the prognosis of the proposed author's methodology for assessing the risk of developing an IT of the first toe in patients at the outpatient stage and of the program developed on its basis, allows us to recommend their use in everyday clinical practice.