CLINICAL PREDICTORS OF BLASTOCYSTOSIS TREATMENT EFFICACY.

Improvement of methods and criteria for diagnostics of human blastocystosis and selection of drugs for etiotropic therapy remain urgent scientific tasks of the present day. Clinical manifestations of blastocystosis in humans are diverse, and in this regard, their influential role on the effectiveness of blastocystosis therapy cannot be excluded. To assess the predictor properties of clinical symptoms, we retrospectively analyzed them in 300 patients with blastocystosis in alternative groups: group A - with good therapeutic effect (n=162) and group B - with satisfactory effect (n=138). Group A included patients who had an overall regression rate of clinical symptoms ≥33% after 1 month from the start of treatment and ≤32% in group B. Preliminary analysis showed that reliable differences between the groups were established for most clinical symptoms, which provided the basis for the development of a prognostic algorithm using a heterogeneous sequential Wald-Henkin procedure. It was found that high prognostic informativity was shown by liver enlargement (J=3,38), the presence of heaviness in the right subcostal area (J=2,45), tachycardia (J=1,76), decreased efficiency (J=1,63), the degree of manifestation of clinical symptoms in general (J=1,74), the nature of body temperature (J=1,56), the presence of dizziness (J=1,46) and duration of the disease (J=1,07). Moderate predictive properties were characterized for the patient's age (J=0,67) and the presence of vegeto-visceral dysfunction (J=0,96), joint pain (J=0,87), headache (J=0,67), heart pain (J=0,64), and muscle pain (J=0,53). Low prognostic informativeness was established with regard to the presence of seizures (J=0,42), lymphoadenopathy (J=0,33) and spleen enlargement (J=0,30). No predictor value was found for symptoms such as abdominal pain syndrome (J=0,09), memory impairment (J=0,008), skin manifestations (J=0,04), jaundice (J=0,02), and patient gender (J=0,01). The testing of the predictive algorithm on the training group (n=300) established its high efficiency, as the erroneous predictions amounted to 3% and did not exceed the specified (5%; p<0.05) level of reliability. Correct predictions amounted to 77% and uncertain predictions amounted to 20% of cases. Using the heterogeneous sequential Wald-Genkin procedure, an algorithm was created to predict the effectiveness of therapy for patients with blastocystosis, which determined the values of the prognostic coefficients of each gradation of indicators and their general prognostic information content. The highest prognostic informativeness was found for clinical symptoms such as liver enlargement (J=3,38), right subcostal heaviness (J=2,45), tachycardia (J=1,76), degree of manifestation of clinical symptoms (J=1,74) and body temperature (J=1,56). Testing the algorithm at 95% reliability level on the training group (n=300) revealed its high efficiency, as the erroneous predictions amounted to 3% and did not exceed the specified (5%; p<0,05) reliability level. Correct predictions amounted to 77% and uncertain predictions amounted to 20% of cases.