RU2531346C1 - Method for prediction of bronchopulmonary pathology in ventilated infants in neonatal period - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: clinical findings are assessed, and laboratory diagnostics is performed. An index ŷ of probability of the bronchopulmonary pathology is additionally calculated by formula: $$\widehat{y}$$

=exp(-5.32+0.77x₁+0.95x₂+1.07x₃+1.55x₄+2.56x₅+0.75x₆+0.6x₇/(1+exp((-5.32+0.77x₁+0.95x₂+1.07x₃+1.55x₄+2.56x₅+0.75x₆+0.6x₇)), wherein $$\widehat{y}$$

is the probability of the bronchopulmonary pathology, x₁ is a gestational age: mature - 0, pre-mature - 1; x₂ is the length of oxygen therapy more than 336 hours (14 days) - 1, less - 0; x₃ is the length of stringent parameters of artificial pulmonary ventilation more than 144 hours (6 days) - 1, less - 0; x₄ is the length of stringent parameters of artificial pulmonary ventilation at peak inspiratory pressure more than 72 hours (3 days) - 1, less - 0; x₅ is the presence of the burdened hereditary bronchopulmonary system: no - 0, yes - 1; x₆ is a degree of connective tissue dysplasia: mild - 0, moderate and severe - 1; x₇ is the presence of complications (pneumothorax, atelectasis, pneumonia): no - 0, yes - 1, and if the index of the probability of the bronchopulmonary pathology $$\widehat{y}$$

>0.5 the bronchopulmonary pathology is predicted.

EFFECT: method enables the high-reliability prediction of the bronchopulmonary pathology in the ventilated children in the neonatal period.

1 tbl

Description

The invention relates to the field of medicine (pediatrics, pulmonology, neonatology) and can be used to predict the formation of bronchopulmonary pathology in children who were on mechanical ventilation (IVL) in the neonatal period.

A known method for predicting bronchopulmonary diseases at an early age in children undergoing mechanical ventilation in the neonatal period (Sokol EB Formation of bronchopulmonary pathology in children undergoing mechanical ventilation in the neonatal period. Author's abstract. Cand. Cand. - Voronezh. - 2011 .-- 21 p.). The method consists in the fact that on the basis of examination and study of the catanesis of 221 children, the most significant factors in the development of respiratory pathology in children who were on mechanical ventilation in the neonatal period were identified.

The disadvantage of this method is the low accuracy of prediction.

The purpose of the invention: to provide a high degree of accuracy in predicting the formation of bronchopulmonary pathology in children who were on mechanical ventilation (mechanical ventilation) in the neonatal period.

по формуле:The goal is achieved by determining the probability index of the development of bronchopulmonary pathology

according to the formula:

Where:

- вероятность положительного эффекта

- probability of a positive effect

x ₁ - gestational age: full-term - 0, premature - 1;

x ₂ - the duration of oxygen therapy more than 336 hours (14 days): no - 0, yes - 1;

x ₃ - the duration of the “hard” mechanical ventilation parameters is more than 144 hours (6 days): no - 0, yes - 1;

x ₄ - the duration of the "hard" parameters of mechanical ventilation according to the peak inspiratory pressure for more than 72 hours (3 days): no - 1, yes - 0;

x ₅ - the presence of a heredity burdened by diseases of the bronchopulmonary system of heredity: no - 0, yes - 1;

x ₆ - the degree of connective tissue dysplasia: light - 0, moderate and severe - 1;

прогнозируют формирование бронхолегочной патологии.x ₇ - the presence of complications (pneumothorax, atelectasis, pneumonia): no - 0, yes - 1, and with

predict the formation of bronchopulmonary pathology.

по формуле:The method is implemented as follows: determine the probability index of the development of bronchopulmonary pathology

according to the formula:

Where:

- вероятность положительного эффекта

- probability of a positive effect

x ₁ - gestational age: full-term - 0, premature - 1;

x ₂ - the duration of oxygen therapy more than 336 hours (14 days): no - 0, yes - 1;

x ₃ - the duration of the “hard” mechanical ventilation parameters is more than 144 hours (6 days): no - 0, yes - 1;

x ₄ - the duration of the "hard" parameters of mechanical ventilation according to the peak inspiratory pressure for more than 72 hours (3 days): no - 1, yes - 0;

x ₅ - the presence of a heredity burdened by diseases of the bronchopulmonary system of heredity: no - 0, yes - 1;

x ₆ - the degree of connective tissue dysplasia: light - 0, moderate and severe - 1;

прогнозируют формирование бронхолегочной патологии.x ₇ - the presence of complications (pneumothorax, atelectasis, pneumonia): no - 0, yes - 1, and with

predict the formation of bronchopulmonary pathology.

Examples of the prognostic implementation of the method (classification of children of the main group on the basis of no pathology of the bronchopulmonary system (Zd.) - with a disease of the bronchopulmonary system (B.) using a logistic regression model in comparison with that observed in the experiment) are shown in table 1:

Table 1

Baby initials Gestation period Duration of O2 therapy over 336 hours. For “Hard” ventilation parameters for more than 144 hours Pl. IVL with "hard" PIP parameters for more than 72 hours Burdened by heredity Degree of dysplasia The presence of complications The diagnosis in the long term Model Forecast one Vasya M. one 0 0 one 0 one one Rear 0.04 2 Sergey S. one 0 0 one 0 one 0 Rear 0.02 3 Dima R. one 0 0 one one 0 0 Rear 0.12 four Pasha M. one 0 0 one 0 one 0 Rear 0.02 5 Vika V. one 0 0 one 0 0 0 Rear 0.01 6 Dima S. one one 0 one 0 one one Rear 0.10 7 Alik M. one one 0 one one one 0 B. 0.78 8 Zhenya S. one 0 0 one one one one B. 0.71 9 Ivan G. one one one 0 one one one B. 0.95 10 Alena P. one one one 0 0 one one B 0.59 Notes: gestational age: 0 - 37-40 weeks, 1 - 30-36 weeks. Duration of O2-therapy is more than 336 hours, “hard” ventilation parameters are more than 144 hours, the presence of complications (atelectasis, pneumothorax, pneumonia), burdened heredity: yes-1, no-0. Duration of ventilation parameters “rigid” in PIP over 72 hours: yes-0, no-1. The degree of dysplasia: 0-mild, 1-moderate, severe. Diagnosis: Zd.-healthy, B.-sick.

This formula is the result of studying the follow-up history of 122 children who were on mechanical ventilation in the early neonatal period. Data on the hereditary history of the family, somatic and obstetric-gynecological history of mothers, obtained on the basis of medical documentation and interviews with parents. The assessment of the neonatal period was carried out retrospectively according to the data of the newborn’s case histories. The health status of children in the postnatal period after discharge from the hospital was evaluated on the basis of an interview with parents, data from outpatient development cards and other medical documents. When examining a child, an assessment was made of the features of physical development, the presence of small developmental abnormalities, signs of connective tissue dysplasia (classification by Abakumova L.N., 2006). Laboratory and instrumental examination was carried out by medical specialists on the basis of SPbGBUZ KDTSD. Then, using a logistic regression method, a mathematical model was built to predict the formation of bronchopulmonary pathology in children who were on mechanical ventilation in the neonatal period. Such a model is an express forecast, because built on the basis of a minimum sufficient number of known features. As a predicted response indicator, a diagnosis was defined at the age of 8 (11) years (favorable - a child without pathology of the bronchopulmonary system - 0 and adverse - a child with lung pathology - 1). As signs included in the model as independent factors-causes, a set of anamnesis data reliably associated with outcomes is determined. 11 features were included in the original training matrix. After a logical analysis and assessment of the relationships of the source data using the correlation analysis for further research in the training matrix, there are 7 signs that were strong (r> 0.7) or moderate (0.27 <r <0.7) and statistically significant (p <0.05) correlation with outcome. The solution to the problem of logistic regression analysis was implemented using the Logistic Regression procedure from Statistica statistical software packages. According to the results of calculations using the Logistic Regression module, 7 features with a statistical reliability of at least 70% are included in the model. The result was a statistically significant (p <0.001) and 76.2% classifiable mathematical model. In the subgroup of children without lung pathology, the studied model ensures the coincidence of the predicted result with the real one in 77.4% of cases (65 out of 84); in the subgroup of children with pathology of the bronchopulmonary system, a coincidence of predicted and real outcomes was noted in 73.6% of cases (in 28 of 38 cases); the informational ability of the model as a whole is 76.2% (the diagnosis of the child coincides with the predicted in 93 out of 122).

Claims (1)

A method for predicting the formation of bronchopulmonary pathology in children who were on mechanical ventilation in the neonatal period by studying clinical data and laboratory studies, characterized in that it determines the probability index of the development of bronchopulmonary pathology

according to the formula:

.
Where

- the likelihood of developing bronchopulmonary pathology,
x ₁ - gestational age: full-term - 0, premature - 1;
x ₂ - the duration of oxygen therapy more than 336 hours (14 days) - 1, less than 0;
x ₃ - the duration of the “hard” mechanical ventilation parameters is more than 144 hours (6 days) - 1, less - 0;
x ₄ - the duration of the "hard" mechanical ventilation parameters according to the peak inspiratory pressure for more than 72 hours (3 days) - 1, less - 0;
x ₅ - presence of heredity burdened by diseases of the bronchopulmonary system: no - 0, yes - 1;
x ₆ - the degree of connective tissue dysplasia: light - 0, moderate and severe - 1;
x ₇ - the presence of complications (pneumothorax, atelectasis, pneumonia): No - 0, Yes - 1,
and with the value of the probability index of the development of bronchopulmonary pathology

predict the formation of bronchopulmonary pathology.