Інформація призначена тільки для фахівців сфери охорони здоров'я, осіб,
які мають вищу або середню спеціальну медичну освіту.

Підтвердіть, що Ви є фахівцем у сфері охорони здоров'я.

Журнал «Медицина неотложных состояний» 2 (57) 2014

Вернуться к номеру

Predicting the risk of abdominal compartment syndrome in patients with surgical abdominal pathology

Авторы: Shano V.P., Gladkaya S.V., Gur''yanov V.G., Gumenuk I.V., Gordienko I.V. - «Institute of Emergency and Reconstructive Surgery named. V.K.Gusaka NAMS of Ukraine, Donetsk, Department of Anesthesiology and Intensive Therapy; Grin V.K. - Director Academician, Academy of Medical Sciences of Ukraine

Рубрики: Медицина неотложных состояний

Разделы: Клинические исследования

Версия для печати

Introduction. Abdominal compartment syndrome, as a component of multiple organ disorders observed not less than 36% of patients in intensive care units [7]. Thus mortality reaches 68 % [14]. Among the pathogenetic mechanisms of abdominal compartment syndrome is considered the most important: the development of microcirculatory failure due to a decrease in perfusion pressure [10, 14], endothelial dysfunction, an imbalance between endothelium derived relaxing factor and vasoconstrictors [6, 11]. This largely determines the severity of intra-abdominal hypertension development in operations on the abdominal organs [1, 3, 11]. Unresolved issues are: identifying severity and risk factors for abdominal compartment syndrome [4, 9, 14]. The multiple-factor mechanism of development of ACS demands not only timely identification and elimination, but also the prevention of influence of endogenous mechanisms of formation of ACS [5, 8, 13].

Objective. Identify the role of risk factors and their impact on the development of ACS for algorithm development of its diagnostics and prevention.

Material and methods. To identify risk factors for the development of ACS, orientation and evaluation of their influence were used methods for design and analysis of multivariate mathematical models based on the analysis of factorial data 30, 155 patients operated for abdominal organs - 92, of the aorta and great vessels - 63 .

 Of the 155 patients, 57 (36.7 %) developed ACS that was designated as result variable Y = 1 in 98 (63.3 %) cases had not developed ACS that was designated as result variable Y = 0 .

All patients were divided into groups depending on the nature of surgical treatment: the abdominal cavity and the main vessels. In each group allocated subgroups depending on the type of treatment: a — the standard treatment, b — optimized intensive therapy, including I stage — diagnosis severity ACS: I, II , III , IV on the basis of indicators IAP, PP, the nature of the evacuation — motor function GIT indicators SAPS; II stage — to define how intensive therapy — conservative when I and II severity — nasogastric intubation gastroenterosorption, gastrointestinal lavage, drug stimulation of the gastrointestinal tract, postoperative analgesia, differentiated infusion therapy in accordance with the amount of content of a nasogastric tube. Correction of metabolic parameters, including delivery and consumption of oxygen, nitric oxide, von Willebrand factor, endothelin-1, blood sugar. Studied the performance in terms of before and after treatment.

Given the data, the algorithm of diagnosis and prevention of ACS that takes into account the studied syndrome pathognomonic for leading features that determine the severity of the ACS.

Results and discussion. To check the quality prediction model, all observations (using a random number generator) were divided into three sets: a training (used to calculate the parameters of the model, 115 cases), the reference set (used retraining control model, 10 cases) and confirmation (used for verify the adequacy of the model in the evaluation of new cases, 30 cases).

At the first stage was built prediction model based on all 30 signs. Sensitivity of the model to the training set was 90.2% (95% CI 79.0–97.5 %), specificity — 89.2% (95% CI 81.0–95.3%). On confirming the set sensitivity of the model was 54.5 % (95 % CI, 22.6–84.6 %), specificity — 84.2 % (95 % CI 63.497.2 %).

To identify the factors that are most associated with the risk of ACS was conducted to select the most significant features using the GA. The analysis was selected 6 factor variables age (X1), violation of the motor-evacuation function of the digestive tract: stomach contents of 500 ml or more, hiccups, lack of bowel sounds, abdominal distension, lack of bowel movement and flatus (X2), endothelin-1 (X3), WBD (X4), VO2 (X5), optimized IT (X6). On a dedicated set of factor variables was constructed linear neural network model predicting the risk of developing ACS. The model expressed by the equation (1):

Y = 0,0056´X1+0,25´X2+0,082´X3+0,15´X4–0,0065´X5–0,186´X6–0,060 (1).

Using the method of analysis of ROC-curves (9), the optimal threshold was chosen acceptance/rejection, Ycrit = 0,486, in the case of Y<Ycrit test predicts a positive outcome, with Y³Ycrit forecast to develop ACS. Sensitivity of this model in the training set was 85.4% (95% CI 72.694.6 %), specificity  83.8% (95% CI 74.491.3%), in the confirmation set sensitivity of the model  72.7 % (95 % CI 40.195.5%), specificity  94.7 % (95 % CI 79.3 % - 100 %).

The sensitivity and specificity on the training sets and confirmation were not significantly different (p = 0,59 and p = 0,39, respectively, when compared by c2), indicating the adequacy of the constructed model.

To assess the adequacy of the model and the choice of the significant factors predicting the risk of ACS was used method of comparing ROC-curves constructed models.

In the analysis, found that the area under the ROC-curve model based on all 30 signs factor was AUC1=0,91±0,03, for a model that is built on a 6-factor and dedicated signs — AUC= 0,90 ± 0,03, revealed no statistically significant difference (p = 0,86).

Found that sufficient for the development of prediction ACS is the significance of the 6 factors and dedicated features which include: age (X1), violation of motor-evacuation function of the gastrointestinal tract (X2), endothelin-1 (X3), IAP (X4), VO2 (X5), optimized IT (X6).

To identify the strength and direction of the impact factor of six selected features was built logistic regression model, the model is adequate (χ2 = 92,7 with six degrees of freedom, p < 0,001).

From the analysis of the coefficients of the logistic regression model shows that age is statistically significant (p = 0,006) increases the risk of ACS, OR = 1.06 (95% CI 1.01–1.08) for each year. Found that the risk of ACS significantly (p = 0,003) increased in the presence of violations of motor-evacuation function of the gastrointestinal tract, OR = 6.0 (95% CI 1.8–19.5). Also revealed an increase (p = 0,013) risk of ACS by increasing IAP, OR = 6.5 (95% CI 1.5–28.6). From the analysis of the logistic model also implies that the standardization of all significant risk factors for the application of the proposed method reduces the intensive care unit (p = 0,001) risk of ACS, OR = 0.2 (95% CI 0.1–0.5 ) compared with the control group .

The data obtained allow to determine the role of these factors in the development of aggression operating abdominal compartment syndrome: abdominal pressure greater than 10 mm. Hg and decreased perfusion pressure less than 60 mm.Hg reflected the presence of intra-abdominal hypertension in violation of the microcirculation in the abdominal cavity; reduction of nitrogen oxide less than 4 mmol/L in combination with increased levels of endothelin-1 by more than 3 times and increased the percentage of von Willebrand factor of over 100% indicating the presence of endothelial dysfunction, lack of relaxation factors in the microcirculation system; inappropriate oxygen transport indicated a decrease in oxygen delivery 700ml/min.m2 less and less reduction in oxygen consumption 140ml/min.m2, reflecting the presence of a combination — the circulatory and tissue hypoxia.

Conclusions. 1. To evaluate the effectiveness of the proposed method optimized intensive therapy built a multifactorial mathematical model predicting treatment ACS. Allocated 6 factorial signs associated with the risk of ACS (age, impaired motor-evacuation function of the gastrointestinal tract, endothelin-1, IAP, VO2, optimized IT). Sensitivity of the test is 85.4% (95% CI 72.6–94.6 %), specificity — 83.8% (95% CI 74.4–91.3%).

2. Found that the risk of developing age- game statistically significant (p = 0,006) increased, OR = 1.06 (95% CI 1.01–1.08) for each year. Risk of ACS significantly (p = 0,003) increased in the presence of violations of motor-evacuation function of the gastrointestinal tract, OR = 6.0 (95% CI 1.8–19.5). Also revealed an increase (p = 0,013) risk of ACS with increasing IAP, OR = 6.5 (95% CI 1.5–28.6) .

3. Revealed that the standardization of all significant risk factors for the application of the proposed method reduces the intensive care unit (p = 0,001) risk of ACS, OR = 0.2 (95% CI 0.1–0.5) compared with the control group.



Вернуться к номеру