Non-invasive prediction model for high-risk esophageal varices in the Chinese population.

BACKGROUND

There are two types of esophageal varices (EVs): high-risk EVs (HEVs) and low-risk EVs, and HEVs pose a greater threat to patient life than low-risk EVs. The diagnosis of EVs is mainly conducted by gastroscopy, which can cause discomfort to patients, or by non-invasive prediction models. A number of non-invasive models for predicting EVs have been reported; however, those that are based on the formula for calculation of liver and spleen volume in HEVs have not been reported.

AIM

To establish a non-invasive prediction model based on the formula for liver and spleen volume for predicting HEVs in patients with viral cirrhosis.

METHODS

Data from 86 EV patients with viral cirrhosis were collected. Actual liver and spleen volumes of the patients were determined by computed tomography, and their calculated liver and spleen volumes were calculated by standard formulas. Other imaging and biochemical data were determined. The impact of each parameter on HEVs was analyzed by univariate and multivariate analyses, the data from which were employed to establish a non-invasive prediction model. Then the established prediction model was compared with other previous prediction models. Finally, the discriminating ability, calibration ability, and clinical efficacy of the new model was verified in both the modeling group and the external validation group.

RESULTS

Data from univariate and multivariate analyses indicated that the liver-spleen volume ratio, spleen volume change rate, and aspartate aminotransferase were correlated with HEVs. These indexes were successfully used to establish the non-invasive prediction model. The comparison of the models showed that the established model could better predict HEVs compared with previous models. The discriminating ability, calibration ability, and clinical efficacy of the new model were affirmed.

CONCLUSION

The non-invasive prediction model for predicting HEVs in patients with viral cirrhosis was successfully established. The new model is reliable for predicting HEVs and has clinical applicability.