Nomogram for predicting pulmonary embolism in gynecologic inpatients with isolated distal deep venous thrombosis

OBJECTIVE

To investigate the incidence of isolated distal deep venous thrombosis (IDDVT) concurrent with pulmonary embolism (PE) in gynecologic inpatients, analyze the risk factors for IDDVT with PE, and establish a nomogram model for IDDVT patients with PE.

METHODS

A total of 260 patients were diagnosed with IDDVT between December 2017 and November 2020. The incidence of PE in these patients was determined using computed tomography pulmonary angiography. Logistic regression analysis was used to identify the related risk factors. On this basis, nomogram risk prediction models were established.

RESULTS

Among 260 patients with IDDVT, 106 (40.8%) had concurrent PE, of whom 74 (28.5%) experienced silent PE. Univariate logistic analysis demonstrated statistical significance for body mass index (BMI; P = 0.044), glucocorticoid therapy (P = 0.009), hypertension (P < 0.001), and diabetes (P < 0.001). Multivariate logistic analysis revealed that these were independent risk factors for IDDVT with PE that retained statistical significance. A nomogram based on these factors was constructed to predict PE in patients with IDDVT. Its receiver operating characteristic (ROC) showed an area under the curve of 0.710 (95% confidence interval 0.642-0.779), with prediction sensitivity of 64.2% and prediction specificity of 76.6%.

CONCLUSIONS

In the present study, a high prevalence of PE was found in gynecologic inpatients with IDDVT. Glucocorticoid therapy, hypertension, diabetes, and BMI were independent risk factors for IDDVT patients with PE. Taking these risk factors into account, a nomogram risk prediction model was developed to help facilitate early detection of concurrent PE.