Abstract
OBJECTIVE: This study aims to explore the association between serum creatinine (Scr), uric acid (UA), D-dimer (D-D) levels and disease severity in patients with hypertensive disorders of pregnancy (HDP), and to establish a nomogram model for risk stratification and prediction of adverse pregnancy outcomes (APO). METHODS: We retrospectively analyzed 230 HDP patients, categorizing them into two groups based on the occurrence of APO: the APO group (n=75) and the non-APO group (n=155). The predictive value of biomarkers for disease severity was evaluated, and a composite risk model incorporating both severity and biomarkers was constructed to assess APO risk. Then we checked different clinical indicators between the two groups to find which ones might be linked to APO. We used this information to make a nomogram model that showed the risk of APO in HDP patients. We tested how well the model worked. RESULTS: Out of 230 HDP patients, 75 had bad pregnancy results (32.61%). The univariate logistic regression analysis showed several factors linked to APO: maternal age, disease severity, D-D levels, Scr, and UA (all P<0.05). Further multivariate logistic regression identified four independent risk factors: disease severity, D-D, Scr, and UA (all P<0.05). Using these, we built a nomogram model. The model exhibited good calibration and goodness-of-fit (P=0.230). A receiver operating characteristic analysis showed the model worked well, with an area under the curve of 0.888 (95% confidence interval: 0.844-0.932). The model had a sensitivity of 89.0%, a specificity of 74.7%, and an overall accuracy of 84.35%. A decision curve analysis showed that the model was helpful for doctors in making clinical decisions. CONCLUSION: We developed a nomogram model to help predict APO in women with HDP. The results showed that the model performed well in terms of accuracy and consistency.