To investigate the value of platelet lymphocyte ratio (PLR) in predicting the prognosis of patients with peritoneal dialysis associated peritonitis (PDAP), and to establish a nomogram model.

Total of 129 hospitalized patients with the PDAP in the Department of Nephrology, Huizhou Central People’s Hospital from January 2019 to June 2021 were included as modeling group, the patients were divided into cured group (83 cases) and poor prognosis group (46 cases) according to the final treatment effect. There were 113 patients with PDAP admitted to Huizhou the First People’s Hospital in the same period served as the validation group. The differences in clinical data, laboratory indicators and pathogenic bacteria composition between the two groups were compared, respectively. The independent predictors of poor prognosis patients with PDAP were obtained by multivariate Logistic regression model. According to the independent predictors, the nomogram model for predicting the risk of poor prognosis of patients with PDAP was established. The Bootstrap method and calibration curve were used for the internal and external verification of the nomogram model. The decision curve was drawn, and the independent predictors and the combined prediction model were analyzed to predict the net return of poor prognosis of patients with PDAP.

The dialysis period of patients [11.23 (8.44, 14.57) months vs. 7.23 (5.31, 10.41) months] in poor prognosis group was significantly higher than that of the cure group (Z = 5.735, P < 0.001), and the proportion of diabetes nephropathy of patients in poor prognosis [69.57% vs. 30.12%] was significantly higher than that of the cured group (χ² = 6.165, P = 0.007), the platelet count (t = 5.687, P < 0.001), procalcitonin (Z = 6.945, P < 0.001), PLR (t = 7.267, P < 0.001) and C-reactive protein (CRP) (t = 8.221, P < 0.001) of patients in poor prognosis group were significantly higher than those of the cured group, and the lymphocyte count of patients [1.01 (0.78, 1.15) × 10⁹/L vs. 1.42 (1.11, 1.67) × 10⁹/L] was significantly lower than that of patients in the cured group (Z = 4.467, P < 0.001). The results of multivariate Logistic regression analysis showed as that the primary disease (χ² = 7.564, P = 0.014), PLR (χ² = 9.786, P = 0.005) and dialysis period (χ² = 8.967, P = 0.009) were all independent predictors of poor prognosis in patients with PDAP, the risk of poor prognosis increased 1.568 times with 1 unit increase in PLR (OR = 2.568, 95%CI: 2.117-2.926, P = 0.005); the risk of poor prognosis of diabetes nephropathy patients was 3.265 times that of chronic nephritis patients (OR = 3.265, 95%CI: 2.196-6.457, P = 0.014); the risk of poor treatment prognosis in patients increases by 1.733 times for every one month increase in dialysis period (OR = 2.733, 95%CI: 2.245-3.214, P = 0.009). A nomogram model was established to predict the risk of poor prognosis of patients with PDAP based on the three independent predictors obtained from the results of multivariate analysis. The H-L test results showed that the predicted value of the risk of poor prognosis of patients with PDAP in the modeling group and the validation group was in good agreement with the actual observation value (χ² = 0.134, P = 0.867; χ² = 0.214, P = 0.785). The analysis results of decision curve showed that within 0-0.79 threshold probability ranges, the three independent predictors had good net benefits for predicting the risk of poor prognosis for patients with PDAP, and the overall net benefits of joint prediction were higher than those of single index.

The nomogram model combined with PLR has high clinical value in predicting the poor prognosis of patients with PDAP.