Nomogram for assessment of risk factors for early mortality after laparotomy in neonatal gastric perforation from a tertiary care unit

Abstract

Purpose Summarize the clinical features of neonatal gastric perforation and establish a nomogram model to predict the risk factors of early mortality after laparotomy in neonatal gastric perforation (NGP) from a tertiary care unit. Methods Retrospective analysis was performed on NGP diagnosed in our hospital between May 2003 and October 2021. All patients underwent laparotomy, and according to the prognosis, they were divided into non-survival and survival groups. All clinical characteristics, preoperative laboratory features, intraoperative situation and outcomes were collected from electronic medical records. We conducted logistic regression analyses to identify the independent factors that contribute to early neonatal death after laparotomy and the nomogram prediction model was constructed. Results A total of 111 patients with NGP were included in our study, the mortality was 23.42% (26/111). Six independent mortality risk factors were identified: APTT(OR,1.015;95% CI,1.001,1.138;p = 0.039), PaO2 (OR,0.977;95%CI,0.957,0.997;p = 0.022), Hco3−(OR,0.777;95%CI,0.616,0.979;p = 0.032),PLT(OR,0.989;95%CI,0.979,0.998;p = 0.022), SPC(OR,2.693;95%CI,1.221,5.942;p = 0.014), Combined with NEC (OR,0.040; 95% CI,0.004,0.421 ; p = 0.007). A nomogram model was constructed based independent prognostic risk factors, and its AUC under ROC curve was 0.886(95%CI,0.806,0.965;P = 0.000), which had a good degree of discrimination. Patients who had a nomogram score of more than 178.548 was considered to take high risks of mortality, and the sensitivity and specificity of identifying postoperative NGP mortality were 91.8% and 76.9%. The correction curves evaluation results showed a high consistency of the model. Conclusion APTT, PaO2, Hco3−,PLT,SPC and combined with NEC change, are independently associated with an increased risk of early mortality in neonates with neonatal gastric perforation after laparotomy. The nomogram model constructed in this study can be used as a tool to predict the risk of mortality, predict postoperative survival, and help to develop an individualized treatment plan.