Background <p><i>Acinetobacter baumannii</i> (<i>A. baumannii</i>) is one of the main pathogens that causesbloodstream infection (BSI) in elderly patients, with high morbidity and mortality rates once infected; new inflammatory indicators, such as the neutrophil–lymphocyte ratio (NLR) and platelet–lymphocyte ratio(PLR), have been proposed in recent years, and the prognostic effects of these new inflammatory indicators have not yet been adequately investigated in <i>A.baumannii</i> BSI in elderly patients. Therefore, we verified the effects of these inflammatory indicators on <i>A.baumannii</i> BSIprognosis in elderly patients by constructing a nomogram model.</p> Methods <p>The clinical data of 126 elderly patients with <i>A. baumannii</i> BSIwere retrospectively analyzed, and they were divided into a survival group (87 patients) and a death group (39 patients) according to survival status 28 days after infection. Variables were screened by univariate Cox regression analysis and LASSO regression, respectively, and different prognostic models were constructed, and the final models were screened by cross-validation and other means, and the performance of the final models, such as differentiation, was evaluated. Finally, 47 exceptions of data were used to validate the prognostic model of <i>A.baumannii</i>BSI in elderly patients.</p> Results <p>Out of 126 patients, 39 died, for a mortality rate of 31.0%. A high neutrophil–plateletratio(NPR)(hazard ratio [HR] of 35.948,95% confidence interval [CI], 6.890–187.548) and a high C-reactive protein (CRP)-to-lymphocyte ratio (CLR) (HR,1.004;95% CI, 1.002–1.006) are independent risk factors for death in elderly patients with <i>A. baumannii</i> BSI. The model constructed by LASSO regression screening variables avoided the overfitting situation of the model and performed better overall and was considered as the final model.In the final model, the nomogram model predicted the highest discriminatory 7-day prognosis of <i>A. baumannii</i>BSI in elderly patients, with an area under the working curve (AUC) of 0.821 for subjects, 0.777 and 0.783 for 14 and 28 days, respectively, and a standardized model with good agreement; The clinical decision curve revealed that the model provided good net benefit, ranging from 20 to 100%.</p> Conclusion <p>The NPR and CLR are closely associated with the prognosis of <i>A. baumannii</i>BSI in elderly patients, and in clinical practice, a focus should be placed on these new indicators of inflammation, especially the NPR and CLR, to help physicians better assess the prognosis of <i>A. baumannii</i>BSI in elderly patients and to develop a more effective therapeutic regimen to improve the survival rate of patients.</p>