Performance of a Predictive Model for Long-Term Hemoglobin Response to Darbepoetin and Iron Administration in a Large Cohort of Hemodialysis Patients

Abstract

Anemia management, based on erythropoiesis stimulating agents (ESA) and iron supplementation, has become an increasingly challenging problem in hemodialysis patients. Maintaining hemodialysis patients within narrow hemoglobin targets, preventing cycling outside target, and reducing ESA dosing to prevent adverse outcomes requires considerable attention from caregivers. Anticipation of the long-term response (i.e. at 3 months) to the ESA/iron therapy would be of fundamental importance for planning a successful treatment strategy. To this end, we developed a predictive model designed to support decision-making regarding anemia management in hemodialysis (HD) patients treated in center. An Artificial Neural Network (ANN) algorithm for predicting hemoglobin concentrations three months into the future was developed and evaluated in a retrospective study on a sample population of 1558 HD patients treated with intravenous (IV) darbepoetin alfa, and IV iron (sucrose or gluconate). Model inputs were the last 90 days of patients' medical history and the subsequent 90 days of darbepoetin/iron prescription. Our model was able to predict individual variation of hemoglobin concentration 3 months in the future with a Mean Absolute Error (MAE) of 0.75 g/dL. Error analysis showed a narrow Gaussian distribution centered in 0 g/dL; a root cause analysis identified intercurrent and/or unpredictable events associated with hospitalization, blood transfusion, and laboratory error or misreported hemoglobin values as the main reasons for large discrepancy between predicted versus observed hemoglobin values. Our ANN predictive model offers a simple and reliable tool applicable in daily clinical practice for predicting the long-term response to ESA/iron therapy of HD patients.

Conflict of interest statement

Competing Interests: Carlo Barbieri, Elena Bolzoni, Flavio Mari, Isabella Cattinelli, Francesco Bellocchio, Claudia Amato, Andrea Stopper, Stefano Stuard and Bernard Canaud are employees of Fresenius Medical Care and may hold stock in the company. Emanuele Gatti may hold stock in Fresenius Medical Care. José Martin and Iain C MacDougall have no competing interests. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.

Figures

Fig 1. Patients’ inclusion criteria.

Fig 2. Predictive anemia modeling based on ANN.

At time t the model predicts the Hb variation between time t and time t+3 months using the patient past history and the subsequent 3 months of darbepoetin and iron prescription.

Fig 3. Steps involved in the development and validation of the Artificial Neural Network anemia modeling.

Fig 4

Histogram of the Hb error distribution in the training phase (left panel). Histogram of the Hb error distribution in the test phase (right panel).

Fig 5. Bland-Altman analysis of observed/predicted Hb values.

Fig 6. Predicted vs. Actual Hb variations for a patient characterized by a prediction error close to the mean absolute error on test set.

At each time step (corresponding to monthly lab tests, on the x-axis), the Hb variation predicted by the model over the next three months (i.e., Hb(t+3)–Hb(t), on the y-axis) is plotted in solid line and compared with the actual variation observed over the same time interval (plotted in dashed line). Time steps are counted starting from the first month for which an Hb prediction for the patient was possible.

Fig 7. Predicted vs. Actual Hb variations for a patient characterized by low prediction error.

Fig 8. Predicted vs. Actual Hb variations for a patient characterized by high prediction error.