An Automated Telephone Monitoring System to Identify Patients with Cirrhosis at Risk of Re-hospitalization

Mary Thomson, Michael Volk, Hyungjin Myra Kim, John D. Piette

Hospitalizations for cirrhosis are costly and associated with increased mortality. Disease management outside of clinic, such as the use of interactive voice response (IVR) calls, may identify signs to prevent hospitalization. The aim of this study was to investigate whether IVR monitoring can predict hospitalization and mortality in cirrhosis. One hundred patients with decompensated cirrhosis were enrolled in this observational study, of which 79 patients were included in the final analysis. Participants were followed until death, transplant, or last clinical follow-up (range 7-874 days). Analysis focused on potential predictors identified during the first month of IVR calls: presence of jaundice, abdominal/leg swelling, weakness, paracentesis requirement, medication changes, and weight change. The primary outcome was time to first hospital admission; secondary outcomes included hospitalization and time to death. Potential predictors with a p value <0.1 were further analyzed after adjustment for covariates (Model for End-stage Liver Disease score, serum sodium, number of medications). Twenty (25 %) patients died, and 49 (62 %) were hospitalized at least once. Fifty-six (70 %) patients completed >80 % of their IVR calls. After adjustment for covariates, weakness was associated with an increased risk of first hospitalization (HR 2.14, CI 1.13-4.05, p = 0.02) and hospitalization rate (HR 2.1, CI 1.0-4.3, p = 0.048). Weight change of ≥five pounds (2.3 kg) in a week increased the rate of hospitalization by 2.7 (CI 1.0-7.1, p = 0.045). No variable predicted death after covariate adjustment. These results suggest IVR calls can be used to predict hospitalization in cirrhosis.