Mathematical modeling of hepatitis C virus transmission in hemodialysis

A deterministic mathematical model is developed to explain nontransfusion nosocomial transmission of hepatitis C virus (HCV) from patient to patient during hemodialysis sessions.

The model requires 4 sequential steps for cross-transmission: (1) The dialysis session contains at least 1 patient infected with HCV; (2) a hemodialysis staff member connects an uninfected patient to dialysis after having connected an infected patient; (3) the hemodialysis staff member does not change gloves between an infected patient and an uninfected patient; and (4) the uninfected patient is contaminated after exposure to the blood of an infected patient.

We tested the model by comparing observed incidences of HCV infection from epidemiologic studies with calculated incidences. Calculated incidences are closed to observed incidences. We assessed the impact of prevalence of HCV infection, no glove change between patients, and nurse:patient ratio on the incidence of HCV infection. We found linear relationships between incidence and prevalence and between incidence and no glove change, and an increasing logarithmic relationship between incidence and nurse:patient ratio.

Our model should be able to estimate the likely incidence of infection in hemodialysis centers. Compliance with recommended hand hygiene and glove use practices, especially glove changes between patients, is essential to prevent HCV infection in hemodialysis centers, particularly those with high HCV prevalence. Mathematical modeling can used as a tool for control.