Prediction of binary response using multivariate longitudinal profiles: Study on chronic hepatitis B patients
Abstract
Nowadays several treatment options are available for patients with chronic hepatitis B. However, the hepatitis B virus remains quite difficult to eliminate and only 2-36% of the treated patients are cured.
Peg-interferon (PEG-INF) has proven effective but also has its limitations regarding multiple and possible serious side-effects. Hence, it is important to predict as early as possible whether the patient will respond positively to the treatment.
During therapy the patient is monitored at frequently scheduled follow-up visits and several markers are measured to anticipate continuation of therapy. In this contribution, we developed and compared several prediction models based on the past history of the patient for the treatment response (cured/not cured).