Nonlinear Mixed Effects Models in the Prediction of S. Pneumoniae Antimicrobial Resistance Rates

• Activity Number: 87
• Type: Contributed
• Date/Time: Monday, August 9, 2004 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Robertino M. Mera*+
• Companies: GlaxoSmithKline
• Keywords: mixed models ; antimicrobial resistance ; epidemiology ; prediction models

Abstract

Streptococcus pneumoniae infections are the most significant bacterial cause of respiratory illness in young children, the elderly and persons with chronic medical conditions. High-level penicillin resistance rates in the U.S. have been increasing from levels of 5.6% in 1992 to 24.3% in 2001. Although compartment-transmission mathematical models have improved our understanding of resistance dynamics at the individual level, prediction of resistance in broad populations requires a different approach. The present study implements a nonlinear mixed effects model that is consistent with the insights provided by the transmission models. The model takes advantage of a large amount of surveillance data; antibiotic prescriptions and other established risk factors at the state and regional level in the U.S. in the period 1996 to 2001. The application shows that penicillin resistance rates will reach a plateau from 2006 to 2008. The two most important factors that affect prediction are antibiotic consumption and the uptake of the conjugate pneumococcal vaccine. Penicillin consumption changes the timing of the plateau, and 90% plus vaccine coverage immediately reduces the resistance rate.