Treatments for infectious diseases are unique in that they produce externalities, spillover effects that are felt by individuals who are currently not infected. While these externalities are recognized, their magnitude is not well measured. Similarly, the impact of nosocomial infections on health care cost and utilization has not been quantified accurately. These two elements (externalities and costs) are important inputs to an economic analysis of any infectious disease treatment strategy.
The goal of this research is to apply cutting edge empirical methods of estimating the costs and benefits of prevention efforts for infectious disease transmission to the specific case of decolonization of MRSA within the VA. This research has 3 aims: (1) quantify the positive externalities associated with decolonization of MRSA positive veterans, (2) estimate the cost and healthcare use associated with hospital-acquired MRSA infections in the VA, and (3) perform economic analyses of a MRSA strategy that adds decolonization to the VA's current VA strategy of contact isolation.
In Aim (1), we will use a nationwide database of VA patient data obtained from Patient Care Services (PCS) to estimate the association between hospital-level decolonization use and resistance to this decolonization and the risk of hospital-acquired MRSA infection using multi-level logistic regressions. Next we will link this PCS data to VA Decision Support System (DSS) data to estimate the impact of MRSA infection on inpatient length of stay using inverse probability weights in a longitudinal logistic regression and on post-discharge health care costs and utilization using generalized linear models that are appropriate for skewed data, in Aim (2). In addition, we will link this PCS and DSS data to Medicare and VA Fee Basis data to capture post-discharge health care costs from outside the VA. Finally, in Aim (3), we will use the results for Aims (1) and (2) as inputs to an agent-based model of MRSA transmission to conduct dynamic economic analyses.
Aim 1: We found that MRSA colonization is strongly associated with subsequent MRSA infection and a large proportion of infection occur in post-discharge period. A paper based on this analysis is currently under review at Clinical Infectious Diseases
Aim 2: Excess cost and mortality of HAIs can be substantial both in the pre- and post-discharge period. However, these estimates are biased upward when the time-dependent nature of HAIs is not taken into account. Papers based on these analyses have been published in American Journal of Infection Control, Infection Control and Hospital Epidemiology, Medical Care, and others are currently under review at Antimicrobial Agents and Chemotherapy and Health Services Research.
Aim 3: We found that interventions to prevent HAIs are cost-effective at standard willingness-to-pay thresholds. Papers based on these analyses have been published in PLOS One, Infection Control and Hospital Epidemiology, and American Journal of Preventive Medicine.
By examining the associations between facility-level decolonization usage and MRSA infection rates, as well as the impact of MRSA infections on length of inpatient stay and post-discharge costs, we will provide valuable insight into the specific question of whether to add decolonization to the VA-wide initiative aimed at controlling Methicillin-resistant Staphylococcus aureus (MRSA) infections within VA hospitals. In addition, estimates of the cost of hospital-acquired MRSA infections can be used to determine the budget impact to the VA of the successful reduction in such infections resulting from the national VA MRSA Prevention Initiative.
Through this research and training experience, I will be prepared to conduct similar research in other healthcare-acquired and community-acquired infections that could impact infection control practice throughout the VA healthcare system.
Many of the methods that I am developing while conducting these analyses can be applied to non-infectious disease-related healthcare-associated events as well, such as venous thromboembolism. In addition, I am applying these methods to estimate the causal effect of a time-varying exposure on patient outcomes within the context of homelessness programs in the VA. This exercise is part of a merit review proposal that was recently funded by HSR&D.
The methods utilized in these analyses can be extended to identify variation in the effect of healthcare-associated infections by facility and by type of infection. In addition, variation in the process measures of the VA MRSA Prevention Initiative can be measured and combined with these other estimates in order to decompose the effect of the initiative. A better understanding of the MRSA Prevention Initiative can inform the ongoing efforts to reduce healthcare-associated infections due to other organisms in the VA such as Clostridium difficile and Gram-negative bacteria.
External Links for this Project
- Nelson RE, Deka R, Khader K, Stevens VW, Schweizer ML, Rubin MA. Dynamic transmission models for economic analysis applied to health care-associated infections: A review of the literature. American journal of infection control. 2017 Dec 1; 45(12):1382-1387. [view]