Infection Control and Epidemic Respiratory Virus | CDC EID

EID Journal Home > Volume 15, Number 12–December 2009

Volume 15, Number 12–December 2009
Research
Cost-effectiveness Analysis of Hospital Infection Control Response to an Epidemic Respiratory Virus Threat
Yock Young Dan, Paul A. Tambyah, Joe Sim, Jeremy Lim, Li Yang Hsu, Wai Leng Chow, Dale A. Fisher, Yue Sie Wong, and Khek Yu Ho
Author affiliations: National University Health System, Singapore (Y.Y. Dan, P.A. Tambyah, J. Sim, L.Y. Hsu, D.A. Fisher, K.Y. Ho); and Singapore General Hospital, Singapore (J. Lim, W.L. Chow, Y.S. Wong)

Suggested citation for this article

Abstract
The outbreak of influenza A pandemic (H1N1) 2009 prompted many countries in Asia, previously strongly affected by severe acute respiratory syndrome (SARS), to respond with stringent measures, particularly in preventing outbreaks in hospitals. We studied actual direct costs and cost-effectiveness of different response measures from a hospital perspective in tertiary hospitals in Singapore by simulating outbreaks of SARS, pandemic (H1N1) 2009, and 1918 Spanish influenza. Protection measures targeting only infected patients yielded lowest incremental cost/death averted of $23,000 (US$) for pandemic (H1N1) 2009. Enforced protection in high-risk areas (Yellow Alert) and full protection throughout the hospital (Orange Alert) averted deaths but came at an incremental cost of up to $2.5 million/death averted. SARS and Spanish influenza favored more stringent measures. High case-fatality rates, virulence, and high proportion of atypical manifestations impacted cost-effectiveness the most. A calibrated approach in accordance with viral characteristics and community risks may help refine responses to future epidemics.

Pandemic influenza A (H1N1) 2009 virus is a new influenza virus of swine origin that was first detected in April 2009. Within 4 months of its appearance in Mexico, it had spread to >100 countries, with >200,000 confirmed cases globally, including >2,000 deaths (1). When the World Health Organization (WHO) raised its global influenza pandemic alert to phase 5 (imminent pandemic) on April 27, 2009, many countries followed suit and activated their pandemic preparedness plans, although this varied between countries. Many countries with direct experience of the 2003 severe acute respiratory syndrome (SARS) outbreak tended toward more stringent measures.

Singapore was one of the countries most affected by SARS and experienced a disproportionate impact of the spread of the disease in hospitals (2,3). A total of 98 healthcare workers in Singapore were infected with SARS, 6 of whom died (4). After the SARS experience, Singapore's Ministry of Health (MOH) developed a pandemic influenza plan with several levels of response that correlated roughly with the WHO Pandemic Alert Response system (5). The Disease Outbreak Response System (DORSCON)-FLU system that MOH devised requires progressively higher levels of infection control in hospitals in addition to border screening, restrictions on visitors to hospitals, and community-based syndromic surveillance for acute febrile illnesses (Table 1).

In accordance with the progressive elevation of WHO pandemic alert levels, Singapore raised its own pandemic alert level to Yellow on April 27, 2009, and further elevated it to Orange 2 days later. At this level, all hospital staff were required to wear N95 masks when dealing with all patients. Patients were restricted to 1 registered and screened visitor, all medical and nursing student rotations and local medical conferences were cancelled, leave restrictions for healthcare workers (HCWs) were put in place, interhospital movement of patients and HCWs was banned, and further limitations were placed on elective surgery. These measures were aimed primarily at avoiding a repeat of the SARS epidemic where nosocomial transmission originated with patients whose SARS infections were undiagnosed in hospital, and because influenza may be contagious before symptoms develop in infected patients. In fact, nosocomial influenza has been well documented since the 1957 Asian influenza pandemic (6). Based on studies conducted primarily in the United States, it has been estimated that 1 nosocomial case of influenza in a pediatric unit can cost up to $7,500 (US) (7). A recent review (8) of 28 nosocomial outbreaks of seasonal influenza summarized the evidence for nosocomial transmission of influenza in hospitals with accompanying illness and death (8,9).

When it subsequently became apparent that the case-fatality rate for pandemic (H1N1) 2009 was much lower than previously thought, especially in settings of industrialized countries, the alert level in Singapore was lowered to Yellow on May 11, 2009, even as WHO moved to alert level 6 after the pandemic was declared.

The risks and impacts of an outbreak will no doubt depend on the transmissibility, virulence, and clinical severity of illness. Thus, the benefits of a high alert status response at the onset of an outbreak as a "safe rather than sorry" strategy is not unreasonable when faced with an unknown novel potentially lethal virus. Yet, on the other hand, preventive measures from a hospital perspective come with a price. Direct costs include activation as well as ongoing administrative, manpower, and logistic resources, such as use of enhanced personal protective equipment, as part of the alert response measures.

We made use of this unique opportunity to evaluate the real costs of our primary prevention interventions and their potential cost-effectiveness against different models of influenza virulence and transmissibility in a simulated outbreak in our 1,000-bed tertiary teaching hospital to understand the relative incremental cost per additional death averted at different alert status levels. The key variables that affected the cost-effectiveness ratio the most were identified and studied. The same analysis was subsequently repeated for a parallel 1,500-bed tertiary teaching hospital. Using the outcome variables of disease cases, deaths, and incremental cost per death averted, we sought to determine if a calibrated and measured response plan based on characteristics of the virus in the outbreak could be better defined.

abrir aquí para acceder al documento CXDC EID completo del cual se reproduce un 5%:
Infection Control and Epidemic Respiratory Virus | CDC EID