Outbreak of Antiviral Drug–Resistant Influenza A in Long-Term Care Facility, Illinois, USA, 2008

An outbreak of oseltamivir-resistant influenza A (H1N1) occurred in a long-term care facility. Eight (47%) of 17 and 1 (6%) of 16 residents in 2 wards had oseltamivir-resistant influenza A virus (H1N1) infections. Initial outbreak response included treatment and prophylaxis with oseltamivir. The outbreak abated, likely because of infection control measures.


The Study
The LTCF in Illinois provides housing, healthcare services, and recreational activities for residents with neu-rologic and developmental medical conditions. During the outbreak, the LTCF housed 583 residents. Building A, the main site of the infl uenza outbreak, housed 108 residents in 6 wards; 104 (96%) received the 2007-08 infl uenza vaccine. Of the 685 LTCF employees involved in direct patient care, 385 (56%) received the 2007-08 infl uenza vaccine on site.
We defi ned a confi rmed case as a positive rapid or reverse transcription-PCR result for infl uenza virus from January 20 through February 8, 2008, in a resident of the LTCF. Surveillance for new case-patients included obtaining a nasopharyngeal specimen from all residents with new onset of fever or respiratory symptoms or any unusual behavior within 24 hours after illness onset. All specimens were tested by using the QuickVue A and B Infl uenza Test (Quidel, San Diego, CA, USA). A second specimen was obtained from all persons with positive rapid test results and some (57%) from persons with negative results for confi rmation of infl uenza virus infection and virus subtyping by reverse transcription-PCR. Medical records, vaccination records, resident activity, and visitor logs were reviewed.
Testing for antiviral drug resistance was conducted directly on clinical specimens by pyrosequencing as described (6,7), including identifi cation of the oseltamivir resistanceconferring H274Y mutation in the neuraminidase gene of infl uenza viruses (H1N1) (H275Y in N1 numbering) and the adamantane resistance-conferring mutations in the matrix 2 protein (7,8). The HA1 portion of the hemagglutinin (HA) gene of the outbreak viruses was sequenced and compared with those of epidemiologically relevant viruses.
Phylogenetic analysis of HA1 was performed by using MEGA version 4.0.1 software (9). A phylogenetic tree was inferred by using maximum composite likelihood available in MEGA version 4.0.1. The outbreak investigation was considered a public health response and granted exemption from review by the Institutional Review Board of the Centers for Disease Control and Prevention.
On January 27, the fi rst 3 residents with fever or respiratory symptoms in ward 1 within building A were positive for infl uenza A virus infection by rapid test (Figure 1). On January 28, outbreak infection control measures were initiated in all 6 wards, including surveillance for new cases, 5 days of treatment with oseltamivir for confi rmed cases, and 14 days of prophylaxis with oseltamivir for all healthy residents in wards with confi rmed case-patients (2). Confi rmed case-patients were quarantined in their rooms for 10 days; all residents in all 6 wards were quarantined for 10 days, and visitor movement was restricted. Staff and visitors were required to use personal protective equipment and practice respiratory and hand hygiene. Prescriptions for prophylactic courses of oseltamivir and infl uenza vaccinations were offered to all staff of building A; uptake was not recorded.
From January 28 through January 31, 2008, a total of 6 additional confi rmed case-patients were identifi ed. Eight (47%) of 17 residents in ward 1 and 1 (6%) of 16 residents in ward 2 were infected with infl uenza A viruses (H1N1) that contained the H274Y mutation but did not have markers of resistance to adamantanes or zanamivir.
On January 30, high fever developed in a male resident in ward 3 while on the fi rst day of a home visit ( Figure 1). He returned to building A on January 31, was positive for infl uenza by rapid test, and was placed in ward 2 in an attempt to group him with other already ill residents. Because of an ongoing outbreak in other nearby wards, oseltamivir prophylaxis was initiated for all residents in ward 3 who were not ill. On February 1, symptoms developed in 2 other residents in ward 3 who were positive for infl uenza by rapid test. Three (18%) confi rmed cases of infl uenza A virus (H3N2) resistant to adamantanes but sensitive to oseltamivir were detected among 17 residents in ward 3. Additional cases, but no clusters, were detected in other buildings 1-2 weeks later.
Characteristics of case-patients are shown in the Table. Establishing a fi rm epidemiologic link between cases, other than ward of residency, was not possible. Antiviral drug resistance results became available on February 7 when all case-patients had completed their treatment courses. Ongoing prophylaxis courses were changed: oseltamivir was replaced with rimantadine in ward 1, and rimantadine was added to oseltamivir in ward 2. Prophylaxis with oseltamivir alone was continued in ward 3. Zanamivir could not be used by most residents because of underlying conditions. Sequence analysis of the HA1 gene in outbreak infl uenza A viruses (H1N1) showed identical or nearly identical sequences, differing by only 1 or 2 nt (Figure 2). These viruses were phylogenetically more closely related to A/Brisbane/59/2007 (H1N1) than to the A/Solomon Islands/3/2006, the infl uenza A virus (H1N1) strain in the 2007-08 infl uenza vaccine. GenBank accession numbers of HA (HA1) sequences for the 9 oseltamivir-resistant infl uenza A viruses (H1N1) are FJ231752-FJ231760.

Conclusions
The attack rate of illness caused by oseltamivir-resistant infl uenza A viruses (H1N1) in ward 1 was within the range (20%-80%) reported for other facility infl uenza outbreaks (1,10,11), indicating effective person-to-person transmission of oseltamivir-resistant infl uenza A viruses (H1N1). Nosocomial transmission of oseltamivir-resistant infl uenza A viruses (H1N1), with possible healthcare worker involvement, has been described (12). We were un-  Figure 2. Phylogenetic analysis of the hemagglutinin gene (HA1 portion) of infl uenza A viruses (H1N1) isolated during an infl uenza A outbreak in a long-term care facility, Illinois, USA, 2008. Viruses from buildings A and B shared nearly identical sequences. One of the viruses from building B was more similar in sequence to 1 virus from building A. However, this fi nding could refl ect natural variance in circulating viruses. Red indicates outbreak viruses, boldface italics indicates vaccine strain for 2008-09, boldface indicates vaccine strain for 2007-08, and arrows indicate nucleotide differences in HA1 subunit. Scale bar indicates nucleotide substitutions per site.
able to assess staff illness in this investigation. Before the 2007-08 infl uenza season, transmission of neuraminidaseresistant infl uenza viruses had rarely been reported (13).
Although we documented a relatively high attack rate in 1 ward (ward 1), and despite resistance to the antiviral agent initially used, the outbreak abated quickly. High annual vaccination rates among residents and relatively high rates among employees (2) may have played a role in limiting the spread of the outbreak viruses. However, the A/Brisbane/59/2007 (H1N1)-like outbreak viruses were not optimally matched to the A/Solomon Islands/3/2006 (H1N1) vaccine strain (14). Also, infection control measures, such as isolation and quarantine, likely played a role in controlling this outbreak.
The proportion of circulating infl uenza viruses resistant to oseltamivir increased from 12% during the 2007-08 season to 99% during the 2008-09 season in the United States, and new interim guidelines for use of antiviral agents were released in December 2008 (15). These guidelines were updated for the 2009-10 season to account for the emergence of pandemic (H1N1) 2009 virus in September 2009 (www. cdc.gov/h1n1fl u/recommendations.htm). This outbreak underscores the possibility of 2 infl uenza A viruses, with different antiviral susceptibilities, in the same facility. During a facility outbreak of infl uenza, providers should consult antiviral recommendations of the Centers for Disease Control and Prevention and obtain infl uenza virus typing and subtyping to guide appropriate antiviral drug choices.  (11)  0 *All values are no. (%) case-patients except as indicated. NA, not applicable. †One case-patient was hospitalized for parotitis to rule out infection with mumps and was discharged in stable condition after 1 day. ‡One case-patient who had a diagnosis of end-stage lung disease and a do not resuscitate/do not intubate directive died.