|Home | About | Journals | Submit | Contact Us | Français|
The 2009 pandemic influenza A (H1N1) outbreak was associated with an increased use of antiviral agents and highlighted the role of population-based monitoring for related adverse drug events (ADEs). An ongoing, nationally representative emergency department-based surveillance system was used to identify and characterize ADEs during the pandemic. Active surveillance for ADEs successfully provided timely, population-based data during the pandemic. Increases in antiviral ADEs paralleled increases in prescribing. Type and severity of ADEs were similar across all seasons.
On April 26, 2009, the U.S. Department of Health and Human Services (HHS) declared a public health emergency for 2009 pandemic influenza A (H1N1) (hereafter, pH1N1).1,2 Influenza activity remained at higher-than-normal levels throughout the spring and summer, peaking in late October and early November 2009, and disproportionately affected younger people.3 The pH1N1 influenza virus was determined to be susceptible to the neuraminidase inhibitors (oseltamivir and zanamivir), which were recommended for treatment and chemoprophylaxis of pH1N1 infection.4 Given the pandemic spread and potential severity of infection with the pH1N1 virus, the Centers for Disease Control and Prevention (CDC) requested an Emergency Use Authorization to expand neuraminidase inhibitor use to include children younger than one year of age (oseltamivir) and patients of any age with severe disease (oseltamivir and zanamivir).5
Adverse events from medications are typically monitored through passive surveillance, primarily the U.S. Food and Drug Administration Adverse Event Reporting System (AERS), which relies on voluntary reporting of adverse drug events (ADEs) by health-care professionals and consumers and reporting by pharmaceutical manufacturers.6 Because the early stages of a novel influenza pandemic are associated with many unknowns, including possible delays in availability of vaccine, a 2008 Institute of Medicine report recommended that HHS consider options, in addition to AERS, for capturing antiviral adverse events, as use of antivirals, particularly among vulnerable populations, is likely to increase.7 We used data from an active, population-based surveillance system to monitor emergency department (ED) visits for antiviral ADEs during pH1N1 and compared these data with those from previous influenza seasons.
The Drug Abuse Warning Network (DAWN) is an ongoing public health surveillance system administered by the Substance Abuse and Mental Health Services Administration (SAMHSA) that monitors drug-related hospital ED visits. DAWN uses a national probability sample of approximately 250 nonfederal, short-stay general hospitals that operate 24-hour EDs. On a regular basis, DAWN reporters at each facility actively review and record data from ED medical charts for the full spectrum of use, misuse, and abuse for any drug-related visit, including adverse reactions to drugs and vaccines, as documented by the treating physician. Unweighted case data are available in near real time; national estimates are calculated annually. Direct patient identifiers are not collected.8
We defined a case as an ED visit made between January 1, 2006, and December 31, 2009, for an adverse event associated with an influenza antiviral (oseltamivir, zanamivir, amantadine, or rimantadine). We identified cases through a Web-based portal to near real-time DAWN data. We obtained additional narrative descriptions of ED visits from SAMHSA (Unpublished data, SAMHSA, Center for Behavioral Health Statistics and Quality, DAWN; unweighted ED data from DAWN, January 2006–December 2009) and coded them using the Medical Dictionary for Regulatory Activities version 9.1.9 Adverse events were categorized by implicated drug, patient demographics, disposition, and adverse event manifestation. We conducted analyses using SAS®.10
Influenza antiviral prescription data were obtained from CDC BioSense, which receives real-time data from RelayHealth™ (www.relayhealth.com), an electronic prescriptions claims service provider for 20,000 to 30,000 pharmacies in all 50 states and Washington, D.C., as well as U.S. territories.11
In the previous three influenza seasons, the number of ED visits for antiviral adverse events peaked in February. In 2009, however, following a peak in February, ED visits peaked again in June and November. The number of antiviral medication prescriptions followed a similar temporal pattern (Figure).
During the 2008–2009 influenza season (and through September 2009 to account for early pH1N1 activity), most adverse event visits were among females (59.4%) and people aged 18–44 years (41.0%) and were related to oseltamivir (71.3%). Most ED visits resulted in the patient being discharged home (77.8%). In comparison with previous influenza seasons, a higher proportion of adverse events involved children younger than 6 years of age during the extended 2008–2009 season and the first half of the 2009–2010 season (Table 1). In 2009, the most common adverse events associated with neuraminidase inhibitor use were gastrointestinal symptoms (48.1%), allergic reactions (31.3%), and central nervous system effects (19.9%). Compared with previous years, type and severity of ADEs were similar in 2009 (Table 2).
This article provides the first description of antiviral-related adverse events during the 2009 H1N1 influenza pandemic based on active, population-based surveillance. Antiviral safety monitoring using DAWN was conducted on an ongoing and regular basis throughout the pH1N1 outbreak and successfully contributed to response efforts by providing timely data to public health authorities on potential safety concerns resulting from widespread use of antiviral agents. We identified an increase in the number of ED visits for antiviral ADEs shortly after an emergency was declared for pH1N1,12 and we identified another peak in visits with the second phase of the outbreak.13 The types of adverse events were consistent with the known safety profile of these agents and similar to data from previous seasons.14–16 These surveillance data reflected the nature of the pandemic and the antiviral recommendations for the 2009–2010 season in that a substantially higher proportion of visits involved neuraminidase inhibitors than adamantanes and were made by people in younger age groups compared with previous seasons.5 Because the severity and types of adverse event manifestations were comparable to previous seasons, these data did not trigger further investigation.
These data had a number of limitations. First, only adverse events that resulted in an ED visit were included. While the ED is an appropriate setting to identify many serious adverse events, rare, previously unreported ADEs or ADEs difficult to diagnose in the ED setting are less likely to be identified. Thus, these types of ADEs may have been underreported.17 Other types of adverse events may have been overreported. For example, in some cases, disease manifestations (e.g., underlying influenza infection) may have been confused with an antiviral adverse event. If more than one medication was reported, drugs other than influenza antivirals may have contributed to the adverse event. None of these limitations, however, would be expected to affect trends over time.
Second, the number of hospitals reporting to DAWN varies over the years; however, this variation would not account for the substantial increase in the number of cases reported in 2009 compared with previous years and should not have affected the types of ADEs reported. Third, BioSense antiviral use data represent prescriptions approved by third-party payers (e.g., private insurance, Medicare, or Medicaid) and do not include prescriptions that were paid for out-of-pocket, denied coverage, or ordered from online pharmacies. However, RelayHealth estimates that CDC receives data for approximately 50% of anti-infective prescription transactions in the U.S.
Surveillance using DAWN data successfully complemented passive surveillance by providing timely, population-based data for the detection and characterization of ADEs associated with influenza antivirals during a pandemic. Continuing these surveillance efforts would be useful for future influenza seasons.
The authors thank Victor Johnson, Northrop Grumman, contractor for the Centers for Disease Control and Prevention, and Inga Allred of RTI International for assistance with programming; and Kathy Ghanouni, PharmD, for assistance with data collection. No individuals named herein received compensation for their contributions.
The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the funding agencies.