Use of Influenza Antiviral Medications Among Outpatients at High Risk for Influenza-Associated Complications During the 2013–2014 Influenza Season

Abstract

During the 2013–2014 influenza season, we analyzed data from 6004 outpatients aged ≥6 months with acute respiratory illness (ARI). Among the 2786 ARI patients at higher risk for influenza complications, 835 (30%) presented to care ≤2 days from symptom onset; among those, 126 (15%) were prescribed an antiviral medication.

Influenza viruses cause substantial morbidity and mortality each year [1]. Older adults, young children, persons with chronic medical conditions, and pregnant women are at increased risk for influenza-associated complications such as pneumonia and hospitalization [2]. Metaanalyses of randomized controlled trials suggest that early treatment, ie, within 2 days of illness onset, with a neuraminidase-inhibitor antiviral medication reduced the risk of secondary complications such as clinically diagnosed pneumonia that requires antibiotics [3–5]. Therefore, empiric antiviral treatment is recommended for all outpatients at higher risk for influenza-associated complications when influenza is suspected, regardless of illness severity, and for all patients with suspected influenza who are hospitalized or who present with severe illness [2, 6]. However, limited evidence suggests that ambulatory care clinicians prescribe antiviral medications infrequently to outpatients for whom therapy would be most beneficial [7]. Also, patients may not present for care early enough for optimal treatment. We describe the use of neuraminidase inhibitors (oseltamivir and zanamivir) in outpatients at high risk for influenza complications at sites comprising the US Flu Vaccine Effectiveness (VE) Network during the 2013–2014 influenza season.

METHODS

Children and adults seeking care for acute respiratory illness (ARI) at ambulatory care centers were enrolled at 5 sites in the US Flu VE Network during 2013–2014, described in detail elsewhere [7]. Patients aged ≥6 months seeking outpatient medical care for an ARI, defined as a new illness with cough ≤7 days duration, were recruited at study clinics affiliated with academic medical centers or large healthcare organizations. Enrollees were tested for influenza with real-time reverse transcription polymerase chain reaction (RT-PCR). Patients or their guardians provided informed consent. Each site's institutional review board approved the study protocols.

Interviewers recorded symptoms, onset date, and demographic characteristics. High-risk patients were defined as those with a chronic medical condition that placed them at higher risk of influenza-associated complications, per the US Advisory Committee on Immunization Practices [2]. Patients were considered to be high risk if they had a healthcare encounter listed in their electronic medical record (EMR) that resulted in an International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code corresponding to a high-risk medical condition during the previous year. Other high-risk categories included persons aged ≤2 years or ≥65 years of age, pregnant women, those who were morbidly obese (body mass index ≥40 kg/m²), and those who reported being of American Indian, Alaska Native, native Hawaiian, or other Pacific Islander race [2]. “Early” presentation to care was defined as ≤2 days between symptom onset and date of outpatient enrollment visit. Prescription of oseltamivir and zanamivir ≤7 days from date of enrollment was verified by pharmacy, insurance, and/or EMR. Two sites (B and E) recorded self-reported patient symptoms. At those sites, we assessed antiviral prescribing among those with ARI compared with those with influenza-like illness (ILI), defined as with fever/feverishness plus cough.

At 1 network site (B), clinicians received study laboratory test results, usually within 24–48 hours of enrollment. Information was not collected regarding clinical diagnostic influenza testing, and clinicians may have used rapid influenza diagnostic tests or other tests to inform prescribing.

RESULTS

Among 6004 outpatients with ARI enrolled between 2 December 2013 and 20 April 2014, 2786 (46%) were classified as high risk. Among high-risk patients, 336 (12%) were aged 6 months to 2 years, 647 (23%) were aged ≥65 years, 2059 (74%) had ≥1 chronic medical condition, 434 (16%) were morbidly obese, 127 (5%) were of American Indian or Alaska Native race, and 17 (0.06%) were pregnant; 1538 (55%) were vaccinated. Also, 796 (29%) were classified as being in >1 high-risk category (Supplementary Table).

Among all enrollees, 419/6004 (7%) were prescribed neuraminidase inhibitors. Among high-risk patients, 199/2786 (7%) received an antiviral prescription, including 126/835 (15%) high-risk patients who presented early. At sites B and E among high-risk patients with ARI, 148/1044 (14%) were prescribed an antiviral compared with 118/583 (20%) patients with ILI. A similar proportion of high-risk patients presented to care early (835/2786; 30%) compared with patients not at high risk (1075/3218; 33%). The proportion of ARI patients who presented early was highest among pregnant women (41%) and children aged <2 years (37%; Figure 1A). The lowest proportion who sought care early was among those aged ≥65 years; 25% presented early, and 40% waited 5–7 days to seek care. Among the high-risk patients who presented early, the proportion treated with antiviral medications was highest among pregnant women (3/7; 43%) and those who were morbidly obese (31/125; 25%). In contrast, 8/124 (6%) children aged <2 years who presented early received an antiviral prescription (Figure 1B). None of the 66 children aged <2 years with chronic medical conditions were prescribed an antiviral medication, including 20 (30%) who presented early. Vaccination status was not significantly associated with receipt of an antiviral prescription among high-risk patients (χ², P = .14).

Figure 1.

US Flu Vaccine Effectiveness Network, 2013–2014 influenza season. A, Proportion of outpatients with acute respiratory illness (ARI; n = 6004), by time from symptom onset to presentation to their outpatient provider and by category of patients at higher risk for complications from influenza (high-risk category). B, Proportion of outpatients with ARI prescribed influenza antiviral medications, by high-risk category, early presentation (presented to care ≤2 days after symptom onset), and laboratory-confirmed influenza status. C, Proportion of outpatients with prescribed antiviral medications among high-risk outpatients with ARI and reverse transcription polymerase chain reaction (RT-PCR)–confirmed influenza who presented to care ≤2 days after symptom onset, by week, 22 December 2013 to 8 March 2014 (left axis). Line indicates proportion of all RT-PCR tests positive for influenza (right axis). *Early Presentation: Sought care from their outpatient provider ≤2 days after symptom onset. **Body mass index ≥40 kg/m². AI/AN^†American Indian, Alaska Native, Native Hawaiian, or Pacific Islander. ^‡There was a small number of pregnant patients with PCR-confirmed influenza (6), among whom 3 presented early. NOTE: Clinicians at 1 of 5 sites had access to study-related influenza PCR testing results.

Among all ARI patients, 1223/6004 (21%) were found to have laboratory-confirmed influenza, including 452/2786 (16%) high-risk patients. At 2 sites, among the high-risk patients with ILI, 177/583 (30%) had laboratory-confirmed influenza; 55/228 (22%) of those with laboratory-confirmed influenza did not meet ILI criteria. Among all patients with laboratory-confirmed influenza, the proportion presenting early was similar among high-risk (192/452; 43%) and non–high-risk patients (348/781; 45%). While 110/ 452 (24%) high-risk patients with laboratory-confirmed influenza were prescribed an antiviral medication, the proportion was higher among those who presented early (82/192; 43%) and was significantly higher than among high-risk patients without laboratory-confirmed influenza who presented early (44/643; 7%; χ² P < .01).

The proportion of patients prescribed an antiviral medication varied over the course of the influenza season and increased with increasing influenza activity (Figure 1C). The proportion of specimens that tested positive for influenza was highest (38%) during the week of 29 December 2013 to 4 January 2014. During that week, among high-risk patients who presented early, regardless of influenza laboratory status, 19/62 (31%) were prescribed an antiviral medication; this is a significantly higher proportion than were prescribed to high-risk patients who presented early in all other weeks (107/773; 14%; χ² P < .01). Among those high-risk patients who presented early during the peak week and who also had laboratory-confirmed influenza, 14/24 (58%) were prescribed an antiviral medication.

DISCUSSION

During the 2013–2014 influenza season, fewer than half of the influenza-infected outpatients seeking care for an ARI and at high risk for influenza-associated complications presented to care early enough for optimal neuraminidase-inhibitor treatment. In addition, among those high-risk patients who did present to care early, only a small fraction (15%) were prescribed an antiviral medication. This proportion was significantly higher among those high-risk patients who presented early and who also had laboratory-confirmed influenza (43%) or who presented during the peak week of influenza season (31%). However, even at the influenza season's peak, 42% of high-risk patients who presented early and had laboratory-confirmed influenza did not receive antiviral treatment.

Current guidance recommends that clinicians treat high-risk outpatients with suspected influenza empirically, without waiting for confirmatory laboratory testing [2, 6], as neuraminidase inhibitors are most effective if prescribed as soon as possible after symptom onset [8–10]. We found that 70% of high-risk patients presented to care >2 days after symptom onset, including 75% of adults aged ≥65 years with ARI symptoms and 58% of those with laboratory-confirmed influenza. Thus, to optimize care and potentially reduce the risk of influenza-associated complications, high-risk patients should be instructed to contact their provider promptly after the onset of ARI symptoms during the influenza season. Other strategies, such as provider-implemented phone triage lines, could be used to reduce the time between illness onset and initiation of antiviral therapy.

As the influenza season progressed, clinicians increased the use of antiviral drugs in high-risk patients. Our data suggest that for many high-risk patients, clinicians correctly identified influenza among ARI patients, either by clinical presentation or diagnostic tests. The proportion of high-risk patients prescribed antiviral medications increased somewhat when we examined patients with ILI rather than ARI. ILI is more specific for influenza than ARI but less sensitive; 22% of high-risk patients with laboratory-confirmed influenza would have been missed if ILI had been the study enrollment criterion. However, even at the peak of influenza season, a substantial proportion (42%) of high-risk ARI patients who presented to care early and also had laboratory-confirmed influenza were not prescribed an antiviral medication. While antiviral drugs are likely underutilized, antibiotics are likely inappropriately prescribed to many influenza patients. During the 2012–2013 influenza season, 3 antibiotic drugs, which were agents with no anti-influenza activity, were empirically prescribed more often than neuraminidase inhibitors to ARI outpatients with laboratory-confirmed influenza (30% vs 15%, respectively) [7]. Further efforts are needed to understand the barriers to antiviral prescribing for high-risk patients in outpatient settings.

Our study has several limitations. Sites varied widely in prescribing practices and may not be representative of other ambulatory care settings for multiple reasons. Ongoing influenza vaccine research at each site may have increased awareness about influenza, increasing the likelihood that clinicians would prescribe antiviral medications. Some sites may have institutional policies that affect prescribing. We also were unable to examine antiviral prescribing by insurance status. One site gave timely study RT-PCR results to providers, and we did not have information on clinician-ordered testing; these factors may have affected prescribing practices. In addition, our study population was restricted to ARI patients who presented for outpatient care within 7 days of illness onset; we have no information regarding those who were directly hospitalized, presented >7 days after illness onset, or did not seek care.

During the current 2014–2015 influenza season, A(H3N2) is the predominant circulating virus. In addition, as of January 2015, 68% of the H3N2 viruses tested at the Centers for Disease Control and Prevention were antigenically drifted from the H3N2 vaccine virus, and vaccine effectiveness is low [11, 12]. Thus, prompt antiviral use for recommended persons at high risk for influenza-associated complications, including adults aged ≥65 years, children aged <2 years, those with underlying medical conditions, and pregnant women, is especially important during the current influenza season.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online (http://cid.oxfordjournals.org). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.