In the United States, annual vaccination against seasonal influenza is recommended
for all persons aged ≥6 months (
1
). Each influenza season since 2004–05, CDC has estimated the effectiveness of seasonal
influenza vaccine to prevent influenza-associated, medically attended, acute respiratory
illness (ARI). This report uses data, as of February 4, 2017, from 3,144 children
and adults enrolled in the U.S. Influenza Vaccine Effectiveness Network (U.S. Flu
VE Network) during November 28, 2016–February 4, 2017, to estimate an interim adjusted
effectiveness of seasonal influenza vaccine for preventing laboratory-confirmed influenza
virus infection associated with medically attended ARI. During this period, overall
vaccine effectiveness (VE) (adjusted for study site, age group, sex, race/ethnicity,
self-rated general health, and days from illness onset to enrollment) against influenza
A and influenza B virus infection associated with medically attended ARI was 48% (95%
confidence interval [CI] = 37%–57%). Most influenza infections were caused by A (H3N2)
viruses. VE was estimated to be 43% (CI = 29%–54%) against illness caused by influenza
A (H3N2) virus and 73% (CI = 54%–84%) against influenza B virus. These interim VE
estimates indicate that influenza vaccination reduced the risk for outpatient medical
visits by almost half. Because influenza activity remains elevated (
2
), CDC and the Advisory Committee on Immunization Practices recommend that annual
influenza vaccination efforts continue as long as influenza viruses are circulating
(
1
). Vaccination with 2016–17 influenza vaccines will reduce the number of infections
with most currently circulating influenza viruses. Persons aged ≥6 months who have
not yet been vaccinated this season should be vaccinated as soon as possible.
Methods used by the U.S. Flu VE Network have been published previously (
3
). At five study sites, patients aged ≥6 months seeking outpatient medical care for
an ARI with cough, within 7 days of illness onset, were enrolled.* Study enrollment
began after ≥1 laboratory-confirmed cases of influenza were identified through local
surveillance for ≥2 consecutive weeks. Patients were eligible for enrollment if they
1) were aged ≥6 months on September 1, 2016, and thus eligible for vaccination; 2)
reported an ARI with cough and onset ≤7 days earlier; and 3) had not been treated
with influenza antiviral medication (e.g., oseltamivir) during this illness. After
obtaining informed consent from patients or parents/guardians for their children,
participants or their proxies were interviewed to collect demographic data, general
and current health status, symptoms, and 2016–17 influenza vaccination status. Respiratory
specimens were collected from each patient using nasal and oropharyngeal swabs, which
were placed together in a single cryovial with viral transport medium. Only nasal
swabs were collected for patients aged <2 years. Specimens were tested at U.S. Flu
VE Network laboratories using CDC’s real-time reverse transcription – polymerase chain
reaction (rRT-PCR) protocol for detection and identification of influenza viruses.
Participants (including children aged <9 years who require 2 vaccine doses during
their first vaccination season) were considered vaccinated if they received ≥1 dose
of any seasonal influenza vaccine ≥14 days before illness onset, according to medical
records and registries (at Wisconsin site), medical records and self-report (at Texas
and Washington sites), or self-report only (Michigan and Pennsylvania sites). VE was
estimated as 100% x (1 - odds ratio).
†
Estimates were adjusted for study site, age group, sex, race/ethnicity, self-rated
general health, and number of days from illness onset to enrollment using logistic
regression. Interim VE estimates for the 2016–17 season were based on patients enrolled
through February 4, 2017.
Among the 3,144 children and adults with ARI enrolled at the five study sites from
November 28, 2016, through February 4, 2017, 744 (24%) tested positive for influenza
virus by rRT-PCR; 656 (88%) of these viruses were influenza A, and 90 (12%) were influenza
B viruses (Table 1). Among 606 subtyped influenza A viruses, 595 (98%) were A (H3N2)
viruses. The proportion of patients with influenza differed by study site, sex, age
group, race/ethnicity, and interval from illness onset to enrollment (Table 1). The
proportion vaccinated ranged from 46% to 61% across sites and differed by sex, age
group, and interval from illness onset to enrollment.
TABLE 1
Selected characteristics for enrolled patients with medically attended acute respiratory
illness, by influenza test result and seasonal influenza vaccination status — U.S.
Influenza Vaccine Effectiveness Network, United States, November 28, 2016–February
4, 2017
Characteristic
Influenza test result
p-value†
Vaccination status
p-value†
No. positive (%)
No. negative (%)
No. enrolled
No. vaccinated*(%)
Overall
744 (24)
2,400 (76)
3,144
1,650 (52)
State of study site
Michigan
92 (26)
267 (74)
<0.001
359
206 (57)
<0.001
Pennsylvania
176 (30)
416 (70)
592
271 (46)
Texas
56 (8)
646 (92)
702
341 (49)
Washington
374 (38)
613 (62)
987
598 (61)
Wisconsin
46 (9)
458 (91)
504
234 (46)
Sex
Male
350 (26)
990 (74)
0.005
1,340
665 (50)
0.006
Female
394 (22)
1410 (78)
1,804
985 (55)
Age group
6 mos–8 yrs
97 (14)
614 (86)
<0.001
711
362 (51)
<0.001
9–17 yrs
122 (33)
247 (67)
369
128 (35)
18–49 yrs
208 (21)
783 (79)
991
452 (46)
50–64 yrs
189 (31)
425 (69)
614
337 (55)
≥65 yrs
128 (28)
331 (72)
459
371 (81)
Race/Ethnicity
§
White
532 (23)
1,744 (77)
<0.001
2,276
1,231 (54)
0.001
Black
81 (35)
153 (65)
234
99 (42)
Other race
72 (24)
222 (76)
294
163 (55)
Hispanic
47 (15)
274 (85)
321
150 (47)
Self-rated health status
Fair or poor
55 (22)
200 (78)
0.52
255
142 (56)
0.04
Good
179 (23)
599 (77)
778
436 (56)
Very good
301 (25)
902 (75)
1,203
622 (52)
Excellent
209 (23)
699 (77)
908
450 (50)
Illness onset to enrollment (days)
<3
284 (29)
693 (71)
<0.001
977
473 (48)
0.003
3–4
304 (25)
933 (75)
1,237
654 (53)
5–7
156 (17)
774 (83)
930
523 (56)
Influenza test result
Negative
—
2,400
—
2,400
1,317 (55)
—
Influenza B positive¶
90
—
90
23 (26)
B/Yamagata
83
—
83
20 (24)
B/Victoria
4
—
4
1 (25)
B lineage pending
3
—
3
2 (67)
Influenza A positive¶
656
—
656
310 (47)
A (H1N1)pdm09
11
—
11
3 (27)
A (H3N2)
595
—
595
282 (47)
A subtype pending
50
—
50
25 (50)
* Defined as having received ≥1 dose of influenza vaccine ≥14 days before illness
onset. A total of 89 participants who received the vaccine ≤13 days before illness
onset were excluded from the study sample.
† The chi-square statistic was used to assess differences between the numbers of persons
with influenza-negative and influenza-positive test results, in the distribution of
enrolled patient and illness characteristics, and in differences between groups in
the percentage vaccinated.
§ Enrollees were categorized into one of four mutually exclusive racial/ethnic populations:
white, black, other race, and Hispanic. Persons identified as Hispanic might have
been of any race. Persons identified as white, black, or other race were non-Hispanic.
Race/ethnicity data were missing for 19 enrollees.
¶ Two patients had coinfection with influenza A and influenza B, making the sum 746,
or two greater than the total number of influenza positives.
The proportion of ARI patients vaccinated with 2016–17 seasonal influenza vaccine
was 45% among influenza patients compared with 55% among influenza-negative participants
(Table 2). After adjusting for study site, age group, sex, race/ethnicity, self-rated
general health, and number of days from illness onset to enrollment, VE against medically
attended ARI because of influenza was 48% (CI = 37%–57%). VE for all ages was 43%
(CI = 29%–54%) against medically attended ARI because of A (H3N2) virus infection
and 73% (CI = 54%–84%) against influenza B virus infection. VE point estimates against
H3N2-related illness varied by age group; statistically significant protection was
found against H3N2-related illness among children aged 6 months through 8 years (VE
= 53%; CI = 16%–74%) and adults aged 50–64 years (VE = 50%; CI = 23%–67%), whereas
protection in other age groups did not reach statistical significance.
TABLE 2
Number and percentage receiving 2016–17 seasonal influenza vaccine among 3,144 outpatients
with acute respiratory illness and cough, by influenza test result status, age group,
and vaccine effectiveness against all influenza A and B and against virus types A
(H3N2) and B — U.S. Influenza Vaccine Effectiveness Network, United States, November
28, 2016–February 4, 2017
Influenza type/Age group
Influenza-positive
Influenza-negative
Vaccine effectiveness*
Total
No. (%) vaccinated
Total
No. (%) vaccinated
Unadjusted % (95% CI)
Adjusted % (95% CI)
Influenza A and B
Overall
744
333 (45)
2,400
1,317 (55)
33 (21 to 44)†
48 (37 to 57)†
Age group
6 mos–8 yrs
97
32 (33)
614
330 (54)
58 (33 to 73)†
53 (22 to 72)†
9–17 yrs
122
36 (30)
247
92 (37)
29 (−12 to 56)
32 (−20 to 61)
18–49 yrs
208
89 (43)
783
363 (46)
13 (−18 to 36)
19 (−17 to 43)
50–64 yrs
189
76 (40)
425
261 (61)
58 (40 to 70)†
58 (38 to 72)†
≥65 yrs
128
100 (78)
331
271 (82)
21 (−31 to 52)
46 (4 to 70)†
Influenza A (H3N2)
Overall
595
282 (47)
2,400
1,317 (55)
26 (11 to 38)†
43 (29 to 54)†
Age group
6 mos–8 yrs
68
24 (35)
614
330 (54)
53 (21 to 72)†
53 (16 to 74)†
9–17 yrs
94
28 (30)
247
92 (37)
29 (−19 to 57)
23 (−43 to 59)
18–49 yrs
168
73 (43)
783
363 (46)
11 (−24 to 36)
13 (−30 to 41)
50–64 yrs
154
70 (45)
425
261 (61)
48 (24 to 64)†
50 (23 to 67)†
≥65 yrs
111
87 (78)
331
271 (82)
20 (−37 to 53)
44 (−3 to 69)
Influenza B
Overall
90
23 (26)
2,400
1,317 (55)
72 (54 to 83)†
73 (54 to 84)†
Abbreviation: CI = confidence interval.
* Vaccine effectiveness was estimated as 100% x (1 - odds ratio [ratio of odds of
being vaccinated among outpatients with influenza-positive test results to the odds
of being vaccinated among outpatients with influenza-negative test results]); odds
ratios were estimated using logistic regression.
† Statistically significant at the p<0.05 level.
As of February 10, 2017, a total of 13 influenza A (H3N2) viruses from U.S. Flu VE
Network participants had been characterized by CDC; 11 (85%) belonged to genetic group
3C.2a or the related group 3C.2a1, and all of those characterized antigenically were
similar to the reference virus representing the 2016–17 A (H3N2) vaccine component.
Discussion
Interim influenza vaccine effectiveness estimates for the 2016–17 season indicate
that vaccination reduced the risk for influenza-associated medical visits by approximately
half. Influenza activity is likely to continue for several more weeks in the United
States, and vaccination efforts should continue as long as influenza viruses are circulating.
Persons aged ≥6 months who have not yet received the 2016–17 influenza vaccine should
be vaccinated as soon as possible.
§
As of February 3, 2017, approximately 145 million doses of influenza vaccine had been
distributed in the United States for the 2016–17 season.
Interim VE estimates indicate improved protection during the 2016–17 influenza season
against the predominant influenza A (H3N2) virus belonging to genetic group 3C.2a,
which emerged in early 2014 and was predominant during the 2014–15 influenza season
in the United States. During 2014–15, these influenza A (H3N2) 3C.2a viruses were
antigenically different from the recommended A (H3N2) vaccine component, and this
resulted in low (1%) vaccine effectiveness against illness caused by influenza A (H3N2)
3C.2a viruses (
4
). Low effectiveness of the 2014–15 vaccines likely contributed to high rates of influenza-associated
hospitalizations that season, especially among adults aged ≥65 years. In contrast,
rates of influenza-associated hospitalizations observed to date have been substantially
lower during the 2016–17 season (
2
). Virologic surveillance indicates that the majority of influenza A (H3N2) viruses
collected by U.S. laboratories during the 2016–17 season remain antigenically similar
to the A/Hong Kong/4801/2014–like cell propagated reference virus belonging to genetic
group 3C.2a, which is the recommended influenza A (H3N2) component of the 2016–17
Northern Hemisphere vaccine.
Since the 2009 influenza A (H1N1) pandemic, VE estimates for A (H3N2) viruses have
been lower than VE estimates against A (H1N1) and influenza B viruses. Interim VE
estimates against illness caused by influenza A (H3N2) viruses during the 2016–17
influenza season are similar to U.S. VE estimates against A (H3N2)-related illness
during the 2011–12 and 2012–13 seasons (VE = 39%) (
5
,
6
). Also, a meta-analysis of VE studies using the test-negative design conducted from
the 2007–08 through the 2014–15 influenza seasons reported a pooled VE estimate against
A (H3N2)-related illness of 33% (CI = 26%–39%), compared with 61% (CI = 57%–65%) against
influenza A (H1N1)pdm09 and 54% (CI = 46%–61%) against influenza B virus–related illness
(
7
). These results reflect properties unique to A (H3N2) viruses that pose special challenges.
Influenza A (H3N2) viruses undergo more frequent and extensive genetic changes than
do influenza A (H1N1) and influenza B viruses, and require more frequent updates to
the A (H3N2) vaccine virus components to maintain activity against evolving circulating
strains. In addition, A (H3N2) viruses continue to undergo changes in their receptor-binding
specificity, which might result in genetic changes during growth in eggs. Most influenza
vaccines are manufactured using egg-based production processes. These genetic changes
(referred to as egg-adapted changes) alter the antigenic properties of candidate vaccine
viruses (CVVs) as they are grown in eggs and potentially during the vaccine production
process (
8
). The egg-adapted changes might contribute to the lower vaccine effectiveness seen
with A (H3N2) viruses compared with A (H1N1) and B viruses. Efforts are ongoing to
improve influenza vaccine effectiveness against A (H3N2) viruses in CVV development
and in manufacturing.
As of February 10, 2017, influenza activity remained elevated nationally and was widespread
across most of the United States. During recent A (H3N2) virus predominant–seasons,
persons aged ≥65 years and young children experienced higher rates of severe illness
and influenza-associated hospitalization compared with other age groups. With vaccine
effectiveness of 48%, some vaccinated persons will become infected with influenza.
Clinicians should maintain a high index of suspicion for influenza infection among
persons with acute respiratory illness while influenza activity is ongoing, especially
among older adults. Early antiviral treatment can reduce severity and complications
of influenza-associated illness (
9
). Early antiviral treatment is recommended for persons with suspected influenza with
severe or progressive illness (e.g., hospitalized persons) and persons at high risk
for complications from influenza, such as children aged <2 years, adults aged ≥65
years and persons with underlying health conditions,
¶
even if illness is less severe. Antiviral medications should be used as recommended
for treatment in patients with suspected influenza, regardless of vaccination status.
The decision to initiate antiviral treatment should not be delayed while waiting for
laboratory confirmation of influenza and should not be dependent on insensitive assays,
such as rapid influenza diagnostic tests.
The findings in this report are subject to at least four limitations. First, vaccination
status included self-report at four of five sites. End-of-season VE estimates based
on updated documentation of vaccination status might differ from interim estimates.
Second, information from medical records and immunization registries is needed to
evaluate VE by vaccine type and for fully vaccinated compared with partially vaccinated
children (children aged <9 years require 2 vaccine doses during their first vaccination
season), as well as to evaluate the effects of prior season vaccination and timing
of vaccination; end-of-season analysis of VE by vaccine type and effects of partial
or prior season vaccination is planned. Third, an observational study design has greater
potential for confounding and bias relative to randomized clinical trials. However,
the test-negative design is widely used in VE studies and has been used by the U.S.
Flu VE Network to estimate VE for the past several influenza seasons. Finally, small
sample sizes in some age groups resulted in wide confidence intervals, and end-of-season
VE estimates could change as additional patient data become available or if there
is a change in circulating viruses late in the season. It is also important to note
that the VE estimates in this report are limited to the prevention of outpatient medical
visits, rather than more severe illness outcomes, such as hospitalization or death;
data from studies measuring VE against more severe outcomes will be available at a
later date.
Annual vaccination against circulating influenza viruses remains the best strategy
for preventing illness from influenza. As of early November 2016, only 37% of children
aged 6 months–17 years, 37% of adults aged 18–64 years, and 57% of adults aged ≥65
years had received influenza vaccine this season (
10
). Among pregnant women, early estimates for 2016–17 indicated that only 47% had been
vaccinated by early November 2016 (
10
). In addition to ongoing vaccination efforts, antiviral medications continue to be
an important adjunct to the treatment and control of influenza and should be used
as recommended, regardless of patient vaccination status.