Vaccine Effectiveness Against Influenza-Associated Hospitalizations Among Adults, 2018–2019, US Hospitalized Adult Influenza Vaccine Effectiveness Network

Abstract

We estimated vaccine effectiveness (VE) for prevention of influenza-associated hospitalizations among adults during the 2018–2019 influenza season. Adults admitted with acute respiratory illness to 14 hospitals of the US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN) and testing positive for influenza were cases; patients testing negative were controls. VE was estimated using logistic regression and inverse probability of treatment weighting. We analyzed data from 2863 patients with a mean age of 63 years. Adjusted VE against influenza A(H1N1)pdm09–associated hospitalization was 51% (95% confidence interval [CI], 25%–68%). Adjusted VE against influenza A(H3N2) virus–associated hospitalization was −2% (95% CI, −65% to 37%) and differed significantly by age, with VE of −130% (95% CI, −374% to −27%) among adults 18 to ≤56 years of age. Although vaccination halved the risk of influenza A(H1N1)pdm09–associated hospitalizations, it conferred no protection against influenza A(H3N2)–associated hospitalizations. We observed negative VE for young and middle-aged adults but cannot exclude residual confounding as a potential explanation.

Influenza vaccination halved the risk of influenza A(H1N1)pdm09–associated hospitalizations but conferred no protection against influenza A(H3N2)–associated hospitalizations among adults enrolled in the HAIVEN study during the 2018–2019 US influenza season.

The 2018–2019 United States (US) influenza season was characterized by extensive circulation of influenza A viruses, with influenza A(H1N1)pdm09 virus circulation beginning in October, followed by a wave of influenza A(H3N2) virus circulation from February to May. Influenza A(H1N1)pdm09 viruses were well-matched to the vaccine virus strain, but the majority of influenza A(H3N2) viruses were of the 3C.3a genetic group and antigenically distinct from the genetic group 3C.2a1 vaccine virus strain. Influenza vaccination was moderately protective against outpatient influenza illness caused by influenza A(H1N1)pdm09 viruses, with estimated vaccine effectiveness (VE) of 44%, but offered no protection against the predominant drifted influenza A(H3N2) viruses [1]. Several studies reported negative VE against influenza A(H3N2) infections among young to middle-aged adults during the 2018–2019 Northern Hemisphere influenza season [2–4], including preliminary findings of the US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN) [5]. However, other studies from the US and England did not observe negative VE [1, 6].

It is important to understand vaccine protection against serious outcomes such as hospitalizations in addition to protection offered against milder, outpatient illnesses. The observational test-negative design is the standard design for studies of influenza VE and estimates VE by comparing odds of vaccination among patients with acute respiratory illness (ARI) who test positive for influenza with those who test negative for influenza [7, 8]. However, interpreting VE estimates for prevention of influenza hospitalization from this nonrandomized design is complicated by the complexity of patients typically enrolled in hospital-based studies, who often have comorbidities and other features that give rise to systematic differences between vaccinated patients and the unvaccinated comparison group [9, 10]. Additionally, the use of controls with noninfluenza ARI may lead to confounding of VE if these patients are more likely to be vaccinated due to their chronic conditions [11–13]; hence, adequate adjustment and control for these characteristics is particularly important for estimation of VE in the inpatient setting.

Here, we report final VE estimates for prevention of influenza-associated hospitalization among adults for the HAIVEN study during the 2018–2019 US influenza season, with a focus on reducing potential bias in VE estimates by improving control of confounding arising from baseline differences in vaccinated and unvaccinated study participants.

METHODS

Participants

Participants were adults (aged ≥18 years) hospitalized for ARI presenting with new or worsening cough or sputum production of ≤10 days’ duration at 14 hospitals comprising the HAIVEN study during the 2018–2019 influenza season. Eligibility criteria, enrollment procedures, and ARI definitions have been described previously [9]. In brief, adults admitted to the hospital with ARI were identified by daily review of electronic medical records (EMRs). Eligible participants or surrogates gave written consent, and institutional review boards at participating hospitals and the Centers for Disease Control and Prevention (CDC) approved study procedures. Participants or proxies were interviewed to collect information about demographics, influenza vaccination, illness characteristics, and subjective assessment of frailty. Diagnosis codes from the index hospitalization and medical encounters in the year prior to enrollment were extracted from EMRs and used to calculate the Charlson Comorbidity Index [14] and identify presence of specific comorbidities known to increase risk of serious influenza complications (“high-risk conditions”) [15]. Clinical outcomes were extracted from EMRs. Influenza vaccine receipt was documented using EMRs, state immunization registries, or plausible patient self-report from the enrollment interview (self-report was considered plausible if timing and location of vaccination were provided).

Specimen Collection and Laboratory Methods

Nasal and throat specimens (at 7 hospitals) or nasopharyngeal specimens (at 7 hospitals) were obtained from patients and tested for influenza and respiratory syncytial virus by molecular assays. A subset of specimens for which respiratory viral panel testing was ordered by the treating clinical team was tested for additional respiratory viruses, including human rhinovirus, human coronaviruses (HKU1, OC43, NL63, and 229E), parainfluenza viruses 1–4, human metapneumovirus, enterovirus, and adenovirus.

Statistical Methods

Patients who tested positive for influenza were cases and patients who tested negative for influenza were controls. Characteristics of cases and controls were compared using the absolute standardized mean difference (SMD) between groups [16] and conventional tests of differences using χ ² test or Fisher exact test for categorical variables and t test or Wilcoxon rank-sum test for continuous variables. Characteristics of vaccinated and unvaccinated participants were compared similarly. For subgroup analyses, we defined age groups by tertiles of age among the sample overall (18 to ≤56, >56 to ≤69, >69 years).

Patients were defined as vaccinated if they had documented or plausible self-report of vaccination ≥14 days before illness onset; we excluded individuals who received the vaccine 1–13 days before illness onset or who could not report either location and approximate timing of vaccination and had no documented vaccination. Frailty score was defined as the sum of the dichotomized subjective assessments of frailty as previously described [9, 17].

Estimation of VE

VE was estimated separately for influenza A(H1N1)pdm09 and influenza A(H3N2). Participants enrolled outside the period of local subtype-specific circulation were excluded, leaving differing numbers of controls in the analytic datasets. Patients coinfected with influenza and another respiratory virus were excluded. The small number of cases precluded estimation of VE against influenza B–associated hospitalizations. VE was estimated in all subgroups using multivariate logistic regression with influenza case status as the outcome and vaccination as the predictor of interest, with VE = 1 – (adjusted odds ratio [OR] for vaccination) × 100%.

To adjust for confounding, for all analyses, the data were first balanced by baseline characteristics that differed substantially between vaccinees and nonvaccinees using propensity score models and inverse probability of treatment weighting (IPTW) [18]. Details of the propensity score model and VE regression model building strategies are provided in the Supplementary Data. In brief, characteristics with large baseline differences between vaccinated and unvaccinated participants (SMD > ≈ 0.20) and also associated with case status were considered for inclusion in the propensity score model. Because the largest baseline difference was in regard to influenza vaccination habit (proportion of participants reporting receipt of influenza vaccine always/almost always) and prior season vaccination may confound current season VE [19], this characteristic was prioritized and balanced for all analyses. Additional characteristics associated with both vaccination and case status were balanced as feasible, while maintaining balance on vaccination habit. Time-varying characteristics (including calendar time of illness onset relative to peak of subtype-specific case onset date and days between onset and specimen collection) and baseline characteristics that remained unbalanced after weighting were included in the regression model as adjustment variables. Goodness of fit between alternative models was compared with the Akaike information criterion; 95% confidence intervals (CIs) excluding the null value were considered statistically significant. In subgroup analyses, we examined VE by tertiles of age (18 to ≤56, >56 to ≤69, >69 years) and by influenza vaccination habit (always/almost always vs never/rarely). In addition, we examined VE against influenza A(H3N2) as a function of age specified as a natural cubic spline. Analyses were performed in SAS version 9.4 (SAS Institute, Cary, North Carolina) and R version 3.4 (R Group, Vienna, Austria) software.

In addition to the primary analyses, we conducted a bias indicator analysis to assess if VE results could be attributable to additional bias [20, 21]. For this analysis, we replicated the primary analysis with identical methods and using as “cases” those patients who tested positive for ≥1 noninfluenza respiratory virus as a negative control outcome. The bias indicator analysis was restricted to patients who received clinically ordered testing for multiple respiratory viral pathogens and excluded influenza-positive cases. Because influenza vaccination should not influence infection with a noninfluenza virus, observing an association between vaccination and this outcome would suggest that residual confounding may have affected results. Failure to observe an association would suggest (but not prove) that residual confounding was minimal. Significantly negative or positive bias-indicator estimates would suggest underestimation or overestimation of VE in our primary analysis, respectively. To increase comparability between primary and bias analysis datasets, bias indicator analyses were conducted separately for influenza A(H3N2) and influenza A(H1N1)pdm09.

RESULTS

Enrollment

From 25 November 2018 to 27 April 2019, 3975 patients were enrolled. Of these, 1112 were excluded, mostly due to enrollment outside the period of local influenza circulation (n = 747 exclusions) or because they lacked variables needed to estimate VE (n = 218 exclusions, primarily due to indeterminate vaccination status), leaving 2863 participants. Of these, 247 participants tested positive for influenza A(H3N2) virus infection, 225 tested positive for influenza A(H1N1)pdm09 virus infection, and 13 tested positive for influenza B viruses (Supplementary Figure 1). One hundred forty-seven influenza A(H3N2)–infected patients and 129 influenza A(H1N1)pdm09–infected patients had viruses sequenced at CDC. Of sequenced viruses, 93% of influenza A(H3N2) viruses belonged to the 3c.3a genetic group, and 100% influenza A(H1N1)pdm09 viruses belonged to the 6B.1A genetic group, similar to national trends [22].

Patient Characteristics by Case Status

Average participant age was 63 years (range, 18–102 years). Fifty-six percent and 26% of participants were female and of black race, respectively. Influenza A(H3N2) cases tended to be older than controls (mean age, 64 vs 61 years; P = .06), and this pattern was similar regardless of vaccination status (Table 1). Comorbidities were common among both influenza A(H3N2) cases and controls, with cardiopulmonary conditions and metabolic disorders (including diabetes) the most prevalent; the mean number of high-risk conditions was >4 in both groups. However, influenza A(H3N2) cases tended to have fewer comorbid conditions and fewer hospitalizations in the past year than controls. Specifically, control patients had significantly more cardiovascular, lung, and immunosuppressive disorders than did influenza A(H3N2) cases. Controls were slightly more likely than cases to have been admitted to tertiary care hospitals (56% vs 48%). Influenza A(H3N2) cases in the youngest age group (18 to ≤56 years) had considerable comorbidity and, although less frail, were more likely to have been hospitalized in the year prior to enrollment than older influenza A(H3N2) cases (55% vs 46%; Supplementary Table 1).

Table 1.

Characteristics of Influenza A(H3N2) Cases and Influenza-Negative Controls, US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018–2019 (n = 2304)

Characteristic	Influenza A(H3N2) Cases	Influenza-Negative Controls	P Valuea	Standardized Differenceb
No.	247	2057
Site			<.001	0.33
Central Texas	59 (23.9)	716 (34.8)
Southeast Michigan	56 (22.7)	472 (22.9)
Western Pennsylvania	60 (24.3)	521 (25.3)
Central Tennessee	72 (29.1)	348 (16.9)
Baseline characteristics
Demographics and behavioral risk factors
Age groupc			.012	0.20
18 to ≤56 y	66 (27)	705 (34)
>56 to ≤69 y	80 (32)	689 (34)
>69 y	101 (41)	663 (32)
Age, y, mean (SD)	63.47 (17.95)	61.31 (16.64)	.056	0.13
Female sex	138 (55.9)	1158 (56.3)	.953	0.01
Black race	60 (24.3)	538 (26.2)	.579	0.04
Community dwelling	231 (93.5)	1943 (94.5)	.648	0.04
Insured	237 (96.0)	1940 (94.3)	.358	0.08
Current tobacco smoking	50 (20.2)	557 (27.1)	.026	0.16
Current vaping/e-cigarette use	7 (2.8)	105 (5.1)	.158	0.12
Comorbid health conditionsd
Obstructive lung disease	92 (37.2)	1038 (50.5)	<.001	0.27
Nonobstructive lung disease	68 (27.5)	895 (43.5)	<.001	0.34
Heart disease	134 (54.3)	1280 (62.2)	.018	0.16
Congestive heart failure	53 (21.5)	666 (32.4)	.001	0.25
Neurologic disorder	71 (28.7)	710 (34.5)	.082	0.12
Renal disorders	90 (36.4)	852 (41.4)	.151	0.10
Malignancy	57 (23.1)	482 (23.4)	.964	0.01
Hematologic disorders	24 (9.7)	230 (11.2)	.557	0.05
Metabolic disorders	133 (53.8)	1187 (57.7)	.275	0.08
Diabetes/obesity/endocrine disorders	169 (68.4)	1530 (74.4)	.05	0.13
Health status indicators
No. of ACIP high-risk conditions, mean (SD)	4.59 (3.22)	5.42 (3.27)	<.001	0.26
Charlson Comorbidity Index, mean (SD)	2.61 (3.00)	3.09 (3.00)	.018	0.16
Frailty score (0–5), mean (SD)	1.72 (1.50)	2.01 (1.46)	.003	0.20
Home oxygen use	56 (23.0)	565 (27.6)	.139	0.11
≥1 hospitalization in prior year (self-report)	119 (48.2)	1309 (63.6)	<.001	0.32
Self-reported chemotherapy/radiation, immunosuppressive medications	80 (32.4)	823 (40.0)	.024	0.16
Immunosuppression by ICD-10 code and/or self-report	103 (41.7)	1045 (50.8)	.008	0.18
Influenza vaccination habite
Never/rarely	38 (15.4)	452 (22.0)	.004	0.24
Sometimes	9 (3.6)	140 (6.8)
Always/almost always	200 (81.0)	1465 (71.2)
Vaccination habit score (0–2), mean (SD)	1.66 (0.73)	1.49 (0.83)	.003	0.21
Influenza vaccination status
Vaccinated by documentation or plausible self-reportf	194 (78.5)	1455 (70.7)	.010	0.18
Vaccinated by self-reportg	190 (76.9)	1426 (69.3)	.014	0.17
Admission characteristics
Days from illness onset to admission, mean (SD)	2.78 (2.15)	2.86 (2.37)	.582	0.04
Admitted to tertiary care hospital	118 (47.8)	1141 (55.5)	.026	0.15
Presented with influenza-like illnessh	152 (61.5)	823 (40.0)	<.001	0.44
Admitted to intensive care unit	20 (8.1)	248 (12.1)	.082	0.13
Length of hospital stay, d, mean (SD)	3.97 (3.46)	4.91 (4.60)	.002	0.23
Died in hospital	2 (0.8)	25 (1.2)	.805	0.04

Data are presented as No. (%) unless otherwise indicated.

Abbreviations: ACIP, Advisory Committee on Immunization Practices; ICD-10, International Classification of Diseases, 10th Edition; SD, standard deviation.

^a  P value for test of difference across case and control groups based on χ ² statistic for categorical variables,

t test for difference of means of normally distributed continuous variables, and nonparametric log-rank test for nonnormally distributed continuous variables.

^bStandardized absolute mean differences >0.10 shown in bold; standardized differences ≤0.10 are generally considered negligible differences between groups.

^cAge groups defined by tertile of age distribution of all subjects.

^dDerived from ICD-10 codes associated with inpatient and outpatient medical encounters in the year prior to enrollment admission.

^eDefined by patient self-report during enrollment interview.

^fPlausible self-report defined as affirmative self-report of current season vaccination including known or approximate date and location of vaccination.

^gSelf-report defined as affirmative self-report of current season vaccination including known or approximate date.

^hInfluenza-like illness defined as self-reported subjective fever/feverishness plus cough or sore throat.

Influenza A(H1N1)pdm09 cases (mean age, 60 years) tended to be younger than controls (mean age, 62 years) (Table 2). As with influenza A(H3N2), cardiorespiratory and metabolic comorbidities were common among influenza A(H1N1)pdm09 cases, but cases tended to have fewer comorbid conditions and hospitalizations than controls. Controls were more likely than influenza A(H1N1)pdm09 cases to have been admitted to tertiary care hospitals. Cases in the youngest age group (18 to ≤56 years) were more likely to have been hospitalized in the prior year and to have been admitted to tertiary care hospitals compared to older cases (Supplementary Table 2).

Table 2.

Characteristics of Influenza A(H1N1)pdm09 Cases and Influenza-Negative Controls, US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018–2019 (n = 2447)

Characteristic	Influenza A(H1N1)pdm09 Cases	Influenza-Negative Controls	P Valuea	Standardized Differenceb
No.	225	2222
Site			<.001	0.54
Central Texas	31 (13.8)	794 (35.7)
Southeast Michigan	57 (25.3)	500 (22.5)
Western Pennsylvania	96 (42.7)	630 (28.4)
Central Tennessee	41 (18.2)	298 (13.4)
Baseline characteristics
Demographics and behavioral risk factors
Age groupc			.116	0.15
18 to ≤56 y	87 (38.7)	745 (33.5)
>56 to ≤69 y	79 (35.1)	752 (33.8)
>69 y	59 (26.2)	725 (32.6)
Age, y, mean (SD)	60.11 (15.38)	61.66 (16.49)	.178	0.10
Female	118 (52.4)	1246 (56.1)	.33	0.07
Black race	61 (27.1)	585 (26.3)	.861	0.02
Community dwelling	213 (94.7)	2095 (94.3)	.932	0.02
Insured	209 (92.9)	2106 (94.8)	.298	0.08
Current tobacco smoking	50 (22.2)	602 (27.1)	.135	0.11
Current vaping/e-cigarette use	14 (6.2)	116 (5.2)	.63	0.04
Comorbid health conditionsd
Obstructive lung disease	89 (39.6)	1188 (53.5)	<.001	0.28
Nonobstructive lung disease	71 (31.6)	1001 (45.0)	<.001	0.28
Heart disease	107 (47.6)	1431 (64.4)	<.001	0.34
Congestive heart failure	48 (21.3)	772 (34.7)	<.001	0.30
Neurologic disorder	59 (26.2)	790 (35.6)	.006	0.20
Renal disorders	77 (34.2)	935 (42.1)	.027	0.16
Malignancy	48 (21.3)	540 (24.3)	.362	0.07
Hematologic disorders	27 (12.0)	244 (11.0)	.724	0.03
Metabolic disorders	117 (52.0)	1308 (58.9)	.055	0.14
Diabetes/obesity/endocrine disorders	157 (69.8)	1673 (75.3)	.083	0.12
Health status indicators
Number of ACIP high-risk conditions, mean (SD)	4.52 (3.09)	5.62 (3.23)	<.001	0.35
Charlson Comorbidity Index, mean (SD)	2.50 (2.91)	3.28 (3.00)	<.001	0.26
Frailty score (0–5), mean (SD)	1.70 (1.46)	2.02 (1.46)	.002	0.22
Home oxygen use	39 (17.5)	629 (28.5)	.001	0.26
≥1 hospitalization in prior year (self-report)	103 (45.8)	1412 (63.5)	<.001	0.36
Self-reported chemotherapy/radiation, immunosuppressive medications	68 (30.2)	904 (40.7)	.003	0.22
Immunosuppression by ICD-10 code and/or self-report	90 (40.0)	1153 (51.9)	.001	0.24
Influenza vaccination habite			<.001	0.35
Never/rarely	84 (37.3)	484 (21.8)
Sometimes	10 (4.4)	150 (6.8)
Always/almost always	131 (58.2)	1588 (71.5)
Vaccination habit score (0–2), mean (SD)	1.21 (0.96)	1.50 (0.83)	<.001	0.32
Influenza vaccination status
Vaccinated by documentation or plausible self-reportf	118 (52.4)	1568 (71.5)	<.001	0.38
Vaccinated by self-reportg	113 (50.2)	1540 (69.3)	<.001	0.40
Admission characteristics
Days from illness onset to admission, mean (SD)	2.83 (2.13)	2.92 (2.40)	.556	0.04
Admitted to tertiary care hospital	98 (43.6)	1226 (55.2)	.001	0.23
Presented with influenza-like illnessh	145 (64.4)	888 (40.0)	<.001	0.51
Admitted to intensive care unit	30 (13.3)	267 (12.0)	.646	0.04
Length of hospital stay, d, mean (SD)	4.37 (3.77)	4.92 (4.51)	.074	0.13
Died in hospital	4 (1.8)	26 (1.2)	.637	0.05

Data are presented as No. (%) unless otherwise indicated.

Abbreviations: ACIP, Advisory Committee on Immunization Practices; ICD-10, International Classification of Diseases, 10th Edition; SD, standard deviation.

^a  P value for test of difference across case and control groups based on χ ² statistic for categorical variables,

t test for difference of means of normally distributed continuous variables, and nonparametric log-rank test for nonnormally distributed continuous variables.

^bStandardized absolute mean differences >0.10 shown in bold; standardized differences ≤0.10 are generally considered negligible differences between groups.

^cAge groups defined by tertile of age distribution of all subjects.

^dDerived from ICD-10 codes associated with inpatient and outpatient medical encounters in the year prior to enrollment admission.

^eDefined by patient self-report during enrollment interview.

^fPlausible self-report defined as affirmative self-report of current season vaccination including known or approximate date and location of vaccination.

^gSelf-report defined as affirmative self-report of current season vaccination including known or approximate date.

^hInfluenza-like illness defined as self-reported subjective fever/feverishness plus cough or sore throat.

Patient Characteristics by Vaccination Status

Compared to unvaccinated participants, vaccinees were older (64 vs 56 years) and more likely to be of non-black race (78% vs 63%), to have most specific types of comorbidities and more comorbidities overall (mean Charlson Comorbidity Index of 3.2 vs 2.7), to use home oxygen (30% vs 19%), and to have been hospitalized in the past year (63% vs 55%) (P < .05 for all; Table 3). Differences in prevalence of comorbidities between the vaccinated and unvaccinated groups were more pronounced in the youngest age group (18 to ≤56 years). Unvaccinated participants were more likely to have been admitted to tertiary care hospitals (58% vs 53% among vaccinees). Vaccinated participants were less likely to be current tobacco users (22% vs 36%) and far more likely than unvaccinated participants to report always or almost always receiving seasonal influenza vaccine (90% vs 28%).

Table 3.

Characteristics of Vaccinated and Unvaccinated Study Participants by Age Group, US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018–2019 (N = 2863)

Characteristic	All (N = 2863)				Patients Aged 18 to ≤56 y (n = 968)				Patients Aged >56 to ≤69 y (n = 968)				Patients Aged >69 y (n = 927)
	Vaccinateda	Unvaccinated	P Valueb	Standardized Differencec	Vaccinateda	Unvaccinated	P Valueb	Standardized Differencec	Vaccinateda	Unvaccinated	P Valueb	Standardized Differencec	Vaccinateda	Unvaccinated	P Valueb	Standardized Differencec
No.	1991	872			563	405			660	308			768	159
Site			.801	0.04			.205	0.14			.225	0.15			.123	0.22
Central Texas	609 (30.6)	277 (31.8)			112 (19.9)	102 (25.2)			178 (27.0)	96 (31.2)			319 (41.5)	79 (49.7)
Southeast Michigan	479 (24.1)	217 (24.9)			160 (28.4)	116 (28.6)			176 (26.7)	81 (26.3)			143 (18.6)	20 (12.6)
Western Pennsylvania	562 (28.2)	234 (26.8)			177 (31.4)	117 (28.9)			206 (31.2)	78 (25.3)			179 (23.3)	39 (24.5)
Central Tennessee	341 (17.1)	144 (16.5)			114 (20.2)	70 (17.3)			100 (15.2)	53 (17.2)			127 (16.5)	21 (13.2)
Baseline characteristics
Demographics and behavioral risk factors
Age group			<.001	0.50
18 to ≤56 y	563 (28.3)	405 (46.4)			563 (100.0)	405 (100.0)			…	…			…	…
>56 to ≤69 y	660 (33.1)	308 (35.3)			…	…			660 (100.0)	308 (100.0)			…	…
>69 y	768 (38.6)	159 (18.2)			…	…			…	…			768 (100.0)	159 (100.0)
Age, y, mean (SD)	63.79 (16.48)	56.03 (15.63)	<.001	0.48	43.20 (10.84)	42.77 (10.51)	.535	0.04	63.05 (3.61)	62.33 (3.62)	.004	0.20	79.52 (6.91)	77.63 (6.45)	.002	0.28
Female sex	1112 (55.9)	487 (55.8)	1	<0.001	323 (57.4)	224 (55.3)	.567	0.04	358 (54.2)	168 (54.5)	.985	0.01	431 (56.1)	95 (59.7)	.452	0.07
Black race	441 (22.1)	324 (37.2)	<.001	0.33	184 (32.7)	176 (43.5)	.001	0.22	174 (26.4)	110 (35.7)	.004	0.20	83 (10.8)	38 (23.9)	<.001	0.35
Community dwelling	1871 (94.0)	830 (95.2)	.229	0.05	539 (95.7)	384 (94.8)	.605	0.04	636 (96.4)	301 (97.7)	.354	0.08	696 (90.6)	145 (91.2)	.94	0.02
Insured	1920 (96.4)	790 (90.6)	<.001	0.24	521 (92.5)	350 (86.4)	.003	0.20	635 (96.2)	283 (91.9)	.007	0.18	764 (99.5)	157 (98.7)	.609	0.08
Current tobacco smoking	429 (21.5)	314 (36.0)	<.001	0.32	185 (32.9)	167 (41.2)	.009	0.17	173 (26.2)	120 (39.0)	<.001	0.28	71 (9.2)	27 (17.0)	.006	0.23
Current vaping/e-cigarette use	92 (4.6)	59 (6.8)	.023	0.09	51 (9.1)	41 (10.1)	.655	0.04	33 (5.0)	15 (4.9)	1	0.01	8 (1.0)	3 (1.9)	.622	0.07
Comorbid health conditionsd
Obstructive lung disease	1002 (50.3)	419 (48.1)	.28	0.05	276 (49.0)	186 (45.9)	.375	0.06	384 (58.2)	148 (48.1)	.004	0.20	342 (44.5)	85 (53.5)	.049	0.18
Nonobstructive lung disease	866 (43.5)	314 (36.0)	<.001	0.15	229 (40.7)	115 (28.4)	<.001	0.26	296 (44.8)	123 (39.9)	.171	0.10	341 (44.4)	76 (47.8)	.486	0.07
Heart disease	1291 (64.8)	460 (52.8)	<.001	0.25	291 (51.7)	164 (40.5)	.001	0.23	436 (66.1)	171 (55.5)	.002	0.22	564 (73.4)	125 (78.6)	.207	0.12
Congestive heart failure	674 (33.9)	227 (26.0)	<.001	0.17	130 (23.1)	72 (17.8)	.054	0.13	223 (33.8)	82 (26.6)	.031	0.16	321 (41.8)	73 (45.9)	.386	0.08
Neurologic disorder	722 (36.3)	226 (25.9)	<.001	0.23	191 (33.9)	85 (21.0)	<.001	0.29	238 (36.1)	82 (26.6)	.005	0.21	293 (38.2)	59 (37.1)	.875	0.02
Renal disorders	846 (42.5)	305 (35.0)	<.001	0.16	219 (38.9)	115 (28.4)	.001	0.22	281 (42.6)	110 (35.7)	.05	0.14	346 (45.1)	80 (50.3)	.261	0.11
Malignancy	493 (24.8)	170 (19.5)	.002	0.13	68 (12.1)	40 (9.9)	.332	0.07	162 (24.5)	77 (25.0)	.942	0.01	263 (34.2)	53 (33.3)	.898	0.02
Hematologic disorders	225 (11.3)	88 (10.1)	.374	0.04	82 (14.6)	36 (8.9)	.01	0.18	74 (11.2)	39 (12.7)	.584	0.05	69 (9.0)	13 (8.2)	.862	0.03
Metabolic disorders	1233 (61.9)	392 (45.0)	<.001	0.35	246 (43.7)	135 (33.3)	.001	0.21	438 (66.4)	154 (50.0)	<.001	0.34	549 (71.5)	103 (64.8)	.112	0.14
Diabetes/obesity/endocrine disorders	1539 (77.3)	562 (64.4)	<.001	0.29	385 (68.4)	227 (56.0)	<.001	0.26	532 (80.6)	208 (67.5)	<.001	0.30	622 (81.0)	127 (79.9)	.83	0.03
Health status indicators
Number of ACIP high-risk conditions, mean (SD)	5.66 (3.27)	4.54 (3.13)	<.001	0.35	5.08 (3.35)	3.73 (2.95)	<.001	0.43	5.98 (3.18)	4.88 (3.17)	<.001	0.35	5.81 (3.22)	5.92 (2.90)	.692	0.04
Charlson Comorbidity Index, mean (SD)	3.23 (3.02)	2.69 (2.90)	<.001	0.18	2.69 (2.78)	2.13 (2.63)	.001	0.21	3.37 (2.96)	2.90 (2.95)	.021	0.16	3.50 (3.19)	3.72 (3.13)	.412	0.07
Frailty score (0–5), mean (SD)	2.01 (1.45)	1.86 (1.52)	.016	0.10	1.81 (1.38)	1.64 (1.44)	.064	0.12	2.07 (1.48)	1.93 (1.51)	.191	0.09	2.10 (1.45)	2.30 (1.61)	.122	0.13
Home oxygen use	600 (30.4)	162 (18.7)	<.001	0.28	150 (26.7)	59 (14.7)	<.001	0.30	228 (35.1)	68 (22.1)	<.001	0.29	222 (29.1)	35 (22.0)	.086	0.16
≥1 hospitalization in prior year (self-report)	1259 (63.2)	477 (54.7)	<.001	0.17	393 (69.8)	223 (55.1)	<.001	0.31	437 (66.2)	160 (51.9)	<.001	0.29	429 (55.9)	94 (59.1)	.505	0.07
Self-reported chemotherapy/radiation, immunosuppressive medications	824 (41.4)	290 (33.3)	<.001	0.17	242 (43.0)	113 (27.9)	<.001	0.32	305 (46.2)	116 (37.7)	.015	0.17	277 (36.1)	61 (38.4)	.648	0.05
Immunosuppression by ICD-10 code and/or self-report	1043 (52.4)	375 (43.0)	<.001	0.19	318 (56.5)	157 (38.8)	<.001	0.36	380 (57.6)	141 (45.8)	.001	0.24	345 (44.9)	77 (48.4)	.47	0.07
Influenza vaccination habite
Never/rarely	106 (5.3)	543 (62.3)	<.001	1.69	45 (8.0)	246 (60.7)	<.001	1.54	41 (6.2)	198 (64.3)	<.001	1.61	20 (2.6)	99 (62.3)	<.001	1.84
Sometimes	91 (4.6)	88 (10.1)			38 (6.7)	54 (13.3)			36 (5.5)	22 (7.1)			17 (2.2)	12 (7.5)
Always/almost always	1794 (90.1)	241 (27.6)			480 (85.3)	105 (25.9)			583 (88.3)	88 (28.6)			731 (95.2)	48 (30.2)
Vaccination habit score (0–2), mean (SD)	1.85 (0.49)	0.65 (0.88)	<.001	1.68	1.77	0.65	<.001	1.52	1.82	0.64	<.001	1.61	1.93	0.68	<.001	1.81
Admission characteristics
Days from illness onset to admission, mean (SD)	2.87 (2.35)	2.93 (2.37)	.513	0.03	2.96 (2.35)	2.90 (2.35)	.656	0.03	2.98 (2.37)	2.97 (2.41)	.97	0.00	2.71 (2.33)	2.95 (2.37)	.236	0.10
Admitted to tertiary care hospital	1055 (53.0)	501 (57.5)	.03	0.09	363 (64.5)	243 (60.0)	.176	0.09	366 (55.5)	179 (58.1)	.479	0.05	326 (42.4)	79 (49.7)	.113	0.15
Presented with influenza-like illnessf	852 (42.8)	416 (47.7)	.017	0.10	280 (49.7)	234 (57.8)	.016	0.16	280 (42.4)	135 (43.8)	.732	0.03	292 (38.0)	47 (29.6)	.054	0.18
Admitted to intensive care unit	219 (11.0)	119 (13.7)	.049	0.08	62 (11.0)	54 (13.3)	.319	0.07	73 (11.1)	42 (13.7)	.287	0.08	84 (11.0)	23 (14.6)	.251	0.11
Length of hospital stay, d, mean (SD)	4.76 (4.30)	4.73 (4.70)	.867	0.01	4.65 (4.77)	4.60 (5.43)	.867	0.01	4.71 (4.40)	4.61 (3.65)	.715	0.03	4.88 (3.83)	5.30 (4.51)	.218	0.10
Died in hospital	26 (1.3)	7 (0.8)	.332	0.05	4 (0.7)	1 (0.2)	.591	0.07	7 (1.1)	4 (1.3)	1	0.02	15 (2.0)	2 (1.3)	.787	0.06

Data are presented as No. (%) unless otherwise indicated.

Abbreviations: ACIP, Advisory Committee on Immunization Practices; ICD-10, International Classification of Diseases, 10th Edition; SD, standard deviation.

^aVaccination defined as documented and/or plausible self-report.

^b  P value for test of difference across case and control groups based on χ ² statistic for categorical variables, t test for difference of means of normally distributed continuous variables, and nonparametric log-rank test for nonnormally distributed continuous variables.

^cStandardized absolute mean differences >0.10 shown in bold; standardized differences ≤0.10 are generally considered negligible differences between groups.

^dDerived from ICD-10 codes associated with inpatient and outpatient medical encounters in the year prior to admission.

^eDefined by patient self-report during enrollment interview.

^fInfluenza-like illness defined as self-reported subjective fever/feverishness plus cough or sore throat.

Vaccine Effectiveness Against Influenza A(H3N2)–Associated Hospitalizations

One-hundred ninety four of 247 (79%) influenza A(H3N2) cases were vaccinated compared with 1455 of 2057 (71%) controls (Table 4). After balancing baseline characteristics of vaccinated and unvaccinated participants by age, number of high-risk conditions, hospitalizations within the past year, and influenza vaccination habit and adjusting for time-varying characteristics (calendar time of illness onset, days between onset, and specimen collection) and remaining unbalanced baseline characteristics (age, black race, frailty, current tobacco smoking, and history of metabolic/endocrine disorders), adjusted VE against influenza A(H3N2) virus–associated hospitalization was −2% (95% CI, −65% to 37%).

Table 4.

Estimated Vaccine Effectiveness Against Influenza A(H3N2)-Associated Hospitalization, US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018-2019

Characteristic	No.	Vaccinated Influenza-Positive Cases/ Total Influenza-Positive Cases (%)	Vaccinated Influenza-Negative Controls/ Total Influenza-Negative Controls (%)	Unadjusted VE (95% CI)	Adjusted VE (95% CI)
Alla,b	2304	194/247 (79)	1455/2057 (71)	–51 (–108 to –10)	–2 (–65 to 37)
By age groupc
18 to ≤56 yd,e	771	50/66 (76)	420/705 (60)	–112 (–279 to –18)	–130 (–374 to –27)
>56 to ≤69 yf,g	769	54/80 (68)	483/689 (70)	11 (–45 to 46)	42 (–23 to 72)
>69 yh,i	764	90/101 (89)	552/663 (83)	–65 (–218 to 15)	–1 (–182 to 64)
By influenza vaccination historyj
Never/rarelyk,l	490	7/38 (18)	82/452 (18)	–4 (–132 to 60)	–1 (–191 to 65)
Always/almost alwaysm,n	1665	182/200 (91)	1301/1465 (89)	–27 (–118 to 22)	–16 (–105 to 34)

Vaccination was defined as affirmative self-report with known or approximate date and location of vaccination provided and/or documented evidence of vaccination.

Abbreviations: CI, confidence interval; VE, vaccine effectiveness.

^aVE adjusted for site, tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection (<5 vs ≥5 days), age group, black race, frailty, current tobacco smoking (yes/no), diabetes/obesity/endocrine disorders, and influenza vaccination habit.

^bData balanced on vaccination habit, age, number of high-risk conditions, home oxygen use, and ≥1 hospitalization in past year.

^cDefined as tertiles of age in years in full dataset.

^dVE adjusted for site, age, tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection (<5 vs ≥5 days), black race, frailty, diabetes/obesity/endocrine disorders, and obstructive lung disorders.

^eData balanced on vaccination habit, number of high-risk conditions, home oxygen use, ≥1 hospitalization in past year, and renal disorders.

^fVE adjusted for age, tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection (<5 vs ≥5 days), black race, frailty, and malignancy.

^gData balanced on vaccination habit, number of high-risk conditions, metabolic disorder, home oxygen use, ≥1 hospitalization in past year, immunosuppression, and site.

^hVE adjusted for age (continuous) tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection, community dwelling, renal disorders, neurologic disorders, and home oxygen use.

ⁱData balanced on vaccination habit, black race, current tobacco smoking (yes/no), site, and frailty.

^jInfluenza vaccination history defined by patient self-report during enrollment interview; patients reporting vaccination history “sometimes” are excluded (n = 149).

^kVE adjusted for tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection, and frailty.

^lData balanced on site, number of high-risk conditions, home oxygen use, ≥1 hospitalization in past year, black race, and immunosuppressive conditions.

^mVE adjusted for tertile of illness onset date relative to peak of influenza A(H3N2) cases, days between onset and specimen collection, age (tertile), home oxygen use, site, number of high-risk conditions, ≥1 hospitalization in past year, black race, and immunosuppressive conditions.

ⁿData balanced on age, black race, current tobacco smoking (yes/no), metabolic disorders and endocrine disorders, heart disease, and obstructive lung disorders.

We observed variation in estimated VE by age group (P = .04 for interaction), with VE against influenza A(H3N2) of −130% (95% CI, −374% to −27%) among participants in the youngest age tertile (18 to ≤56 years), VE of 42% (95% CI, −23% to 72%) among those >56 to ≤69 years of age, and VE of −1% (95% CI, −182% to 64%) among those >69 years of age. VE as a function of continuous age in years is shown in Figure 1, with point estimates negative between ages 18 and 40 years and at >60 years of age (Figure 1). Among participants who self-reported never or rarely receiving influenza vaccine, estimated VE against influenza A(H3N2) was −1% (95% CI, −191% to 65%). Among those who reported always/almost always receiving influenza vaccine, estimated VE was −16% (95% CI, −105% to 34%).

Figure 1.

Adjusted vaccine effectiveness against influenza A(H3N2)–associated hospitalizations among adults by participant age (years), US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018–2019 (n = 2304). Abbreviations: CI, confidence interval; VE, vaccine effectiveness.

Fully adjusted VE against influenza A(H3N2)–associated hospitalization was –15% (95% CI, –79% to 25%) when estimated using baseline characteristics as model covariates instead of IPTW-balanced data (Supplementary Table 3). The largest confounders of the association between vaccination status and influenza risk were age, study site, and influenza vaccination habit. Regression models using IPTW-balanced datasets demonstrated better goodness of fit compared to analogous models using the same baseline characteristics as model covariates.

Bias Indicator Analysis for VE Against Influenza A(H3N2)

The bias indicator analysis included 1386 participants, including 459 (33%) participants positive for ≥1 noninfluenza respiratory viral pathogen. Of these “cases,” 352 (77%) were vaccinated. Of 927 participants negative for all respiratory viral pathogens, 647 (70%) were vaccinated. Most characteristics of individuals included in the bias indicator analysis were similar to those in primary analysis, except that individuals in the bias indicator analysis were more likely to be from the central Texas site and of black race (Supplementary Table 5). After balancing baseline characteristics and adjusting for covariates as in the primary analysis, VE against hospitalization associated with noninfluenza respiratory viral pathogens was −47% (95% CI, −115% to 0). Bias indicator analysis results were similar when stratified by age tertile, with the exception of the oldest age tertile (>69 years of age), for which the bias analysis indicated a null result.

Vaccine Effectiveness Against Influenza A(H1N1)pdm09–Associated Hospitalizations

One hundred eighteen of 225 (52%) influenza A(H1N1)pdm09 cases were vaccinated compared with 1568 of 2222 (71%) controls (Table 5). After balancing baseline characteristics by number of high-risk conditions, hospitalizations within the past year, home oxygen use, current tobacco use, history/presence of metabolic disorders (including diabetes), and influenza vaccination habit and adjusting for time-varying characteristics (calendar time of illness onset, days between onset, and specimen collection) and remaining unbalanced baseline characteristics (black race, frailty), adjusted VE against influenza A(H1N1)pdm09–associated hospitalization was 51% (95% CI, 25%–68%). We observed lower VE among participants in the oldest age group, with VE against influenza A(H1N1)pdm09 of 63% (95% CI, 22%–82%) among participants in the youngest age tertile (18 to ≤56 years), VE of 44% (95% CI, −7% to 71%) among those >56 to ≤69 years, and VE of 16% (95% CI, −147% to 71) among those >69 years of age; however, this interaction was not statistically significant. Among participants who self-reported never or rarely receiving influenza vaccine, estimated VE against influenza A(H1N1) was 58% (95% CI, −7% to 84%). Among those who reported always/almost always receiving influenza vaccine, estimated VE was 40% (95% CI, −4% to 65%). When estimated using baseline characteristics as model covariates instead of IPTW-balanced data, fully adjusted VE against influenza A(H1N1)pdm09–associated hospitalization was 50% (95% CI, 25%–67%) (Supplementary Table 4).

Table 5.

Vaccine Effectiveness Estimates Against Influenza A(H1N1)pdm09, US Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN), 2018-2019

Characteristic	No.	Vaccinated Cases/Total Cases (%)	Vaccinated Influenza-Negative Controls/ Total Influenza-Negative Controls (%)	Unadjusted VE (95% CI)	Adjusted VE (95% CI)
Alla,b	2447	118/225 (52)	1568/2222 (71)	54 (39–65)	51 (25–68)
By age groupc
18 to ≤56 yd,e	832	32/87 (37)	443/745 (59)	60 (37–75)	63 (22–82)
>56 to ≤69 yf,g	831	40/79 (51)	527/752 (70)	56 (30–73)	44 (–7 to 71)
>69 yh,i	784	46/59 (78)	598/725 (82)	25 (–48 to 60)	16 (–147 to 71)
By influenza vaccination historyj
Never/rarelyk,l	568	7/84 (8)	87/484 (18)	58 (11–83)	58 (–7 to 84)
Always/almost alwaysm,n	1719	107/131 (82)	1405/1588 (88)	42 (6–63)	40 (–4 to 65)

Vaccination was defined as affirmative self-report with known or approximate date and location of vaccination provided and/or documented evidence of vaccination.

Abbreviations: CI, confidence interval; VE, vaccine effectiveness.

^aVE adjusted for age (tertiles), site, tertile of illness onset date relative to peak of influenza A(H1N1) cases, days between onset and specimen collection (<5 vs ≥5 days), black race, and frailty.

^bData balanced on vaccination habit, number of high-risk conditions, home oxygen use, ≥1 hospitalization in past year, current tobacco smoking (yes/no), and metabolic disorders.

^cDefined as tertiles of age in years in full dataset.

^dVE adjusted for site, tertile of illness onset date relative to peak of influenza A(H1N1)pdm09 cases, days between onset and specimen collection (<5 vs ≥5 days), and obstructive lung disorders.

^eData balanced on vaccination habit, number of high-risk conditions, home oxygen use, and ≥1 hospitalization in past year.

^fVE adjusted for tertile of illness onset date relative to peak of influenza A(H1N1)pdm09 cases and days between onset and specimen collection (<5 vs ≥5 days).

^gData balanced on vaccination habit, number of high-risk conditions, home oxygen use, ≥1 hospitalization in past year, heart disease, obstructive lung disorders, and site.

^hVE adjusted for tertile of illness onset date relative to peak of influenza A(H1N1)pdm09 cases, days between onset and specimen collection, and home oxygen use.

ⁱData balanced on vaccination habit, current tobacco smoking (yes/no), and obstructive lung disorders.

^jInfluenza vaccination history defined by patient self-report during enrollment interview; patients reporting vaccination history “sometimes” are excluded (n = 160).

^kVE adjusted for tertile of illness onset date relative to peak of influenza A(H1N1)pdm09 cases and days between onset and specimen collection.

^lData balanced on number of high-risk conditions, home oxygen use, ≥1 hospitalization in past year, heart disease, neurologic disorders, congestive heart failure, and site.

^mVE adjusted for tertile of illness onset date relative to peak of influenza A(H1N1)pdm09 cases, days between onset and specimen collection, and sex.

ⁿData balanced on age, number of high-risk conditions, home oxygen use, heart disease, current tobacco use (yes/no), and site.

Bias Indicator Analysis for VE Against Influenza A(H1N1)pdm09

The bias indicator analysis included 1586 participants, including 526 (33%) participants positive for ≥1 noninfluenza respiratory viral pathogen. Of these, 402 (76%) were vaccinated. Of 1060 participants negative for all respiratory viral pathogens, 741 (70%) were vaccinated. After balancing and covariate adjustment as in the primary analysis, VE against hospitalization associated with noninfluenza respiratory viral pathogens was −46% (95% CI, −108% to −2%).

Discussion

During the 2018–2019 US influenza season, we found that vaccination halved the risk of influenza A(H1N1)pdm09–associated hospitalizations. Vaccination conferred no protection against influenza A(H3N2)–associated hospitalizations among adults, most of which were attributable to infection with influenza A(H3N2) viruses of genetic group 3C.3a. This finding is consistent with laboratory analyses indicating antigenic difference between the 2018–2019 A(H3N2) vaccine virus belonging to genetic group 3C.2a1 and circulating 3C.3a viruses. Our observed VE against influenza A(H3N2) of −2% (95% CI, −65% to 37%) is consistent with that observed by the US Influenza Vaccine Effectiveness Network, which found VE against outpatient influenza illness of −2% (95% CI, −24% to 15%) among adults aged ≥18 years [1] (personal communication, J. Chung, July 2020).

Estimated VE differed significantly by age group, and we observed a statistically significant VE of −130% against influenza A(H3N2) among the youngest age group (18 to ≤56 years); VE was most markedly negative for participants between the ages of 18 and 40 years. Our findings are broadly similar to those reported from studies conducted in Canada and Europe [2–4], which reported negative VE against influenza A(H3N2) of genetic group 3C.3a. A potential explanation for this finding is provided by Skowronski and colleagues’ hypothesis that childhood priming following the 1968 influenza A(H3N2) pandemic provided immunity among imprinted individuals, protecting them as adults from subgroup 3C.3a viruses that were similar in 2018–2019 to the imprinting childhood viruses, and that receipt of antigenically mismatched vaccine negatively interfered with this immunity, leaving vaccinees at greater risk of 3C.3a infection [2]. However, the immunologic processes underlying this interference are unknown, and a limitation of our study is the lack of serologic markers with which to examine correlation of antibody titers with disease risk or protection.

An alternative explanation is that uncontrolled confounding, selection bias, or chance was responsible for the observed elevation in risk among adult vaccinees, which was most pronounced among the youngest age group. Our relatively small sample of 66 young adult cases suffered as much baseline comorbidity as older cases, alluding to their medical complexity and suggesting they likely had characteristics associated with baseline differences in vaccination status that may confound VE estimates. A strength of our study is controlling for this confounding by systematically balancing baseline participant characteristics within each subgroup. Despite this, we observed that influenza vaccinees had higher risk of hospitalization with noninfluenza respiratory pathogens in our bias indicator analyses, suggesting that residual confounding may have affected our primary VE results, biasing them lower.

VE against influenza A(H3N2) in both the primary analysis and bias indicator analysis was negative in the youngest age group and essentially null in the oldest age group. This may suggest that bias influencing the primary VE results was more pronounced among the younger age group and implying greater confidence in the results for the older age group, for which a finding of null VE was not unexpected given that most circulating viruses were antigenically drifted from the vaccine virus. A potential source of this bias could be a selection bias that allowed preferential recruitment of vaccinated patients with respiratory viral infections in the younger age groups. To manifest as the bias we observed, preferential recruitment would have occurred among vaccinated individuals relatively more than unvaccinated individuals, in the younger age groups relative to the older groups. It is possible that the threshold for hospitalization may have been lower among vaccinated individuals, who tended to have greater underlying comorbidity. If influenza and other respiratory virus infection was also associated with greater likelihood of hospitalization, it is possible, although speculative, that selection bias related to threshold for hospital admission contributed to the bias we observed. For example, immunosuppressed patients account for about a third of all HAIVEN influenza-positive participants and rates of vaccination tend to be higher in this population. Because fever is a common presenting sign of influenza, febrile immunosuppressed patients with influenza, who also have high rates of vaccination, may have a lower threshold for admission [23]. Additional research is needed to further examine the potential for unmeasured confounding and selection bias, and methodologic development of techniques for controlling for unmeasured confounding in the context of influenza VE studies is needed [24]. In particular, the use of a negative control outcome as a bias indicator method and the attributes of an adequate negative control outcome should be further investigated [25]. The utility of positivity for another respiratory virus as a negative control outcome could be limited if influenza infection provided nonspecific protection against noninfluenza respiratory viruses, which has been observed for some viruses [26–28]. Nonetheless, the bias indicator analysis raises concerns that confounding could not be adequately controlled in this study. Simulation and sensitivity analyses [29, 30] to examine the potential magnitude of bias could also be useful next steps.

Although we do not know with certainty the cause of our observed finding of negative VE among the youngest age group, it is worth noting that negative VE results for one virus are unlikely to influence current recommendations for yearly vaccination since vaccination is protective against other influenza viruses and the negative VE result is confined to a specific subgroup and possibly due to measurement bias.

Our results highlight the benefit conferred by currently available influenza vaccines for prevention of influenza A(H1N1)pdm09–associated hospitalizations and provide additional evidence that vaccine-induced protection against influenza A(H3N2) is consistently lower [31]. Biological mechanisms related to early childhood imprinting may potentially contribute to this diminished VE [32], especially among older individuals who are repeatedly vaccinated [33]. While current vaccines can prevent substantial morbidity and mortality from influenza, particularly for influenza A(H1N1)pdm09 [34, 35], better vaccines are urgently needed for prevention of influenza A(H3N2)–associated illness and hospitalization.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

HAIVEN Study Investigators. Shoshona Le, Juliana DaSilva, Lisa M. Keong, Thomas J. Stark, Joshua G. Petrie, Lois E. Lamerato, Anurag Malani, Adam Lauring, Ryan E. Malosh, Dayna Wyatt, Yuwei Zhu, Zhouwen Liu, Stephanie Longmire, Kellie Graves, Emily Sedillo, Alina Simion, Karen Speer, Bethany Alicie, Briana Krantz, Donna Carillo, Laura Adams, Amelia Drennan, Jan Orga, Lynn Peterson, Natasha Halasa, Rendi McHenry, Claudia Guevara Pulido, Kempapura Murthy, Kelsey Bounds, Tnelda Zunie, Lydia Clipper, Shekhar Ghamande, Heath White, Chandni Raiyani, Kevin Chang, Arundhati Rao, Manohar Mutnal, Alejandro Arroliga, Mary Patricia Nowalk, GK Balasubramani, Heather Eng, Sean G. Saul, Kailey Hughes, Nicole Wheeler, Lori Stiefel, Mohamed Yassin, and John V. Williams.

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the United States Centers for Disease Control and Prevention (CDC).

Financial support. This work was supported by the CDC (cooperative agreement number IP15-002); the National Institutes of Health Clinical and Translational Science Award (CTSA) program (grant number UL1 TR001857); and the National Center for Advancing Translational Sciences at Vanderbilt University Medical Center (CTSA award number UL1 TR002243).

Potential conflicts of interest. J. M. F. reports nonfinancial support from the Institute for Influenza Epidemiology (funded in part by Sanofi Pasteur), outside the submitted work. M. G. reports grants from the CDC during the conduct of the study and CDC-Abt Associates, outside the submitted work. E. T. M. reports personal fees from Pfizer and grants from Merck, outside the submitted work. A. S. M. reports personal fees from Sanofi Pasteur and Seqirus, outside the submitted work. F. P. S. reports grants from the CDC during the conduct of the study and grants from Shire, Qiagen, and Novartis, outside the submitted work. R. K. Z. reports grants from the CDC during the conduct of the study, and grants from Merck & Co and Sanofi Pasteur, outside the submitted work. D. B. M. reports personal fees from Seqirus, Pfizer, and Sanofi Pasteur, and grants from Pfizer, outside the submitted work. All other authors report no potential conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.