Abstract
Background:
Annual influenza vaccination has been recommended for persons with high-risk conditions since the 1960s. However, few estimates of influenza vaccine effectiveness (VE) for persons with high-risk conditions are available.
Methods:
Data from the U.S. Influenza Vaccine Effectiveness Network from 2012‒2016 were analyzed to compare VE of standard-dose inactivated vaccines against medically-attended influenza among patients aged ≥6 months with and without high-risk medical conditions. Patients with acute respiratory illness were tested for influenza by RT-PCR. Presence of high-risk conditions and vaccination status were obtained from medical records. VE by influenza virus type/subtype and age group was calculated for patients with and without high-risk conditions using the test-negative design. Interaction terms were used to test for differences in VE by high-risk conditions.
Results:
Overall, 9,643 (38%) of 25,369 patients enrolled during four influenza seasons had high-risk conditions; 2,213 (23%) tested positive for influenza infection. For all ages, VE against any influenza was lower among patients with high-risk conditions (41%, 95% CI: 35%‒47%) than those without (48%, 95% CI: 43%‒52%; P-for-interaction = 0.02). For children aged <18 years, VE against any influenza was 51% (95% CI: 39%‒61%) and 52% (95% CI: 39%‒61%) among those with and without high-risk conditions, respectively (P-for-interaction = 0.54). For adults aged ≥18 years, VE against any influenza was 38% (95% CI: 30%‒45%) and 44% (95% CI: 38%‒50%) among those with and without high-risk conditions, respectively (P-for-interaction = 0.21). For both children aged <18 and adults aged ≥18 years, VEs against illness related to influenza A(H3N2), A(H1N1)pdm09, and influenza B virus infection were similar among those with and without high-risk conditions.
Conclusions:
Influenza vaccination provided protection against medically-attended influenza among patients with high-risk conditions, at levels approaching those observed among patients without high-risk conditions. Results from our analysis support recommendations of annual vaccination for patients with high-risk conditions.
Keywords: influenza, vaccine, vaccine effectiveness, high-risk medical conditions, acute respiratory illness
INTRODUCTION
Since the 1960s, annual influenza vaccination has been recommended for persons with underlying medical conditions that are associated with an increased risk of influenza-related complications and death [1]. In the 1960 Surgeon General’s report, these conditions included rheumatic heart disease and other cardiovascular diseases, chronic bronchopulmonary disease (including chronic asthma), diabetes mellitus and Addison’s disease [1]. Since that time, recommendations for influenza vaccination have been expanded to all persons aged >6 months regardless of underlying medical conditions. The US Advisory Committee on Immunization Practices (ACIP) has continued to note that vaccination “is especially important” for persons with high-risk medical conditions (hereafter referred to high-risk conditions)[2]. The list of high-risk conditions associated with increased risk of influenza-related complications has expanded to include neurological conditions, immunosuppressive conditions, blood disorders, kidney disorders, liver disorders, metabolic disorders, extreme obesity, and long-term aspirin therapy among children aged <19 years old [2, 3]. However, relatively few studies have evaluated the protection provided by influenza vaccination among persons with high-risk conditions.
Observational studies are widely used to evaluate influenza vaccine effectiveness (VE) for the prevention of medically-attended illness and often include a substantial proportion of patients with high-risk conditions. Many observational studies employ a test-negative design, in which symptomatic patients are tested for influenza and VE is estimated by contrasting the odds of influenza among vaccinated and unvaccinated patients [4, 5]. Test-negative design has the strength in controlling health-seeking behavior comparing with traditional case-control or cohort methods. However, to obtain unbiased VE estimates, variables including calender time when influenza is circulating and high-risk conditions that may affect VE must be considered in analysis [4, 5]; the test-negative design may also be used to estimate VE among groups of patients with high-risk conditions given adequate sample size. In the United States, the Influenza Vaccine Effectiveness (Flu VE) Network routinely collects information on patients’ high-risk conditions in annual VE studies [6–10]. We analyzed data from four influenza seasons to evaluate VE among patients with high-risk conditions, and assess differences between VE in patients with and without high-risk conditions.
Materials and methods
Patients
Methods of the Flu VE Network have been described previously [6–10]. In this analysis, we included patients aged ≥6 months with acute respiratory illness (ARI) with cough and symptom onset ≤7 days prior to enrollment, who were enrolled during four influenza seasons from 2012‒13 through 2015‒16 at outpatient facilities associated with participating institutions in Michigan, Pennsylvania, Texas, Washington and Wisconsin. For each season, we excluded influenza-negative patients enrolled before or after the date of symptom onset of the first and last confirmed influenza case, respectively, at each site. Patient demographics were obtained through interview at enrollment. Influenza virus infection was determined by real-time reverse transcription polymerase chain reaction (RT-PCR) testing of combined nasal and throat swab specimens (for patients aged ≥2 years) or nasal swab specimens only (for patients aged <2 years). Influenza virus type and subtype were determined by RT-PCR for influenza-positive specimens; patients with inconclusive RT-PCR results were excluded.
High-risk conditions
Presence of high-risk conditions was defined as documentation in medical records of ≥1 medical encounter including outpatient visits and inpatient stays during the 12 months before enrollment associated with an International Classification of Diseases, 9th Edition (ICD-9) and 10th Edition (ICD-10), Clinical Modification [11, 12] diagnostic code corresponding to a high-risk condition identified by ACIP (Supplementary Table 1) [2]. High-risk categories included in this analysis were asthma, cerebrovascular diseases, chronic obstructive pulmonary diseases (COPD) and other lung diseases, diabetes, heart diseases, hematologic conditions/blood disorders, liver diseases, neurologic conditions, renal diseases, and immunosuppressive conditions (including rheumatoid arthritis, immune system disorders or immunodeficiency, antineoplastic chemotherapy, HIV, organ transplants). High-risk categories were not mutually exclusive such that individual patients could contribute to more than one category. Patients with no high-risk ICD-9 or ICD-10 codes in medical records during the 12-month period were classified as having no high-risk conditions; patients without 12 months of medical records prior to encounter were excluded.
Vaccination status
Vaccination status was determined at all sites based on electronic immunization records, including medical records, employee health records and state or local immunization registries. Vaccinated patients were those with documented receipt of current season influenza vaccine at least 14 days before illness onset. We included vaccinated patients who received standard-dose inactivated influenza vaccines only; patients who received live-attenuated influenza vaccines, high dose inactivated or adjuvanted vaccine were excluded. In addition, at four sites (excluding Wisconsin), patients aged ≥9 years without documented vaccination who reported plausible location of vaccination ≥14 days prior to illness onset were classified as vaccinated. We excluded patients vaccinated 0‒13 days prior to illness onset and children aged 6 months‒8 years who received only one of two recommended doses of influenza vaccine in the current season [13–16].
Statistical analyses
For patients with and without high-risk conditions, we examined factors associated with influenza positivity and vaccination status using X2 tests for differences in proportions, and estimated VE using a test-negative design [4, 17], where VE = 100 × (1 - adjusted odds ratio [aOR]) from logistic regression models comparing odds of influenza among vaccinated versus unvaccinated patients, with 95% confidence intervals (CI) estimated from odds ratios. Logistic regression models included a priori study site, influenza season, age (in years as linear tail-restricted cubic-spline function with 4 percentile knots), calendar time (four-week intervals), and days from illness onset to enrollment (0–2 days, 3‒4 days, 5‒7 days). Additional variables (sex, race/ethnicity, self-reported general health status [excellent, very good, good, fair, poor], number of children aged <12 years in household [0, 1, ≥1] and self/household exposure to tobacco smoke) were individually assessed but not retained in the final model, as their inclusion did not result in a ≥5% change in VE [7]. For age-stratified estimates (<18 years and ≥18 years), age was modelled as a continuous variable in years [18]. Adjusted ORs were not estimated when the total number of patients in the stratum (denominator) was <50. VE estimates were considered statistically significant if the 95% CI excluded zero.
We estimated VE among patients with and without any high-risk conditions and separately by age group, high-risk category, and infecting influenza virus type. For influenza A(H3N2), we conducted separate analyses of VE against A(H3N2)-related illness among adults aged ≥18 years for the 2014–15 season, when antigenically-drifted A(H3N2) viruses predominated, and for the other three seasons combined [6–10]. To assess the difference of influenza VE among patients with only one high-risk condition with that among patients with multiple high-risk conditions, we repeated the primary analysis for individuals with a single high-risk condition (i.e., individuals with asthma only). We also compared VE among patients with high-risk conditions associated with an inpatient stay in the preceding 12 months assuming that hospital stays were associated with more severe underlying illness. We included interaction terms in logistic regression models to test for statistically significant differences in VE among patients with and without a specific high-risk condition. All reported tests were 2-sided and P values <0.05 indicated statistical significance. Statistical analyses were conducted using SAS for Windows (version 9.3, Cary, NC).
RESULTS
A total of 25,369 patients enrolled in the US Flu VE study between December 12, 2012 and April 14, 2016 were included in analyses; reasons for exclusion from this analysis (n = 4674) are shown in Figure 1. Among those included, 9,643 (38%) had one or more high-risk conditions while 15,726 (62%) had no medical encounters associated with a high-risk condition in the preceding 12 months. Prevalence of any high-risk condition ranged from 25% among children aged 6 months to 17 years, to 76% among patients aged ≥65 years. Among children aged <18 years with high-risk conditions, the most common conditions were asthma (82%), heart diseases (8%) and neurologic conditions (5%). Among adults aged ≥18 years with high-risk conditions, common conditions included asthma (37%), heart diseases (35%), diabetes (26%), immunosuppressive conditions (17%), cerebrovascular diseases (14%), COPD and other lung diseases (12%), renal diseases (11%), neurologic conditions (9%), liver diseases (5%), and blood disorders (1%). Among adults aged ≥18 years with COPD and other lung diseases, 35% also had asthma.
Figure 1.
Flow diagram of study population
Overall, 6,032 (63%) of 9,643 patients with high-risk conditions had received current season influenza vaccination versus 6,224 (40%) of 15,726 patients without high-risk conditions (P <0.01). Compared to patients without high-risk conditions, patients with high-risk conditions were more likely to be vaccinated across categories by study site, age group, sex, race/ethnicity, influenza season, and interval from onset to enrollment (P <0.01 for all, Table 1). In patients with and without high-risk conditions, proportions of vaccinated were lowest among children and increased with age (χ2 test for trend, P <0.01). Among enrolled patients, those with high-risk conditions were less likely to have influenza (23%) than those without high-risk conditions (27%, P <0.01). This difference remained statistically significant (P <0.05) across categories by age group (except aged <65 years), gender, race/ethnicity, influenza season (except during 2014–2015), and interval from illness onset to enrollment (Supplementary Table 2).
Table 1.
Characteristics of population enrolled in The US Influenza Vaccine Effectiveness Network by presence of high-risk conditions and vaccination status, 2012‒2016
| Patients without high-risk conditions |
Patients with any high-risk conditions |
||||
|---|---|---|---|---|---|
| Characteristics | Total | Number of vaccinated (%) | Total | Number of vaccinated (%) | P-value |
| Site | |||||
| Seattle, WA | 3516 | 1671 (48) | 2763 | 1985 (72) | b |
| Marshfield, WI | 3513 | 1370 (39) | 1911 | 1199 (63) | b |
| Temple, TX | 2982 | 989 (33) | 1741 | 964 (55) | b |
| Ann Arbor and Detroit, MI | 2594 | 1046 (40) | 1520 | 892 (59) | b |
| Pittsburgh, PA | 3121 | 1148 (37) | 1708 | 992 (58) | b |
| Age (years) | |||||
| 0.5–17 | 6565 | 2246 (34) | 2227 | 1070 (48) | b |
| 18–49 | 6235 | 2280 (37) | 2679 | 1440 (54) | b |
| 50–64 | 2220 | 1192 (54) | 2465 | 1660 (67) | b |
| >65 | 706 | 506 (72) | 2272 | 1862 (82) | b |
| Sex | |||||
| Female | 9131 | 3858 (42) | 5748 | 3704 (64) | b |
| Male | 6595 | 2366 (36) | 3895 | 2328 (60) | b |
| Race/ethnicity a | |||||
| White, non-Hispanic | 11777 | 4879 (41) | 7296 | 4778 (66) | b |
| Black, non-Hispanic | 1142 | 317 (28) | 877 | 377 (43) | b |
| Hispanic | 1377 | 560 (41) | 752 | 476 (63) | b |
| Other, non-Hispanic | 1391 | 457 (33) | 692 | 389 (56) | b |
| Influenza season | |||||
| 2012–13 | 3790 | 1420 (38) | 2064 | 1282 (62) | b |
| 2013–14 | 3195 | 1262 (40) | 1862 | 1177 (63) | b |
| 2014–15 | 4974 | 2066 (42) | 3307 | 2137 (65) | b |
| 2015–16 | 3767 | 1476 (39) | 2410 | 1436 (60) | |
| Interval from onset to enrollment | |||||
| 0–2 days | 5163 | 1918 (37) | 2884 | 1739 (60) | b |
| 3–4 days | 6198 | 2437 (39) | 3771 | 2372 (63) | b |
| 5–7 days | 4365 | 1869 (43) | 2988 | 1921 (64) | b |
Race/ethnicity of 65 patients was unknown.
P <0.01
Among patients with high-risk conditions, VE against any influenza was 41% (95% CI: 35%‒47%) for all ages combined, 51% (95% CI: 39%‒61%) for those aged <18 years, and 38% (95% CI: 30%‒45%) for those aged ≥18 years (Table 2). Among patients without high-risk conditions, VEs were 48% (95% CI : 43%‒52%) for all ages combined, and 52% (95% CI: 44%‒58%) and 44% (95% CI: 38%‒50%) for the pediatric and adult age categories. Differences in VE between patients with and without high-risk conditions were significant for all ages combined (P for interaction = 0.02), but not among patients aged <18 or ≥18 years (P for interaction >0.05 for both).
Table 2.
Adjusted influenza vaccine effectiveness against any influenza virus by age group and high-risk condition
| Influenza positive patients | Influenza negative patients | Vaccine effectiveness Adjusted a % (95% CI) | ||
|---|---|---|---|---|
| Characteristics | No. vaccinated/total (%) | No. vaccinated/total (%) | P-value for interaction | |
| All patients | ||||
| Patients without high-risk conditions | 1207/4168 (30) | 5017/11558 (43) | 48 (43,52) | - |
| Patients with any high-risk conditions | 1216/2213 (55) | 4816/7430 (65) | 41 (35, 47) | 0.02 |
| <18 years | ||||
| Patients without high-risk conditions | 368/1705 (22) | 1878/4860 (39) | 52 (44, 58) | - |
| Patients with any high-risk conditions | 166/478 (35) | 904/1749 (52) | 51 (39, 61) | 0.54 |
| Asthma | 147/407 (36) | 726/1420 (51) | 48 (34, 60) | 0.31 |
| Heart diseases | 12/35 (34) | 76/143 (53) | - | - |
| Neurologic conditions | 13/20 (65) | 56/88 (64) | - | - |
| >18 years | ||||
| Patients without high-risk conditions | 839/2463 (34) | 3139/6698 (47) | 44 (38, 50) | |
| Patients with any high-risk conditions | 1050/1735 (61) | 3912/5681 (69) | 38 (30, 45) | 0.21 |
| Asthma | 354/579 (61) | 1440/2184 (66) | 27 (10, 41) | 0.02 |
| Cerebrovascular diseases | 50/81 (62) | 206/249 (83) | 63 (26, 81) | 0.20 |
| COPD and other lung diseases | 111/155 (72) | 546/715 (76) | 21 (−21, 48) | 0.22 |
| Diabetes | 315/478 (66) | 1082/1449 (75) | 46 (30, 58) | 0.86 |
| Heart diseases | 381/586 (65) | 1526/1994 (77) | 47 (35, 58) | 0.41 |
| Immunosuppressive conditions | 195/285 (68) | 748/961 (78) | 46 (26, 60) | 0.80 |
| Liver diseases | 55/96 (57) | 200/268 (75) | 61 (31, 78) | 0.30 |
| Neurologic conditions | 93/155 (60) | 413/540 (76) | 49 (22, 66) | 0.30 |
| Renal diseases | 141/182 (77) | 512/620 (83) | 32 (−6, 57) | 0.39 |
| Blood disorders | 7/14 (50) | 43/66 (65) | - | - |
High-risk categories with fewer than 5 vaccinated influenza or non-influenza patients are not listed.
P-value for interaction for each specific high-risk conditions is calculated by comparing patients with the specific high-risk condition with patients without high-risk conditions.
Adjusted for site, age, influenza season, interval from symptom onset to enrollment, calendar time (4 weeks).
Among children aged <18 years, we observed statistically significant VE against any influenza among children with asthma (48%; 95% CI: 34%−60%) that was similar to VE among children without high-risk conditions (P for interaction = 0.31; Table 2). Among adults aged ≥18 years, we observed statistically significant VE against any influenza among those with asthma, cerebrovascular diseases, diabetes, heart diseases, immunosuppressive conditions, liver diseases, and neurologic conditions (Table 2). When analysis was restricted to adults aged ≥18 years with heart diseases associated with an inpatient stay in the preceding 12 months, VE against any influenza was 53% (95% CI: 24%–71%), similar to VE among adults with any heart diseases (VE 47%; 95% CI: 35%–58%; P for interaction >0.05). Among adults aged ≥18 years with high-risk conditions, only those with asthma (with or without other high-risk conditions) had significantly lower VE (27%; 95% CI: 10%–41%) than patients without high-risk conditions (P for interaction = 0.02); all other tests for differences in VE by high risk status were non-significant (P for interaction for all > 0.05; Table 2). However, when analyses were limited to adults aged ≥18 years with asthma as their only high-risk condition, VE was 39% (95% CI, 7%–46%; P for interaction = 0.05), suggesting that presence of other high-risk conditions may contribute to the lower VE observed among adults with asthma.
By influenza virus type, VE was higher against A(H1N1)pdm09 and influenza B than against influenza A(H3N2) among patients with and without high-risk conditions (Table 3). For children aged <18 years and adults aged ≥18 years, those with high-risk conditions had similar VEs against influenza A (H3N2), influenza A (H1N1)pdm09, and influenza B virus infection compared to patients without high-risk conditions (P for interaction >0.05 for all, Table 3). In addition, adults with diabetes and heart diseases had statistically significant VE against illness due to all influenza virus types/subtypes. Adults with immunosuppressive conditions had significant VE against influenza A (H3N2) and influenza B but not against influenza A (H1N1)pdm09. Adults with asthma had similar VEs against influenza A (H3N2) and (H1N1)pdm09 but lower VE against influenza B virus compared with adults without high-risk conditions (P for interaction <0.05 ). During the antigenically-mismatched A(H3N2) season in 2014‒2015, we observed similar, non-significant VE against A(H3N2) among adults with high-risk conditions (7%; 95% CI: ‒19%–28%) and among those without high-risk conditions (2%; 95% CI: ‒20%–19%) (Supplementary Table 3).
Table 3.
Adjusted influenza vaccine effectiveness against influenza type/subtype by age group and high-risk condition
| Influenza positive patients |
Influenza negative patients |
Vaccine effectiveness Adjusted a % (95% CI) | P-value for interaction | |
|---|---|---|---|---|
| Characteristics | Number of vaccinated/total (%) | Number of vaccinated/total (%) | ||
| Influenza A (H3N2) | ||||
| <18 years | ||||
| Patients without high-risk condition | 195/750 (26) | 1878/4860 (39) | 33(18, 45) | - |
| Patients with any high-risk condition | 95/236(40) | 904/1749(52) | 27 (0, 47) | 0.54 |
| >18 years | ||||
| Patients without high-risk condition | 450/1063 (42) | 3139/6698 (47) | 30 (12, 41) | - |
| Patients with any high-risk condition | 599/889(67) | 3912/5681(69) | 25(10, 37) | 0.93 |
| Asthma | 210/324(65) | 1440/2184(66) | 21 (−6, 41) | 0.69 |
| COPD and other lung diseases | 64/83(71) | 546/715(76) | −1 (−91,46) | 0.39 |
| Diabetes | 192/266(72) | 1082/1449(75) | 39 (3, 49) | 0.87 |
| Heart diseases | 239/331(72) | 1526/1994(77) | 36 (13, 54) | 0.34 |
| Immunosuppressive conditions | 97/132(73) | 748/961(78) | 47(24, 68) | 0.26 |
| Neurologic conditions | 45/64(70) | 413/540(76) | 41 (−18,71) | 0.33 |
| Renal diseases | 92/115(80) | 512/620(83) | 36 (−18, 65) | 0.57 |
| Cerebrovascular diseases | 34/47(72) | 206/249(83) | - | - |
| Liver diseases | 29/46 (63) | 200/268(75) | - | - |
| Influenza A (H1N1) pdm09 | ||||
| <18 years | ||||
| Patients without high-risk conditions | 56/315(18) | 1878/4860 (39) | 69 (57,77) | - |
| Patients with any high-risk conditions | 23/77(30) | 904/1749(52) | 67 (44,80) | 0.52 |
| ≥18 years | ||||
| Patients without high-risk conditions | 227/826(28) | 3139/6698 (47) | 56 (47, 63) | - |
| Patients with any high-risk conditions | 183/340(54) | 3912/5681(69) | 44 (31, 55) | 0.3 |
| Asthma | 77/147(52) | 1440/2184(66) | 43(17, 60) | 0.22 |
| Diabetes | 62/110(56) | 1082/1449(75) | 52 (25, 69) | 0.93 |
| Heart diseases | 75/135(56) | 1526/1994(77) | 49 (25,66) | 0.08 |
| Immunosuppressive conditions | 49/78(63) | 748/961(78) | 34 (−14, 61) | 0.56 |
| Neurologic conditions | 32/57(56) | 413/540(77) | 42 (−11,70) | 0.79 |
| Renal diseases | 92/115(80) | 512/620(83) | 11 (−118, 64) | 0.19 |
| COPD and other lung diseases | 34/49(69) | 546/715(76) | - | - |
| Liver diseases | 16/32(50) | 200/268(75) | - | - |
| Cerebrovascular diseases | 11/18(61) | 206/249(83) | - | - |
| Influenza B | ||||
| <18 years | ||||
| Patients without high-risk conditions | 115/635(18) | 1878/4860(39) | 62 (52, 69) | - |
| Patients with any high-risk conditions | 46/160(29) | 904/1749(52) | 61 (55, 79) | 0.71 |
| ≥18 years | ||||
| Patients without high-risk conditions | 152/533 (29) | 3139/6698 (47) | 59 (50, 67) | - |
| Patients with any high-risk conditions | 183/340(54) | 3912/5681(69) | 52 (39, 62) | 0.58 |
| Asthma | 60/97(62) | 1440/2184(66) | 18(−29, 48) | 0.02 |
| Diabetes | 53/92(58) | 1082/1449(75) | 57 (31, 78) | 0.95 |
| Heart diseases | 59/107(55) | 1526/1994(77) | 64 (45, 76) | 0.23 |
| Immunosuppressive conditions | 44/66(67) | 748/961(78) | 49 (9, 71) | 0.39 |
| COPD and other lung diseases | 12/20(60) | 546/715(76) | - | - |
| Liver diseases | 9/16 (56) | 200/268(75) | - | - |
| Neurologic conditions | 16/32(50) | 413/540(77) | - | - |
| Renal diseases | 20/28(71) | 512/620(83) | - | - |
High-risk categories with fewer than 5 vaccinated influenza or non-influenza patients are not listed.
P-value for interaction for each specific high-risk conditions is calculated by comparing patients with the specific high-risk condition with patients without high-risk conditions.
Adjusted for site, age, influenza season, interval from symptom onset to enrollment, calendar time (4 weeks).
DISCUSSION
In this analysis of data from four influenza seasons, 2012–2013 through 2015–2016, high-risk conditions were common among ambulatory patients seeking care for ARI in the U.S. Influenza vaccination coverage (limited to standard-dose inactivated influenza vaccines) was higher among patients with high-risk conditions than those without high-risk conditions, but coverage in both groups was below the Healthy People 2020 goal of 70% in all age groups except among patients aged ≥65 years[19]. Among patients with and without high-risk conditions, influenza vaccination was associated with statistically significant protection against any influenza virus and for each influenza virus type/subtype. Further, we observed consistent trends suggesting protection against medically-attended influenza for most categories of high-risk conditions, although several estimates did not reach statistical significance. While VE against any influenza was statistically lower among patients with high risk conditions (41%) compared to those without high-risk conditions (48%), this analysis did not suggest large deficits in vaccine-induced protection among people with high-risk conditions. In addition, the differences in age strata were not significantly different by high-risk status, suggesting that protection among patients with high-risk conditions was not substantially lower than that observed among patients without high-risk conditions. Results support the benefit of influenza vaccination among patients with high-risk medical conditions, who are at increased risk of influenza-associated severe complications and death.
Ascertainment of high-risk medical conditions continues to be important for observational studies of influenza VE. In test-negative studies, high-risk conditions are often associated with both likelihood of vaccination and severe influenza (Supplemental Table 2), making it important to include high-risk conditions in VE analyses [5]. In the US Flu VE Network, patients with high-risk conditions were more likely to be vaccinated and to be enrolled with non-influenza ARI than patients without high-risk conditions. In many countries, presence of high-risk conditions may determine eligibility for publicly financed vaccination [5]. It is reassuring that for most comparisons, VE by influenza virus type/subtype did not differ substantially among patients with and without high-risk conditions, with higher VE against influenza A(H1N1)pdm09 and B, and lower VE against A(H3N2), similar to the results from a recent meta-analysis of test-negative studies [20].
One of the strengths of this analysis is the large number of patients with high-risk conditions, allowing some exploration of VE by high-risk category. Although many estimates lacked precision due to small numbers of patients in individual high-risk categories, trends in point estimates and confidence intervals suggested benefit of vaccination in most high-risk categories. Among the more common high-risk categories, we observed statistically significant VE among patients with asthma, adults with cerebrovascular diseases, diabetes, heart diseases, immunosuppressive conditions, liver diseases, and neurologic conditions. Another strength of this study is the use of laboratory-confirmed influenza outcomes to investigate VE among high-risk patients. Studies in children and adults with asthma, and in persons with immunosuppressive conditions have shown statistically significant VE against laboratory-confirmed influenza and severe outcomes [21–25]. In addition, studies that used non laboratory-confirmed outcomes, e.g., influenza-like illness, physician-diagnosed pneumonia, and hospitalization associated with influenza diagnostic codes, have shown significant reductions in these outcomes associated with influenza vaccination among persons with cardiovascular diseases, COPD, diabetes mellitus, liver diseases, and renal diseases [26–33].
Our study is subject to several limitations. First, these findings are limited to ambulatory patients and effectiveness against more severe influenza illness may differ. However, VE against medically-attended influenza in ambulatory settings has been a good proxy for VE against more severe outcomes measured among hospitalized patients [34]. Second, the US Flu VE Network did not enroll patients from specialty clinics where people with more severe conditions might receive most of their care. In addition, although a large proportion of patients enrolled in the US Flu VE Network had high-risk medical conditions, stratification by high-risk category resulted in small numbers in certain high-risk conditions, limiting statistical power to detect differences in VE by high-risk category or disentangle associations between the increasing age and high-risk conditions [35]. Trends support evidence of protection for most high-risk categories but additional evidence is needed for several high-risk categories, including children with neurologic conditions. Further, identification of high-risk conditions from medical records is limited by completeness and accuracy of diagnostic codes, and does not differentiate by level of severity or immunosuppression [36]. Older patients may have had high-risk conditions without corresponding diagnostic codes, making it less likely to observe differences between patients with and without documented high-risk conditions [35]. Broad categories with varying severity of high-risk conditions may also hide differences in VE among patients with specific conditions. Patients with multiple high-risk conditions or those who are more immunocompromised may have lower VE. Finally, analyses were limited to standard-dose inactivated vaccines due to small numbers of people vaccinated with other vaccine types during the study period. Increased use of vaccines such as high-dose vaccine among patients aged ≥65 years and recombinant vaccine may provide an opportunity to evaluate VE for more immunogenic vaccines, but special studies maybe needed to evaluate use of different types of influenza vaccine for patients with high-risk conditions.
CONCLUSIONS
Patients with high-risk conditions continue to be an important population for influenza vaccination. Given their increased risk of severe complications and death associated with influenza infection, influenza vaccination is especially important among patients with high-risk conditions. While more effective vaccines are needed, the current analysis suggests the benefit of vaccination for the prevention of medically-attended influenza does not differ substantially between patients with and without high-risk conditions.
Supplementary Material
Highlights.
We analyzed influenza VE among outpatients by high risk condition, age, and flu type during in U.S.
VE among patients with high risk conditions are approaching the levels as it among patients without.
VE among children and adults with high risk wasn’t different from it among those without.
VEs against flu A(H3N2), (H1N1)pdm09, and B were similar in patients with and without high-risk.
ACKNOWLEDGEMENTS
The authors would like to thank the research staff at all study sites and individuals who participated in this study.
FUNDING
This work was supported by the Centers for Disease Control and Prevention (CDC) through cooperative agreements with the University of Michigan (U01 IP000474 and U01 IP001034), Group Health Research Institute (U01 IP000466 and U01 IP001037), Marshfield Clinic Research Institute (U01 IP000471 and U01 IP001038), University of Pittsburgh (U01 IP000467 and U01 IP001035), and Baylor Scott and White Health (U01 IP000473 and U01 IP001039) and by the National Institutes of Health (NIH) (grant UL1TR001857 to the University of Pittsburgh).
Footnotes
DISCLAIMER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
CONFLICT OF INTEREST
MPN has research funding from Merck & Co., Inc., RKZ has research funding from Merck & Co., Inc. and Sanofi Pasteur, Inc. EAM and HQM report past research support from MedImmune.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- [1].Burney LE. Influenza immunization: Statement. Public health reports (Washington, DC : 1974). 1960;75:944. [PMC free article] [PubMed] [Google Scholar]
- [2].Grohskopf LA, Sokolow LZ, Broder KR, Olsen SJ, Karron RA, Jernigan DB, et al. Prevention and Control of Seasonal Influenza with Vaccines. MMWR Recomm Rep. 2016;65:1–54. [DOI] [PubMed] [Google Scholar]
- [3].Centers for Disease Control and Prevention: People at High Risk of Developing Flu–Related Complications. Available at https://www.cdc.gov/flu/about/disease/high_risk.htm. Accessed at March 4, 2018. [Google Scholar]
- [4].Jackson ML, Nelson JC. The test-negative design for estimating influenza vaccine effectiveness. Vaccine. 2013;31:2165–8. [DOI] [PubMed] [Google Scholar]
- [5].Sullivan SG, Feng S, Cowling BJ. Potential of the test-negative design for measuring influenza vaccine effectiveness: a systematic review. Expert review of vaccines. 2014;13:1571–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Flannery B, Zimmerman RK, Gubareva LV, Garten RJ, Chung JR, Nowalk MP, et al. Enhanced Genetic Characterization of Influenza A(H3N2) Viruses and Vaccine Effectiveness by Genetic Group, 2014–2015. J Infect Dis. 2016;214:1010–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].McLean HQ, Thompson MG, Sundaram ME, Kieke BA, Gaglani M, Murthy K, et al. Influenza vaccine effectiveness in the United States during 2012–2013: variable protection by age and virus type. J Infect Dis. 2015;211:1529–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Gaglani M, Pruszynski J, Murthy K, Clipper L, Robertson A, Reis M, et al. Influenza Vaccine Effectiveness Against 2009 Pandemic Influenza A(H1N1) Virus Differed by Vaccine Type During 2013–2014 in the United States. J Infect Dis. 2016;213:1546–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Jackson ML, Chung JR, Jackson LA, Phillips CH, Benoit J, Monto AS, et al. Influenza Vaccine Effectiveness in the United States during the 2015–2016 Season. The New England journal of medicine. 2017;377:534–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Zimmerman RK, Nowalk MP, Chung J, Jackson ML, Jackson LA, Petrie JG, et al. 2014–2015 Influenza Vaccine Effectiveness in the United States by Vaccine Type. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2016;63:1564–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].International Classification of Diseases, Clinical Modification. 9th ed 2015.
- [12].International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM). 10th ed2015.
- [13].Centers for Disease C, Prevention. Prevention and control of influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP)--United States, 2012–13 influenza season. MMWR Morb Mortal Wkly Rep. 2012;61:613–8. [PubMed] [Google Scholar]
- [14].Centers for Disease C, Prevention. Prevention and control of seasonal influenza with vaccines. Recommendations of the Advisory Committee on Immunization Practices--United States, 2013–2014. MMWR Recomm Rep. 2013;62:1–43. [PubMed] [Google Scholar]
- [15].Grohskopf LA, Olsen SJ, Sokolow LZ, Bresee JS, Cox NJ, Broder KR, et al. Prevention and control of seasonal influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP) -- United States, 2014–15 influenza season. MMWR Morb Mortal Wkly Rep. 2014;63:691–7. [PMC free article] [PubMed] [Google Scholar]
- [16].Grohskopf LA, Sokolow LZ, Olsen SJ, Bresee JS, Broder KR, Karron RA. Prevention and Control of Influenza with Vaccines: Recommendations of the Advisory Committee on Immunization Practices, United States, 2015–16 Influenza Season. MMWR Morb Mortal Wkly Rep. 2015;64:818–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Foppa IM, Haber M, Ferdinands JM, Shay DK. The case test-negative design for studies of the effectiveness of influenza vaccine. Vaccine. 2013;31:3104–9. [DOI] [PubMed] [Google Scholar]
- [18].Greenland S, Daniel R, Pearce N. Outcome modelling strategies in epidemiology: traditional methods and basic alternatives. International journal of epidemiology. 2016;45:565–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Healthy People 2020. Increase the percentage of children and adults who are vaccinated annually against seasonal influenza. Available at: https://wwwhealthypeoplegov/2020/topics-objectives/topic/Immunization-and-Infectious-Diseases/objectives#4659. Accessed May 18, 2018.
- [20].Belongia EA, Simpson MD, King JP, Sundaram ME, Kelley NS, Osterholm MT, et al. Variable influenza vaccine effectiveness by subtype: a systematic review and meta-analysis of test-negative design studies. Lancet Infect Dis. 2016;16:942–51. [DOI] [PubMed] [Google Scholar]
- [21].Vasileiou E, Sheikh A, Butler C, El Ferkh K, von Wissmann B, McMenamin J, et al. Effectiveness of Influenza Vaccines in Asthma: A Systematic Review and Meta-Analysis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2017;65:1388–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Suarez-Varela MM, Llopis A, Fernandez-Fabrellas E, Sanz F, Perez-Lozano MJ, Martin V, et al. Asthma and influenza vaccination in elderly hospitalized patients: Matched case-control study in Spain. The Journal of asthma : official journal of the Association for the Care of Asthma. 2017:1–11. [DOI] [PubMed] [Google Scholar]
- [23].Remschmidt C, Wichmann O, Harder T. Influenza vaccination in HIV-infected individuals: systematic review and assessment of quality of evidence related to vaccine efficacy, effectiveness and safety. Vaccine. 2014;32:5585–92. [DOI] [PubMed] [Google Scholar]
- [24].Sykes A, Gerhardt E, Tang L, Adderson EE. The Effectiveness of Trivalent Inactivated Influenza Vaccine in Children with Acute Leukemia. The Journal of pediatrics. 2017;191:218–24.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Vinograd I, Eliakim-Raz N, Farbman L, Baslo R, Taha A, Sakhnini A, et al. Clinical effectiveness of seasonal influenza vaccine among adult cancer patients. Cancer. 2013;119:4028–35. [DOI] [PubMed] [Google Scholar]
- [26].Restivo V, Costantino C, Bono S, Maniglia M, Marchese V, Ventura G, et al. Influenza vaccine effectiveness among high-risk groups: A systematic literature review and meta-analysis of case-control and cohort studies. Human vaccines & immunotherapeutics. 2017:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Hak E, Nordin J, Wei F, Mullooly J, Poblete S, Strikas R, et al. Influence of high-risk medical conditions on the effectiveness of influenza vaccination among elderly members of 3 large managed-care organizations. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2002;35:370–7. [DOI] [PubMed] [Google Scholar]
- [28].Udell JA, Zawi R, Bhatt DL, Keshtkar-Jahromi M, Gaughran F, Phrommintikul A, et al. Association between influenza vaccination and cardiovascular outcomes in high-risk patients: a meta-analysis. Jama. 2013;310:1711–20. [DOI] [PubMed] [Google Scholar]
- [29].Mohseni H, Kiran A, Khorshidi R, Rahimi K. Influenza vaccination and risk of hospitalization in patients with heart failure: a self-controlled case series study. European heart journal. 2017;38:326–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Lau D, Eurich DT, Majumdar SR, Katz A, Johnson JA. Effectiveness of influenza vaccination in working-age adults with diabetes: a population-based cohort study. Thorax. 2013;68:658–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Vamos EP, Pape UJ, Curcin V, Harris MJ, Valabhji J, Majeed A, et al. Effectiveness of the influenza vaccine in preventing admission to hospital and death in people with type 2 diabetes. CMAJ : Canadian Medical Association journal = journal de l’Association medicale canadienne. 2016;188:E342–e51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Wang IK, Lin CL, Lin PC, Liang CC, Liu YL, Chang CT, et al. Effectiveness of influenza vaccination in patients with end-stage renal disease receiving hemodialysis: a population-based study. PLoS One. 2013;8:e58317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Bekkat-Berkani R, Wilkinson T, Buchy P, Dos Santos G, Stefanidis D, Devaster JM, et al. Seasonal influenza vaccination in patients with COPD: a systematic literature review. BMC pulmonary medicine. 2017;17:79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Feng S, Cowling BJ, Sullivan SG. Influenza vaccine effectiveness by test-negative design - Comparison of inpatient and outpatient settings. Vaccine. 2016;34:1672–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Barker WH. Excess pneumonia and influenza associated hospitalization during influenza epidemics in the United States, 1970–78. Am J Public Health. 1986;76:761–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Mathai SC, Mathew S. Breathing (and Coding?) a Bit Easier: Changes to International Classification of Disease Coding for Pulmonary Hypertension. Chest. 2018. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.