Highlights
-
•
High-dose influenza vaccine reduces laboratory-confirmed influenza, but its effectiveness against severe cardio-respiratory outcomes in the older general population has not been assessed in a fully powered, individually randomized trial.
-
•
DANFLU-2 is a pragmatic, individually randomized trial to assess the relative effectiveness of high-dose vs standard-dose influenza vaccine against cardio-respiratory hospitalizations in older adults ≥65 years.
-
•
Utilizing innovative recruitment and online informed consent methods, the trial has randomized 332,438 participants over 3 influenza seasons.
-
•
DANFLU-2 is the largest individually randomized influenza vaccine trial ever conducted.
-
•
The trial will provide high-quality data on the effectiveness of high-dose influenza vaccine in preventing hospitalizations to inform vaccination policies on a global level.
Background
The high-dose inactivated influenza vaccine (HD-IIV) has been shown to reduce the incidence of influenza infection compared with standard-dose inactivated influenza vaccine (SD-IIV); however, its effectiveness in preventing severe respiratory and cardiovascular outcomes in the older general population has not yet been assessed in a fully powered individually randomized trial.
Methods
DANFLU-2 is an ongoing pragmatic, registry-based, open-label, active-controlled, individually randomized trial conducted in Denmark during the 2022/2023, 2023/2024, and 2024/2025 influenza seasons. Utilizing innovative electronic recruitment strategies and an online informed consent process, the trial has enrolled 332,438 older adults ≥65 years; enrollment was completed in October 2024. Participants were randomly allocated in a 1:1 ratio to either HD-IIV or SD-IIV. Baseline, endpoint, and safety data are primarily obtained from the nationwide Danish administrative health registries. The primary endpoint is hospitalization for influenza or pneumonia with the trial also powered to assess the first secondary endpoint of hospitalization for any cardio-respiratory disease. The additional secondary endpoints of all-cause hospitalization and mortality will be tested hierarchically. Key ancillary analyses include cost-effectiveness and health care resource consumption assessments as well as a home self-swab sub-cohort.
Discussion
DANFLU-2 is the largest individually randomized influenza vaccine trial ever conducted and will provide critical, high-quality evidence on the effectiveness of HD-IIV against cardio-respiratory hospitalizations and mortality in the older general population.
Trial registration
Clinicaltrials.gov: NCT05517174, registered August 24, 2022, https://clinicaltrials.gov/study/NCT05517174 .
Background
Seasonal influenza causes an estimated 290,000-650,000 respiratory deaths worldwide per year and is associated with a wide range of complications including pulmonary and cardiovascular (CV) events. ,, At-risk individuals such as older adults, persons with chronic conditions such as heart or lung disease, and persons who are immunocompromised are at increased risk of influenza infection and its associated complications. Influenza vaccination reduces the risk of both acute infection and downstream complications. ,,,,, However, vaccine effectiveness against influenza infection typically ranges between 30% and 50%, leaving room for further improvement ; and even lower VE is often observed among older or otherwise immunocompromised individuals who are unable to mount a sufficient antibody response. ,
The high-dose inactivated influenza vaccine (HD-IIV) was developed specifically to address the attenuated immune responses observed in older adults vaccinated with the conventional standard-dose inactivated influenza vaccine (SD-IIV). Accomplished by quadrupling the dose of hemagglutinin antigen compared with SD-IIV (60 µg vs 15 µg), the HD-IIV was approved in 2009 based on a phase 3 trial demonstrating a significant increase in antibody response compared with SD-IIV among older adults ≥65 years. Superior efficacy of HD-IIV compared with SD-IIV was demonstrated in a North American trial among 31,989 adults ≥65 years, which showed that HD-IIV significantly reduced the incidence of laboratory-confirmed influenza illness by 24.2% (95% confidence interval [CI]: 9.7%-36.5%) compared with SD-IIV with no increase in serious adverse events. A subsequent secondary analysis showed higher efficacy for HD-IIV vs SD-IIV in preventing adjudicated serious cardio-respiratory events (defined as pneumonia, asthma/chronic obstructive pulmonary disease/bronchial events, influenza, coronary artery events, congestive heart failure, cerebrovascular events, and other respiratory events) with a relative vaccine effectiveness (rVE) of 17.7% (95% CI: 6.6%-27.4%).
The effectiveness of HD-IIV in preventing serious clinical events such as hospitalizations and mortality has been investigated in randomized trials in select high-risk groups. In a cluster-randomized trial, HD-IIV reduced the risk of respiratory-related hospital admissions by approximately 13% among US nursing home residents aged ≥65 years compared with SD-IIV, while in the INVESTED (Influenza Vaccine to Effectively Stop Cardio Thoracic Events and Decompensated Heart Failure) trial, HD-IIV did not reduce the risk of the primary composite endpoint of all-cause death or cardiopulmonary hospitalization compared with SD-IIV in very high-risk patients with CV disease. A recent meta-analysis of randomized controlled trials of HD-IIV vs SD-IIV including a total of 105,685 individuals found significant reductions in the incidence of pneumonia and influenza hospitalization and all-cause hospitalization with HD-IIV. While the superior efficacy of HD-IIV against laboratory-confirmed influenza has been previously demonstrated, no individually randomized, adequately powered trial has ever investigated the effectiveness of HD-IIV vs SD-IIV against severe clinical outcomes including hospitalizations and mortality in the older general population. An effort was made to conduct such a trial in Finland, but the trial had to be terminated due to the COVID-19 pandemic.
Most current vaccine effectiveness studies are based on observational data with inherent limitations including healthy vaccinee bias, indication bias, and other sources of uncontrolled confounding. While randomized controlled trials (RCTs) are the gold standard for unbiased assessment of treatment effects, data from traditional explanatory RCTs often suffer from reduced generalizability due to a highly controlled trial environment and restrictive eligibility criteria as well as sample sizes that may not allow for assessment of broader outcomes such as hospitalizations and mortality, which are less frequent but often of greater importance to public health. An individually randomized trial powered for severe clinical outcomes such as hospitalizations requires a very large sample size, but could help elucidate the full public health impact of HD-IIV and provide high-quality evidence for future influenza vaccine recommendations and cost-effectiveness analyses. Any improvements in the real-world effectiveness of influenza vaccination would be expected to result in significant public health benefits due to the large, worldwide scale on which influenza vaccinations are administered each year. Pragmatic individually randomized trials can combine the benefit of delivering the highest quality of evidence due to randomization, with the improved generalizability conferred by a pragmatic approach to recruitment and trial conduct.
The Danish health infrastructure contains several features which make Denmark a suitable setting for conducting large-scale RCTs with pragmatic elements across recruitment, trial logistics, and data collection, resulting in representative trial populations, improved generalizability of trial results, and the opportunity for large sample sizes allowing for assessment of broad clinical endpoints. Denmark has a public tax-funded universal health system which provides most healthcare services including hospital care free-of-charge. All activities in the Danish health system are captured in administrative health registries including but not limited to data on duration of healthcare encounters, diagnosis codes, medical procedures, filled prescriptions, laboratory results, causes of death, and vaccinations. The health registries can be used for both nationwide identification of potential participants and for subsequent data collection including baseline, endpoint, and safety data.
Feasibility of conducting such a large-scale, pragmatic, registry-based, individually randomized trial of HD-IIV vs SD-IIV in Denmark was demonstrated in the DANFLU-1 (Feasibility of Randomizing Danish Citizens Aged 65-79 Years to High-Dose Quadrivalent Influenza Vaccine vs Standard-Dose Quadrivalent Influenza Vaccine in a Pragmatic Registry-Based Setting) pilot trial. Although clinical outcomes were explored, DANFLU-1 was designed to assess operational feasibility and was not powered for clinical outcomes. We therefore designed the DANFLU-2 (A Pragmatic Randomized Trial to Evaluate the Effectiveness of High-Dose Quadrivalent Influenza Vaccine vs. Standard-Dose Quadrivalent Influenza Vaccine in Older Adults) trial to accurately determine the incremental effectiveness of HD-IIV against severe clinical outcomes among community-dwelling older adults across multiple seasons.
Purpose
The purpose of the DANFLU-2 trial is to evaluate the relative effectiveness of HD-IIV vs SD-IIV against severe cardio-respiratory outcomes and mortality in the general older adult population ≥65 years.
Methods
Study design and organization
DANFLU-2 is a pragmatic, registry-based, open-label, active-controlled, individually randomized trial conducted in Denmark during the 2022/2023, 2023/2024, and 2024/2025 influenza seasons with participants randomly allocated in a 1:1 ratio to either HD-IIV or SD-IIV.
The trial is conducted as a collaboration between the academic research group at the Center for Translational Cardiology and Pragmatic Randomized Trials (CTCPR) at the Department of Cardiology, Copenhagen University Hospital– Herlev and Gentofte, Copenhagen, Denmark and the private vaccination provider Danske Lægers Vaccinations Service/European LifeCare Group (ELCG). CTCPR acts as sponsor-investigator, serves as the central trial site, and is responsible for overall study design and conduct, registry-based recruitment and data collection, and safety monitoring. ELCG is a private company which has been contracted to administer vaccinations in the public Danish vaccination program during numerous influenza seasons since 2002. In this trial, ELCG is responsible for organizing vaccination sessions, obtaining informed consent, randomizing participants, and administering the study vaccines. During the 2024/2025 season, ELCG organized vaccination sessions at >500 unique locations across Denmark.
In each study season, approximately 1,000,000 Danish citizens were invited to participate through electronic invitation letters. To minimize participant burden, trial participation required only 1 study visit at which both randomization and vaccine administration took place ( Figure 1 ). An innovative informed consent procedure was implemented, allowing participants to provide informed consent online prior to their study visit. Data collection including baseline, endpoint, and safety data is carried out by CTCPR primarily using the nationwide Danish administrative health registries.
SPIRIT figure. HD-IIV, high-dose inactivated influenza vaccine; SD-IIV, standard-dose inactivated influenza vaccine.
Trial Population and Recruitment
To ensure a broad representative older general population sample, the trial enrolled older adults ≥65 years regardless of comorbidity. The trial had no formal exclusion criteria, but only required that vaccines were administered according to routine clinical practice guidelines including assessing contraindications such as allergy/hypersensitivity towards the vaccines and fever/illness on the day of vaccination.
Participants were primarily recruited through electronic invitation letters delivered via the Danish governmental electronic letter system (Digital Post). The system is primarily used for high-priority communication from eg, governmental authorities, hospitals, and banks, but is also available for use in research projects to facilitate participant recruitment and communication. From the age of 15, all Danish citizens automatically receive all official correspondence through this system. Citizens can apply for an exemption and instead receive physical letters. When an individual receives a letter through the system, notifications are sent via email and text message, and the individual can then log onto the system to read the letter through several web portals and smartphone apps. In Denmark, all permanent residents are assigned a unique personal identifier (“CPR” number) at birth or immigration– this identifier allows for unambiguous cross-linkage of administrative health registries and is also used as the “delivery address” in the governmental electronic letter system.
Through application to the Danish Health Data Authority, researchers at authorized Danish institutions can obtain contact information for potential trial participants identified based on data contained in the Danish registries. In each season, CTCPR received data extracts allowing for the delivery of electronic invitation letters to up to 1,000,000 individuals per season. The invitation letters contained a link to a website where further information on the trial was presented in both written and video form, and after reviewing the material, participants were able to book a visit for trial participation. After booking their trial visit, participants were asked to provide informed consent online. Under an experimental program on informed consent organized by the Danish Medical Research Ethics Committee for low-intervention clinical trials, online informed consent did not require any real-time contact with the study staff (“nonsynchronous” consent); however, in case of any questions, participants were able to contact study staff through phone or email. Informed consent could also be provided in-person at the trial visit.
In addition to electronic invitation letters, participants were also recruited through ELCG’s databases of prior vaccines, patient organizations, and social media advertising.
Enrollment for the trial was completed in October 2024. A total of 332,438 participants were randomized: 19,699 in the 2022/2023 season, 160,451 in the 2023/2024 season, and 152,288 in the 2024/2025 season ( Figure 2 ). A total of 166,218 participants were allocated to the HD-IIV group and 166,220 participants were allocated to the SD-IIV group. Participants were allowed to enroll in >1 season; a new informed consent was required for each season and participants re-enrolling in additional seasons were re-randomized. A total of 226,876 unique individuals underwent randomization in the trial, of which 87,826 participated in 2 seasons and 8,868 participated in all 3 seasons. In each season, randomization and vaccination typically took place from late September through early November. Of all randomized participants, 231,345 (69.6%) provided online informed consent, whereas the remainder provided consent in-person.
Enrollment and randomization across study seasons. Participants were enrolled over 3 consecutive influenza seasons and randomly assigned in a 1:1 ratio to HD-IIV or SD-IIV. Participants were allowed to re-enroll in subsequent seasons; they were then re-randomized. HD-IIV, high-dose inactivated influenza vaccine; SD-IIV, standard-dose inactivated influenza vaccine.
Randomization and blinding
Participants were randomly allocated in a 1:1 ratio to either HD-IIV or SD-IIV. Randomization was performed at the vaccination sites using tablet computers linked to a centralized randomization algorithm. Separate unstratified randomization lists with random permuted blocks (block sizes: 6, 8, and 10) were generated for each season using R software (R Foundation for Statistical Computing, Vienna, Austria).
The trial is not blinded. Neither investigators nor trial participants are blinded to treatment allocation. However, several key unique design elements in this pragmatic trial ensure a low risk of bias despite the lack of blinding. The centralized randomization algorithm ensures a consistent, unbiased randomization process across all study sites. Investigators are not involved in subsequent clinical treatment of participants, and investigators reviewing safety events are blinded to treatment assignment. Endpoints are ascertained using passively collected data from administrative health registries with prespecified endpoint definitions based on globally recognized International Classification of Disease, 10th Edition (ICD-10) codes. Severe outcomes such as hospitalizations and deaths would not be expected to be affected by knowledge of treatment assignment by participants, and the decision to hospitalize or not is made by working physicians without involvement in the trial, who also perform diagnosis coding for each hospitalization/death.
Intervention and comparator
Participants received a single intramuscular injection of either the HD-IIV or a SD-IIV. The HD-IIV administered was Fluzone® High-Dose Quadrivalent (Sanofi) in the 2022/2023 and 2023/2024 seasons, and Efluelda Tetra® (Sanofi) in the 2024/2025 season. The protocol allowed for any SD-IIV vaccine to be used as comparator; however, all administered SD-IIV were VaxigripTetra (Sanofi). HD-IIV contains 60 µg of hemagglutinin (HA) antigen for each strain included in the vaccine whereas SD-IIV contains 15 µg for each strain. Fluzone® High-Dose Quadrivalent and Efluelda Tetra® are identical in formulation and manufacturing, representing the exact same high-dose quadrivalent influenza vaccine; brand name designations only differ based on their region of marketing (Fluzone® in North America vs Efluelda® in the European Union). Given the pragmatic nature of the trial, differently labeled products were used due to logistical reasons (supply, timelines etc.). All vaccines administered in the trial were quadrivalent and, in each season, contained the same 4 strains as recommended by the World Health Organization for the northern hemisphere.
Co-administration with other seasonal vaccines including COVID-19 vaccines/boosters was allowed in the trial as per local guidelines and recommendations.
Data collection and data sources
Only minimal data collection was performed by vaccination site personnel who, besides routine vaccination procedures, were only required to confirm trial participation, perform randomization, and administer the vaccine. Participants who had not provided online consent prior to their visit provided consent at the vaccination site. These data were registered using a tablet computer linked to a centralized medical record system administered by the vaccination provider and subsequently transferred to the central trial site.
Almost all other trial data were sourced from the nationwide Danish administrative health registries. An individual’s CPR number enables unambiguous linkage between both healthcare-related and other administrative registries. By uploading the CPR numbers of study participants to a secure remote-access server environment (“Forskermaskinen”), registry data access can be obtained and linked.
The Danish National Patient Registry contains information on all in- and outpatient visits in the Danish public health system. The Danish National Prescription Registry contains data on filled prescriptions from all pharmacies in Denmark. The Danish Civil Registration System contains information on date of birth, deaths, immigration, and emigration. The Danish Registry of Causes of Death can be used to retrieve specific causes of death. All vaccinations including prior influenza vaccinations can be retrieved from the Danish Vaccination Registry. The Danish Microbiology Database contains microbiology results from routinely performed tests including influenza tests. Several registries containing socioeconomic data including personal income, pension, unemployment and sickness benefits exist as well enabling accurate assessment of the sociodemographic characteristics of the study population. Education status can be retrieved from the Population Education Registry. Baseline information will be collected from the registries from the date of randomization up to 10 years prior to the randomization date as has been done in most previous Danish registry-based studies. , Safety monitoring uses additional data collected through manual review of electronic health records.
The registry-based data collection in the trial uses prespecified definitions to minimize bias. Prespecified definitions for assessment of baseline conditions, medication use, and prior vaccinations can be found in Supplementary Tables 1-3 and Appendix 1 in the Supplementary Material document. We expect the risk of imbalance in any baseline variable to be extremely low due to individual-level randomization and a very large sample size.
In each season, quality-of-life questionnaires (EQ-5D-5L) are administered to participants through the governmental electronic letter system at baseline, at the peak of the influenza season, and at the end of the follow-up period (May 31) to facilitate detailed healthcare resource consumption and cost-effectiveness analyses. Response rates have so far been very satisfactory at >85%.
During the 2023/2024 and 2024/2025 seasons, 12,116 participants were enrolled in a sub-cohort in which participants were asked to perform home self-swabs and answer symptom questionnaires (Respiratory Infection Intensity and Impact Questionnaire [RiiQ®]) in case of influenza-like illness (ILI) to enable further assessment of the impact of HD-IIV vs SD-IIV on patient-reported outcomes (PROs) among individuals with ILI and laboratory-confirmed influenza. Participation was offered at the trial visit prior to randomization and vaccination at select high-volume sites. Self-swabs are mailed to a laboratory facility and subsequently undergo PCR testing for influenza A, influenza B, RNAseP, and additional respiratory pathogens. Sub-cohort participants report daily questionnaires for at least 7 days and up to 14 days in case of persisting symptoms.
Endpoints
Endpoints will be assessed primarily using registry data including all primary and secondary endpoints. Hospitalization-based endpoints are defined using ICD-10 codes. As the trial was conducted during and following the COVID-19 pandemic, any hospitalization-based events with an associated COVID-19 ICD-10 discharge diagnosis code (B34.2, B97.2) will not be considered an eligible endpoint for assessment, as to improve endpoint specificity. The primary analyses will assess events occurring from 14 days after vaccination and up to May 31 the following year.
The primary endpoint of the trial is hospitalization for influenza or pneumonia defined as an overnight hospital stay with a primary diagnosis of J09-J18. The trial is also powered to assess the first secondary endpoint of hospitalization for any cardio-respiratory disease. A full list of prespecified endpoints is shown in Table 1 . The trial uses prespecified registry-based definitions to minimize bias; these can be found in Supplementary Table 4.
Table 1
List of study endpoints
Primary endpoint
|
Secondary endpoints
|
Exploratory endpoints
|
Healthcare resource consumption endpoints
|
|
Influenza-like illness and laboratory-confirmed influenza endpoints in home self-swab sub-cohortInfluenza-like illness endpoints:
|
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
