Use of digital technology to enhance participation in population-based studies for Public Health & Epidemiology

Advances in digital technologies represent alternative or complementary solution for generating, collecting, storing, processing, and exchanging health information in Public Health and Epidemiological (PHE) research settings. The expanding availability, affordability and accessibility of these tools has enabled the adoption and use of a wide range of hardware, software, and networking services throughout the entire research continuum. Among other capabilities, researchers are now able to collect personal and self-reported health data directly from the research subjects and their own environments.

In particular, new information and communication technologies may be practically integrated as part of pre-analytical research stages, such as recruitment and data collection, with the purpose of improving the external and internal validity of sampled data used for statistical analyses and outcome generation. That is, digital technologies are integrated into the routine procedures of population-based studies (e.g., interviewing, invitation and consent material) with the double aim of increasing confidence over findings from available data and increasing the representativeness of such findings to the wider population under study to other settings beyond the study parameters.

While these technologies certainly offer promising solutions to the current limitations of procedural stages in population-based research, their use is, however, not without caveats. For example, while technologies such as Machine Translation (MT) may support outreach to certain specific linguistically diverse communities, these technologies might exclude or offer less quality communication services to other specific communities due to differential accuracy across language translations. Moreover, while the availability of online survey technologies offers solutions to increase participation and reduce nonresponse (and, as such, to increase study generalizability and representativeness), these solutions may also generate measurement errors and reduce internal reliability of the data collected.

For such reasons, continuous assessments of the technical, socio-economic, and ethical-legal maturity of digital technologies are essential to align their effective and efficient development, adoption and deployment with PHE objectives, such as the World Health Organization (WHO) Essential Public Health Operations. Moreover, the research on their use may also help to clarify their limitations and shed light on their actual potential (e.g., as to not misuse them and further health outcome divides among already disenfranchised communities).

Against this background, this dissertation investigates how different digital technologies may serve to increase response in population-based studies from their research onset, namely at the points of initial study invitation and data collection. The cumulative dissertation presents four peer-reviewed manuscripts: three published prior to submission, and one published after, in March 2026. The dissertation addresses the use of digital technology to enhance participation in population-based studies for public health & epidemiology. Section I of the dissertation consists of three introductory chapter. In Chapter One, the dissertation introduces its reader to the current debate on the use of digital technologies in public health and epidemiological research, presenting the working rationale and objectives of each submitted manuscript. In Chapter Two, the dissertation offers an overview of the research concepts, definitions and working frameworks used throughout these manuscripts. Additionally, this chapter offers an overview and mapping of available literature on the use of digital technology throughout ten general research stages for developing population-based studies. Chapter Three summarizes the methods and results of the four submitted manuscripts, while presenting short commentaries and ancillary analyses.

Section II presents the four manuscripts submitted as part of the cumulative dissertation. Chapter Four presents a scoping review of peer-reviewed literature published after 2016 on the use of use of MT technologies for disseminating PHE material to specific target audiences in population-based research and surveillance efforts (Herrera-Espejel and Rach 2023). Culturally and linguistically diverse (CALD) populations are often underrepresented in PHE studies. Such underrepresentation may lead to biased study results, limiting their generalizability in practice. Modern MT tools might offer a cost-effective solution to this problem. The review identifies 46 articles addressing the use of MT to facilitate dissemination of public health information to CALD communities. The review indicates that current commercial MT solutions, such as Google Translate and DeepL, may provide sufficiently accurate translations, when used along with pre- and post-editing efforts, namely for non-legally or ethically sensitive materials. As such, the Chapter discusses the limitations of current MT solutions to one-way communication between public health staff and addressed audiences. Questions for future research include: assessing how machine-translated texts are received by different segments of target populations; and understanding the effectiveness of MT for specific two-way communication material, such as participant enrollment, informed consent, and response materials in general.

Chapter Five exhibits a cluster randomized control experiment (Herrera-Espejel et al. 2025), which was embedded in the World Health Organization European Childhood Obesity Surveillance Initiative in second-and third-grade classrooms in the cities of Bremen and Bremerhaven, Germany. The study offers an evaluation of the effectiveness of a co-developed physical-digital multilingual outreach solution, namely, flyer with a QR-code linking to multiple language translation of the original study invitation. Based on multilevel mixed-effects logistic regression models, the study measures the extent to which the inclusion of the flyer in the study invitation was positively associated with the response and eventual study participation proportions of
linguistically diverse eligible participants, supporting more equitable participation. While the presence of the flyer was associated with increased participation and response of linguistically diverse households from medium-level socioeconomic backgrounds, its effect was not significant for children studying in schools in low- nor high-levels. In view of these results, the study in Chapter Five underscores the importance of tailored approaches to increase engagement among underrepresented communities in population-based health research, as well as the importance of using appropriate statistical methods to investigate the differential outcomes of digitally-based interventions across socio-economic and linguistic subgroups within heterogenous populations.

Chapter Six offers an exploratory analysis and evaluation of the effect of using two concurrent interviewing modes on the collection of self-reported symptoms in population-based studies. As was the case of the CoVerlauf study (Rach et al. 2023), two inherently different interviewing modes were used to collect participant data on post-Covid symptoms at their time of infection and interview: Computer-Assisted Website Interviewing and Computer-Assisted Telephone Interviewing. The Chapter explores and discusses the extent of selection effects and measurements effects cause by the use of two modes. The use of the two modes was in itself advantageous to reduce nonresponse, as well broaden the coverage and response of eligible study
participants which would have otherwise (e.g., in a single-mode design) not engaged in the study activities. However, due to the inherent characteristics of each mode, the mixed-mode design potentially introduced parallel mode-specific measurement errors, blurring the comparability of items collected through the different information exchange methods at the point of data collection. [This chapter presents the submitted manuscript prior to its publication. The published version is available online (DOI: 10.2196/80631)].

Chapter Seven explores how digital tools might unintentionally increase health inequalities, especially among different socio-economic, gender, and underserved groups. It highlights how the unequal access, use and effectiveness of digital technologies may widen differential health outcomes. The chapter underscores the importance of involving targeted subgroups in the design and development of health interventions to reduce nonresponse or differential response due to the use of digital technology.

Section III consists of three conclusive chapters. In Chapter Eight, the dissertation discusses the overall research contributions and limitations for each of the previous chapters. In Chapter Nine, the dissertation offers overall research reflections after completion of the writing of the manuscripts included. Specifically, the dissertation shares final reflections on the importance of linguistic inclusiveness and fairness in PHE participant material (e.g., invitations, consent forms, questionnaires, etc.). It also discusses insights that might be gained from language demographics in Germany based on available data from the Statistische Bundesamt from August 2024, and the foreseeable paradoxes of using digital technology for inclusion. The final Chapter Ten summarizes the overall research conclusions of the dissertation, concluding that, while digital technologies may significantly enhance participation and response in population-based studies, they also risk exacerbating existing inequalities and creating new forms of divides due to limitations in their technological readiness. The dissertation argues that the evaluation and continuous investment in training PHE researchers to understand intersectionality theory and adopt quantitative methods that reflect the superdiversity of populations under study is crucial to improve both generalizability and the reliability of findings in the field. A last section IV organizes the appendixes from Chapter Four, Five and Six.