A multinational research initiative developing precision medicine tools for personalised tinnitus therapy.
Overview
Tinnitus affects approximately 15% of the global population, yet no universally effective treatment exists. The UNITI project addresses this critical gap by integrating clinical, neurobiological, and audiological data from across Europe to identify patient subgroups and match them to optimal therapies.
By combining large-scale cohort data with cutting-edge machine learning, UNITI moves tinnitus care from a one-size-fits-all approach towards truly personalised medicine — predicting which interventions will benefit which patients before treatment begins.
The consortium brings together audiologists, neuroscientists, data scientists, clinicians, and patient advocates from leading European institutions, creating an unprecedented collaborative framework for tinnitus research.
Key Objectives
Study Design
UNITI employs a three-phase design that first mines retrospective data, then builds predictive models, and finally validates them in a prospective clinical trial — ensuring that findings are both scientifically robust and clinically actionable.
Harmonisation and pooling of existing tinnitus datasets from European clinical centres. Standardised outcome measures (THI, TFI, TQ, PSI) are aligned to create a unified repository enabling cross-cohort analysis.
Application of machine learning algorithms to identify clinically distinct patient subgroups based on audiological, psychological, and neuroimaging biomarkers. Predictive models are trained to forecast treatment response.
A multinational prospective study recruiting over 1,000 newly-presenting tinnitus patients. The DSS recommendations are evaluated against real-world clinical outcomes at 12-month follow-up.
Decision Support System
The UNITI Decision Support System (DSS) is the project's central clinical deliverable — a web-based tool designed for use by audiologists and ENT clinicians at the point of care.
Drawing on the harmonised UNITI dataset and validated predictive models, the DSS analyses an individual patient's profile and generates evidence-based recommendations for tinnitus interventions, including sound therapy, cognitive behavioural therapy (CBT), hearing aids, and combined approaches.
Scientific Publications
UNITI has produced a substantial body of peer-reviewed research spanning methodology, subtyping, outcomes, and the DSS. Selected key publications are listed below.
For a complete and current list of UNITI publications, please visit uniti.tinnitusresearch.net/publications.
Deliverables
UNITI's work packages collectively generate tools, datasets, and frameworks designed to have lasting impact on tinnitus research and clinical practice.
A standardised, GDPR-compliant repository integrating retrospective data from seven European clinical sites, enabling unprecedented cross-cohort analysis.
Evidence-based taxonomy of clinically relevant tinnitus subtypes derived through unsupervised and supervised machine learning methods.
A validated, open-access web application enabling clinicians to generate personalised treatment recommendations from structured patient input data.
Integration with the established ecological momentary assessment platform, enabling real-world tinnitus variability monitoring to inform DSS predictions.
A recommended minimum dataset and harmonised protocol for tinnitus clinical trials, made available to the international research community.
All major UNITI findings published in peer-reviewed journals under open-access terms, maximising scientific impact and public benefit.
Educational resources for clinicians on the use of the DSS and subtyping framework, distributed through professional audiological societies.
Complete report on the 1,000-patient prospective clinical trial, including DSS performance evaluation, safety data, and recommendations for implementation.
Project Results
The UNITI project concluded in September 2023, having successfully met all its objectives and milestones. Its overall aim was to deliver a predictive computational model based on existing and longitudinal data to address the central clinical question: which treatment approach is optimal for which patient, and why?
Clinical, epidemiological, medical, genetic and audiological data — including signals reflecting ear-brain communication — were analysed from existing databases. Predictive factors for different patient groups were extracted and their prognostic relevance tested in a randomised controlled trial (RCT) in which different groups of patients underwent single and combination therapies targeting the auditory and central nervous systems.
A list of candidate genes related to tinnitus was developed. Genes showing an overload of rare variants were identified and integrated into the DSS algorithm. Shared genes across analyses provided insights into potential genetic factors contributing to tinnitus susceptibility and severity. An analysis of 92 neurology proteins found no significant associations, indicating that expanded proteomics platforms are needed for future biomarker discovery.
RCT data from five UNITI clinical centres was unified, harmonised and fully anonymised. All data will be made publicly accessible via the UNITI Zenodo community alongside the RCT manuscript, together with open-source analysis scripts. An additional metadata paper will be published to enable other researchers to re-use the dataset for further study.
A real-time data recording mobile application was developed for each study participant, enabling ecological momentary assessment of tinnitus variability throughout the trial period. The app captured fluctuations in tinnitus loudness, distress, and associated symptoms in daily life, providing a richer picture of patient experience than clinic visits alone.
The UNITI-RCT was a multicenter, randomised controlled clinical trial with ten treatment arms designed to compare single and combination therapies for tinnitus. A total of 461 patients were included in the analysis. The trial is one of the world's largest tinnitus trials and the first to directly compare established standard treatments performed alone or in combination. The study protocol was pre-registered and the statistical analysis plan published prior to unblinding.
A systematic literature review and an empirical study with 679 European participants revealed a major gap in knowledge about the economic burden of tinnitus. The study found high out-of-pocket expenditures with variations based on tinnitus severity. Total annual out-of-pocket expenditures for tinnitus in the studied countries exceeded 17 billion euros. Findings confirmed a clear correlation between increased tinnitus severity and decreased quality of life.
The CDSS was developed with two complementary cores. Core 1 builds a predictor for treatment improvement based on UNITI data, optimised for predictive performance. Core 2 is inspired by historical centre-level data analysis and additionally prioritises the ranking of treatments by applicability for individual patients. The system supports missing values during prediction generation, enabling real-world use even when data collection is still ongoing. The CDSS is available to UNITI partners via the EU Tinnitus Database Platform.
To ensure that UNITI knowledge is sustained and widely exploited across European stakeholders, a dedicated Working Group (UNITI WG) has been established under the umbrella of the project. The group brings together individuals with scientific and practical interest in the topics addressed by UNITI — continuing the work on tinnitus subtyping, DSS development, and clinical translation beyond the funded project period.
EU Funding
UNITI is funded by the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No. 848261. Full project documentation, consortium details, and official reporting are available through the EU CORDIS database.