Patient-Initiated Biobanks and Data Resources
The DecodeME study represents the largest co-produced genetics study in ME/CFS, with a target enrollment of 25,000 participants. The study was co-designed with patient input from inception: patients contributed to the study protocol, the recruitment strategy, the communication materials, and the operational decisions about how data would be collected and shared. The participatory approach was not an afterthought β it was the studyβs design principle (Devereux-Cooke et al. 2022).
The DecodeME model is significant beyond its scientific output (a 2025 preprint reports 8 genome-wide significant loci and estimates SNP-based heritability at ~9.5%, confirming the feasibility of large-scale ME/CFS GWAS History of ME/CFS). It demonstrates that large-scale co-production is operationally feasible β that patient communities can participate in study design at scale, not just in token advisory roles. The studyβs recruitment strategy β leveraging patient organizations, social media, and community networks β reached a population that traditional clinical recruitment would have missed, particularly severe and homebound patients who cannot attend clinic-based research visits. (Certainty: 0.80 β mid-tier journal; 25K target enrollment; co-produced from inception; ME/CFS-specific. Diagnostic criteria: self-reported ME/CFS diagnosis with validation against symptom criteria. Publication date: 2022 β recruitment methodology paper; 2025 GWAS preprint available.)
Consequence: DecodeME demonstrates that co-production and large-scale data collection are not in tension β a study can be both large enough to power GWAS and participatory enough to represent patient priorities. This has implications for every disease community that argues for involvement in research design: DecodeME provides the operational template. For co-production applied to healthcare service design rather than research, see Co-Production as a Healthcare Design Principle in Healthcare Systems, Policy, and Disability. Severity applicability: the remote, community-based recruitment model makes the study more accessible to severe and homebound patients than clinic-based recruitment β a design choice that directly addresses the underrepresentation of the most severely ill in research.
The UK ME/CFS Biobank, based at the London School of Hygiene and Tropical Medicine, was the worldβs first disease-specific biobank for ME/CFS, with 543 participants recruited across the UK and supported by patient charities including Action for ME and ME Research UK. Its establishment represented a collaboration between patients, charities, and academic researchers β a model in which patient organizations contributed funding, recruitment, and clinical characterization to a formal biobanking initiative.
The Open Medicine Foundation has funded the collection and centralization of biospecimens β blood, plasma, serum, PBMCs β from ME/CFS patients across multiple clinical sites, supporting multi-site collaborative studies with standardized collection protocols. The DecodeME study, while not a biobank per se, collected DNA samples from over 15,000 participants recruited through patient organizations and community networks, creating a genetic data resource co-produced with patient communities.
The significance of patient-initiated biobanks extends beyond the samples themselves. A biobank determines what gets studied: which tissues are collected, what metadata is captured, which patient subgroups are recruited, and how samples are distributed to researchers. When patients and patient organizations participate in biobank design β contributing to decisions about sample types, longitudinal collection schedules, and severity-level inclusion β the biobankβs architecture reflects priorities that institution-only biobanks may miss (e.g., longitudinal sample collection to capture PEM-related changes, inclusion of severe patients via home blood draws, collection of pre- and post-exertional samples).
Slavin and colleaguesβ analysis of 119 ME/CFS Common Data Elements (CDEs) demonstrated the importance of patient input in data collection design: the existing CDEs were found to be strong on biomedical measures (laboratory values, vital signs) but weak on activity, participation, and quality-of-life domains β precisely the domains that patients identify as most meaningful (Slavin et al. 2023). (Certainty: 0.70 β CDE analysis from PLOS ONE; OMF co-authored; descriptive, not interventional.)
Consequence: The architecture of a biobank determines twenty years of research downstream. A biobank that collects samples from ambulatory patients at a single time point will produce twenty years of research on ambulatory patients at a single time point β missing PEM dynamics, severity stratification, and longitudinal trajectory. Patient involvement in biobank design changes what can be discovered by shaping what is collected before any experiment is run. Severity applicability: recruitment methods that include severe patients (via home blood draws, community-based outreach) address the most significant sampling bias in ME/CFS research.