
Background: Clinical research and biobanking workflows require retrieval and redaction of protected health information (PHI) from clinical documents before sharing with sponsors and investigators, often alongside biospecimens. Manual redaction is time-intensive, error-prone, and may increase organizational risk through privacy exposure. Automated redaction tools could reduce human burden, improve completeness, and minimize exposure. Objective: This exploratory pilot study aimed to measure the human time burden of manual document retrieval and PHI redaction across common clinical document types, quantify accuracy based on missed PHI (false-negative) patterns, and benchmark results against a fully automated redaction system applied to the same source documents. Methods: A prospective time-motion and error characterization study with six experienced research staff were performed. Each participant redacted nine documents four discharge reports, four pathology reports, and one lab-results report. Automated time was defined as zero human time, as machine runtime does not represent human effort; missed PHI was measured using the same criteria in both conditions. Results: Six participants each completed nine documents (54 total). Mean manual session time was 058:54. Mean document-level time was ∼0:06:05, excluding Research Electronic Data Capture search (0:00:51) and miscellaneous time (0:00:48). Manual redaction produced 55 missed PHI instances (∼1.02/document); the automated system missed 3 instances across 18 unique documents (∼0.17/document), all medical staff names. Conclusion: Automated redaction was associated with fewer missed PHI instances with no measurable human time burden. These exploratory results suggest automation may reduce privacy risk and eliminate the nonredaction workflow steps comprising nearly 40% of manual session time, supporting further evaluation in biospecimen research and biobanking operations.
BACKGROUND Hospital-based biobanks are vital for biomedical research, translational medicine, and personalized health care. However, in Indonesia, they remain underdeveloped, limiting access to well-annotated biospecimens and underrepresenting Indonesian cohorts. Challenges include limited infrastructure, fragmented data, regulatory complexities, insufficient funding, and a lack of standardized procedures and trained staff. These hinder biospecimen collection, storage, and use. This study proposes a hospital biobank model tailored to Indonesia's health care environment to foster sustainable biobank development. METHOD Modeling occurred between 2020 and 2022 as a collaboration between the Biobank Unit, Faculty of Medicine, Public Health and Nursing UGM, and UGM Academic Hospital. The process involved four steps (1) planning; (2) protocol preparation; (3) modeling; and (4) evaluation, recommendation, and dissemination. Planning involved a desk review and stakeholder discussions, resulting in a business plan and a protocol. Four workstreams-governance, ethics, sample collection, and data management-developed in parallel. The modeling included team training, a dry trial, and model implementation with 24 participants from UGM Hospital's general surgery. Results were discussed and shared at a national workshop with hospital biobank stakeholders. RESULTS The hospital-based biobank modeling successfully established and evaluated an integrated framework encompassing governance; specimen collection; ethics, legal, and social implications; and data management. A governance structure, standardized workflows and SOPs, a dual-consent process, and a SIMBIOX 1.2-based data management system were developed and implemented. The pilot enrolled 24 participants, demonstrating the feasibility of integrating biobanking activities into routine clinical services, with most participants completing the process successfully and expressing positive acceptance of biobank participation. The modeling process also identified operational, logistical, and communication challenges that informed subsequent improvements. CONCLUSION This modeling strategy lays the groundwork for establishing biobanks, enabling health care professionals and policymakers to address challenges and maximize the benefits of biobanking in Indonesia.
Background: Peripheral blood mononuclear cells (PBMC) are essential for immunological research and biomarker discovery in various diseases. They are widely used to monitor immune responses and evaluate treatment efficacy, including immunotherapies. Biobanking PBMCs enables longitudinal studies and batch analyses, but cryopreservation may affect cell viability, phenotype, and function. While most studies assess PBMC quality after ≤5 years of storage, little is known about samples preserved for >20 years, which is relevant for slow-progressing diseases. Materials and Methods: PBMC samples cryopreserved for over 20 years (very long-term cryopreserved, VLTC; n = 22) in liquid nitrogen were evaluated and compared with recently frozen samples (<2 months; short-term cryopreserved, STC; n = 6). PBMCs were isolated using Ficoll-based protocols, cryopreserved and stored in liquid nitrogen. After thawing, cell viability, recovery, and immune subset composition were assessed by multiparametric flow cytometry. Functional integrity was evaluated by baseline activation marker expression and viability after 24-hour culture. Results: VLTC samples showed reduced recovery rates compared with STC (mean 54.1% vs. 82.4% cells recovered after thawing; p = 0.021), but viability remained high (≥79%). Flow cytometry revealed preserved staining quality and comparable frequencies of major immune subsets (T cells, B cells, monocytes, and dendritic cells) between VLTC and STC samples. Differences were observed in rarer populations such as intermediate and non-classical monocytes. Memory T cells were also decreased in VLTC samples. Baseline Th activation markers (IFN-γ, IL-4, and IL-13) were similar across groups. After 24 hours of culture, cell viability decreased similarly across VLTC and STC samples, but remained above 75%, thus supporting suitability for functional assays. Conclusion: PBMCs cryopreserved for over 20 years retain sufficient viability and immunophenotypic integrity for flow cytometry and short-term culture. Despite reduced recovery, these samples are suitable for immunological analyses, provided viability and storage duration are reported. Long-term PBMC biobanking is feasible for research on diseases with extended clinical timelines.
ISO 20387 has become an important framework for quality, competence, and governance in biobanking. Published experiences have frequently addressed implementation pathways, accreditation processes, and measurable organizational outcomes associated with the adoption of quality systems. Comparatively less attention has been devoted to the long-term post-accreditation phase, during which quality systems transition from explicit organizational projects into normalized operational infrastructures embedded within routine practice. Drawing on the experience of the first Italian biobank accredited according to ISO 20387 in 2022 and successfully renewed in 2026, this perspective article reflects on what may be defined as the “invisible phase” of mature quality systems. Rather than revisiting implementation strategies or quantitative indicators, it explores three organizational paradoxes that may emerge during long-term accreditation maintenance: visibility versus invisibility, standardization versus meaning, and stability versus vitality. The proposed framework suggests that the normalization of quality-oriented practices may represent both a strength and a potential vulnerability of mature accredited systems. As quality processes become increasingly integrated into everyday organizational behavior, potential risks include procedural ritualization, reduced critical awareness, and reduced adaptive capacity. In this context, accreditation renewal may represent more than a technical reassessment of compliance, providing an opportunity to reflect on organizational resilience, sustainability, and cultural integration over time. Understanding this “invisible phase” may contribute to developing more reflexive and sustainable models of quality governance in biobanking, extending current discussions beyond implementation and initial accreditation alone.
OBJECTIVES Biological sample documentation is operationally demanding in clinical research centers running concurrent trials. Paper-based traceability workflows are commonly perceived as slow and prone to transcription errors, but quantitative evidence is limited. This work reports the first 12 months of operational experience with TraceLab, an in-house web-based platform designed to centralize sample traceability at a high-volume site. METHODS A single-site observational implementation study was conducted at the Bradford Hill Clinical Research Center in Santiago, Chile (April 17, 2025-April 16, 2026). TraceLab was connected to the existing laboratory information system via a read-only REST API and covered three documentation stages: sample collection, laboratory processing, and shipment. Data entry time, transcription error rate, collection-to-registration time, monthly kit volumes, and audit outcomes were evaluated. Collection-to-registration time was compared across implementation phases using exploratory inferential tests. RESULTS The system processed 5466 kits across 49 protocols and 398 patients, generating 8743 forms, 12,000 tubes, 18,000 aliquots, and 90 smears. Median data entry time was 6 seconds per form (interquartile range 4-10), substantially shorter than the estimated 2-4 minutes required by the former paper-based workflow. No errors were found in the audited 10% sample (0/874; 95% confidence interval [CI]: 0.00%-0.42%, exact binomial [Clopper-Pearson]). The mean collection-to-registration time was 15.08 minutes (standard deviation [SD]: 3.01) when excluding the initial learning phase (May 2025). After the initial learning phase, mean collection-to-registration time was lower during optimization than during stabilization, with an exploratory mean difference of 4.25 minutes (95% CI: 1.10-7.41; p = 0.014). Monthly kit volume averaged 420.5 (SD: 93.4), peaking at 523 in December 2025. No reported data loss or critical performance failures occurred; external audits yielded no traceability-related findings. CONCLUSIONS In this single-site observational implementation study, TraceLab was associated with short documentation times, clean audited records, and the handling of variable workloads without reported critical performance failures. These findings support operational feasibility in this setting but do not establish causal effects or universal transferability.