Chapter 2 - Data, People, and Systems

The Practitioner's Guide to Data Quality Improvement

High-quality health information is critical for quality health care and for effective and efficient management of the health care system. This chapter highlights the Canadian experience in the capture and use of health information including a brief introduction to the Canadian health care system and the Canadian Institute for Health Information (CIHI); an overview of CIHI strategies and programs that support quality health information for key stakeholders, including clinicians, health system managers and policymakers; and four case studies from across the health care continuum illustrating data quality strategies in action—prevention, monitoring, feedback and continuous improvement. This chapter concludes with a discussion of two key data quality opportunities for CIHI—the movement toward interoperable electronic health records and data integration. Further information on CIHI, its data quality program and the case studies in this chapter may be found at www.cihi.ca

BackgroundResearchers have increasingly adopted AI and next-generation sequencing (NGS), revolutionizing genomics and high-throughput screening (HTS), and transforming our understanding of cellular processes and disease mechanisms. However, these advancements generate vast datasets requiring effective data stewardship and curation practices to maintain data integrity, privacy, and accessibility. This review consolidates existing knowledge on key aspects, including data governance, quality management, privacy measures, ownership, access control, accountability, traceability, curation frameworks, and storage systems.MethodsWe conducted a systematic literature search up to January 10, 2024, across PubMed, MEDLINE, EMBASE, Scopus, and additional scholarly platforms to examine recent advances and challenges in managing the vast and complex datasets generated by these technologies. Our search strategy employed structured keyword queries focused on four key thematic areas: data governance and management, curation frameworks, algorithmic bias and fairness, and data storage, all within the context of AI applications in genomics and microscopy. Using a realist synthesis methodology, we integrated insights from diverse frameworks to explore the multifaceted challenges associated with data stewardship in these domains. Three independent reviewers, who systematically categorized the information across critical themes, including data governance, quality management, security, privacy, ownership, and access control conducted data extraction and analysis. The study also examined specific AI considerations, such as algorithmic bias, model explainability, and the application of advanced cryptographic techniques. The review process included six stages, starting with an extensive search across multiple research databases, resulting in 273 documents. Screening based on broad criteria, titles, abstracts, and full texts followed this, narrowing the pool to 38 highly relevant citations.ResultsOur findings indicated that significant research was conducted in 2023 by highlighting the increasing recognition of robust data governance frameworks in AI-driven genomics and microscopy. While 36 articles extensively discussed data interoperability and sharing, AI-model explain ability and data augmentation remained underexplored, indicating significant gaps. The integration of diverse data types—ranging from sequencing and clinical data to proteomic and imaging data—highlighted the complexity and expansive scope of AI applications in these fields. The current challenges identified in AI-based data stewardship and curation practices are lack of infrastructure and cost optimization, ethical and privacy considerations, access control and sharing mechanisms, large scale data handling and analysis and transparent data-sharing policies and practice. Proposed solutions to address issues related to data quality, privacy, and bias management include advanced cryptographic techniques, federated learning, and blockchain technology. Robust data governance measures, such as GA4GH standards, DUO versioning, and attribute-based access control, are essential for ensuring data integrity, security, and ethical use. The study also emphasized the critical role of Data Management Plans (DMPs), meticulous metadata curation, and advanced cryptographic techniques in mitigating risks related to data security and identifiability. Despite advancements, significant challenges persisted in balancing data ownership with research accessibility, integrating heterogeneous data sources, ensuring platform interoperability, and maintaining data quality. Ongoing risks of unauthorized access and data breaches underscored the need for continuous innovation in data management practices and stricter adherence to legal and ethical standards.ConclusionsThese findings explored the current practices and challenges in data stewardship, offering a roadmap for strengthening the governance, security, and ethical use of AI in genomics and microscopy. While robust governance frameworks and ethical practices have established a foundation for data integrity and transparency, there remains an urgent need for collaborative efforts to develop interoperable platforms and transparent data-sharing policies. Additionally, evolving legal and ethical frameworks will be crucial to addressing emerging challenges posed by AI technologies. Fostering transparency, accountability, and ethical responsibility within the research community will be key to ensuring trust and driving ethically sound scientific advancements.

OBJECTIVE: To describe all stages of developing strategies and quality control in the Aging and Brain Working Study linked to the Estudo Longitudinal de Saúde do Adulto (ELSA-Brasil – Brazilian Longitudinal Study of Adult Health), a prospective cohort of 15,105 civil servants (aged 35–74) followed up since 2008. METHODS: Based on the sample of participants alive at the beginning of the sample selection in September 2022 (n = 4,566) from ELSA-Brasil (Centro Investigação São Paulo), a representative sample of 2,165 individuals was calculated according to the following criteria and divided into two subsamples: (1) Sample of participants < 70 years old in Wave 3 (2017–2019) randomized according to cognitive trajectory from Wave 1 to 3, n = 1,670; (2) SuperAgers case-control sample, individuals aged ≥ 70 years in Wave 3, based on episodic memory in Wave 3, n = 495 participants (171 cases, 324 controls). The processes implemented to ensure quality control of information prior to data collection were selection of data collection instruments, training and certification of teams, preparation of a manual of definitions and standardized operating procedures, pilot studies, data collection logistics, harmonization of imaging protocols, and creation of an anti-noise device. Given the scope of the study, its multicenter nature, and the diversity of measures involved, this project required the implementation of effective quality assurance and control protocols, ensuring standardization, methodological consistency, and data integrity at all stages. RESULTS: After implementing the quality assurance and control processes, the final stages related to clinical and imaging data collection, management, and continuous monitoring of f ield activities were conducted, with weekly reports and statistical monitoring of the sample. CONCLUSIONS: The adoption of systematic stages of development and quality control was fundamental to ensure the production and reliability of information from brain neuroimaging studies.

The Data Standards Initiatives, the SDDS and the GDDS, have achieved the goals the Executive Board set in its Fifth Review of July 2003. The staff sees the next three years as a period of consolidating these gains by maintaining the credibility of the SDDS through improved monitoring of countries’ observance of its requirements, and further integrating both the SDDS and GDDS under the Fund’s Data Quality Program (DQP) by aligning their structure with the Fund’s Data Quality Assessment Framework (DQAF). The staff proposes to include no new data categories in the SDDS and GDDS. Instead, the staff proposes to deepen descriptive information on how countries cover oil and gas activities and products in selected existing data categories.

• High-quality rangeland data are critical to supporting adaptive management. However, concrete, cost-saving steps to ensure data quality are often poorly defined and understood. • Data quality is more than data management. Ensuring data quality requires 1) clear communication among team members; 2) appropriate sample design; 3) training of data collectors, data managers, and data users; 4) observer and sensor calibration; and 5) active data management. Quality assurance and quality control are ongoing processes to help rangeland managers and scientists identify, prevent, and correct errors in past, current, and future monitoring data. • We present 10 guiding data quality questions to help managers and scientists identify appropriate workflows to improve data quality by 1) describing the data ecosystem, 2) creating a data quality plan, 3) identifying roles and responsibilities, 4) building data collection and data management workflows, 5) training and calibrating data collectors, 6) detecting and correcting errors, and 7) describing sources of variability. • Iteratively improving rangeland data quality is a key part of adaptive monitoring and rangeland data collection. All members of the rangeland community are invited to participate in ensuring rangeland data quality.

Ribonucleoprotein (RNP) granules transport, store, or degrade messenger RNAs, thereby indirectly regulating protein synthesis. Normally, RNP granules are highly dynamic compartments. However, because of aging or severe environmental stress, RNP granules, in particular stress granules (SGs), convert into solid, aggregate-like inclusions. There is increasing evidence that such RNA-protein inclusions are associated with several age-related neurodegenerative diseases, such as amyotrophic lateral sclerosis (ALS), fronto-temporal dementia (FTD) and Alzheimer’s disease (AD). Thus, understanding what triggers the conversion of RNP granules into aggregates and identifying the cellular players that control RNP granules will be critical to develop treatments for these diseases. In this review article, we discuss recent insight into RNP and SG formation. More specifically, we examine the evidence for liquid-liquid phase separation (LLPS) as an organizing principle of RNP granules and the role of aggregation-prone RNA-binding proteins (RBPs) in this process. We further discuss recent findings that liquid-like SGs can sequester misfolded proteins, which promote an aberrant conversion of liquid SGs into solid aggregates. Importantly, very recent studies show that a specific protein quality control (PQC) process prevents the accumulation of misfolding-prone proteins in SGs and, by doing so, maintains the dynamic state of SGs. This quality control process has been referred to as granulostasis and it relies on the specific action of the HSPB8-BAG3-HSP70 complex. Additional players such as p97/valosin containing protein (VCP) and other molecular chaperones (e.g., HSPB1) participate, directly or indirectly, in granulostasis, and ensure the timely elimination of defective ribosomal products and other misfolded proteins from SGs. Finally, we discuss recent findings that, in the stress recovery phase, SGs are preferentially disassembled with the assistance of chaperones, and we discuss evidence for a back-up system that targets aberrant SGs to the aggresome for autophagy-mediated clearance. Altogether the findings discussed here provide evidence for an intricate network of interactions between RNP granules and various components of the PQC machinery. Molecular chaperones in particular are emerging as key players that control the composition and dynamics of RNP granules, which may be important to protect against age-related diseases.

The capabilities of a new microcomputer system are reviewed. This system allows easy data management for both geologists and processing geophysicists. Currently available peripherals include a digitizer, pen plotter, raster graphics printer, and graphics video. Communications allow hardwired or modern access to other computers. End_Page 1704------------------------------ The application of software centers around a unified data management system and is extremely menu oriented, allowing easy use by personnel unfamiliar with computers. Applications software serves to assist in data acquisition, quality control, and computation for both seismic processing and interpretation. These applications include handling of such data as geometry, velocities, and muting, as well as geologic applications such as digitizing horizons, storing and plotting regional data, and digitizing and processing well log data. End_of_Article - Last_Page 1705------------

This article examines the impact of technological innovations on employee retention in the manufacturing sectors, focusing on a case study of quality control. It considers how technology, by reducing the monotony of manual labor and increasing efficiency, can enhance employee experiences. The authors explore the application of innovative technology in the context of quality control, highlighting the challenges associated with its implementation, such as safety, costs, and the need for employee training. The article emphasizes how the proper use of technology can lead to increased job satisfaction, reduced employee turnover, and improved productivity, by presenting the case of a company that has successfully implemented technologies into its quality control processes. The authors note that the effective implementation of technological innovations requires a strategic approach that considers both technological and human aspects.

Industry 4.0 is generally defined as a development system that compels the digitalization of processes to create integrated and autonomous systems. The process tracking of parts is very important in terms of detecting missed faulty products. Some defects that escape from quality control directly affect the end-user. Machine learning algorithms have been used to predict changes in the quality control processes and defective products, toward real-time and effective data processing. Thus, the highest quality of the final product will be delivered to the customer and to reduce the defective production coming out of the manufacturing chain. In this article, the study aims to establish a predictive quality model that can detect defect-free approved but faulty products overlooked during the quality inspection operations. Machine learning methods are used to analyze the relationship between quality control data and customer complaints. For this purpose, we use the last quality stage data of an automobile manufacturer’s brake system from 2018 to 2020. Machine learning models are constructed using logistic regression, ridge regression, support vector machine, random forest classification tree, gradient boost, XGBoost, LightGBM, and CatBoost algorithms. The results of specificity and negative prediction value show that the Gradient Boost and CatBoost algorithms have the best classification benefit for detecting the rare events.KeywordsQuality 4.0Industry 4.0Predictive qualityAutomobile industryMachine learningFault detectionRare event detection

As we have presented, it is evident that cohort studies are confronted with their own special, non-trivial issues of quality assurance and quality control. Such studies are typically large-scale designs and involve an extensive amount of data to be collected and processed, the quality of which depends on a variety of factors related to study personnel and equipment. The fact that data are collected over an extended period of time and at several centers greatly increases the magnitude of the data processing task, significantly increasing the likelihood of discrepancies and measurement error in the data. As presented in tables 1 and 2, the quality assurance and quality control procedures span the entire course of the study and include a multitude of tasks. Such tasks are delegated to various committees and/or are undertaken by participating centers, all of which must take responsibility for understanding, implementing, and following through on all procedures that maximize data quality. The quality of the quality assurance/quality control process is highly correlated with the quality of the communication within and between centers and all researchers. Maintaining standardization of procedures across centers and long-term stability of equipment and analytic procedures are integral components of quality control. In conclusion, the magnitude of the quality control process in a multicenter longitudinal study should not be underestimated, requiring a significant commitment of study resources. The quality control process is key to the integrity of the study, and an integral part of the design of the study. In a well-designed study, with a good quality control process and dedication to the process by the research team, the validity of the conclusions of the cohort study can be established.

Data quality plays a vital role in scientific research and decision-making across industries. Thus, it is crucial to incorporate the data quality control (DQC) process, which comprises various actions and operations to detect and correct data errors. The increasing adoption of machine learning (ML) techniques in different domains has raised concerns about data quality in the ML field. Conversely, ML’s capability to uncover complex patterns makes it suitable for addressing challenges involved in the DQC process. However, supervised learning methods demand abundant labeled data, while unsupervised learning methods heavily rely on the underlying distribution of the data. Active learning (AL) provides a promising solution by proactively selecting data points for inspection, thus reducing the burden of data labeling for domain experts. Therefore, this survey focuses on applying AL to DQC. Starting with a review of common data quality issues and solutions in the ML field, we aim to enhance the understanding of current quality assessment methods. We then present two scenarios to illustrate the adoption of AL into the DQC systems on the anomaly detection task, including pool-based and stream-based approaches. Finally, we provide the remaining challenges and research opportunities in this field.

A proposed methodology for quick assessment of timestamp and quality control results of solar radiation data

Objective To explore the application effect of the whole process quality control on the Da Vinci surgical instruments cleaning. Methods Using objective sampling method, 312 Da Vinci surgical instruments handled from July 2016 to September 2016 were selected as the control group, while 513 instruments handled from October 2016 to December 2016 were recruited into the observation group. The control group used routine cleaning and sterilization method. Meanwhile, the observation group received the whole process quality control. The rate of re-cleaning for Da Vinci surgical instruments were compared between two groups. Results The rate of re-cleaning for Da Vinci surgical instruments was 3.8% in the control group, which was higher than that (1.8%) in the observation group. The difference between two groups was statistically significant (χ2=3.423, P<0.05) . Conclusions The application of the whole process quality control on the Da Vinci surgical instruments cleaning can improve the service quality of central sterile supply and guarantee the smooth process of Da Vinci-assisted surgery. Key words: Robotics; Da Vinci; Whole process quality control; Cleaning quality

BackgroundHigh-quality immunization data is critical for making evidence-based decisions, ensuring effective program implementation, and achieving intended outcomes. Conversely, low-quality data hinders program monitoring and evaluation activities. Despite efforts to improve data quality in the health sector, the status of immunization data quality in Ethiopia is not well-known. This national study examined immunization data quality status and its challenges in Ethiopia.MethodsA mixed-method study was conducted on health facilities and all levels of administrative units, including the Ministry of Health, 12 regional health bureaus, 51 zonal health departments, 233 district health offices, 395 health facilities, and 267 health posts from April 15 to July 30, 2023. In-depth interviews were conducted with 74 purposefully selected key informants across these levels. The quality of immunization data was assessed based on accuracy, timeliness, and completeness. Open Data Kit (ODK) was used to collect quantitative data, and STATA statistical software was used for analysis. Qualitative data were analyzed using the thematic analysis approach with ATLAS.ti.ResultsOut of 4062 immunization data elements expected to be reported in the months under review, only 2413 (59.4%) were available. One-third of health facilities had acceptable reporting accuracy, while 27% had inaccurate reports. 41% of health facilities had inadequate documentation for verifications. Complete documentation and accuracy of reporting were better in hospitals than in health centers and health posts, but lower in pastoralist settings and some regions. Report timeliness ranged from 94% in the Oromia region to 37% in Gambella. The main reasons for low immunization data quality were inadequate skilled data management personnel, lack of standard data management and data quality assessment tools, and irregular data quality assessments at health facilities.ConclusionsThe study identified a significant immunization data quality problem in Ethiopia’s health institutions, with data accuracy, timeliness and completeness rates falling far below the standard. To improve immunization data quality and informed decision-making, it is crucial to strengthen data quality improvement strategies starting from the point of data collection.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13690-025-01758-2.