This dataset was designed and created to enable advancements in healthcare-focused large language models, particularly in the context of retrieval-augmented clinical question-answering capabilities. Developed using a self-constructed pipeline based on the 13-billion parameter Meta Llama 2 model, this dataset encompasses 21466 medical discharge summaries extracted from the MIMIC-IV-Note dataset, with 156599 synthetically generated question-and-answer pairs, a subset of which was verified for accuracy by a physician. These pairs were generated by providing the model with a discharge summary and instructing it to generate question-and-answer pairs based on the contextual information present in the summaries. This work aims to generate data in support of the development of compact large language models capable of efficiently extracting information from medical notes and discharge summaries, thus enabling potential improvements for real-time decision-making processes in clinical settings. Additionally, accompanying the dataset is code facilitating question-and-answer pair generation from any medical and non-medical text. Despite the robustness of the presented dataset, it has certain limitations. The generation process was confined to a maximum context length of 6000 input tokens, owing to hardware constraints. The large language model's nature in generating these question-and-answer pairs may introduce an underlying bias or a lack in diversity and complexity. Future iterations should focus on rectifying these issues, possibly through diversified training and expanded verification procedures as well as the employment of more powerful large language models.

Synthetic EHR Data Generation Platforms Market Outlook

According to our latest research, the global synthetic EHR data generation platforms market size reached USD 412.5 million in 2024, demonstrating robust momentum with a CAGR of 21.8% from 2025 to 2033. The market is projected to grow significantly to reach USD 2.98 billion by 2033, driven by the escalating demand for privacy-preserving data in healthcare analytics, research, and clinical trials. This growth is primarily fueled by the increasing adoption of artificial intelligence (AI) and machine learning (ML) technologies, stringent regulatory requirements for patient data privacy, and the growing need for high-quality, diverse datasets to train healthcare algorithms.

One of the primary growth factors for the synthetic EHR data generation platforms market is the surging emphasis on data privacy and security in the healthcare sector. As healthcare organizations transition to digital health records, concerns over patient confidentiality and compliance with regulations such as HIPAA and GDPR have intensified. Synthetic EHR data generation platforms offer a compelling solution by producing realistic, statistically accurate, yet entirely artificial datasets that eliminate the risk of exposing real patient information. This capability not only facilitates secure data sharing and collaboration among healthcare stakeholders but also supports the development of advanced analytics, AI-driven diagnostics, and personalized medicine initiatives without compromising patient privacy. The capacity to generate tailored datasets further empowers healthcare providers and researchers to address specific research questions, model rare diseases, and conduct robust clinical trials, all while adhering to the highest standards of data protection.

Another significant driver of market expansion is the growing integration of synthetic data within healthcare analytics, medical research, and clinical trial workflows. The traditional reliance on real-world patient data often encounters barriers such as incomplete records, data silos, and limited accessibility due to ethical and regulatory constraints. Synthetic EHR data generation platforms overcome these limitations by providing scalable, customizable, and bias-free datasets that enhance the accuracy and generalizability of predictive models and research findings. This is particularly valuable in the context of AI and ML, where large, diverse, and high-quality datasets are essential for algorithm training and validation. The ability to simulate a wide range of clinical scenarios, demographic profiles, and disease patterns accelerates innovation in drug discovery, epidemiological studies, and population health management, ultimately contributing to improved patient outcomes and healthcare system efficiency.

The market is further bolstered by the increasing adoption of cloud-based solutions, which enable seamless integration, scalability, and remote accessibility for organizations of all sizes. Cloud deployment models facilitate real-time collaboration among geographically dispersed teams, reduce infrastructure costs, and support continuous updates and improvements in synthetic data generation algorithms. Additionally, the rise of value-based care models and the proliferation of digital health ecosystems have heightened the demand for interoperable, high-fidelity synthetic EHR data that can be leveraged across multiple platforms and applications. As healthcare providers, pharmaceutical companies, and academic institutions seek to optimize their research and operational workflows, synthetic EHR data generation platforms are emerging as a cornerstone technology for driving data-driven decision-making, regulatory compliance, and innovation.

Regionally, North America dominates the synthetic EHR data generation platforms market, accounting for the largest share due to its advanced healthcare infrastructure, strong regulatory framework, and early adoption of digital health technologies. Europe follows closely, propelled by stringent data protection laws and a vibrant research ecosystem. The Asia Pacific region is expected to witness the fastest growth over the forecast period, driven by increasing investments in healthcare IT, expanding clinical research activities, and rising awareness of data privacy issues. Latin America and the Middle East & Africa are also poised for steady growth, supported by ongoing digital transformation initiatives and efforts to enhance healthcare data interoperability and security. Overall, the global market is character

🏥 Clinical Tabular Dataset (Non-PII, Realistic, Research-Ready)

Subtitle

Dataset mimicking real-world patient records for AI research.

Overview

This dataset is a synthetically generated clinical tabular dataset designed to closely mimic real-world patient health records while ensuring zero personally identifiable information (PII). It was created using statistical distributions, clinical guidelines, and publicly available medical references to replicate patterns typically observed in hospital and outpatient settings.

Unlike real EHR datasets, this synthetic dataset is free from privacy restrictions, making it safe to use for AI/ML model training, benchmarking, academic research, and prototyping healthcare applications.

Dataset Details

• Number of Records: 10000 patients (scalable to millions)
• Data Type: Structured, tabular
• Format: CSV (Comma-separated values)
• Domain: Healthcare / Clinical informatics
• PII Free: ✅ (No names, IDs, or sensitive personal details)

🔍 Columns & Clinical Context Age, Sex, BMI — basic demographics Vitals: Systolic/Diastolic BP, Glucose, Cholesterol, Creatinine Comorbidities: Diabetes, Hypertension Diagnosis: Normal, Pneumonia, Heart Failure, Sepsis Outcomes: 30-day Readmission, Mortality

Applications

This dataset can be used for:

• Machine Learning: Classification, clustering, regression models
• Healthcare AI: Predictive modeling for risk factors and disease detection
• Data Science Education: Hands-on exercises for students
• Synthetic Data Research: Benchmarking synthetic data generation approaches
• Fairness & Bias Testing: Evaluating ML models across age, gender, and lifestyle groups

Why This Dataset?

• Realistic: Matches clinical ranges and distributions found in actual healthcare data
• Safe to Share: 100% synthetic, no HIPAA/GDPR concerns
• Flexible: Can be scaled, modified, or extended with more medical variables
• High Impact: Fills a major gap in openly available clinical tabular datasets

Disclaimer

This dataset is synthetic and for research/educational purposes only. It should not be used for medical decision-making or clinical care.

Citation

If you use this dataset, please cite as:

Synthetic Clinical Tabular Dataset (2025). Generated for ML research and benchmarking.

Introduction: Electronic health records (EHR) and administrative healthcare data (AHD) are frequently used in geriatric mental health research to answer various health research questions. However, there is an increasing amount and complexity of data available that may lend itself to alternative analytic approaches using machine learning (ML) or artificial intelligence (AI) methods. We performed a systematic review of the current application of ML or AI approaches to the analysis of EHR and AHD in geriatric mental health.Methods: We searched MEDLINE, Embase, and PsycINFO to identify potential studies. We included all articles that used ML or AI methods on topics related to geriatric mental health utilizing EHR or AHD data. We assessed study quality either by Prediction model Risk OF Bias ASsessment Tool (PROBAST) or Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) checklist.Results: We initially identified 391 articles through an electronic database and reference search, and 21 articles met inclusion criteria. Among the selected studies, EHR was the most used data type, and the datasets were mainly structured. A variety of ML and AI methods were used, with prediction or classification being the main application of ML or AI with the random forest as the most common ML technique. Dementia was the most common mental health condition observed. The relative advantages of ML or AI techniques compared to biostatistical methods were generally not assessed. Only in three studies, low risk of bias (ROB) was observed according to all the PROBAST domains but in none according to QUADAS-2 domains. The quality of study reporting could be further improved.Conclusion: There are currently relatively few studies using ML and AI in geriatric mental health research using EHR and AHD methods, although this field is expanding. Aside from dementia, there are few studies of other geriatric mental health conditions. The lack of consistent information in the selected studies precludes precise comparisons between them. Improving the quality of reporting of ML and AI work in the future would help improve research in the field. Other courses of improvement include using common data models to collect/organize data, and common datasets for ML model validation.

BackgroundMedication error (MedE) is a leading global cause of harm in human healthcare with significance both in patient morbidity and mortality, and consequent legal and financial issues. Despite this, MedEs are a poorly explored area in veterinary medicine. Research has so far focussed on survey work and errors spontaneously reported to third parties, such as professional indemnity providers.AimDetermine if MedEs can be successfully identified in first opinion electronic health records (EHRs).AnimalsEHRs pertaining to animals treated in UK first opinion practice.Materials and methodsRegular expressions (REGEX) were designed (with assistance from a domain expert) to identify explicit reference to MedEs in the SAVSNET EHR dataset. Identified MedEs were then classified by the linear sequence of medication therapy, the degree of harm caused, the role of the person who made the error, and the medication type involved.ResultsIn total, 6,665 EHRs were identified by the REGEX, of which a random 2,847 were manually reviewed, with 1,023 (35.9%) matching the MedEs case definition. Of these MedEs, 29.5% (n = 302) caused mild harm to the patient, 2.8% (n = 27) moderate harm and 0.2% (n = 2) severe harm. MedEs were most frequent during the “drug administered” phase (51.4%) and within this phase, “dosing errors” were most common (68.1%). The most common medication types, associated with “drug administered” phase MedEs were vaccinations (27.1%) and non-steroidal anti-inflammatory drugs (19.0%).ConclusionEHRs are a useful source of data on MedEs. MedEs are a common cause of patient harm in veterinary practice. The data provided here highlights drug classes at higher risk of problems for which mitigating action and/or education interventions are indicated.

ObjectiveTo perform a systematic review examining the variation in methods, results, reporting and risk of bias in electronic health record (EHR)-based studies evaluating management of a common musculoskeletal disease, gout.MethodsTwo reviewers systematically searched MEDLINE, Scopus, Web of Science, CINAHL, PubMed, EMBASE and Google Scholar for all EHR-based studies published by February 2019 investigating gout pharmacological treatment. Information was extracted on study design, eligibility criteria, definitions, medication usage, effectiveness and safety data, comprehensiveness of reporting (RECORD), and Cochrane risk of bias (registered PROSPERO CRD42017065195).ResultsWe screened 5,603 titles/abstracts, 613 full-texts and selected 75 studies including 1.9M gout patients. Gout diagnosis was defined in 26 ways across the studies, most commonly using a single diagnostic code (n = 31, 41.3%). 48.4% did not specify a disease-free period before ‘incident’ diagnosis. Medication use was suboptimal and varied with disease definition while results regarding effectiveness and safety were broadly similar across studies despite variability in inclusion criteria. Comprehensiveness of reporting was variable, ranging from 73% (55/75) appropriately discussing the limitations of EHR data use, to 5% (4/75) reporting on key data cleaning steps. Risk of bias was generally low.ConclusionThe wide variation in case definitions and medication-related analysis among EHR-based studies has implications for reported medication use. This is amplified by variable reporting comprehensiveness and the limited consideration of EHR-relevant biases (e.g. data adequacy) in study assessment tools. We recommend accounting for these biases and performing a sensitivity analysis on case definitions, and suggest changes to assessment tools to foster this.

The acute-care pathway (from the emergency department (ED) through acute medical units or ambulatory care and on to wards) is the most visible aspect of the hospital health-care system to most patients. Acute hospital admissions are increasing yearly and overcrowded emergency departments and high bed occupancy rates are associated with a range of adverse patient outcomes. Predicted growth in demand for acute care driven by an ageing population and increasing multimorbidity is likely to exacerbate these problems in the absence of innovation to improve the processes of care.

Key targets for Emergency Medicine services are changing, moving away from previous 4-hour targets. This will likely impact the assessment of patients admitted to hospital through Emergency Departments.

This data set provides highly granular patient level information, showing the day-to-day variation in case mix and acuity. The data includes detailed demography, co-morbidity, symptoms, longitudinal acuity scores, physiology and laboratory results, all investigations, prescriptions, diagnoses and outcomes. It could be used to develop new pathways or understand the prevalence or severity of specific disease presentations.

PIONEER geography: The West Midlands (WM) has a population of 5.9 million & includes a diverse ethnic & socio-economic mix.

Electronic Health Record: University Hospital Birmingham is one of the largest NHS Trusts in England, providing direct acute services & specialist care across four hospital sites, with 2.2 million patient episodes per year, 2750 beds & an expanded 250 ITU bed capacity during COVID. UHB runs a fully electronic healthcare record (EHR) (PICS; Birmingham Systems), a shared primary & secondary care record (Your Care Connected) & a patient portal “My Health”.

Scope: All patients with a medical emergency admitted to hospital, flowing through the acute medical unit. Longitudinal & individually linked, so that the preceding & subsequent health journey can be mapped & healthcare utilisation prior to & after admission understood. The dataset includes patient demographics, co-morbidities taken from ICD-10 & SNOMED-CT codes. Serial, structured data pertaining to process of care (timings, admissions, wards and readmissions), physiology readings (NEWS2 score and clinical frailty scale), Charlson comorbidity index and time dimensions.

Available supplementary data: Matched controls; ambulance data, OMOP data, synthetic data.

Available supplementary support: Analytics, Model build, validation & refinement; A.I.; Data partner support for ETL (extract, transform & load) process, Clinical expertise, Patient & end-user access, Purchaser access, Regulatory requirements, Data-driven trials, “fast screen” services.

EHR downtime impacts an estimated 13.2% of the U.S. population, disrupting access to patient health records and creating delays in clinical decision-making. Raw medical transcripts are often unstructured, inconsistent, and sensitive, making them difficult to use directly for research or AI applications. This leads to wasted time on preprocessing and limits the potential for advanced analytics.

This dataset provides cleaned and de-identified medical transcripts from MIMIC-IV, allowing researchers to focus on NLP, predictive modeling, and knowledge graph applications without the burden of raw data cleaning. By reducing barriers to analysis, it supports the development of tools that can improve healthcare efficiency and patient outcomes.

Applications: - Healthcare NLP (Named Entity Recognition, text classification) - Predictive modeling for admission/discharge outcomes - Analysis of patient demographics and clinical severity - AI-driven knowledge graph construction from structured + unstructured hospital data

Notes Data is de-identified to ensure HIPAA compliance Intended for research and educational purposes only Source: MIMIC-IV, MIT Laboratory for Computational Physiology

Synthesizing information from various data sources plays a crucial role in the practice of modern medicine. Current applications of artificial intelligence in medicine often focus on single-modality data due to a lack of publicly available, multimodal medical datasets. To address this limitation, we introduce INSPECT, which contains de-identified longitudinal records from a large cohort of pulmonary embolism (PE) patients, along with ground truth labels for multiple outcomes. INSPECT contains data from 19,402 patients, including 23,248 CT images, sections of radiology reports, and structured electronic health record (EHR) data (including demographics, diagnoses, procedures, and vitals). Using our provided dataset, we develop and release a benchmark for evaluating several baseline modeling approaches on a variety of important PE related tasks. We evaluate image-only, EHR-only, and fused models. Trained models and the de-identified dataset are made available for non-commercial use under a data use agreement. To the best our knowledge, INSPECT is the largest multimodal dataset for enabling reproducible research on strategies for integrating 3D medical imaging and EHR data. EHR modality data is uploaded to Stanford Redivis website (https://redivis.com/Stanford).

Digitization of healthcare data along with algorithmic breakthroughts in AI will have a major impact on healthcare delivery in coming years. Its intresting to see application of AI to assist clinicians during patient treatment in a privacy preserving way. While scientific knowledge can help guide interventions, there remains a key need to quickly cut through the space of decision policies to find effective strategies to support patients during the care process.

Offline Reinforcement learning (also referred to as safe or batch reinforcement learning) is a promising sub-field of RL which provides us with a mechanism for solving real world sequential decision making problems where access to simulator is not available. Here we assume that learn a policy from fixed dataset of trajectories with further interaction with the environment(agent doesn't receive reward or punishment signal from the environment). It has shown that such an approach can leverage vast amount of existing logged data (in the form of previous interactions with the environment) and can outperform supervised learning approaches or heuristic based policies for solving real world - decision making problems. Offline RL algorithms when trained on sufficiently large and diverse offline datasets can produce close to optimal policies(ability to generalize beyond training data).

As Part of my PhD, research, I investigated the problem of developing a Clinical Decision Support System for Sepsis Management using Offline Deep Reinforcement Learning.

MIMIC-III ('Medical Information Mart for Intensive Care') is a large open-access anonymized single-center database which consists of comprehensive clinical data of 61,532 critical care admissions from 2001–2012 collected at a Boston teaching hospital. Dataset consists of 47 features (including demographics, vitals, and lab test results) on a cohort of sepsis patients who meet the sepsis-3 definition criteria.

we try to answer the following question:

Given a particular patient’s characteristics and physiological information at each time step as input, can our DeepRL approach, learn an optimal treatment policy that can prescribe the right intervention(e.g use of ventilator) to the patient each stage of the treatment process, in order to improve the final outcome(e.g patient mortality)?

we can use popular state-of-the-art algorithms such as Deep Q Learning(DQN), Double Deep Q Learning (DDQN), DDQN combined with BNC, Mixed Monte Carlo(MMC) and Persistent Advantage Learning (PAL). Using these methods we can train an RL policy to recommend optimum treatment path for a given patient.

Data acquisition, standard pre-processing and modelling details can be found here in Github repo: https://github.com/asjad99/MIMIC_RL_COACH

BackgroundRoutine Data Quality Assessments (RDQAs) were developed to measure and improve facility-level electronic medical record (EMR) data quality. We assessed if RDQAs were associated with improvements in data quality in KenyaEMR, an HIV care and treatment EMR used at 341 facilities in Kenya.MethodsRDQAs assess data quality by comparing information recorded in paper records to KenyaEMR. RDQAs are conducted during a one-day site visit, where approximately 100 records are randomly selected and 24 data elements are reviewed to assess data completeness and concordance. Results are immediately provided to facility staff and action plans are developed for data quality improvement. For facilities that had received more than one RDQA (baseline and follow-up), we used generalized estimating equation models to determine if data completeness or concordance improved from the baseline to the follow-up RDQAs.Results27 facilities received two RDQAs and were included in the analysis, with 2369 and 2355 records reviewed from baseline and follow-up RDQAs, respectively. The frequency of missing data in KenyaEMR declined from the baseline (31% missing) to the follow-up (13% missing) RDQAs. After adjusting for facility characteristics, records from follow-up RDQAs had 0.43-times the risk (95% CI: 0.32–0.58) of having at least one missing value among nine required data elements compared to records from baseline RDQAs. Using a scale with one point awarded for each of 20 data elements with concordant values in paper records and KenyaEMR, we found that data concordance improved from baseline (11.9/20) to follow-up (13.6/20) RDQAs, with the mean concordance score increasing by 1.79 (95% CI: 0.25–3.33).ConclusionsThis manuscript demonstrates that RDQAs can be implemented on a large scale and used to identify EMR data quality problems. RDQAs were associated with meaningful improvements in data quality and could be adapted for implementation in other settings.

BackgroundApproximately 28% of adults have ≥3 chronic conditions (CCs), accounting for two-thirds of U.S. healthcare costs, and often having suboptimal outcomes. Despite Institute of Medicine recommendations in 2001 to integrate guidelines for multiple CCs, progress is minimal. The vast number of unique combinations of CCs may limit progress.Methods and findingsTo determine whether major CCs segregate differentially in limited groups, electronic health record and Medicare paid claims data were examined in one accountable care organization with 44,645 Medicare beneficiaries continuously enrolled throughout 2015. CCs predicting clinical outcomes were obtained from diagnostic codes. Agglomerative hierarchical clustering defined 13 groups having similar within group patterns of CCs and named for the most common CC. Two groups, congestive heart failure (CHF) and kidney disease (CKD), included 23% of beneficiaries with a very high CC burden (10.5 and 8.1 CCs/beneficiary, respectively). Five groups with 54% of beneficiaries had a high CC burden ranging from 7.1 to 5.9 (descending order: neurological, diabetes, cancer, cardiovascular, chronic pulmonary). Six groups with 23% of beneficiaries had an intermediate-low CC burden ranging from 4.7 to 0.4 (behavioral health, obesity, osteoarthritis, hypertension, hyperlipidemia, ‘other’). Hypertension and hyperlipidemia were common across groups, whereas 80% of CHF segregated to the CHF group, 85% of CKD to CKD and CHF groups, 82% of cancer to Cancer, CHF, and CKD groups, and 85% of neurological disorders to Neuro, CHF, and CKD groups. Behavioral health diagnoses were common only in groups with a high CC burden. The number of CCs/beneficiary explained 36% of the variance (R2 = 0.36) in claims paid/beneficiary.ConclusionsIdentifying a limited number of groups with high burdens of CCs that disproportionately drive costs may help inform a practical number of integrated guidelines and resources required for comprehensive management. Cluster informed guideline integration may improve care quality and outcomes, while reducing costs.

Cloud Electronic Health Records Market Outlook

According to the latest research conducted in 2025, the global cloud electronic health records (EHR) market size stands at USD 7.3 billion in 2024. The market is exhibiting robust momentum, driven by the accelerating digital transformation in healthcare, with a compound annual growth rate (CAGR) of 13.2% projected through the forecast period. By 2033, the market is anticipated to reach approximately USD 22.3 billion, highlighting the increasing adoption of cloud-based solutions across healthcare organizations globally. The primary growth factor fueling this expansion is the urgent need for interoperable, scalable, and cost-effective health information management systems, as healthcare providers strive to enhance patient care, streamline clinical workflows, and comply with evolving regulatory mandates.

The surge in demand for cloud EHR solutions is fundamentally underpinned by the global shift toward value-based healthcare and the growing emphasis on patient-centric care models. Healthcare organizations are increasingly recognizing the necessity of real-time access to patient data, not only for improving clinical decision-making but also for enhancing care coordination among multidisciplinary teams. The cloud-based architecture offers unparalleled advantages in terms of data accessibility, scalability, and integration capabilities, which are crucial for supporting telemedicine, population health management, and remote patient monitoring initiatives. Furthermore, the increasing prevalence of chronic diseases and the aging global population necessitate robust data management platforms, further fueling the adoption of cloud EHR systems.

Another significant growth driver is the rapid advancement in cloud computing technologies and the proliferation of health IT infrastructure. The integration of artificial intelligence (AI), machine learning, and advanced analytics into cloud EHR platforms is transforming the way healthcare data is captured, analyzed, and utilized. These technological innovations enable healthcare providers to derive actionable insights from vast datasets, optimize resource allocation, and personalize treatment plans. Additionally, the growing adoption of mobile health applications and wearable devices is generating a wealth of patient-generated health data, which can be seamlessly integrated into cloud EHR systems for holistic patient management. The flexibility and cost-efficiency offered by cloud deployment models are compelling even small and medium healthcare organizations to transition from legacy on-premises systems to cloud-based EHR solutions.

On the regulatory front, governments and healthcare authorities worldwide are implementing stringent data protection and interoperability standards, such as the Health Insurance Portability and Accountability Act (HIPAA) in the United States and the General Data Protection Regulation (GDPR) in Europe. These regulatory frameworks are compelling healthcare providers to adopt secure, compliant, and interoperable EHR solutions, further propelling market growth. However, the market's expansion is not uniform across all regions. North America continues to dominate the global landscape, owing to its advanced healthcare IT ecosystem and favorable reimbursement policies, while Asia Pacific is emerging as a high-growth market driven by healthcare digitization initiatives and rising investments in health infrastructure.

Product Type Analysis

The cloud electronic health records market is segmented by product type into standalone EHR and integrated EHR solutions. Standalone EHR systems are designed to function independently, offering core functionalities such as patient record management, appointment scheduling, and basic reporting. These solutions are particularly attractive to smaller healthcare facilities and clinics that require a cost-effective and easy-to-deploy platform without the complexities of broader system integration. However, standalone systems often face limitations in terms of scalability and interoperability, which can hinder their long-term viability as healthcare organizations grow or seek to connect with external partners and health information exchanges.

Integrated EHR solutions, on the other hand, are rapidly gaining traction due to their ability to seamlessly connect with other healthcare information systems, including laboratory information systems (LIS), radiology information systems (RIS), billing platforms, a

ONC established the SHARP program to support innovative research and to address well-documented problems that impede the adoption and use of health IT. The program covers four subject areas managed by four distinct project groups: health IT security (SHARPS), patient-centered cognitive support (SHARPc), health care application and network design (SMART), and secondary use of EHR information (SHARPn). This dataset provides the full inventory of project outputs from the SHARP program, ranging from presentations and manuscripts to APIs and software.

Abstract The Medical Information Mart for Intensive Care (MIMIC)-IV database is comprised of deidentified electronic health records for patients admitted to the Beth Israel Deaconess Medical Center. Access to MIMIC-IV is limited to credentialed users. Here, we have provided an openly-available demo of MIMIC-IV containing a subset of 100 patients. The dataset includes similar content to MIMIC-IV, but excludes free-text clinical notes. The demo may be useful for running workshops and for assessing whether the MIMIC-IV is appropriate for a study before making an access request.

Background The increasing adoption of digital electronic health records has led to the existence of large datasets that could be used to carry out important research across many areas of medicine. Research progress has been limited, however, due to limitations in the way that the datasets are curated and made available for research. The MIMIC datasets allow credentialed researchers around the world unprecedented access to real world clinical data, helping to reduce the barriers to conducting important medical research. The public availability of the data allows studies to be reproduced and collaboratively improved in ways that would not otherwise be possible.

Methods First, the set of individuals to include in the demo was chosen. Each person in MIMIC-IV is assigned a unique subject_id. As the subject_id is randomly generated, ordering by subject_id results in a random subset of individuals. We only considered individuals with an anchor_year_group value of 2011 - 2013 or 2014 - 2016 to ensure overlap with MIMIC-CXR v2.0.0. The first 100 subject_id who satisfied the anchor_year_group criteria were selected for the demo dataset.

All tables from MIMIC-IV were included in the demo dataset. Tables containing patient information, such as emar or labevents, were filtered using the list of selected subject_id. Tables which do not contain patient level information were included in their entirety (e.g. d_items or d_labitems). Note that all tables which do not contain patient level information are prefixed with the characters 'd_'.

Deidentification was performed following the same approach as the MIMIC-IV database. Protected health information (PHI) as listed in the HIPAA Safe Harbor provision was removed. Patient identifiers were replaced using a random cipher, resulting in deidentified integer identifiers for patients, hospitalizations, and ICU stays. Stringent rules were applied to structured columns based on the data type. Dates were shifted consistently using a random integer removing seasonality, day of the week, and year information. Text fields were filtered by manually curated allow and block lists, as well as context-specific regular expressions. For example, columns containing dose values were filtered to only contain numeric values. If necessary, a free-text deidentification algorithm was applied to remove PHI from free-text. Results of this algorithm were manually reviewed and verified to remove identified PHI.

Data Description MIMIC-IV is a relational database consisting of 26 tables. For a detailed description of the database structure, see the MIMIC-IV Clinical Database page [1] or the MIMIC-IV online documentation [2]. The demo shares an identical schema and structure to the equivalent version of MIMIC-IV.

Data files are distributed in comma separated value (CSV) format following the RFC 4180 standard [3]. The dataset is also made available on Google BigQuery. Instructions to accessing the dataset on BigQuery are provided on the online MIMIC-IV documentation, under the cloud page [2].

An additional file is included: demo_subject_id.csv. This is a list of the subject_id used to filter MIMIC-IV to the demo subset.

Usage Notes The MIMIC-IV demo provides researchers with the opportunity to better understand MIMIC-IV data.

CSV files can be opened natively using any text editor or spreadsheet program. However, as some tables are large it may be preferable to navigate the data via a relational database. We suggest either working with the data in Google BigQuery (see the "Files" section for access details) or creating an SQLite database using the CSV files. SQLite is a lightweight database format which stores all constituent tables in a single file, and SQLite databases interoperate well with a number software tools.

Code is made available for use with MIMIC-IV on the MIMIC-IV code repository [4]. Code provided includes derivation of clinical concepts, tutorials, and reproducible analyses.

Release Notes Release notes for the demo follow the release notes for the MIMIC-IV database.

Ethics This project was approved by the Institutional Review Boards of Beth Israel Deaconess Medical Center (Boston, MA) and the Massachusetts Institute of Technology (Cambridge, MA). Requirement for individual patient consent was waived because the pr...

This dataset contains 10,000 synthetic patient records representing a scaled-down US Medicare population. The records were generated by Synthea ( https://github.com/synthetichealth/synthea ) and are completely synthetic and contain no real patient data. This data is presented free of cost and free of restrictions. Each record is stored as one file in HL7 FHIR R4 ( https://www.hl7.org/fhir/ ) containing one Bundle, in JSON. For more information on how this specific population was created, or to generate your own at any scale, see: https://github.com/synthetichealth/populations/tree/master/medicare

Synthetic Data in Healthcare Market Outlook

According to our latest research, the global synthetic data in healthcare market size was valued at USD 420 million in 2024. The market is expected to grow at a robust CAGR of 32.4% from 2025 to 2033, reaching USD 4.7 billion by 2033. This rapid expansion is primarily driven by the increasing demand for privacy-preserving data solutions, the rising adoption of artificial intelligence (AI) and machine learning (ML) in healthcare, and the need to overcome data scarcity and bias in medical research and clinical applications. As per our latest research, the proliferation of digital health technologies and stringent regulatory frameworks for patient data privacy are further fueling the adoption of synthetic data solutions across the global healthcare ecosystem.

The growth of the synthetic data in healthcare market is underpinned by the urgent need for high-quality, diverse, and privacy-compliant datasets. Traditional healthcare data is often limited by privacy concerns, regulatory restrictions, and inherent biases that can hinder AI model training and clinical research. Synthetic data addresses these challenges by generating artificial datasets that mimic the statistical properties of real patient data without exposing sensitive information. This capability is particularly valuable for training AI algorithms in diagnostic imaging, drug discovery, and predictive analytics, where access to large, unbiased datasets is critical. The increasing adoption of electronic health records (EHRs) and the digitization of healthcare workflows have further amplified the demand for synthetic data, enabling healthcare organizations to accelerate innovation while maintaining compliance with data protection regulations such as HIPAA and GDPR.

Another key growth driver is the expanding application of AI and ML technologies in healthcare. As organizations strive to develop robust predictive models for disease diagnosis, treatment planning, and patient management, the need for large and diverse datasets has become paramount. Synthetic data not only supplements real-world data but also enables the simulation of rare disease cases and edge scenarios, enhancing the generalizability and reliability of AI models. This is particularly relevant in medical imaging, clinical trials, and drug development, where data variability and sample size limitations can impede progress. By leveraging synthetic data, healthcare stakeholders can accelerate R&D timelines, reduce costs, and improve the accuracy of AI-driven solutions, ultimately leading to better patient outcomes and operational efficiencies.

The global push for interoperability and data sharing across healthcare systems is also contributing to the market's growth. Synthetic data enables secure data exchange between institutions, facilitating collaborative research and multi-center studies without compromising patient privacy. This is especially important in the context of global health crises, such as pandemics, where rapid data sharing and analysis are essential for effective response and decision-making. Moreover, the integration of synthetic data with real-world evidence is helping pharmaceutical companies, research institutions, and regulatory bodies to enhance the design and execution of clinical trials, optimize resource allocation, and streamline drug approval processes. As the healthcare industry continues to embrace digital transformation, the role of synthetic data in enabling secure, scalable, and innovative solutions is expected to expand significantly.

Regionally, North America leads the synthetic data in healthcare market, accounting for the largest revenue share in 2024, followed by Europe and Asia Pacific. The dominance of North America can be attributed to the presence of advanced healthcare infrastructure, robust investments in AI and digital health, and supportive regulatory frameworks for data privacy and security. Europe is also witnessing significant growth, driven by increasing adoption of AI in medical research and strong emphasis on patient data protection under GDPR. Meanwhile, Asia Pacific is emerging as a high-growth region, fueled by rapid healthcare digitization, government initiatives to promote AI adoption, and a burgeoning pharmaceutical sector. Latin America and the Middle East & Africa are gradually catching up, with growing awareness of the benefits of synthetic data and increasing investments in healthcare IT.

Overview

New York, NY – Aug 08, 2025 : Global Cloud-based EHR Market is expected to grow significantly over the next decade. It is projected to reach approximately US$ 82.49 Billion by 2034, up from US$ 35.82 Billion in 2024. This growth reflects a steady compound annual growth rate (CAGR) of 8.7% from 2025 to 2034.

The rising demand for digital healthcare infrastructure and scalable IT solutions continues to drive this expansion. Healthcare providers increasingly prefer flexible, cloud-hosted systems that offer secure and efficient management of electronic health records without heavy upfront investment.

Cloud computing refers to delivering hosted services over the internet, either via public or private models. Public cloud services, typically offered by third-party providers, allow organizations to access resources on a pay-as-you-go basis. These services include computing power, data storage, and bandwidth. Healthcare organizations often choose public cloud options due to their cost efficiency and scalability. This model enables medical facilities to allocate IT budgets more effectively by avoiding the large capital expenditure required for traditional, in-house IT systems.

One of the main advantages of cloud-based EHR systems is cost reduction. Unlike conventional systems that require costly infrastructure, cloud EHRs are subscription-based. Some providers offer services starting as low as US$ 100 per month. This makes cloud EHRs attractive, especially to smaller healthcare practices. A survey by Black Book revealed that 83% of small practices found cloud-based EHR implementation to be a top business decision. These systems help practices stay current with evolving health IT standards without straining their financial resources.

Scalability is another major factor boosting the adoption of cloud-based EHRs. These platforms allow healthcare systems to expand or adjust IT infrastructure as needed. For organizations with multiple locations, cloud solutions enable real-time access to patient data across sites. This supports improved clinical decision-making and interoperability. Seamless updates and centralized data management enhance coordination among care teams. As healthcare networks grow more complex, cloud systems offer a reliable and adaptable solution to manage patient information efficiently.

In October 2024, Oracle showcased its next-generation EHR at the Oracle Health Summit. This new system is built on Oracle Cloud Infrastructure (OCI) and incorporates AI to enhance clinical workflows. Features include automation of documentation, appointment preparation, and follow-up tasks. With military-grade security and high performance, the platform aims to improve provider efficiency and care delivery. Oracle’s investment highlights a broader industry shift toward intelligent, secure, cloud-based systems. As AI and cloud technologies converge, more healthcare providers are expected to adopt similar advanced EHR platforms.

BackgroundThe ability to apply standard and interoperable solutions for implementing and managing medical registries as well as aggregate, reproduce, and access data sets from legacy formats and platforms to advanced standard formats and operating systems are crucial for both clinical healthcare and biomedical research settings.PurposeOur study describes a reproducible, highly scalable, standard framework for a device registry implementation addressing both local data quality components and global linking problems.Methods and ResultsWe developed a device registry framework involving the following steps: (1) Data standards definition and representation of the research workflow, (2) Development of electronic case report forms using REDCap (Research Electronic Data Capture), (3) Data collection according to the clinical research workflow and, (4) Data augmentation by enriching the registry database with local electronic health records, governmental database and linked open data collections, (5) Data quality control and (6) Data dissemination through the registry Web site. Our registry adopted all applicable standardized data elements proposed by American College Cardiology / American Heart Association Clinical Data Standards, as well as variables derived from cardiac devices randomized trials and Clinical Data Interchange Standards Consortium. Local interoperability was performed between REDCap and data derived from Electronic Health Record system. The original data set was also augmented by incorporating the reimbursed values paid by the Brazilian government during a hospitalization for pacemaker implantation. By linking our registry to the open data collection repository Linked Clinical Trials (LinkedCT) we found 130 clinical trials which are potentially correlated with our pacemaker registry.ConclusionThis study demonstrates how standard and reproducible solutions can be applied in the implementation of medical registries to constitute a re-usable framework. Such approach has the potential to facilitate data integration between healthcare and research settings, also being a useful framework to be used in other biomedical registries.

ObjectiveTo develop a non-invasive, radiation-free model for early colorectal adenoma prediction using clinical electronic medical record (EMR) data, addressing limitations in current diagnostic approaches for large-scale screening.DesignRetrospective analysis utilized 92,681 cases with EMR, spanning from 2012 to 2022, as the training cohort. Testing was performed on an independent test cohort of 19,265 cases from 2023. Several classical machine learning algorithms were applied in combination with the BGE-M3 large-language model (LLM) for enhanced semantic feature extraction. Area under the receiver operating characteristic curve (AUC) is the major metric for evaluating model performance. The Shapley additive explanations (SHAP) method was employed to identify the most influential risk factors.ResultsXGBoost algorithm, integrated with BGE-M3, demonstrated superior performance (AUC = 0.9847) in the validation cohort. Notably, when applied to the independent test cohort, XGBoost maintained its strong predictive ability with an AUC of 0.9839 and an average advance prediction time of 6.88 hours, underscoring the effectiveness of the BGE-M3 model. The SHAP analysis further identified 16 high-impact risk factors, highlighting the interplay of genetic, lifestyle, and environmental influences on colorectal adenoma risk.ConclusionThis study developed a robust machine learning-based model for colorectal adenoma risk prediction, leveraging clinical EMR and LLM. The proposed model demonstrates high predictive accuracy and has the potential to enhance early detection, making it well-suited for large-scale screening programs. By facilitating early identification of individuals at risk, this approach may contribute to reducing the incidence and mortality associated with colorectal cancer.