Automated Versus Semi-Automated Lab Value Extraction for the VA Cardiac Surgical Quality Improvement Program.

INTRODUCTION: The Veterans Affairs Surgical Quality Improvement Program (VASQIP) trains surgical quality nurses (SQNs) at each Veterans Affairs (VA) hospital to extract or verify 187 variables from the medical record for all cardiac surgical cases. For ten preoperative laboratory values, VASQIP has a semiautomated (SA) system in which local lab values are automatically extracted, verified by SQNs, and lab values recorded at other VA facilities are manually extracted. The objective of this study was to develop and validate a method to automate the extraction of these ten preoperative laboratory values and compare results with the current SA method. MATERIALS AND METHODS: We developed methods to extract ten preoperative laboratory values and measurement dates from the VA Corporate Data Warehouse using Logical Observation Identifiers Names and Codes. Automated (A) versus SA information extraction was compared in terms of agreement, conformance to data definitions, proximity to surgery, and missingness. RESULTS: For surgeries with both A and SA lab values, the intraclass correlation coefficients for the ten variables ranged from 0.90 to 0.98. For several variables, the A method resulted in much lower rates of missing data (e.g., 2.4% versus 22.5% missing data for high-density lipoprotein) and eliminated out-of-date-range entries. CONCLUSIONS: Although SQN-extracted data are widely considered the gold standard within National Surgical Quality Improvement Programs, there may be advantages to fully automating extraction of lab values, including high congruence with SA SQN-extracted or verified values and lower rates of missingness and out-of-date-range data.

Response rates in clinical quality registries and databases that collect patient reported outcome measures: a scoping review.

BACKGROUND: Patient Reported Outcome Measures (PROMs) are being increasingly introduced in clinical registries, providing a personal perspective on the expectations and impact of treatment. The aim of this study was to describe response rates (RR) to PROMs in clinical registries and databases and to examine the trends over time, and how they change with the registry type, region and disease or condition captured. METHODS: We conducted a scoping literature review of MEDLINE and EMBASE databases, in addition to Google Scholar and grey literature. All English studies on clinical registries capturing PROMs at one or more time points were included. Follow up time points were defined as follows: baseline (if available), < 1 year, 1 to < 2 years, 2 to < 5 years, 5 to < 10 years and 10 + years. Registries were grouped according to regions of the world and health conditions. Subgroup analyses were conducted to identify trends in RRs over time. These included calculating average RRs, standard deviation and change in RRs according to total follow up time. RESULTS: The search strategy yielded 1,767 publications. Combined with 20 reports and four websites, a total of 141 sources were used in the data extraction and analysis process. Following the data extraction, 121 registries capturing PROMs were identified. The overall average RR at baseline started at 71% and decreased to 56% at 10 + year at follow up. The highest average baseline RR of 99% was observed in Asian registries and in registries capturing data on chronic conditions (85%). Overall, the average RR declined as follow up time increased. CONCLUSION: A large variation and downward trend in PROMs RRs was observed in most of the registries identified in our review. Formal recommendations are required for consistent collection, follow up and reporting of PROMs data in a registry setting to improve patient care and clinical practice. Further research studies are needed to determine acceptable RRs for PROMs captured in clinical registries.

Addressing priorities for surgical research in Africa: implementation of a multicentre cloud-based peri-operative registry in Ethiopia.

In resource-constrained settings, where inequalities in access to and quality of surgical care results in excess mortality, peri-operative care registries are uncommon. A south-south collaboration supported the implementation of a context specific, clinician-led, multicentre real-time peri-operative registry in Ethiopia. Peri-operative information, including the Ethiopian Ministry of Health's national 'Saving Lives through Safe Surgery initiative', was linked to real-time dashboards, providing clinicians and administrators with information on service utilisation, surgical access, national surgical key performance indicators and measures of quality of care. We recruited four hospitals representing 285 in-patient beds from the Amhara and Southern Nations Nationalities and Peoples regions and Addis Ababa city, and reported on 1748 consecutive surgical cases from April 2019 to April 2020. Key performance indicators included: compliance with the World Health Organization's Surgical Safety Checklist in 1595 (92.1%) surgical cases; adverse events during anaesthesia in 33 (3.1%) cases; and surgical site infections in 21 (2.0%) patients. This collaboration has successfully implemented a multicentre digital surgical registry that can enable measurement of key performance indicators for surgery and evaluation of peri-operative outcomes. The peri-operative registry is currently being rolled out across the Amhara region and Addis Ababa city administration. It will provide continuous granular healthcare information necessary to empower clinicians to drive context-specific priorities for service improvement and research, in collaboration with national stakeholders and international research consortiums.

Cardiac Networks United: an integrated paediatric and congenital cardiovascular research and improvement network.

Optimising short- and long-term outcomes for children and patients with CHD depends on continued scientific discovery and translation to clinical improvements in a coordinated effort by multiple stakeholders. Several challenges remain for clinicians, researchers, administrators, patients, and families seeking continuous scientific and clinical advancements in the field. We describe a new integrated research and improvement network - Cardiac Networks United - that seeks to build upon the experience and success achieved to-date to create a new infrastructure for research and quality improvement that will serve the needs of the paediatric and congenital heart community in the future. Existing gaps in data integration and barriers to improvement are described, along with the mission and vision, organisational structure, and early objectives of Cardiac Networks United. Finally, representatives of key stakeholder groups - heart centre executives, research leaders, learning health system experts, and parent advocates - offer their perspectives on the need for this new collaborative effort.

The Preterm Clinical Network (PCN) Database: a web-based systematic method of collecting data on the care of women at risk of preterm birth.

BACKGROUND: Despite much research effort, there is a paucity of conclusive evidence in the field of preterm birth prediction and prevention. The methods of monitoring and prevention strategies offered to women at risk vary considerably around the UK and depend on local maternity care provision. It is becoming increasingly recognised that this experience and knowledge, if captured on a larger scale, could be a utilized as a valuable source of evidence for others. The UK Preterm Clinical Network (UKPCN) was established with the aim of improving care and outcomes for women at risk of preterm birth through the sharing of a wealth of experience and knowledge, as well as the building of clinical and research collaboration. The design and development of a bespoke internet-based database was fundamental to achieving this aim. METHOD: Following consultation with UKPCN members and agreement on a minimal dataset, the Preterm Clinical Network (PCN) Database was constructed to collect data from women at risk of preterm birth and their children. Information Governance and research ethics committee approval was given for the storage of historical as well as prospectively collected data. Collaborating centres have instant access to their own records, while use of pooled data is governed by the PCN Database Access Committee. Applications are welcomed from UKPCN members and other established research groups. The results of investigations using the data are expected to provide insights into the effectiveness of current surveillance practices and preterm birth interventions on a national and international scale, as well as the generation of ideas for innovation and research. To date, 31 sites are registered as Data Collection Centres, four of which are outside the UK. CONCLUSION: This paper outlines the aims of the PCN Database along with the development process undertaken from the initial idea to live launch.

Health outcomes measurement and organizational readiness support quality improvement: a systematic review.

BACKGROUND: Using outcome measures to advance healthcare continues to be of widespread interest. The goal is to summarize the results of studies which use outcome measures from clinical registries to implement and monitor QI initiatives. The second objective is to identify a) facilitators and/or barriers that contribute to the realization of QI efforts, and b) how outcomes are being used as a catalyst to change outcomes over time. METHODS: We searched the PubMed, EMBASE and Cochrane databases for relevant articles published between January 1995 and March 2017. We used a standardized data abstraction form. Studies were included when the following three criteria were fulfilled: 1) they relied on structural data collection, 2) when a structural and comprehensive QI intervention had been implemented and evaluated, and 3) impact on improving clinical and/or patient-reported outcomes was described. Data on QI strategies, QI initiatives and the impact on outcomes was extracted using standardized assessment tools. RESULTS: We included 21 articles, of which eight showed statistically significant improvements on outcomes using data from clinical registries. Out of these eight studies, the Chronic Care Model, IT application as feedback, benchmarking and the Collaborative Care Model were used as QI methods. Encouraging trends in realizing improved outcomes through QI initiatives were observed, ranging from improving teamwork, implementation of clinical guidelines, implementation of physician alerts and development of a decision support system. Facilitators for implementing QI initiatives included a high quality database, audits, frequent reporting and feedback, patient involvement, communication, standardization, engagement, and leadership. CONCLUSION: This review suggests that outcomes collected in clinical registries are supportive to realize QI initiatives. Organizational readiness and an active approach are key in achieving improved outcomes.

Patient data and patient rights: Swiss healthcare stakeholders' ethical awareness regarding large patient data sets - a qualitative study.

BACKGROUND: There is a growing interest in aggregating more biomedical and patient data into large health data sets for research and public benefits. However, collecting and processing patient data raises new ethical issues regarding patient's rights, social justice and trust in public institutions. The aim of this empirical study is to gain an in-depth understanding of the awareness of possible ethical risks and corresponding obligations among those who are involved in projects using patient data, i.e. healthcare professionals, regulators and policy makers. METHODS: We used a qualitative design to examine Swiss healthcare stakeholders' experiences and perceptions of ethical challenges with regard to patient data in real-life settings where clinical registries are sponsored, created and/or used. A semi-structured interview was carried out with 22 participants (11 physicians, 7 policy-makers, 4 ethical committee members) between July 2014 and January 2015. The interviews were audio-recorded, transcribed, coded and analysed using a thematic method derived from Grounded Theory. RESULTS: All interviewees were concerned as a matter of priority with the needs of legal and operating norms for the collection and use of data, whereas less interest was shown in issues regarding patient agency, the need for reciprocity, and shared governance in the management and use of clinical registries' patient data. This observed asymmetry highlights a possible tension between public and research interests on the one hand, and the recognition of patients' rights and citizens' involvement on the other. CONCLUSIONS: The advocation of further health-related data sharing on the grounds of research and public interest, without due regard for the perspective of patients and donors, could run the risk of fostering distrust towards healthcare data collections. Ultimately, this could diminish the expected social benefits. However, rather than setting patient rights against public interest, new ethical approaches could strengthen both concurrently. On a normative level, this study thus provides material from which to develop further ethical reflection towards a more cooperative approach involving patients and citizens in the governance of their health-related big data.

Conceptualizing patient-reported outcome measures for use within two Danish psychiatric clinical registries: description of an iterative co-creation process between patients and healthcare professionals.

BACKGROUND: Denmark has national clinical indicator programs for adult patients diagnosed with depression and schizophrenia, respectively. Within each program, the responsible steering group (SG) decided to add some indicators based upon patient-reported outcome measures (PROMs). AIMS: The primary aim was to describe the process of selecting PROMs and defining a national measurement concept for use in clinical practice and for indicator monitoring and the secondary aim s to collect patient recommendations for implementation. METHODS: An interdisciplinary SG of healthcare professionals and a Patient Peer Board (PPB) representing both patient groups co-created the output in an iterative process. The work included literature search, PPB workshops, SG meetings, ratings of PROM topics and items, and a pilot. The PPB discussed the following: item relevance, mode of data collection, graphical format of the online PROMs, and display of results. Finally, requirements for PROM patient information were identified. Based upon input from the PPB, the SG selected the items and specified the measurement concept. RESULTS: The PPB prioritized 20 of 53 suitable items and suggested alternative wording and answer categories. A pilot was performed and 19 items covering well-being, lack of well-being, impairment of functioning, and overall health were selected for clinical testing. The patients recommended concrete, unambiguous, easily understandable information and procedures for data collection and display of results. CONCLUSIONS: The iterative co-creation process based upon a high degree of patient involvement resulted in a set of PROMs, a national measurement concept, and patient recommendations for implementation. The cooperation between patients and professionals was successful.

[Biobanking requirements from the perspective of the clinician : Experiences in hematology and oncology]. / Biobankinganforderungen aus Sicht des Klinikers : Erfahrungen aus der Hämatologie und Onkologie.

The requirements for optimal biobanking from the point of view of the clinical partner can be highly variable. Depending on the material, processing, storage conditions, clinical data, and involvement of external partners, there will be special requirements for the participating clinician and specialist areas. What they all have in common is that the goal of any biobanking must be to improve clinical, translational, and basic research. While in the past biomaterials often had to be individually stored for each research project, modern biobanking offers decisive advantages: a comprehensive ethics vote fulfilling state-of-the-art data safety requirements, standardized processing and storage protocols, specialized biobank software for pseudonymization and localization, protection against power failures and defects of the equipment, centralized and sustainable storage, easy localization and return of samples, and their destruction or anonymization after completion of an individual project. In addition to this important pure storage function, central biobanking can provide a link to clinical data as well as the anonymous use of samples for project-independent research. Both biobank functions serve different purposes, are associated with specific requirements, and should be pursued in parallel. If successful, central biomaterial management can achieve a sustainable improvement of academic and non-academic biomedical research and the optimal use of resources. The close collaboration between clinicians and non-clinicians is a crucial prerequisite for this.

Rare disease registries: a call to action.

When registries collect accurate clinical data over time, they can act as fundamental support structures for patients and their families and powerful cost-effective instruments to support clinical trials and translational research to improve quality of care, quality of life and survival. Registries are critical for rare diseases (RD) with low prevalence and propensity for variation in treatment and outcomes. Rare Voices Australia is leading a call for action to the research and clinical community to prioritise RD data collection and develop an integrated RD Registry strategy for Australia. Financial, operational and governance challenges exist for establishing and maintaining RD registries. As a multidisciplinary team whose interests converge on RD, we highlight the need for the establishment of an Australian RD Registry Alliance. This 'umbrella' organisation will: (i) bring together existing RD registries across Australia; (ii) establish National RD Registry Standards to support interoperability and cohesion across registries; (iii) develop strategies to attract sustainable funding from government and other sources to maximise the utility of existing RD registries and support the development of new RD registries. The most important role for the Alliance would be to use the RD registries for translational research to address current knowledge gaps about RD and to improve the care for the over 1.4 million Australians estimated to live with RD.

Clustering and Residual Confounding in the Application of Marginal Structural Models: Dialysis Modality, Vascular Access, and Mortality.

In the application of marginal structural models to compare time-varying treatments, it is rare that the hierarchical structure of a data set is accounted for or that the impact of unmeasured confounding on estimates is assessed. These issues often arise when analyzing data sets drawn from clinical registries, where patients may be clustered within health-care providers, and the amount of data collected from each patient may be limited by design (e.g., to reduce costs or encourage provider participation). We compared the survival of patients undergoing treatment with various dialysis types, where some patients switched dialysis modality during the course of their treatment, by estimating a marginal structural model using data from the Australia and New Zealand Dialysis and Transplant Registry, 2003-2011. The number of variables recorded by the registry is limited, and patients are clustered within the dialysis centers responsible for their treatment, so we assessed the impact of accounting for unmeasured confounding or clustering on estimated treatment effects. Accounting for clustering had limited impact, and only unreasonable levels of unmeasured confounding would have changed conclusions about treatment comparisons. Our analysis serves as a case study in assessing the impact of unmeasured confounding and clustering in the application of marginal structural models.

Heart failure registries - Future directions.

Heart failure (HF) is a growing, global public health issue. Despite advances in HF care, many challenges remain and HF outcomes are poor. Some of the major reasons for this are the lack of understanding and treatment for certain HF sub-types as well as the lack of implementation of treatment in areas where effective treatment exists. HF registries provide the opportunity to transform clinical research and patient care. Recently the registry-based randomized clinical trial has emerged as a pragmatic and inexpensive alternative to the gold standard in clinical trial design, the randomized controlled trial. Registries may also provide platforms for strategy trials, implementation trials, and screening. Using examples from the Swedish Heart Failure Registry and others, the present review provides insights into how registry-based research can address many of the unmet needs in HF.

Automatic inference of ICD-10 codes from German ophthalmologic physicians' letters using natural language processing.

Physicians' letters are the optimal source of diagnoses for registries. However, most registries demand for diagnosis codes such as ICD-10. We herein describe an algorithm that infers ICD-10 codes from German ophthalmologic physicians' letters. We assess the method in three German eye hospitals. Our algorithm is based on the nearest-neighbor method as well as on a large thesaurus for ICD-10 codes. This thesaurus was embedded into a Word2Vec space created from anonymized physicians' reports of the first hospital. For evaluation, each of the three hospitals sent all diagnoses taken from 100 letters. The inferred ICD-10 codes were evaluated for correctness by the senders. A total of 3332 natural language terms had been sent in (812 hospital one, 1473 hospital two, 1047 hospital three). A total of 526 non-diagnoses were excluded upfront. 2806 ICD-10 codes were inferred (771 hospital one, 1226 hospital two, 809 hospital three). In the first hospital, 98% were fully correct and 99% correct at the level of the superordinate disease concept. The percentages in hospital two were 69% and 86%. The respective numbers for hospital three were 69% and 91%. Our simple method is capable of inferring ICD-10 codes for German natural language diagnoses, especially when the embedding space has been built with physicians' letters from the same hospital. The method may yield sufficient accuracy for many tasks in the multi-centric setting and can easily be adapted to other languages/specialities.

[Frequency and distribution of the active agent of intravitreal injections in German centers 2015-2021-An oregis study]. / Häufigkeit und Wirkstoffverteilung intravitrealer Injektionen in deutschen Zentren 2015 bis 2021 ­ eine oregis-Studie.

INTRODUCTION: Digitalization in medicine, especially the electronic documentation of patient data, is revolutionizing healthcare systems worldwide. The evaluation of real-world data collected under everyday conditions presents opportunities but also challenges. Electronic medical registries provide a means to compile extensive patient data for scientific inquiries. Oregis is the first nationwide digital registry for health services research established by the German Ophthalmological Society (DOG). Intravitreal operative medicinal injections (IVOM) are among the most frequently performed procedures in ophthalmology. Data on injection numbers and injection frequencies with anti-vascular endothelial growth factor (VEGF) are already available from other countries, whereas data at a national level are not yet available in Germany due to the lack of a nationwide register. It is known that the treatment success of anti-VEGF IVOMs depends largely on the adherence to treatment and thus on the number of injections. There are also differences in cost. In the context of this study, real-world data on the frequency and distribution of intravitreal injections in German centers from 2015 to 2021 were compiled for the first time since the introduction of oregis. The aim of this study is to collect data on the use of anti-VEGF IVOMs in Germany from oregis for the first time and to show the development of injection numbers and anti-VEGF drugs used. At the same time, the possibilities of data retrieval from oregis are demonstrated using a concrete example from daily ophthalmological practice. MATERIAL AND METHODS: An automated query of records was performed for all patients who received IVOM at oregis-affiliated healthcare facilities between 2015 and 2021. The number of treated patients and the use of anti-VEGF medications, including aflibercept, bevacizumab, brolucizumab, and ranibizumab, were determined. The data were collected in a pseudonymized and anonymized manner. RESULTS: At the time of data collection, 9 German ophthalmological healthcare facilities were affiliated with oregis. In total, 309,152 patients were registered during the observation period, with 8474 receiving IVOMs. Over the observation period, the number of participating centers, patients, and intravitreal injections increased. The proportional share of anti-VEGF agents among the total number of injections varied during the observation period. DISCUSSION: Real-world data captured in oregis offer significant potential for enhancing healthcare provision. Oregis enables the depiction of ophthalmological care conditions in Germany and contributes to research and quality assurance. The ability to query the presented data exemplifies the multitude of inquiries through which oregis can contribute to the representation of ophthalmological care in Germany.

Automating surgical procedure extraction for society of surgeons adult cardiac surgery registry using pretrained language models.

Objective: Surgical registries play a crucial role in clinical knowledge discovery, hospital quality assurance, and quality improvement. However, maintaining a surgical registry requires significant monetary and human resources given the wide gamut of information abstracted from medical records ranging from patient co-morbidities to procedural details to post-operative outcomes. Although natural language processing (NLP) methods such as pretrained language models (PLMs) have promised automation of this process, there are yet substantial barriers to implementation. In particular, constant shifts in both underlying data and required registry content are hurdles to the application of NLP technologies. Materials and Methods: In our work, we evaluate the application of PLMs for automating the population of the Society of Thoracic Surgeons (STSs) adult cardiac surgery registry (ACS) procedural elements, for which we term Cardiovascular Surgery Bidirectional Encoder Representations from Transformers (CS-BERT). CS-BERT was validated across multiple satellite sites and versions of the STS-ACS registry. Results: CS-BERT performed well (F1 score of 0.8417 ± 0.1838) in common cardiac surgery procedures compared to models based on diagnosis codes (F1 score of 0.6130 ± 0.0010). The model also generalized well to satellite sites and across different versions of the STS-ACS registry. Discussion and Conclusions: This study provides evidence that PLMs can be used to extract the more common cardiac surgery procedure variables in the STS-ACS registry, potentially reducing need for expensive human annotation and wide scale dissemination. Further research is needed for rare procedural variables which suffer from both lack of data and variable documentation quality.

Data quality assessment of the Dutch Breast Implant Registry by automated data verification using medical billing data.

BACKGROUND: The Dutch Breast Implant Registry (DBIR) provides real-time population-based data to monitor and improve the quality and safety of breast implants and to trace patients in the event of an (inter)national recall. To serve these main goals, the capture rate and percentage of implants that are actually registered should be known and data should be complete. This study aimed to describe an automated verification process for capture rates and data completeness using medical billing data as the external source. METHODS: DBIR-data on reconstructive permanent breast implants and tissue expanders (TEs) from 2019 were compared to medical billing data. At the hospital level, the capture rate of DBIR and accuracy of the selected data points were assessed. RESULTS: In total, 2389 implants (1420 patients) were included from 12 participating hospitals (11% of all healthcare institutions registering in DBIR). DBIR had capture rates of 99% to 114% for inserted permanent implants and TEs and 49% for explanted permanent implants and TEs. Among the 9015 data points analyzed in DBIR, 8861 (98%) matched the medical billing data. Mastopexy had the highest matching percentage (100%) and capsulectomy the lowest (86%). CONCLUSION: This study showed varying capture rates in DBIR depending on the intervention group, indicating that registration of DBIR-data and medical billing codes could be further optimized. Data accuracy was high as only 2% of data points did not correspond to medical billing data. For future data verification, other external data sources could be explored, including sources that enable verification of cosmetic implants.

Implementation of an Atrioventricular Valve Intervention Registry: Comparative Study of REDCap vs. CDR-Based openEHR Registry.

This comparative study examines the transition from isolated registries to a consolidated data-centric approach at University Hospital Schleswig-Holstein, focusing on migrating the Atrioventricular Valve Intervention Registry (AVIR) from REDCap to a Medical Data Integration Center based openEHR registry. Through qualitative analysis, we identify key disparities and strategic decisions guiding this transition. While REDCap has historical utility, its limitations in automated data integration and traceability highlight the advantages of a data-centric approach, which include streamlined data (integration) management at a single-point-of-truth based on e.g., centralized consent management. Our findings lay the groundwork for the AVIR project and a proof-of-concept data-centric registry, reflecting a broader industry trend towards data-centric healthcare initiatives.

Registry-based study comparing health-related quality of life between patients with primary rectal cancer and locally recurrent rectal cancer.

AIM: National clinical registries offer the benefits of a comprehensive dataset, particularly when linked with patient-reported outcome (PRO) data. This aim of this study was to utilise UK registry data to assess cross-sectional differences in health-related quality of life (HrQoL) in patients with primary rectal (PRC) and locally recurrent rectal cancer (LRRC). MATERIALS AND METHODS: Data were extracted from the COloRECTal cancer Repository (CORECT-R) and the Locally Recurrent Rectal Cancer - Quality of Life (LRRC-QoL) datasets. Propensity score matching was undertaken in a 1:1 ratio using two covariates: age and sex. The primary outcome was the FACT-C Colorectal Cancer Subscale (CCS). Statistical significance was determined using p < 0.05 and clinical significance using effect size (ES) and minimally important clinical difference (MCID). RESULTS: A matched cohort with 72 patients in each group was identified. Overall FACT-C CCS scores were worse in patients with LRRC from a statistical (11.80 vs 18.03, p < 0.001) and clinically meaningful perspective (ES 1.63, MCID 6.23). Patients with PRC reported better digestion (p < 0.001, ES 0.85), better control over their bowels (p < 0.001, ES 1.03) and increased appetite (p < 0.001, ES 1.74, MCID 2.08). Patients with LRRC reported worse stomach swelling (p < 0,001, ES 0.97) and more diarrhoea (p < 0.001, ES 0.92), however they reported better body image (p < 0.001, ES 0.80). CONCLUSION: Patients with LRRC reported significantly worse overall scores in the FACT-C CCS from both a statistical and clinical perspective, demonstrating the ability of the FACT-C to distinguish between these patient groups and the benefits of the inclusion of PROs within colorectal cancer registries, specifically including patients with advanced/recurrent disease.

Development of a Bariatric Surgery Core Data Set for an International Registry.

PURPOSE: Bariatric and metabolic surgery is an effective treatment for severe and complex obesity; however, robust long-term data comparing operations is lacking. Clinical registries complement clinical trials in contributing to this evidence base. Agreement on standard data for bariatric registries is needed to facilitate comparisons. This study developed a Core Registry Set (CRS) - core data to include in bariatric surgery registries globally. MATERIALS AND METHODS: Relevant items were identified from a bariatric surgery research core outcome set, a registry data dictionary project, systematic literature searches, and a patient advisory group. This comprehensive list informed a questionnaire for a two-round Delphi survey with international health professionals. Participants rated each item's importance and received anonymized feedback in round 2. Using pre-defined criteria, items were then categorized for voting at a consensus meeting to agree the CRS. RESULTS: Items identified from all sources were grouped into 97 questionnaire items. Professionals (n = 272) from 56 countries participated in the round 1 survey of which 45% responded to round 2. Twenty-four professionals from 13 countries participated in the consensus meeting. Twelve items were voted into the CRS including demographic and bariatric procedure information, effectiveness, and safety outcomes. CONCLUSION: This CRS is the first step towards unifying bariatric surgery registries internationally. We recommend the CRS is included as a minimum dataset in all bariatric registries worldwide. Adoption of the CRS will enable meaningful international comparisons of bariatric operations. Future work will agree definitions and measures for the CRS including incorporating quality-of-life measures defined in a parallel project.